Handbook Tech Standar TV Production

ITC
HANDBOOK OF TECHNICAL STANDARDS

FOR
TELEVISION PROGRAMME PRODUCTION
PART A
ISSUE 2.0 - DECEMBER 1996
Standards and Technology, Engineering Division,

Independent Television Commission,
Kings Worthy Court, Kings Worthy, Winchester, Hants SO23 7QA
ITC Handbook of Technical Standards for Television Programme Production
Issue 2.0-December 1996
Contents & Introduction
On-line version source EBU website
Independent Television Commission, 1992, 1993 and 1996.

All rights reserved. No reproduction, copy or transmission of this publication may be made without the
written permission of the Independent Television Commission.
II
Issue 2.0 December 1996
HANDBOOK OF TECHNICAL STANDARDS

PART A:
STUDIO CENTRES AND OUTSIDE BROADCAST FACILITIES
CONTENTS
CONTENTS
III
INTRODUCTION
VI
Technical Performance Working Party Membership
VI
Past Membership
VI
SECTION 1 VIDEO CIRCUITS AND EQUIPMENT WITHIN THE SIGNAL PATH
1.1
PERFORMANCE FIGURES (COMPOSITE PATHS)
1.2
RECOMMENDED TEST METHODS (COMPOSITE PATHS)
1.3
PERFORMANCE FIGURES (COMPONENT PATHS)
1.4
RECOMMENDED TEST METHODS (COMPONENT PATHS)
SECTION 2 AUDIO CIRCUITS AND EQUIPMENT WITHIN THE SIGNAL PATH
13
2.1.
PERFORMANCE FIGURES
13
2.2
RECOMMENDED TEST METHODS
15
SECTION 3 VIDEO TAPE RECORDERS
19
3.1.
PERFORMANCE FIGURES (COMPOSITE RECORDERS)
19
3.2.
RECOMMENDED TEST METHODS (COMPOSITE RECORDERS)
21
VIDEO MEASUREMENTS
21
AUDIO MEASUREMENTS
23
3.3
PERFORMANCE FIGURES (COMPONENT RECORDERS)
25
3.4
RECOMMENDED TEST METHODS (COMPONENT RECORDERS)
26
SECTION 4 AUDIO RECORDERS
31
4.1.
PERFORMANCE FIGURES
31
4.2.
32
SECTION 5 CAMERAS
35
5.1.
PERFORMANCE FIGURES
35
5.2
36
SECTION 6 TELECINES AND SOUND FOLLOWERS
43
6.1.
43
PERFORMANCE FIGURES
VIDEO TOLERANCES
43
AUDIO TOLERANCE
45
6.2
46
VIDEO MEASUREMENTS
46
SECTION 7 DISC REPRODUCERS
55
7.1.
55
PERFORMANCE FIGURES
III

7.2
56
SECTION 8 WAVEFORMS
59
8.1
SOUND-IN-SYNCS
Fig 8.1 - Sound-in-Syncs Signal
59
59
8.2
VERTICAL BLANKING INTERVAL
59
SECTION 9 PEAK PROGRAMME METERS
63
9.1
PERFORMANCE FIGURES
63
9.2
64
SECTION 10 SATELLITE LINKS PATHS
67
10.1
PERFORMANCE FIGURES (Vision)
67
10.2
RECOMMENDED TEST METHODS (Vision)
69
10.3
PERFORMANCE FIGURES (Sound)
74
10.4
RECOMMENDED TEST METHODS (Sound)
75
SECTION 11 DIGITAL VIDEO CIRCUITS AND EQUIPMENT
77
11.1
RECOMMENDED CRITERIA
77
11.2
81
SECTION 12 DIGITAL AUDIO CIRCUITS AND EQUIPMENT
87
12.1
87
12.2
89
REFERENCE SECTION
93
Ref. 1: Pulse and Bar test signals (ITU-R BT.451-2)

Ref. 2: Staircase test signal (ITU-R BT.451-2)
Ref. 3: Differentiating and shaping network
Ref. 4: Typical K-rating graticule
Ref. 5: 50 Hz Square-wave test signals (ITU-R BT.451-2)
Ref. 6: Chrominance Pulse and Bar test signals
Ref. 7: Characteristics of Weighting Filters for video noise measurements (ITU-R BT.567)
Ref. 8: 100.0.100.0 Colour Bars (100%)
Ref. 9: Characteristics of Filters for audio noise measurements (ITU-R BT.468-4)
Ref. 10: Specification of Wow and Flutter meter (ITU-R BT.409-2)
Ref. 10: Specification of Wow and Flutter meter (continued)
Ref. 11: Specification of Rumble Meter to BS7063
Ref. 12: Test Pattern for measurement of Telecine long-term Streaking
Ref. 13: Test patterns for measurement of Telecine Flare
Ref. 14: Measurement of Film frame steadiness
Ref. 15: List of Test Films and Tapes for Telecine
Ref. 16: Test Patterns for Camera Tests
Ref. 16: Test Patterns for Camera Tests (continued)
Ref. 17: Picture Zones
Ref. 18: 5-point Impairment scale
Ref. 19: Crosstalk and Phase Profiles
Ref. 19: Crosstalk and Phase Profiles (continued)
Ref. 20: Delay Inequality Test Signals
Ref. 21: Non-linearity Test Sawtooth Signals
IV
93
94
94
95
96
97
98
99
100
101
102
103
104
105
106
106
107
108
109
110
111
112
113
113
114
115
116
117
118
119
120
Ref. 22: Non-linearity Test Staircase Signals

Ref. 23: Typical Multiburst Test Signals
Ref. 24: Colour Difference Noise Filter
Ref. 25: Vertical Synchronising and Blanking waveformsError! Bookmark not defined.
Ref. 26: Field interval Blanking of the Colour Burst
Ref. 27: Allocation of VBI Lines
Ref. 28: Insertion Test Signals
Ref. 29: Teletext Data signals in the VBI
Ref. 30: Widescreen signalling in Line 23
Ref. 31: Status bits for Widescreen signalling
Ref. 32: Ghost Cancellation Reference signals
Ref. 33: Parameters for GCR signals
Ref. 34: Analogue and Digital sync. And blanking timing
Ref. 35: Timing reference signals
Ref. 36: Audio channel status
Ref. 37: Ramp signal for Noise Measurement
Ref. 38: Colour Gamut in R, G, B and Y, Cr, Cb domains
Ref. 39: Audio Timing Reference
Ref. 40: Noise in the presence of signal
Ref. 41: DFIM performance profile
Ref. 42: IMD and Harmonic Distortion performance profile
Ref. 43: Group Delay profile
Ref. 44: 16:9 Telecine alignment test film
NOTES:
121
122
123
124
125
125
126
127
128
128
129
130
131
132
133
134
134
135
135
136
136
137
138
141
INTRODUCTION
This handbook contains performance figures for the main elements of the video and audio equipment
and the signal paths used in television programme production, and incorporates recommended test
procedures for checking compliance with these performance criteria. The main body of the ITC
Technical Performance Code requires licensees to use reasonable endeavours to ensure conformance
with these performance targets, stopping short of requiring conformance as an absolute condition.
However, the ITC may at its discretion require conformance with specified targets if tests have
demonstrated poor performance resulting in a less than high standard of technical quality.
Developments in technology, for example resulting in the introduction of a new VTR format or new
camera technology, during the period of the licence might necessitate revision of this handbook. The
ITC intends to discuss such revisions with licensees, as appropriate. Any revisions to existing Sections,
or additional Sections, will be considered by the Technical Performance Working Party which consists
of a membership drawn from the ITC Engineering Division and the licensees.
Technical Performance Working Party Membership

C Girdwood
ITC
(Chairman)
C Hunt
ITC
(Secretary)
P Gray
Anglia Television
M Hughes
Carlton Television
R Hurley
Channel Four Television
R Soczywko -
Granada Television
R White
Meridian Broadcasting
Past Membership
P Ballabon (London Weekend Television), I Dutton (Tyne Tees Television), C Hibbert (Carlton
Television), C Hunter (Scottish Television), P Marshall (Channel Four Television), J Nichol (Tyne
Tees Television), R Pickles (Granada Television), J Rogers (Yorkshire Television), T Ross (Scottish
Television) and S Waring (Thames Television).
The Technical Performance Working Party would like to thank the many people and organisations for
their help in the preparation of the Handbook.
VI
Section 1: Video Circuits & Equipment within the signal path
SECTION 1
VIDEO CIRCUITS AND EQUIPMENT WITHIN THE SIGNAL PATH
1.1
PERFORMANCE FIGURES (COMPOSITE PATHS)
1.1.1
Definitions and Operational Practices
Direct Path
For purposes of measurement the direct path is assumed to comprise the circuit from the agreed
interface with British Telecom or Transmission Operators equipment, through the Presentation and
Master Control switching and processing equipment back to the agreed interface with British Telecom
or Transmission Operator's equipment.
The limits in brackets refer to the situation when a synchroniser is included in the path.
Worst Path
For the purposes of measurement, the worst path is assumed to comprise the following with all
interconnections carried out using the normal equipment routes:(i)
The source studio mixer
(ii)
A looped VTR path
(iii)
A second studio mixer
(iv)
A second looped VTR path
(v)
The Presentation and Master Control Path
The tolerance limits do not include degradations due to signal sources such as cameras, telecines or
video tape recorders, as tolerances for these are separately specified.
The limits in brackets refer to the situation when digital video effects are included in the path.
A measurement of the Worst Path parameters is normally only necessary after the completion of a new
installation or a major re-installation.
Production Path
For the purposes of measurement the production path will comprise that part of the system that starts at
the output of originating equipment (camera or VTR) and that includes assignment switching, mixing
and effects equipment and ends at the interface with a VTR or the Master Control Room. The path
may be in a studio centre or outside broadcast scanner.
The limits in brackets refer to the situation when digital video effects are included in the path.
O.B. Link Paths
OB link tolerances are related to an unspecified number of point-to-point SHF links. Measurement is
made at the final output of the link at which point connection to a permanent circuit would be made.
1.1.2
(a)
(b)
Signal Levels
Signal Level
Adjustment Error
Signal Level Gain Stability
Direct
Path
Worst
Path
0.7V
2%
2%
0.7V
2%
5%
Production O.B. Link

Path
Path
0.7V
2%
2%
0.7V
2%
2%
1
1.1.3
(a)
Linear Waveform Distortion

2T Pulse-to-Bar Ratio
(b)
2T Pulse Response
(c)
2T Bar Response
(d)
50 Hz Square Wave Response
(e)
Chrominance/Luminance Gain
Inequality
Chrominance/Luminance Delay
Inequality
(f)
1.1.4
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
1.1.5
(a)
(b)
(c)
1.1.6
(a)
(b)
1.1.7
(a)
(b)
(c)
(d)
1.1.8
Direct
Path
Worst
Path
% K
(1% K)
% K
(1% K)
% K
(1% K)
% K
(1% K)
3%
1% K
(2% K)
1% K
(2% K)
1% K
(2% K)
1% K
(2% K)
4%
% K
(1% K)
% K
(1% K)
% K
(1% K)
% K
(1% K)
3%
20 ns
40 ns
20 ns
20 ns
5%
5
5
5%
3%
2
2
3%
5%
5
5%
5%
5%
Non-Linearity Distortion
Luminance Line Time Non-Linearity
3%
Differential Phase
2
Burst/Chroma Phase
2
Differential Gain
3%
Transient Gain Change, Luminance
Transient Gain Change,
Chrominance
Transient Gain Change, Sync
Chrominance / Luminance Crosstalk
Input/Output Impedance-Return Loss
Luminance
-30 dB
Chrominance
-30 dB
Low Frequency
-30 dB
VLF Response
First Overshoot
Second Overshoot
Noise
Weighted Luminance (RMS)
-64 dB
(-60 dB)*
Weighted Chrominance (RMS)
-58 dB
Total Low Frequency
-45 dB
Random and Periodic (p-p)
Interchannel Crosstalk
-55 dB
Modulation Derived Distortion

Path
Path
2% K
2% K
2% K
2% K
4%
5%
3%
-30 dB
-30 dB
-30 dB
-30 dB
-30 dB
-30 dB
-30 dB
-30 dB
-30 dB
20%
8%
-58 dB
-55 dB
-52 dB
-45 dB
-64 dB
(-60 dB)*
-58 dB
-45 dB
-45 dB
-52 dB
-52 dB
-40 dB
(Sound to Vision Crosstalk)

2
Direct
Path
-
(a)
(b)
Worst
Path
-

Path
Path
-52 dB
-57 dB
Sound Subcarrier Modulated

Sound Subcarrier Unmodulated
(Level of Intermodulation products
between sound and
chrominance subcarriers)
The figure applies to 8-bit processors. If 9-bit processors are used the figure should be
improved by 3-4 dB.
1.2
RECOMMENDED TEST METHODS (COMPOSITE PATHS)
1.2.1
Test Conditions
Before commencing a measurement, all test equipment should be checked for accuracy. Any
inaccuracies should be corrected if possible, or noted and allowed for in the measurement.
This section gives examples of test methods that use basic techniques. These examples do not preclude
the use of other valid methods. The use of ITS type test signals is also not precluded but the ITU
waveforms referred to in these notes are regarded as the primary standard.
The signals specified below are applied to the path under test; when vision mixers are included in the
path then the route should include the shortest normally used path through each vision mixer and any
processing amplifiers that are normally used. The processing amplifiers should be set to the mode in
which they are normally used operationally.
1.2.2
(a)
Signal Levels
Signal Level Adjustment Error
The test may be carried out using a calibrated television waveform monitor.
The signal level adjustment error may be measured by using a 75 ohm generator of the 2T Pulse and
Bar test signal as shown in Reference Section, Ref. 1. The generator should be adjusted so that the bar
amplitude is 700 mV and the synchronising pulse amplitude is 300 mV. The sine-squared pulse is
ignored in this application. The difference in amplitude of the bar centre at the output, expressed as a
percentage of 700 mV, is taken as the signal level adjustment error.
(b)
Having completed the measurements in 1.2.2 (a), no level adjustments should be made for a period of
one hour. The measurements of 1.2.2 (a) should then be repeated using the identical path and any
change recorded as the parameter for this section.
1.2.3
(a)

The test signal should be the 2T Pulse and Bar Waveform as specified in Reference Section Ref. 1.
The pulse-to-bar K-rating is defined as:
K=
BP
x 100%
4P
Where B and P are the amplitudes of the bar and pulse respectively.
Therefore, in practice, to make the measurement, the pulse will be taken as reference.
Set the pulse amplitude to be 100% on the centre scale of an appropriate graticule (Reference Section
Ref. 4) and divide by four the percentage difference in amplitude between the pulse and the bar
measured at its mid-point, to obtain the K-rating.
When the waveform is subject to line tilt or an extended distortion along the leading edge at the top of
the bar, the amplitude of the bar must be measured at its midpoint after first setting the blanking level
mid-way between two successive bars to 0%.
b)
2T Pulse Response
The test signal should be the 2T Pulse and Bar Waveform as specified in Reference Section, Ref. 1.
Measurement may be made using a graticule such as that shown in Reference Section, Ref. 4.
The vertical gain is adjusted to make the pulse amplitude 100% and then the vertical shift moved to
bring the blanking level onto the base line at 30%. The horizontal gain is advanced and the horizontal
shift adjusted to make the waveform touch the H.A.D. markers on the 80% line. With normal gain the
graticule markers are 2% K and 4% K. For 1% K and 2% K the calibrated vertical gain is advanced by
2. For limits of % K and 1% K the pulse amplitude is first set to 80% and the calibrated vertical
gain then advanced by 5.
If it is desired to measure the K rating exactly, the variable vertical gains should be adjusted until the
worst pulse overshoot just touches the inner limits. The calibrated gain is then returned to normal and
the amplitude of the pulse measured (P%) then
K=
200
%
P x Calibrated gain
This is illustrated in the following table:Pulse Amplitude

100
80
67
57.5
50
(c)
5 Gain
0.4% K
0.5% K
0.6% K
0.7% K
0.8% K
2T Bar Response
The test signal should be the 2T Pulse and Bar waveform as specified in Reference Section, Ref. 1.
The horizontal timebase of the oscilloscope is adjusted so that the half amplitude points of the bar
reach the outer limits marked on a graticule such as that shown in Reference Section, Ref. 4.
Ignoring the first and last 2.5% (0.625 ms) of the bar, the deviation from its mid-point, expressed as a
percentage of its amplitude at that point, is the K rating of the bar. It must be emphasised that
measurements are made using only half the bar, the worst half being quoted as the result. It is wrong to
measure the whole bar and divide by two to obtain the K rating.
(d)
The test signal should be the 50 Hz square wave test signal as specified in Reference Section, Ref. 5.
With the horizontal scan at field rate the 50 Hz signal is adjusted as in 1.2.3. (c). For a stationary
display the signal must contain field synchronising pulses. Again, ignoring the first and last 2.5%
(250 ms) the percentage deviation of the worst half divided by 2 is the K rating of the bar. (It may be
noted that for the same deviation on the display a 4% K figure for the bar response looks the same as a
2% K for 50 Hz).
4
(e)
Chrominance/Luminance Gain Inequality
The measurement is best made using the 2Tc non-composite waveform (Reference Section, Ref. 6b).
The 50% luminance pedestal is used to calibrate the vertical gain of the oscilloscope. The chrominance
amplitude is then measured directly.
NOTE:
(f)
The use of the composite 2Tc waveform with a gain and delay test set will produce an
erroneous result in the presence of chrominance/luminance crosstalk.
Chrominance/Luminance Delay Inequality
The measurement is made using a 2Tc composite Pulse-and-Bar signal (Reference Section, Ref. 6a)
and a delay measuring test set where available.
The output of the test set is viewed on an oscilloscope and the test set adjusted to cancel any path
chrominance/luminance delay inequality. If a test set is not available then the level of distortion should be
estimated by examining the sinusoidal distortions at the bottom of the 2Tc composite pulse on a
waveform monitor or oscilloscope. The method is described in Part B, Guidelines.
1.2.4
(a)
The test signal consists of a 5-step staircase (Reference Section, Ref. 2 occupying one line in every
four, followed by three lines of black or white. Measurements are made with three lines of white (bar
on) and with three lines of black (bar off) and the worst result quoted.
It should be noted that the staircase with added sub-carrier waveform is used to conform with C.C.I.R.
recommended practice.
At the receiving end the test signal is passed through a suitable differentiating network (Reference
Section, Ref. 3) and amplifier and displayed on an oscilloscope. The result is a train of five pulses.
Non-linearity is measured as the difference in amplitude between the largest and the smallest
expressed as a percentage of the largest.
i.e.
(b)
E max E min
x 100%
E max
Differential Phase
The test signal should be a 5 step staircase with added subcarrier (Reference Section, Ref. 2).
The differential phase may be measured by using a vectorscope in the line-time mode. The six sections
of subcarrier are compared for their phase relationships taking the blanking level section as a
reference. The differential phase is defined as the largest departure in phase from that reference.
Measurements are made with the white bar on and with the white bar off and the worst measurement is
quoted.
(c)
Burst/Chroma Phase
Burst/Chroma Phase errors may be measured as follows. Display the output of a colour bar generator
directly on a vectorscope and after aligning the burst on the graticule, carefully measure the phase
displacement (if any) of the BLUE bar. Apply the colour bar signal to the equipment or path under test
and display the output signal on the vectorscope. After aligning the burst on its graticule, measure
again the BLUE bar phase displacement. Phase measurement minus the phase displacement of the
original signal indicates the burst/chroma distortion due to the equipment or path under test.
(d)
Differential Gain
The test signal should be a 5 step staircase with added subcarrier (Reference Section, Ref. 2.)
The differential gain may be measured by using a vectorscope in the line time mode. The six sections
of subcarrier are compared for their amplitude relationship and taking the blanking level section as a
reference, the differential gain is defined as the largest departure in amplitude from that reference.
quoted.
NOTE ON TRANSIENT DISTORTION APPLICABLE TO SUB-SECTIONS (e), (f) & (g)
BELOW
The transient gain change due to a change of APL is defined as the maximum transient departure in the
amplitude of each component from that which existed before the change in APL, expressed as a
percentage of the original amplitude. Separate measurements are made on the five step staircase with
added subcarrier (Reference Section, Ref. 2), with the APL changed from low (intervening lines at
blanking level) to high (intervening lines at white level) and from high to low.
(e)
Transient Gain Change, Luminance (see note above)
Section, Ref. 2), amplified and displayed on an oscilloscope (some commercial filters with amplifiers
overload at normal signal level and require some 10 dB reduction of input signal level).
The oscilloscope should be synchronised by an external source and the black level clamp or dc restorer
should be switched off. Movement of the base line of the waveform when the APL is changed
indicates overload or some other non-standard measuring condition.
Set up the oscilloscope to make the amplitude of each of the spikes corresponding to the steps in turn
equal to 100% with intervening lines at black. Measure the maximum transient departure from 100%
of each of the spike amplitudes when the APL is switched from low to high and vice-versa.
The largest departure from 100% is taken as the result and it should be noted whether the change is
predominantly on only one spike and if so, on which spike.
(f)
Transient Gain Change, Chrominance (See note above)
Set up the oscilloscope using the chrominance filter and measure the maximum transient departure
from 100% of the peak-to-peak subcarrier amplitude on the third step, when the APL is switched from
low to high and vice-versa.
(g)
Transient Gain Change, Sync (See note above)
Using the differentiating network, amplifier and oscilloscope as in (e) above, set the oscilloscope so
that the amplitude of the positive spike corresponding to the trailing edge of sync equals 100% with
intervening lines at black.
Measure the maximum transient departure from 100% of the spike amplitude when the APL is
switched from low to high and vice-versa.
(h)
Chrominance to Luminance Crosstalk
The 2Tc pulse and bar waveform (Reference Section, Ref. 6b) should be used for the test. The
crosstalk, which manifests itself as a change in the mean level of the pedestal during transmission of
the chrominance component, should be expressed as a percentage of the picture level, as determined
by the measurement described in Para 1.2.2. (a), (nominally 700 mV).
1.2.5
(a)(b)(c)
Input/Output Impedance - Return Loss

Return Loss
The measuring point for this test is the same interface as defined in Section 1.1.1. Direct Path.
6
The test is first carried out using a 2T pulse-and-bar waveform in conjunction with a return loss bridge
which should first be calibrated using two very closely matched, 75 0.1% ohm resistors.
In addition, the same leads should be used for calibration and measurement and the reference path lead
should be identical to the main path connection. With one return loss bridge presently available, a
calibration distortion of -40 dB is provided; if this bridge is used the output is displayed on an
oscilloscope and adjusted to give a reference display (5 divisions for example). The bridge is then
rearranged to include the circuit under test and the unbalance output measured. The return loss is then
calculated by linear interpolation. For large mismatches a 10 dB switch is incorporated in the bridge to
allow calibration at -30 dB. When measuring output impedance the input signal should be removed
and the input terminated. For very small return loss measurements an external trigger to the
oscilloscope is often necessary.
The test should be repeated using the 2Tc pulse-and-bar and the 50 Hz waveforms. These results are
respectively recorded as the (a) Luminance, (b) Chrominance and (c) Low Frequency parameters.
1.2.6
V.L.F. Response
The signal used should switch all lines to black and white. The switching should occur at a sufficiently
slow rate to allow the waveform to settle before the following transition. The 1st and 2nd overshoots of
blanking level variation are measured (Fig. 1.1) and expressed as a percentage of standard picture level
(700 mV peak-to-peak).
It should be noted that the dc change ("c" in Fig. 1.1) is not measured since it is a function only of the
test signal.
Both the black to white and the white to black transitions are measured and the worst result quoted.
A dc-coupled oscilloscope with a very slow timebase may be used for these measurements.
Alternatively, if only a television waveform monitor is available, a line rate display should be used
with the Y amplifier switched to dc coupled and the dc restorer switched off.
Fig. 1.1
1.2.7
Noise
Measurement is made using a 10% lift signal. Care should be taken that the noise of the generated
signal is not significant. When measurements are made on paths containing digital processing
equipment the 10% lift signal may be adjusted slightly to minimise the effects of quantisation noise or
a ramp waveform may be used.
(a)
Weighted Luminance
Measurement is made in the band 10 kHz (7.5 kHz) - 5.0 MHz using RMS detection. The
characteristic of the luminance weighting filter is shown in Reference Section, Ref. 7.
(b)
Weighted Chrominance
Measurement is made in the band 3.5 MHz - 5.5 MHz as defined only by the characteristic of the
weighting filter as shown in Reference Section, Ref. 7 and using RMS detection.
(c)
Total L.F. Random and Periodic
Measurement is made peak-to-peak in the band 40 Hz - 10 kHz (7.5 kHz).

(d)
One channel is selected as the one to be measured as the receiver of crosstalk interference. This
channel is fed with a blanking and sync waveform.
Another channel that is considered to be the nearest electrically adjacent channel is used as the hostile
channel. This is fed with Colour Bars (Reference Section, Ref. 8).
The signal-to-crosstalk ratio is defined as the ratio, expressed in decibels, of the normal peak-to-peak
amplitude of the picture signal to the peak-to-peak amplitude of the crosstalk waveform.
1.2.8
a)
Modulation Derived Distortion (Sound to Vision Crosstalk)

Measured with whole-time 5-step staircase, without chrominance sub-carrier, into the vision channel
and +8 dBu at 1 kHz into the sound channel. The crosstalk should be measured unweighted, peak-topeak, in the frequency band 40 Hz to 10 kHz (7.5 kHz) using a noise measuring set. The result is
expressed with reference to standard picture level (700 mV p-p).
(b)
Measured with whole time 5-step staircase, with chrominance subcarrier, into the vision channel and
no sound modulation. The crosstalk should be measured luminance weighted, peak-to-peak in the
frequency band 40 Hz to 5.0 MHz, using a noise measuring set. The result is expressed with reference
to standard picture level (700 mV p-p).
N.B. Other methods of measurement using spectrum analysis are acceptable.
1.3
PERFORMANCE FIGURES (COMPONENT PATHS)
This section gives the performance figures for production component paths (excluding digital effects
equipment).
A Company may reserve the right to test all three component channels to the luminance channel
performance. This may be necessary in ensuring good chroma-keying signals where a wide bandwidth
in the colour difference channels is required. In this case the linear waveform distortions should be
measured using the 2T Pulse and Bar signal and the noise performance should be the same as in the
luminance channel.
1.3.2
Signal Levels
(a)
Signal Level
Adjustment Error
(b)
1.3.3
(a)
(b)
8
0.7 V
2%
2%
Linear Waveform Distortions (2T Pulse and Bar in Luminance

and 5T Pulse and Bar in Colour Difference Channels)
Pulse-to-Bar Ratio
0.5% K
Pulse Response
0.5% K
(c)
(d)
Bar Response
1.3.4
Delay Inequality between all Component Channels

Timing difference
10 ns
Non-Linear Waveform Distortion
Amplitude Non-Linearity
3%
Component Crosstalk
-50 dB
Noise
-64 dB (-60 dB)*
Colour Difference (RMS) - 1.6 MHz low pass filter
-64 dB
Total Low Frequency in Luminance Channel (p-p)
-45 dB
Total Low Frequency in Colour Diff. channels (p-p)
-43 dB
The figure applies to 8-bit processors. If 9-bit processors are used the figure
should be improved by 3-4 dB.
1.3.5
(a)
(b)
1.3.6
(a)
(b)
(c)
(d)
*
1.4
0.5% K
0.5% K
RECOMMENDED TEST METHODS (COMPONENT PATHS)
1.4.1
Test Conditions
The analogue component signals (Y, Pr and Pb) will be related to the colour separation signals (R, G
and B) by the following matrix equations:
Y
Pb
Pr
0.299
-0.169
0.500
0.587
-0.331
-0.419
0.114
0.500
-0.081
R
G
B
The colour separation signals have normal peak amplitudes of 700 mV. Synchronising signals may be
added to or kept separate from the luminance component.
The following methods are relevant when component paths are being measured and it may be
convenient to use a waveform monitor capable of displaying the three components simultaneously:
1.4.2
(a)
Signal Levels
The test may be carried out using a calibrated television waveform monitor and suitable graticule.
The insertion gain may be measured by using a 75 ohm generator of the 2T and 5T Pulse and Bar test
signals as shown in the Reference Section, Ref. 1. The 2T signal is fed to the Luminance Channel and
the 5T to the Colour Difference channels when required. The generator should be adjusted so that the
bar amplitude is 700 mV in both cases and the synchronising pulse amplitude is 300 mV in the
luminance channel. The difference in amplitude of the bar centre at the output, expressed as a
(b)
one hour. The measurements should be repeated using the same channels and any change recorded as
the parameter for this section.
1.4.3
Linear Waveform Distortions
Linear waveform distortions (a) to (c) are measured using the 2T Pulse and Bar test signal in the Y
(luminance) channel and the 5T Pulse and Bar test signal in the Pr and Pb (colour difference) channels.
The waveforms are shown in the Reference Section, Ref. 1.
(a)
Pulse-to-Bar Ratio
The pulse-to-bar K rating is defined as:

K=
BP
x 100%
4P
where B and P are the amplitudes of the bar and pulse respectively.
In practice, the pulse will be taken as the reference during measurement.
Set the pulse amplitude to be 100% on the centre scale of an appropriate graticule (Reference Section,
measured at its mid-point, to obtain the K rating.
the bar, the amplitude of the bar must be measured at its mid-point after first setting the reference level
(blanking for Y) mid-way between two successive bars to 0%.
(b)
Pulse Response
Measurement may be made using a graticule such as that shown in the Reference Section, Ref. 4.
bring the reference level (blanking for Y) onto the baseline at 30%. The horizontal gain is advanced
and the horizontal shift adjusted to make the waveform touch the H.A.D. markers on the 80% line.
With normal gain the graticule markers are 2% K and 4% K. For 1% K and 2% K the calibrated
vertical gain is increased by x2. For 0.5% K and 1% K the pulse amplitude is first set to 80% and the
calibrated vertical gain is increased to x5.
worst pulse overshoot just touches the inner limits. The calibration gain is then returned to normal and
K=
200
%
P x Calibrated Gain
This is illustrated in the following table:

Pulse Amplitude
100
80
67
57.5
50
(c)
5 Gain
0.4% K
0.5% K
0.6% K
0.7% K
0.8% K
Bar Response
The horizontal timebase of the waveform monitor or oscilloscope is adjusted so that the half amplitude
points of the bar reach the outer limits marked on a graticule such as that shown in the Reference
Section, Ref. 4.
Ignoring the first and last 2.5% (0.625ms) of the bar, the deviation from its mid-point, expressed as a
10
(d)
The test signals are the 50 Hz square waves shown in the Reference Section, Ref. 5. The 0V to
700 mV signal is used in the Y channel and the -350 mV to +350 mV signal is used in the Pr and Pb
channels.
With the horizontal scan at field rate the 50 Hz signal is adjusted to coincide with the appropriate bar
markings on the graticule. For a stationary display the signal must contain field synchronising pulses
(Y) or the waveform monitor must be externally triggered from the same synchronising pulses (Pr and
Pb). Ignoring the first and last 2.5% (250 ms) of the bar, the percentage deviation of the worst half
divided by two is the K rating of the bar. (It may be noted that for the same deviation on the display a
4% K figure for the bar response looks the same as a 2% K figure for 50 Hz).
1.4.4
Delay Inequality
The test signal consists of sinusoids as shown in the Reference Section, Ref. 20 where the frequency of
the sinusoid applied to the luminance channel is 500 kHz and that to the colour difference channels is
502 kHz. When these signals are subtracted a null appears half way along active line time giving the
appearance of a "bowtie". A timing difference in the paths gives rise to a positional displacement of
the null by 1ms for about 4ns of timing error. Markers on some of the picture lines enable a direct
reading of any timing errors to be made. The method is not well suited for showing timing differences
which are not constant throughout the line time.
In this situation, or when there is noise or the sinewaves are distorted, a dual channel oscilloscope with
delayed timebase and writing speed in the order of 20 ns per division should compare the coincidence
of rising or falling slopes of the sinewaves shown in the Reference Section, Ref . 20.
The delaying timebase should be triggered from the rising edge of the luminance pedestal so that Pr
and Pb can be compared with Y. High gain should be used in the vertical amplifiers to give a steep
slope to the edges being measured and the oscilloscope delay time multiplier control used to inspect
the full line period. The worst error should be quoted.
The tracking of the oscilloscope input amplifier controls can be checked using the staircase at the start
of the waveform, the step amplitudes being designed to match the 5, 10 and 20 mV per division
sequence of many general purpose instruments.
When this method is used, it is also beneficial if some lines of the test signals are left unmodulated to
give a line through the sinewave crossings, aiding the vertical positioning of the traces at high
magnifications.
If a high speed oscilloscope is not available, an estimate can be made using matched and calibrated
switchable delays in both oscilloscope inputs (to account for insertion delay) and adjusted for a visual
timing null. Again this should be checked across the line period.
1.4.5
(a)

The waveforms used are ramp signals shown as typical examples in the Reference Section, Ref. 21. In
order that the system may be tested under a wide range of APL, the signal should consist of 6 lines of
ramp in every 24, with the intervening lines at black or white. Measurements are made with 18 lines of
white (bar on) and then 18 lines of black (bar off) and the worst result quoted.
To make the measurement, the output of the appropriate component channel is differentiated by a
suitable network as in the Reference Section, Ref. 3 and the mid-point of the resultant signal level
11
made to occupy 100% on an oscilloscope. The error is then the peak-to-peak percentage deviation over
the duration of the differentiated ramp.
If the significant noise is present, making the assessment of non-linearity difficult and the ramp
waveforms have been inspected to ensure that there are no quantisation errors, then staircase
waveforms similar to those shown in the Reference Section, Ref. 22 may be used. The signal should
consist of 6 lines of staircase in every 24 with the intervening lines at black and white. Measurements
are made with 18 lines of white (bar on) and then with 18 lines of black (bar off) and the worst result
quoted.
At the receiving end the test signal is passed through the differentiating network and displayed on an
oscilloscope to show a train of five pulses. Non-linearity is given by the difference in amplitude between
the largest and smallest pulse expressed as a percentage of the largest.
i.e.
(b)
E max E min
x 100%
E max
Component Crosstalk
Two of the component channels are energised with the multiburst test signals shown in the Reference
Section, Ref. 23 and the crosstalk into the "dormant" third component is measured peak-to-peak. The
result is expressed in dB relative to 700 mV
NOTE:
1.4.6
Some waveform monitors may have insufficient gain to achieve 100% amplitude when a
line rate ramp such as that shown in Reference 21 is differentiated. In cases such as these,
the ramp signals provided as `valid' waveforms from Component Test Signal Generators
may be used, as the shorter duration of these ramps give a larger amplitude pedestal when
differentiated. Alternatively, the line rate signal may be used as for the case when noise is
present.
Noise
equipment the 10% lift signal may be adjusted slightly to minimise the effects of quantisation noise.
(a)
Weighted Luminance
Measurement is made on the Y channel in the band 10 kHz (7.5 kHz) - 5.0 MHz using RMS
detection. The characteristic of the luminance weighting filter is shown in the Reference Section,
Ref. 7.
(b)
Colour Difference Noise
Measurement is made on the Pr and Pb channels in the band 10 kHz -1.6 MHz using RMS detection.
The Colour Difference filter having the characteristic shown in the Reference Section, Ref. 24 should
be used. Note that this network has a 6 dB insertion loss and therefore the measured figure should be
corrected accordingly.
(c)
Total Low Frequency Noise in Luminance Channel
Measurement is made unweighted in the band 40 Hz - 10 kHz (7.5 kHz) using peak-to-peak detection.
(d)
Total Low Frequency Noise in Colour Difference Channels
Measured the same as in the Luminance Channel.
12
Section 2: Audio Circuits & Equipment in the Signal Path
SECTION 2
AUDIO CIRCUITS AND EQUIPMENT WITHIN THE SIGNAL PATH
2.1.
PERFORMANCE FIGURES
2.1.1
Direct Path
For purposes of measurement the direct path is assumed to comprise the circuit from the agreed
interface with British Telecom or Transmission Operator's equipment, through the Presentation and
Master Control switching and processing equipment back to the agreed interface with British Telecom
or Transmission Operator's equipment.
Worst Path
For the purposes of measurement, the worst path is assumed to comprise the following, with all
interconnections carried out using the normal equipment routes:(i)
A studio mixer
(ii)
A looped VTR path
(iii)
A second studio mixer
(iv)
A second looped VTR path
(v)
The presentation and Master Control Path.
The input signal may either be an assigned source or commence at a studio wall box at a microphone
input.
The tolerance limits do not include degradations due to signal sources such as tape recorders, as
tolerances for these are separately specified.
A measurement of the Worst Path parameters is normally only necessary after the completion of a new
installation or a major re-installation.
Production Path
For the purposes of measurement the production path will comprise that part of the system that starts at
the output of originating equipment (microphone, disc reproducer, ATR or VTR etc). and that includes
assignment switching and mixing and ends at the interface with recording equipment or the Master
Control Room. The path may be in a studio centre or outside broadcast scanner.
O.B. Link Paths
O.B. Link tolerances are related to an unspecified number of point-to-point SHF links. Measurement is
made at the final output of the link at which point connection to a permanent circuit would be made.
13
2.1.2
(a)
(b)
2.1.3
(a)
(b)
2.1.4
(a)
(b)
(c)
(d)
(e)
2.1.5
(a)
(i)
(ii)
(b)
(i)
(ii)
(c)
2.1.6
(a)
Output Signal Level

Output signal level at agreed
interface after line-up
Gain Stability, variation of
insertion gain during one hour
Amplitude/Frequency Response
40 Hz-15 kHz
w.r.t.1 kHz
125 Hz-10 kHz
w.r.t.1 kHz
Total Harmonic Distortion
1 kHz at -10 dBu
1 kHz at +8 dBu
80 Hz at -10 dBu
80 Hz at +8 dBu
Input Overload
Signal/Noise Ratio
0 dBu input
Weighted, Random Peak
Unweighted, Random Peak
-50 dBu input
Weighted,Random Peak
Unweighted,Random Peak
Interchannel Crosstalk,
Weighted, Peak
Vision to Sound Crosstalk,
Weighted
Direct
Path
Worst
Path
Production
Path
O.B. Link
Path
0 dBm
0.25 dB
0.25 dB
0 dBm
0.5 dB
0.5 dB
0 dBm
0.25 dB
0.25 dB
0 dBm
0.25 dB
0.25 dB
+1
-2
+1
-1
dB
dB
dB
dB
dB
dB
dB
dB
+1
-2
+1
-1
dB
dB
dB
dB
+0.5
-3.0
+0.5
-2.0
dB
dB
dB
dB
0.5%
0.5%
0.5%
0.5%
-
0.5%
1.0%
0.5%
2.0%
-
0.5%
0.5%
0.5%
1.0%
17 dB
1.0%
1.0%
1.0%
1.0%
-
60 dB
-
56 dB
-
60 dB
63 dB
42 dB
47 dB
53 dB
53 dB
53 dB
56 dB
60 dB
53 dB
45 dB
1.5 dB
1.0 dB
1.0 dB
0.5 dB
1.0 dB
0.5 dB
-26 dB
-6 dB/
octave
-44 dB
+6 dB/
octave
-36 dB
-35 dB
-6 dB/
octave
-53 dB
+6 dB/
octave
-45 dB
-35 dB
-6 dB/
octave
-53 dB
+6 dB/
octave
-45 dB
DUAL CHANNEL SOUND PATHS ONLY

2.1.7 Level Difference Between A and B Channels
(a)
40 Hz-15 kHz
1.0 dB
(b)
125 Hz-10 kHz
0.5 dB
2.1.8 Crosstalk Between A and B Channels
(a)
40 Hz
-35 dB
(b)
40 Hz-315 Hz
-6 dB/
octave
(c)
315 Hz-6.3 kHz
-53 dB
(d)
6.3 kHz-15 kHz
+6 dB/
octave
(e)
15 kHz
-45 dB
14
+1
-3
+1
-2
Direct
Worst
Path
Path
Profiles are shown in Reference Section, Ref. 19 (a) and Ref. 19 (b)
Production
Path
O.B. Link
Path
20
Oblique
Segment
10
Oblique
Segment
20
20
Oblique
Segment
10
Oblique
Segment
20
2.1.9
(a)
(b)
Phase Difference Between A and B Channels

40 Hz
20
30
40 Hz-200 Hz
Oblique
Oblique
Segment
Segment
(c)
200 Hz-4 kHz
10
15
(d)
4 kHz-15 kHz
Oblique
Oblique
Segment
Segment
(e)
15 kHz
20
30
Profiles are shown in Reference Section, Ref. 19(f) and Ref. 19 (g)
2.2
2.2.1
Test Conditions
Normally signal levels are measured as voltages irrespective of impedance and are quoted in decibels
with reference to O dBu, where O dBu corresponds to 0.775 volts, RMS. This definition of signal level
applies throughout this Han dBook of Technical Standards for equipment measurements but does not
apply to line measurements or where it is separately defined.
2.2.2
Output Signal Level
The measurements may be made at any overall gain setting. The PPMs, which are used to control the
programme output levels of each mixer, will be used as the indicating meters.
With the input level set constant at -50 dBu for microphone level inputs, or O dBu for line level inputs,
the greatest change occurring in one hour in the output is defined as the gain stability.
2.2.3
This measurement may be made at any gain setting up to the maximum available; the output level
should be O dBu approximately on each output when the measurement is made.
Tests should be made at the following frequencies and the measurements should be referenced to the
level at 1 kHz.
40 Hz,
60 Hz,
125 Hz,
250 Hz,
500 Hz,
1 kHz,
2 kHz,
4 kHz,
6 kHz,
8 kHz,
10 kHz,
12 kHz,
15 kHz,
Additional tests should be made to ensure that the overall response falls off smoothly outside this
frequency band.
It should be noted that, as this test is a measurement of the variation of gain of the equipment with
frequency, corrections should be made for any variations in the input level with frequency.
2.2.4
(i)
For microphone channels, -50 dBu input with normal balance attenuator, channel, group and
main fader settings to achieve O dBu output.
(ii)
For line level channels, O dBu input with normal balance attenuator, channel, group and main
fader settings to achieve O dBu output.
15
The input signal level is varied to give an output level of first -10 dBu and then +8 dBu w.r.t. line-up.
At each level tests are made at 80 Hz and 1 kHz.
For microphone inputs only, additional tests of input overload capability at 80 Hz and 1 kHz are made.
The input signal level is slowly increased and the channel fader adjusted to keep the output level at
+8 dBu (ie peak signal level) until the onset of evident distortion (for the purposes of this measurement
this is defined at 3%). The increase in input signal level above normal peak input level, -42 dBu, is the
input overload capability.
2.2.5
Signal/Noise Ratio
The noise levels are measured using a test set incorporating a standard PPM (to BS 6840), and a low
noise amplifier with calibrated variable gain. The 'unweighted' bandwidth is constrained in accordance
with ITU-R BT.468-4, shown in Reference Section, Ref. 9(a) and the 'weighted' frequency response is
determined by the ITU network as defined in ITU-R BT.468-4 shown in Reference Section, Ref. 9(b).
1 kHz tone at the appropriate level is fed to the path under test and the gain of the test set is adjusted so
that the PPM gives a scale reading of '4' (i.e. O dBu). The input signal is then replaced by a termination
(as defined below) and the gain of the test set is re-adjusted so that the PPM again peaks to the scale
reading of '4'. The signal/noise ratio is the difference between settings of test set gain. The
measurements are made both weighted and unweighted.
(a)
Line Level Path (O dBu)
The input should be terminated in 600ohms.

(b)
Microphone Input (-50 dBu)
Balance attenuator, channel, group and main faders should be set as for normal operation. The input
should be terminated in 200 ohms directly at the injection point.
(c)
Crosstalk
The interfering signal, consisting of a 7 kHz tone, is fed to an adjacent input of each sound desk and
switching matrix in the path under test. The interfering path is lined up, using separate group and
output faders where this is possible without mixing with the path under test. Desk inputs may be at
microphone level (-50 dBu) or line level (O dBu). When the interfering path has been lined up the
input level is raised by 8 dB. The input of the path under test is terminated and the peak, weighted
output level of the path under test is then measured on a noise meter. A bandpass filter may be needed
to separate the crosstalking tone from random noise.
2.2.6
Modulation Derived Distortion (Vision to Sound Crosstalk)
Measured as noise (Para 2.2.5), with vision channel modulated by 100% amplitude, 100% saturated
colour bars.

2.2.7
Level Difference Between A and B Channels
For any signal path one of four possible input conditions will be applicable.
(i)
Microphone inputs to stereophonic channel - stereophonic circuits throughout.
(ii)
Microphone input to monophonic channel - stereophonic signals derived in a 'pan-pot'.
(iii)
Line inputs to stereophonic channel - stereophonic circuits throughout.
(iv)
Line input to monophonic channel - stereophonic signals derived in a 'pan-pot'.

16
With stereophonic input channels, test signals (initially at 1 kHz) from a common source should be
injected into both channel inputs, the channels being lined up in the normal way. The output levels
from the A and B chains should be measured at frequencies between 40 Hz and 15 kHz and the
differences calculated.
2.2.8
Crosstalk Between A and B Channels
A test signal at 1 kHz from a common source should be injected into both channel inputs, the channels
being lined up in the normal way.
In the case of stereophonic channels, the test signal should then be injected into the input of the A
channel, the input of the B channel being terminated in 200 ohms for microphone inputs and 600 ohms
for line inputs. The levels of the signals on the A and B outputs should be measured and the difference
(i.e. crosstalk) calculated. The inputs should then be reversed and the measurements taken again to
ascertain the crosstalk under this configuration.
In the case of monophonic input channels test signals should be injected and the channel routing selector
switched so that the signal is fed to only one output. The levels of the wanted signal on this output and the
unwanted signal on the other should be measured and the difference (i.e. crosstalk) calculated. The
channel routing selector should then be switched so that the input is fed to the other output and the
measurements taken again to ascertain the crosstalk under this configuration.
Measurements should be made at frequencies between 40 Hz and 15 kHz. Profiles for Crosstalk
performance are shown in Reference Section, Ref. 19 (a) and Ref. 19 (b)
2.2.9
For any signal path, one of the four input conditions described in Section 3.2.7 above will be
applicable.
injected into both channel inputs, the channels being lined up in the normal way. The phase difference
between the outputs of the A and B chains should be measured and the tests repeated at frequencies
between 40 Hz and 15 kHz.
In the case of monophonic input channels, a test signal (initially at 1 kHz) should be injected into the
channel input, the channel being lined up in the normal way. The test may be made with the 'pan-pot'
set in any position. The phase difference between the outputs of the A and B chains should be
measured and the tests repeated at frequencies between 40 Hz and 15 kHz. Profiles for Phase
performance are shown in Reference Section, Ref. 19 (f) and Ref. 19 (g).
17
[THIS PAGE INTENTIONALLY BLANK]
18
Section 3: Video Tape Recorders
SECTION 3
VIDEO TAPE RECORDERS
3.1.
PERFORMANCE FIGURES (COMPOSITE RECORDERS)
3.1.1
Tolerances listed for video tape recorders refer to a single recording and replay not necessarily on the
same machine.
The tolerances are based on full field measurements and the most common and straightforward
methods of measurement are given in 3.2. Where alternative methods, giving more accurate results, are
available these are mentioned in the appropriate paragraph.
The tolerances given below apply to both quadruplex and helical recorders, which lay down PAL
tracks. They may also apply to component recorders if these are tested in the PAL domain using
suitable codec pairs.
NOTE:
The use of VTRs not fully meeting this specification should be the subject of discussion
with the ITC where the subjective quality of the recordings justifies this.
VIDEO TOLERANCES
3.1.2
Output Signal Level
(a)
Adjustment Error
(b)
Gain Stability (over 1 hour)
3.1.3
(a)
(b)
2T Pulse Response
(c)
2T Bar Response
(d)
(e)
(f)
3.1.4
(a)
(b)
Differential Phase
(c)
Differential Gain
3.1.5
Noise
(a)
(b)
(c)
Total Low Frequency Random and Periodic (p-p)
(d)
Moire and Chrominance Modulation Noise
AUDIO TOLERANCES
3.1.6
Output Signal Level
(a)
Signal level at output after line-up
(b)
Gain Stability
2.0%
2.0%
1.5% K
1.5% K
1.5% K
1.5% K
3%
20 ns
4%
5
5%
-52 dB
-46 dB
-46 dB
-25 dB
1.0 dB
0.5 dB
19
3.1.7
(a)
40 Hz - 15 kHz w.r.t 1 kHz
(b)
125 Hz - 10 kHz w.r.t. 1 kHz
3.1.8
(a)
1 kHz at +8 dBu
(b)
80 Hz at +8 dBu
3.1.9
Signal/Noise Ratio
(a)
Weighted, Random, Peak
(b)
Unweighted, Random, Peak
3.1.10 Interchannel Crosstalk
(a)
40 Hz
(b)
40 Hz - 125 Hz
(c)
125 Hz - 10 kHz
(d)
10 kHz - 15 kHz
(e)
15 kHz
(f)
15 kHz - 80 kHz
A profile is shown in Reference Section, Ref. 19 (c)
3.1.11 Wow and Flutter
DUAL CHANNEL SOUND RECORDING
(a)
40 Hz - 15 kHz
(b)
125 Hz - 10 kHz
(a)
40 Hz
(b)
40 Hz - 125 Hz
(c)
125 Hz - 10 kHz
(d)
10 kHz - 15 kHz
(e)
15 kHz
A profile is shown in Reference Section, Ref. 19 (d)
3.1.14 Phase Difference Between A and B Channels
average:
(a)
40 Hz
(b)
40 Hz - 200 Hz
(c)
200 Hz - 4 kHz
(d)
4 kHz - 15 kHz
(e)
15 kHz
20
2.0 dB
1.0 dB
2.5%
2.5%
42 dB
46 dB
-45 dB
oblique segment
-55 dB
oblique segment
-45 dB
-35 dB
0.1%
2.0 dB
1.0 dB
-20 dB
oblique segment
-40 dB
oblique segment
-30 dB
30
oblique segment
15
oblique segment
30
peak:
(a)
40 Hz - 15 kHz
(b)
200 Hz - 4 kHz
A profile is shown in Reference Section 2, Ref. 19 (h)
3.2.
40
20
RECOMMENDED TEST METHODS (COMPOSITE RECORDERS)
VIDEO MEASUREMENTS
3.2.2
(a)
Output Signal Level

A recording is made of the 2T Pulse and Bar signal shown in the Reference Section, Ref. 1. after
adjustment of the bar amplitude to 700 mV at the generator output. The tape is replayed and the
amplitude of the bar centre at the output of the VTR, expressed as a percentage of 700 mV, is taken as
the signal level adjustment error.
(b)
Gain Stability
The greatest change occurring in the output level over a period of 1 hour, using the same recording.
3.2.3
(a)
Linear Distortion
The signal should be the 2T Pulse-and-Bar waveform as shown in Reference Section, Ref. 1.
The K rating of the pulse-to-bar ratio is defined as:K=
BP
x 100%
4P
When the waveform is subject to line tilt or an extended distortion of the leading edge at the top of the
bar, the amplitude of the bar must be measured at its mid-point after first setting the blanking level
(b)
2T Pulse Response
The test signal should be the 2T Pulse-and-Bar waveform as shown in Reference Section, Ref. 1.
Measurement may be made using a graticule such as that shown in Reference Section, Ref. 4
If this graticule is used, the vertical gain of the oscilloscope is adjusted to make the pulse amplitude
100% and then the vertical shift moved to bring the banking level onto the base line at 30%. The
horizontal gain is advanced and the horizontal shift adjusted to make the waveform touch the H.A.D
markers on the 80% line. With normal gain the graticule markers are 2% K and 4% K. For 1% K and
2% K the calibrated vertical gain is advanced by x2. For limits of % K and 1% K the pulse amplitude
is first set to 80% and the calibrated vertical gain then advanced by x5.
c)
2T Bar Response
The test signal should be the 2T Pulse-and-Bar waveform as shown in Reference Section, Ref. 1.
21
reach the outer limits marked on a graticule such as that shown in Reference Section, Ref. 4.
percentage of its amplitude at that point, is the K rating of the bar. Measurements are made using only
half the bar, the worse half being quoted as the result. It is incorrect to measure the whole bar tilt and
divide by two to obtain the K rating.
(d)
The test signal should be the 50 Hz square wave test signal as shown in Reference Section, Ref. 5, but
with added field synchronising pulses.
With the horizontal scan at field rate the 50 Hz signal is adjusted as in 3.2.3(c). For a stationary display
the signal must contain field synchronising pulses. Again, ignoring the first and last 2.5% (250 ms), the
percentage deviation of the worse half divided by 2 is the K rating of the bar. (It may be noted that for
the same deviation on the display a 4% K figure for the bar response looks the same as 2% K for
50 Hz).
(e)(f) Chrominance/Luminance Gain and delay Inequality
The measurements are made using a 2Tc composite Pulse-and-Bar signal (Reference Section, Ref. 6a)
with the output of the recorder under test fed to a Gain and Delay test set where available. The output
of the test set is viewed on an oscilloscope and the test set is adjusted to make the envelope of the
chrominance pulse flat along the baseline. If a test set is not available then the level of distortion
should be estimated
By examining the sinusoidal distortions at the bottom of the 2Tc composite pulse on a waveform
monitor or oscilloscope. The method is described in Part B.
It should be noted that if Chrominance/Luminance crosstalk is present the above method for
gain inequality will produce an erroneous result. The measurement is best made using the 2Tc noncomposite waveform (Reference Section, Ref. 6b). The 50% luminance pedestal is used to calibrate
the vertical gain of the oscilloscope and the chrominance amplitude is measured directly.
3.2.4
(a)
The test signal consists of a 5-step staircase (Reference Section, Ref. 2) occupying one line in every
The output signal is passed through a suitable differentiating network (Reference Section, Ref. 3),
amplified and displayed on an oscilloscope. The result is a train of five pulses. Non-linearity is
measured as the difference in amplitude between the largest and the smallest expressed as a percentage
of the largest.
i.e.
(b)
E max E min
x 100%
E max
Differential Phase
The test signal should be a 5 step staircase with added subcarrier (Reference Section, Ref. 2)
reference. The differential phase is defined as the largest departure in phase form that reference.
quoted.
22
(c)
Differential Gain
The test signal should be a 5 step staircase with added subcarrier (Reference Section, Ref. 2)
The differential gain may be measured by using a vectorscope in the line-time mode. The six sections
of subcarrier are compared for their amplitude relationships and taking the blanking level section as a
reference, the differential gain is defined as the largest departure in amplitude from that of the
reference expressed as a percentage. Measurements are made with the white bar on and with the white
bar off and the worst measurement is taken as the result.
The measurements in paragraph 3.2.4 (a) and (b) are difficult to make accurately on a VTR due to the
presence of noise, moire and jitter. More accurate measurements can be made using a suitable nonlinearity test set, preferably one which integrates the measurement and has a line strobe facility. Some
improvement can also be obtained by using a 200 kHz low-pass filter in the display circuit of a
vectorscope.
3.2.5
Noise
Noise measurements should be made using a 50% pedestal test signal.

When measurements are made on equipment incorporating digital timebase correctors the pedestal
level may be adjusted slightly to minimise the effects of quantisation noise.
(a)
Weighted Luminance
(b)
(c)
Total LF Random and Periodic
The total LF noise should be measured peak-to-peak in the frequency band 40 Hz -10 kHz (7.5 kHz).
(d)
Moire and Chrominance Modulation Noise
Noise measurements are made using 100% colour bars as the test signal (Reference Section, Ref. 8).
The VTR replay output is fed to a PAL decoder and the RED output measured on an RMS noise
measuring set over a frequency band of 0 - 3 MHz. The figure obtained is increased by 8 dB to convert
to peak-to-peak and to allow for the weighting of the decoder. Each colour bar is sampled in turn in the
middle of the bar and the worst figure is taken as the result.
100% saturated full field Y, C, G, M, R, B colours may be used instead of colour bars.
Moire may also be measured on a spectrum analyzer. An assessment of the combined effect can be
obtained by a square law addition of the individual components.
AUDIO MEASUREMENTS
3.2.6
Output Signal Level
(a)
Signal level at output after line-up using the EBU alignment tape or a recording made to the
same standard.
(b)
Gain Stability
The greatest change occurring in the output level over a period of 1 hour, using the same recording.
23
3.2.7
The input level to the recorder should be -10 dBu. As this test is a measurement of the variation of gain
of the equipment with frequency, corrections should be made for any variation of the input level with
frequency.
level at 1 kHz:40 Hz,
60 Hz,
125 Hz,
250 Hz,
500 Hz,
1 kHz,
2 kHz,
4 kHz,
6 kHz,
8 kHz,
10 kHz,
12 kHz,
15 kHz,
Additional tests should be made to ensure that the overall response falls off smoothly outside the
frequency band.
3.2.8
The input level to the VTR should be +8 dBu at each frequency.
3.2.9
Signal/Noise Ratio
with ITU-R BT.468-4, shown in Reference Section, Ref. 9(a), and the 'weighted' frequency response is
With the VTR under test lined up to its normal gain setting, it is first supplied with 1 kHz tone at
0 dBu and a recording is made. The input signal is then replaced by a 600 ohm termination and a
further recording is made. The output of the VTR is connected to the test set and the recordings played
back. The gain of the test set is adjusted so that, on the first recording, the PPM gives a scale reading
of '4' (i.e. 0 dBu); with the second recording, the gain of the test set is re-adjusted so that the PPM
again peaks to a scale reading of '4'. The signal/noise ratio is the difference between the two settings of
the test set gain.
3.2.10
This test is intended to measure the crosstalk performance from unrelated tracks such as those used for
timecode and unrelated audio signals.
The test signals at a level of -10 dBu should be fed to tracks likely to cause interference to the track
being measured. The input to the track being measured should be terminated in 600 ohms. Upon replay
of the recorded signals the crosstalk is determined from the difference in measured level of the two
tracks under consideration.
As crosstalk performance can approach, or be better than, the noise performance in tape recorders, it
may be necessary to employ selective filtering in this measurement.
3.2.11
Wow And Flutter
Measurements are made by first recording a test frequency of 3.15 kHz at standard reference level. On
replay wow and flutter amplitudes should be measured using an instrument complying with IEC
Publication 386, the relevant details of which are given in Reference Section, Ref. 10.
24

3.2.12
Test signals (initially at 1 kHz) from a common source at a level of -10 dBu should be fed into the
equipment under test. The output levels from the A and B channels on the replay should be measured
at frequencies between 40 Hz and 15 kHz and the differences calculated.
3.2.13
Test signals (initially at 1 kHz) at a level of -10 dBu should be fed into one input of the recorder, and a
600ohm termination connected to the other. The Crosstalk is calculated from the measured outputs of
the A and B channels. Measurements should be made at frequencies between 40 Hz and 15 kHz.
The measurements should be repeated with the input signals reversed.
A profile is shown in Reference Section, Ref. 19 (d)
3.2.14
Test Signals (initially at 1 kHz) from a common source at a level of -10 dBu should be fed into the
equipment under test. The phase difference between the outputs form the A and B channels on replay
should be measured at frequencies between 40 Hz and 15 kHz. When the difference is not constant,
the mean difference is taken as the result, though a note should be made of the maximum difference as
well.
A profile is shown in Reference Section, Ref. 19 (h)
3.3
PERFORMANCE FIGURES (COMPONENT RECORDERS)
This section gives the performance figures for component VTRs:

The figures refer to a single recording and replay not necessarily on the same machine.
When more than one head is used each head should perform within the given limits.
3.3.2
(a)
(b)
(c)
3.3.3
Signal Levels
Signal Level (Adjustment Error)
Signal Level Gain Stability (over 1 hour)
Interfield Flicker :
Luminance
Chrominance
0.7V (2%)
2%
1%
2%
Linear Waveform Distortions (2T Pulse and Bar in Luminance and 5T Pulse and Bar
in Colour Difference Channels)
Y
1.5% K
1.5% K
1.5% K
1.5% K
(a)
(b)
(c)
(d)
Pulse-to-Bar Ratio
Pulse Response
Bar Response
3.3.4
(a)
(b)
Delay Inequality between all Component Channels

Mean Timing difference
Timing perturbations (p-p)
Pr & Pb
2.0% K
2.0% K
1.5% K
1.5% K
20 ns
10 ns
25
3.3.5
(a)
(b)
Non-Linear Waveform Distortions

Component Crosstalk
4%
-43 dB
3.3.6
(a)
(b)
(c)
(d)
Noise
Colour Difference (RMS) 1.6 MHz low pass filter
Total Low Frequency in Luminance channel (p-p)
Total Low Frequency in Colour Difference Channels (p-p)
-52 dB
-48 dB
-45 dB
-43 dB
3.4
RECOMMENDED TEST METHODS (COMPONENT RECORDERS)
3.4.1
Test Conditions
The analogue component signals (Y, Pr and Pb) will be related to the colour separation signals (R, G
and B) by the following matrix equations.
Y
Pb
Pr
0.299
-0.169
0.500
0.587
-0.331
-0.419
0.114
0.500
-0.081
R
G
B
The colour separation signals have normal peak amplitudes of 700 mV. Synchronising signals may be
added to or kept separate from the luminance component.
The following methods are relevant when component recorders are being measured and it may be
convenient to use a waveform monitor capable of displaying the three components simultaneously:
3.4.2
(a)
Signal Levels
A recording is made of the Pulse and Bar signals shown in the Reference Section, Ref. 1 after
adjustment of the bar amplitudes to 700 mV at the generator output. The 2T signal is fed to the Y
channel and the 5T signal is fed to Pr and Pb. The tape is replayed and the amplitude of the bar centre
at the output of each of the component channels, expressed as a percentage of 700 mV, is taken as the
signal level adjustment error.
In some cases the "Lightning" method of measurement described in Part B Guidelines may be useful.
(b)
Gain Stability
The greatest change occuring in the output level of each channel over a period of 1 hour, using the
same recording.
(c)
Interfield Flicker
The odd and even fields of each component channel are examined separately and the maximum
amplitude difference between the two fields, expressed as a percentage of 700 mV, is taken as the
flicker.
3.4.3
Linear Waveform Distortions
Linear waveform distortions (a) to (c) are measured using the 2T Pulse and Bar test signal in the Y
(luminance) channel and the 5T Pulse and Bar test signal in the Pr and Pb (colour difference) channels.
The waveforms are shown in the Reference Section, Ref. 1.
26
(a)
Pulse-to-Bar Ratio
The pulse-to-bar K rating is defined as:

K=
BP
x 100%
4P
where B and P are the amplitudes of the bar and pulse respectively.
In practice, the pulse will be taken as the reference during measurement.
the bar, the amplitude of the bar must be measured at its mid-point after first setting the reference level
(blanking for Y) mid-way between two successive bars to 0%.
(b)
Pulse Response
Measurement may be made using a graticule such as that shown in the Reference Section, Ref. 4.
bring the reference level (blanking for Y) onto the baseline at 30%. The horizontal gain is advanced
and the horizontal shift adjusted to make the waveform touch the H.A.D. markers on the 80% line.
With normal gain the graticule markers are 2% K and 4% K. For 1% K and 2% K the calibrated
vertical gain is increased by x2. For 0.5% K and 1% K the pulse amplitude is first set to 80% and the
calibrated vertical gain is increased to x5.
worst pulse overshoot just touches the inner limits. The calibration gain is then returned to normal and
K=
200
%
P x Calibrated Gain

Pulse Amplitude
100
80
67
57.5
50
(c)
5 Gain
0.4% K
0.5% K
0.6% K
0.7% K
0.8% K
Bar Response
The horizontal timebase of the waveform monitor or oscilloscope is adjusted so that the half amplitude
points of the bar reach the outer limits marked on a graticule such as that shown in the Reference
Section, Ref. 4.
percentage of its amplitide at that point, is the K rating of the bar. It must be emphasised that
27
(d)
The test signals are the 50 Hz square waves shown in the Reference Section, Ref. 5. The 0V to
700 mV signal is used in the Y channel and the -350 mV to +350 mV signal is used in the Pr and Pb
channels.
With the horizontal scan at field rate the 50 Hz signal is adjusted to coincide with the appropriate bar
markings on the graticule. For a stationary display the signal must contain field synchronising pulses
(Y) or the waveform monitor must be externally triggered from the same synchronising pulses (Pr and
Pb). Ignoring the first and last 2.5% (250 ms) of the bar, the percentage deviation of the worst half
divided by two is the K rating of the bar. (It may be noted that for the same deviation on the display a
4% K figure for the bar response looks the same as a 2% K figure for 50 Hz).
3.4.4
(a)
Delay Inequality
Timing Differences
The test signal consists of sinusoids as shown in the Reference Section, Ref. 20 where the frequency of
the sinusoid applied to the luminance channel is 500 kHz and that to the colour difference channels is
502 kHz. When these signals are subtracted a null appears half way along active line time giving the
appearance of a "bowtie". A timing difference in the components gives rise to a positional
displacement of the null by 1ms for about 4ns of timing error. Markers on some of the picture lines
enable a direct reading of any timing errors to be made. The method is not well suited for showing
timing differences that are not constant throughout the line time.
In this situation, or when there is noise or the sine waves are distorted, a dual channel oscilloscope
with delayed timebase and writing speed in the order of 20 ns per division should be used to compare
the coincidence of rising or falling slopes of the sine waves shown in the Reference Section, Ref. 20.
The delaying timebase should be triggered from the rising edge of the luminance pedestal so that Pr
and Pb can be compared with Y. High gain should be used in the vertical amplifiers to give a steep
slope to the edges being measured and the oscilloscope delay time multiplier control used to inspect
the full line period. The worst error should be quoted.
The tracking of the oscilloscope input amplifier controls can be checked using the staircase at the start
of the waveform, the step amplitudes being designed to match the 5, 10 and 20 mV per division
sequence of many general purpose instruments. When this method is used, it is also beneficial if some
lines of the test signals are left unmodulated to give a line through the sine wave crossings, aiding the
vertical positioning of the traces at high magnifications.
If a high speed oscilloscope is not available, an estimate can be made using matched and calibrated
switchable delays in both oscilloscope inputs (to account for insertion delay) and adjusted for a visual
timing null. Again this should be checked across the line period.
In some cases the "Lightning" method of measurement described in Part B Guidelines may be useful.
(b)
Timing Perturbations
In component analogue television tape recorders, time division multiplex techniques are used to share
the luminance and colour difference signals between the recording channels and, as a consequence of
this, the three component signals are not recorded simultaneously. In replay this gives rise to to timing
differences between the signals in addition to the absolute timing jitter. Even after timebase correction
it is likely that some errors will remain.
Timing perturbations can be measured using any of the methods described for Delay Inequality in
3.4.4 (a). In all cases the oscilloscope should be externally triggered from the studio reference pulses.
28
3.4.5
(a)

The waveforms used are ramp signals shown as typical examples in the Reference Section, Ref. 21. In
order that the system may be tested under a wide range of APL, the signal should consist of 6 lines of
ramp in every 24, with the intervening lines at black or white. Measurements are made with 18 lines of
white (bar on) and then 18 lines of black (bar off) and the worst result quoted.
To make the measurement, the output of the appropriate component channel is differentiated by a
suitable network as in the Reference Section, Ref.3 and the mid-point of the resultant signal level made
to occupy 100% on an oscilloscope. The error is then the peak-to-peak percentage deviation over the
duration of the differentiated ramp.
If significant noise is present, making the assessment of non-linearity difficult and the ramp waveforms
have been inspected to ensure that there are no quantisation errors, then staircase waveforms similar to
those shown in the Reference Section, Ref. 22 may be used. The signal should consist of 6 lines of
staircase in every 24 with the intervening lines at black and white. Measurements are made with 18 lines
of white (bar on) and then with 18 lines of black (bar off) and the worst result quoted.
On replay the signal is passed through the differentiating network and displayed on an oscilloscope to
show a train of five pulses. Non-linearity is given by the difference in amplitude between the largest
and smallest pulse expressed as a percentage of the largest.
i.e.
NOTE:
(b)
E max E min
x 100%
E max
Some waveform monitors may have insufficient gain to achieve 100% amplitude when a line
rate ramp such as that shown in Ref. 21 is differentiated. In cases such as these, the ramp
signals provided as `valid' waveforms from Component Test Signal Generators may be used,
as the shorter duration of these ramps give a larger amplitude pedestal when differentiated.
Alternatively, the line rate signal may be used as for the case when noise is present.
Component Crosstalk
Recordings are made with two of the component channels energised with the multiburst test signals
shown in the Reference Section, Ref. 23 and the crosstalk into the "dormant" third component is
measured peak-to-peak. The result is expressed in dB relative to 700 mV.
3.4.6
Noise
signal is not significant. When measurements are made on VTRs containing digital processing
(a)
Weighted Luminance
Measurement is made on the Y channel in the band 10 kHz (7.5 kHz) - 5.0 MHz using RMS
detection. The characteristic of the luminance weighting filter is shown in the Reference Section,
Ref. 7.
(b)
Colour Difference Noise
Measurement is made on the Pr and Pb channels in the band 10 kHz -1.6 MHz using rms detection. The
Colour Difference filter having the characteristic shown in the Reference Section, Ref. 24 should be
used. Note that this network has a 6 dB insertion loss and therefore the measured figure should be
corrected accordingly.
29
(c)
Total Low Frequency Noise in Luminance Channel
Measurement is made unweighted in the band 40 Hz - 10 kHz (7.5 kHz) using peak-to-peak detection.
(d)
Total Low Frequency Noise in Colour Difference Channels
Measured the same as in the Luminance Channel.
30
Section 4: Audio Recorders
SECTION 4
AUDIO RECORDERS
4.1.
PERFORMANCE FIGURES
4.1.1
This section is applicable to all recording media including hard disc recorders, sprocketed sound
followers, high quality tape recorders and less high quality tape equipments.
"High quality" tolerances apply to equipments, including multi-track recorders, used for the recording
and replaying of significant speech and music.
"Less high quality" applies to audio cartridge equipment for NAB type B audio cartridges or similar
equipment used for effects.
Tolerances listed refer to a single recording and replay not necessarily on the same machine.
Tape recorders and reproducers should preferably employ ITU/IEC equalisation characteristics in
accordance with IEC Publication 94, 3rd Edition 1968.
Related tracks are defined as those which normally carry specific contributions to a composite sound,
such as the orchestral components of a musical balance.
Unrelated tracks are defined as those carrying information which is acoustically dissimilar, such as
time-code or other synchronising signals, effects and foreign language tracks.
Sound
Followers
High
Quality
Less High
Quality
1.0 dB
0.5 dB
1.0 dB
0.5 dB
1.5 dB
1.0 dB
1.5 dB
1.0 dB
1.5 dB
1.0 dB
2.0 dB
1.5 dB
42 dB
46 dB
42 dB
46 dB
38 dB
42 dB
-45 dB
-45 dB
-45 dB
-35 dB
-45 dB
oblique segment
-55 dB
oblique segment
-45 dB
-35 dB
-45 dB
-35 dB
-65 dB
4.1.2
Output Signal Level
(a)
Insertion Gain Adjustment Error
(b)
Gain Stability
4.1.3
(a)
40 Hz to 15 kHz w.r.t. 1 kHz
(b)
125 Hz to 10 kHz w.r.t. 1 kHz
4.1.4
Signal/Noise Ratio
(a)
(b)
4.1.5
(a)
40 Hz
(b)
40 Hz - 125 Hz
(c)
125 Hz - 10 kHz
(d)
10 kHz - 15 kHz
(e)
15 kHz
(f)
15 kHz - 80 kHz
4.1.6
Timecode Crosstalk
500 Hz - 20 kHz
-55 dB
-55 dB
31
4.1.7
(a)
(b)
4.1.8

1 kHz at +8 dBu
80 Hz at +8 dBu
Wow And Flutter, Weighted, peak
Sound
Followers
High
Quality
Less High
Quality
2.5%
2.5%
2.0%
2.0%
3.0%
3.0%
0.12%
0.1%
0.15%
2.0 dB
1.0 dB
3.0 dB
1.5 dB
-20 dB
oblique segment
-40 dB
oblique segment
-30 dB
-20 dB

4.1.9
(a)
40 Hz - 15 kHz
2.0 dB
(b)
125 Hz - 10 kHz
1.0 dB
4.1.10
(a)
40 Hz
-20 dB
(b)
40 Hz - 125 Hz
(c)
125 Hz - 10 kHz
-40 dB
(d)
10 kHz - 15 kHz
(e)
15 kHz
-30 dB
A profile is shown in Reference Section, Ref. 19 (d).
4.1.11
average:
(a)
40 Hz
30
(b)
40 Hz - 200 Hz
(c)
200 Hz - 4 kHz
15
(d)
4 kHz - 15 kHz
(e)
15 kHz
30
peak:
(a)
40 Hz - 15 kHz
60
(b)
200 Hz - 4 kHz
30
Profiles are shown in Reference Section, Ref. 19 (h) and Ref. 19 (i).
4.2.
4.2.2
(a)
-40 dB
-30 dB
30
oblique segment
15
oblique segment
30
50
40
20
60
30
30
50

Output Signal Level
Insertion Gain Adjustment Error
The Insertion Gain Adjustment Error is the difference between the input and output levels at 1 kHz.
The Measurement should be made with an input level of 0 dBu.
(b)
Gain Stability
The greatest change occurring in the output level when the same point on the tape is measured again
after 1 hour.
32
4.2.3
The input level to the recorder should be -10 dBu. As this test is a measurement of a variation of gain
of the equipment with frequency, corrections should be made for any variation in the input level with
frequency.
level at 1 kHz:40 Hz,
60 Hz,
125 Hz,
250 Hz,
500 Hz,
1 kHz,
2 kHz,
4 kHz,
6 kHz,
8 kHz,
10 kHz,
12 kHz,
15 kHz,
frequency band.
4.2.4
Signal/Noise Ratio
With the machine under test lined up to its normal gain setting, it is first supplied with 1 kHz tone at
0 dBu and a recording is made. The input signal is then replaced by a 600 ohm termination and a
further recording is made. The output of the machine is connected to the test set and the recordings are
played back. The gain of the test set is adjusted so that, on the first recording, the PPM gives a scale
reading of '4' (i.e. 0 dBu); with the second recording, the gain of the test set is re-adjusted so that the
PPM again peaks to a scale reading of '4'. The signal/noise ratio is the difference between the two
settings of the test set gain. The measurements are made both weighted and unweighted.
4.2.5
This test is intended to measure the crosstalk performance from unrelated tracks such as those used for
timecode and unrelated audio signals.
The test signals at a level of -10 dBu should be fed to tracks likely to cause interference to the track
being measured. The input to the track being measured should be terminated in 600ohms. Upon replay
of the recorded signals the crosstalk is determined from the difference in measured level of the two
tracks under consideration.
A profile is shown in Reference Section, Ref. 19 (c).
4.2.6
Timecode Crosstalk
The spectrum of the unweighted noise should be examined and the result of any peak due to time code
should not exceed the figure in Section 4.1.6.
4.2.7
The input level to the recorder should be +8 dBu and -10 dBu at each frequency.
33
4.2.8
Wow and Flutter
Wow and Flutter should be measured at a test frequency of 3.15 kHz. On replay wow and flutter
amplitudes should be measured using an instrument complying with IEC Publication 386, the relevant
details of which are given in Reference Section, Ref. 10.
A recorder-reproducer should be measured by recording a 3.15 kHz test frequency, and subsequently
reproducing this recording and measuring the total wow and flutter. Only in the case of tape delay
machines should wow and flutter be measured while simultaneously recording and reproducing.
4.2.9
equipment under test. The output levels from the A and B channels on the replay should be measured
at frequencies between 40 Hz and 15 kHz and the differences calculated.
4.2.10
Test signals (initially at 1 kHz) at a level of -10 dBu should be fed into one input of the recorder, and a
600 ohm termination connected to the other. The Crosstalk is calculated from the measured outputs of
the A and B channels. Measurements should be made at frequencies between 40 Hz and 15 kHz.
The measurements should be repeated with the input signals reversed.
A profile is shown in Reference Section, Ref. 19 (d).
4.2.11
equipment under test. The phase difference between the outputs from the A and B channels on replay
should be measured at frequencies between 40 Hz and 15 kHz. When the difference is not constant, the
mean difference is taken as the result, though a note should be made of the maximum difference as
well.
Profiles are shown in Reference Section, Ref. 19 (h) and Ref. 19 (i).
34
Section 5: Cameras
SECTION 5
CAMERAS
5.1.
PERFORMANCE FIGURES
5.1.1
Definitions And Operational Practices
The tolerances in this section are related to general-purpose studio and OB cameras. Equipment for
specific purposes, such as ENG and clock scanners, is required to meet a subjective performance limit
of grade 4 or better.
The cameras should be set up for the following tests according to the normal procedures of the
programme contractor. Controls should be re-adjusted subsequently only as required to perform the
test.
(a)
Iris - same as needed to produce 100% output from 60% reflectance neutral surface under
normal studio lighting.
(b)
Zoom angle - the test shall be carried out at a non-extreme setting.
(c)
Aperture correction and contours - adjusted as for normal studio use.
(d)
Gamma correction - 0.45 (nominal) except where stated.
(e)
Flare correction - in circuit.
(f)
Matrix - in circuit.
(g)
Measurement zones - in the following tests "inner zone" refers to the area defined in Reference
Section, Ref. 17.
5.1.2
Stability
No parameter should exceed Handbook of Technical Standards limits at any time during two hours
immediately following line-up.
5.1.3
Black Shading
(a)
Inner zone - luminance
3% peak-to-peak
(b)
Overall (whole field)-luminance
5% peak-to-peak
(c)
Overall - colour separation difference
2% peak *
*A change of not more than half of the limit may occur in any 10% of the picture width or height.
5.1.4
White Shading
(a)
Inner Zone - luminance
5% peak-to-peak
(b)
Overall (whole field) - luminance
10% peak-to-peak
(c)
Overall - colour separation signals
6% peak *
* Not more than half of these tolerances may occur in any 10% of the picture width or height.
5.1.5
Resolution
(a)
Centre (0.5 - 5.0 MHz)
100 20%
(b)
Corners (5.0 MHz)
40% to 110% (of Centre)
(c)
Horizontal aliasing (ccd cameras) on 5 MHz grating
Less than 10% difference
in p-p amplitudes
5.1.6
Waveform Response
(a)
Negative pulse maximum overshoot or preshoot
25% of pulse height
35
Section 5: Cameras
(b)
Pre-shoot minus overshoot
5.1.7
(a)
(b)
5.1.8
(a)
(b)
5.1.9
(a)
(b)
5.1.10
(a)
(b)
(c)
(d)
Geometry
Inner zone
Whole field
Registration
Inner zone
Whole field
Latent and Spurious Images
Line Scan Ringing
Blemishes, tube spots, pixel errors, fixed pattern noise,
smearing, shutter and illumination beating, vertical aliasing
and other defects.
Streaking
Short-term luminance
Short-term colour separation difference
Long-term luminance
Long-term colour separation difference
5.1.11
(a)
Flare
Flare correction error,
(b)
Differential Flare,
correction error,
5.1.12
Greyscale
Differential error between any Y,R,G or B signal in a single
studio, OB vehicle or combination of sources used to
contribute to single productions.
Noise
5.1.13
(a)
(b)
(c)
5.2
0 10% of pulse height
iris range 1
iris range 2
iris range 1
iris range 2
1% of picture height
0.15% of picture height
5%
Impairment level 4
(See Ref. 18)
1%
Less than 1%
2%
1%
3%
7%
1%
3%
2%
-48 dB
-43 dB
-45 dB
5.2.3
Black Shading
The measurements are best carried out with the camera lens capped. 5% lift is introduced in the centre
of the picture so that the outputs are clear of black clipping effects.
The luminance shading is measured peak-to-peak on the encoder output (or Luminance (Y) output,
where the camera is used in a component environment. Chrominance may be switched off for a clearer
display. The inner zone is best identified using an active picture area generator that complies with BS
5115. Where this is not available, a white card with the active picture area cut away (Reference
Section, Ref. 16a) mounted on a deep, black felt-lined box can be used. The card is illuminated so that
no light falls on the rear of the box and the iris is adjusted to give approximately 10% output from the
white of the card. This is enough to distinguish the inner zone without incurring significant amounts of
flare or streaking.
36
Section 5: Cameras
The colour separation difference measurement is made on the camera output signals, or on the
monitoring (registration) output where the R, G, and B outputs are not available; the peak change from
the centre, regardless of sign, is measured on the R-G, B-G and R-B (if available) signals. The peak
(rather than peak-to-peak) error is assessed since it relates more closely to the perceived effect.
All results are expressed as a percentage of the nominal peak output (700 mV) of the camera.
5.2.4
White Shading
Great care must be taken to ensure a very even illumination of a 60% reflectance neutral card. Natural
daylight is one of the most reliable sources. In studios, an integrating sphere can provide an area of very
even luminance. Otherwise it is best to light a full size test card but only use part of the area by zooming
in and defocussing (provided that an active picture area generator which complies with BS 5115 can be
used to separate the inner zone.)
The camera is adjusted so that the R, G and B signals are balanced in the centre. The light level and/or iris
should be adjusted so as to achieve 100% (700 mV) on the encoder output at a non-extreme iris setting but
care should be taken to ensure that no part of the picture is affected by white clipping circuits.
The luminance shading is measured peak-to-peak on the encoder output (or Luminance (Y) output, where
the camera is used in a component environment). Chrominance may be switched off for a clearer display.
The inner zone is best identified using an active picture area generator. Where this is not available, a white
card, with the outer zone black, must be used (Reference Section, Ref. 16b). Lamp shading may place
limitations on the accuracy of measurement especially for the whole field result.
The colour separation difference measurement is made on the camera output signals or on the monitoring
(registration) output where the R, G and B outputs are not available; the peak change from the centre,
regardless of sign, is measured on the R-G, B-G and R-B signals if available. The peak (rather than peakto-peak) error is assessed as it relates more closely to the perceived effect.
All results are expressed as a percentage of the nominal peak output (700 mV) of the camera.
5.2.5
(a)
Resolution
Centre
A test transparency having sinusoidal resolution gratings, such as Test Charts 52 or 60 (Reference
Section, Ref. 16c), is evenly illuminated and correctly framed. With gamma correction set to the
normally used value, the black and white reference blocks are adjusted by means of lift and iris
controls to 175 mV and 525 mV above blanking level respectively. On Test Chart 60 the reference
blocks are positioned at the sides of each frequency grating so that compensation is carried out for any
exposure and printing errors across the chart or shading errors in the camera.
Each resolution grating in the range 0.5 MHz - 5.0 MHz is measured at the coder output (or
Luminance (Y) output, where the camera is used in a component environment) and is expressed as a
percentage of the reference black to white transition, after correction circuits have been adjusted for a
flat response.
For 16:9 camera measurements, adjust the camera/lens to frame the (4:3) test transparency on the
horizontal axis only. Tilt the camera to bring the central frequency gratings vertically into the centre of
the picture. The frequencies are now as marked on the test transparency and measurements can be
carried out as above.
NOTE:
(b)
Also see Section 5.2.13, Noise.

Corner
The 5 MHz (400 lines) grating is positioned by means of the camera pan and tilt controls such that it
lies just outside the boundary of the inner zone along the picture diagonals. Thus four measurements
37
Section 5: Cameras
are made and the peak-to-peak amplitudes are recorded as percentages of the value achieved at the
centre.
The use of Test Card 60 will render movement of the camera unnecessary since the 5 MHz corner
gratings on this chart are ideally positioned (i.e. on the picture diagonals and immediately outside the
inner zone).
For 16:9 camera measurements, adjust the camera/lens to frame the (4:3) test transparency on the
vertical axis only. Use the line strobe facility on a waveform minitor to mark the position of the centre
of the upper corner 5 MHz gratings. Re-adjust the camera/lens to frame the test transparency on the
horizontal axis only and tilt the camera to place the centre of the upper corner 5 MHz gratings back on
the line strobe. Carry out the measurements as above. Repeat the process for the lower 5 MHz gratings.
(c)
Horizontal Aliasing (ccd cameras)
The centre 5 MHz grating should be examined at its mid-point and the difference in peak-to-peak
amplitudes of the sinewave due to aliasing is measured. This difference is expressed as a percentage of
the reference black to white transitions.
For 16:9 camera measurements, adjust the camera/lens as described in (a) above.
5.2.6
Waveform Response
The waveform response is determined by measuring the preshoot and overshoot produced when the
camera scans a white and black transient, such as that on Test Chart 60 (between the 2 MHz and
4 MHz resolution gratings).
The camera should be correctly framed and exposed for Test Chart 60 (or equivalent chart) such that
the black and white reference areas produce levels of 175 mV and 525 mV respectively above
blanking level.
Registration controls should be optimised for the best pulse shape as misregistration can severely
affect the results. With ccd cameras the horizontal position could also severely affect the results.
Measurements are made at the coder output (or Luminance (Y) output, where the camera is used in a
component environment.
For 16:9 camera measurements, adjust the camera/lens as in 5.2.5 (a) above.
5.2.7
Geometry
The ITVA Test Card 10 (Reference Section, Ref. 16d) should be placed in front of the camera taking
care that the plane of the test card is orthogonal to the camera lens axis and that the vertical axis is
truly vertical using a plumb line if necessary. The test card should precisely fill the scanned area.
The camera coded output is then combined with an electronically generated dot pattern having the
same number of horizontal and vertical dots and the combination is displayed on a black and white
picture monitor. Operational adjustments should be made to superimpose the two images as closely as
possible.
Geometric errors can be determined directly by reading off the amount of displacement of the dots
from the centre of the test card circles, The circle radii correspond to 0.5%, 1.0%, 1.5% and 2.0% of
picture height.
5.2.8
Registration
After alignment the camera may be moved before the tests are carried out.
A suitable grid pattern chart is used for the tests (Reference Section, Ref. 16e). The camera signal
outputs are displayed differentially on a black and white picture monitor.
38
Section 5: Cameras
In the case of tubed cameras the green channel horizontal and vertical shift controls are used to correct
the largest registration errors in the inner and outer zones in turn. Measurement is made of the change
of each shift control voltage and is expressed as a fraction of the voltage change for one square. The
results are expressed as a percentage of picture height. If both horizontal and vertical errors exist at the
same point the result should be summed vectorially.
In the case of ccd cameras where centring controls are not used it may be necessary to carry out the
measurement by using variable delay lines or by adjusting the centring controls of a picture monitor
and using the change in voltage as with camera centring control adjustment.
5.2.9
Latent And Spurious Images
The intention is to include in this section any defect not covered by other measurements. Where
possible objective measurements should also be used.
Line scan ringing should be measured using the same test conditions as for black shading.
See also the notes on CCD cameras in Guidelines Section 8.
Impairment level 4 on the 5 point scale is defined as "perceptible but not annoying" (Reference
Section, Ref. 18).
5.2.10
Streaking
(a)(b) Short Term Streaking

A test chart such as the one shown in Reference Section, Ref. 16 (f) is used in conjunction with a black
felt lined box. The white part of the chart is set to nominal peak output on the luminance channel and
the black hole is set for 5% lift.
Short term streaking is defined as the disturbance resulting from the white bar to camera output signal
corresponding to the hole in the chart. The camera luminance signal is displayed on a waveform
monitor at line rate and the change in level of the black hole during the lines which scan the white bar
is measured and expressed as a percentage of the nominal peak signal. The first 0.625ms after the
transition from white to black is considered as a function of resolution and is ignored for the purpose
of this measurement.
Colour streaking is measured using the camera output signals. Separate measurements G-R, G-B and
R-B should be made. A line strobe may be useful in separating the wanted signal from the general
background.
(c)(d) Long Term Streaking
A suitable test chart is shown in Reference Section, Ref. 16(f). Black level and iris are adjusted to
make the black holes give an output with 5% lift at black and the white patches 100% of the peak
signal.
The level of the background may be affected by both flare and streaking. The streaking is measured by
comparing a line following the white patch with the equivalent line before the white patch. The
greatest difference in amplitude is expressed as a percentage of the black to white transition. Any
variations during the first two lines after the transition, due to vertical aperture correction, should be
ignored.
5.2.11
Flare
Test Chart 60 carrying a centrally placed 25mm square piece of black felt is placed in front of the
camera and the controls are adjusted to properly frame the Chart.
39
Section 5: Cameras
Adjustment Method:
The iris should be adjusted to be at least 2 f-stops down from fully open and the Test Chart
illumination should be set for correct exposure (i.e. for white level at R, G and B outputs).
The camera is capped and the black level controls are adjusted for 5% lift above blanking level for the
black felt chip. The camera is then uncapped and the flare corrector controls are adjusted to reestablish the 5% lift in the black area of the picture.
For 16:9 camera measurements, adjust the camera/lens as in 5.2.5 (a) above.
Test Method:
The iris is adjusted over the ranges given below and any changes in level in the black area of the
picture, expressed as a percentage of white level, represents the flare correction error.
A performance limit also exists for differential flare correction error which is the difference in black
levels between the R, G and B outputs.
Iris range 1:
From iris fully closed (or camera capped) to a correctly exposed chart.
Iris range 2:
As for 1 but extended to 2 f-stops above that for the correctly exposed condition.
5.2.12
Greyscale
The main object of this test is to establish the similarity of transfer characteristics on cameras used in
the same production and to ensure adequate compensation is made for the electro-optical transfer
characteristics of the system.
A Test Chart such as that shown in Reference Section, Ref. 16 (g) is used in conjunction with a light
source and integrating sphere.
On exposure to the Test Chart, the black level produced by the black felt chip should be set to be just
clear of blanking level using the master black level control (or individual controls if necessary). The
iris control (and individual channel gain controls if necessary) should be adjusted so that the 100%
transmission chip produces an output of 700 mV (100%).
Using a waveform monitor or video level meter the signal amplitudes corresponding to each of the
transmission chips is measured at the coded output of the camera. These values can be plotted against
the corresponding chip transmission values to show the transfer characteristic and the linear gain near
to black level can be calculated. The value of gamma can be determined by taking a least squares best
fit to the logarithms of the datapoints in the gamma region of the curve.
If greater accuracy is required, for example when assessing performance of the camera in its linear
gain region, the above procedure may be repeated for a different set of transmission chips. One way of
achieving this is to use the same chart but with a sheet of neutral density filter placed in front of it. A
x1 ND filter will give a reduction in chip transmission values by a factor of 10.
Measurements of signal amplitudes corresponding to each of the chips at the R, G and B outputs are
used for the calculation of the tracking error. A representative transfer characteristic is shown in the
Reference Section, Ref. 16 (h).
5.2.13
Noise
When the centre resolution measurements are made, an additional test over and above that required in
Section 5.1.5 is made to determine the response of the normal setting of the aperture/contour
correction circuits. In general, the noise measurements are carried out with these normal operating
settings. If, for any reason, the 'normal' settings are not known, then the aperture/contour correction
circuits are adjusted to produce the following response.
40
Section 5: Cameras
The resolution response is adjusted so that the amplitude of the 3 MHz grating is 525 mV peak-to-peak
and that of the 5 MHz grating is as near as possible to 350 mV peak-to-peak, both with respect to
reference black and white areas which are set to 175 mV and 525 mV respectively above blanking
level.
Measurement is then made using a neutral card with the light level reduced to give an output of
approximately 52% of white level. This should correspond to the point on the gamma correction curve
at which the gain is unity. The lens may be zoomed in and defocussed to minimise shading.
(a)
Weighted Luminance
(b)
weighting filter - shown in Reference Section, Ref. 7, using RMS detection.
(c)
Measurement is made peak-to-peak in the band 40 Hz - 10 kHz (7.5 kHz). Particular care should be
taken to avoid field shading on the card. It is suggested that the signal input to the detector should be
examined on an oscilloscope and, if field shading is present, correction should be made to the
illumination or the camera field shading correctors, if fitted, should be adjusted.
NOTE:
Part B, Guidelines contains details of the measurement of Colorimetry and calculation of

camera colour fidelity. It also contains information on operating techniques to minimise
the effects of picture degradations due to CCD technology.
41
Section 5: Cameras
42
Section 6: Telecine & Sound Followers
SECTION 6
TELECINES AND SOUND FOLLOWERS
6.1.
PERFORMANCE FIGURES
6.1.1
Definitions And Test Conditions
Tolerances in this section apply to telecine, slide scanners and replay - only sprocketed sound
followers. Monochrome slide and caption scanners used with colour synthesizers are not covered by
the tests but are required to meet a subjective performance limit of grade 4 or better. Tape reproducers
should preferably employ equalisation characteristics in accordance with IEC Publication 94, 3rd
edition, 1968.
Equipment should be set up for the tests according to the normal procedure of the programme
contractor. Controls should be re-adjusted subsequently only as required to perform the test. Details of
test methods are contained in Part 2, using the appropriate paragraph numbers.
(a)
Gain or light control - should be adjusted to achieve a standard picture level output of 700 mV
p-p for a film density of 0.3.
(b)
Aperture correction - should be adjusted so that the 400 line output on 35 mm and 300 line output on 16 mm test
film is equivalent to that of the low frequency transitions. It is intended to specify the use of sine wave resolution
test film when such film becomes available.
However, it is recognised that film used operationally may have a response less than this test film; therefore, the
aperture corrector should be capable of a correction 6 dB greater than that required to achieve the above response.
(c)
Unless otherwise stated, the percentages in the following tests refer to peak white being
700 mV and equal to 100% where blanking level is 0%.
(d)
Gamma Correction = 0.35(nominal)
(e)
Peak audio signal level (+8 dBu) should correspond to 100% modulation on optical tracks and
to a recorded flux density of 405 nWb/m on magnetic tracks.
VIDEO TOLERANCES
6.1.2
Stability
No parameter should exceed Han dBook of Technical Standards limits at any time during two hours
immediately following line-up.
6.1.3
Black Shading
(a)
3% peak-to-peak
(b)
5% peak-to-peak
(c)
Overall - colour separation difference
2% peak *
* A change of not more than half of the limit may occur in any 10% of the picture width or height.
6.1.4
White Shading
(a)
5% peak-to-peak
(b)
10% peak-to-peak
(c)
6% peak *
*Not more than half of these limits may occur in any 10% of picture width or height.
(d)
White clipper operation
To operate between 100% and 105% with
negligible difference between channels
43
6.1.5
(a)
(b)
(c)
6.1.6
(a)
(b)
(c)
(d)
6.1.7
(a)
(b)
6.1.8
(a)
(b)
(c)
(d)
6.1.9
(a)
(b)
(c)
6.1.10
(a)
(b)
6.1.11
6.1.12
(a)
(b)
6.1.13
(a)
(b)
(c)
(d)
6.1.14
(a)
(b)
44
Flicker (Flying Spot Types)

1%
2%
3%
Resolution (To 400 Lines For Both 35 mm And 16mm Film)
Centre
100 20%
Corners
40% to 110% of centre
amplitude
Difference between fields
(Flying Spot) - centre
10%
Difference between fields
(Flying Spot) - corner
20%
Waveform Response
Negative pulse maximum overshoot or preshoot
25% of pulse height
Preshoot minus overshoot
0 10% of pulse height
Noise
-50 dB
-44 dB
-47 dB
Fixed Pattern Noise
Impairment level 4
[See Ref. 18 and note in
Section 6.2.8 (d)]
Geometry
Inner Zone
Whole field
Difference between adjacent intervals
10% of an interval
Registration (Camera Types)
Inner Zone
Whole field
Interlace
40%:60%
Interfield Registration (Flying Spot Type)
Centre
Errors Imperceptible
Corners
Streaking
Short-term Luminance
1%
Short-term colour separation difference
Less than 1%
Long-term Luminance
2%
Long-term colour separation difference
1%
Flare
D.C. Flare
*
A.C. Flare
*
*At the first inspection of newly installed telecines, the A.C. and D.C. flare should be measured; these
initial figures should be noted in the report documents. At subsequent inspections the change in flare
performance will be measured. Whilst it would be desirable to stipulate an absolute figure for flare
performance, it is considered premature to do this.
6.1.15
Picture Steadiness
p-p Vertical and Horizontal Hop and Weave

This limit should not be exceeded more than
3 times per minute and the hop and weave
should never exceed twice the limit.
AUDIO TOLERANCE
6.1.16
(a)
(b)
Output Signal Level

Tolerance on signal level at the output of the machine after
line-up
Output level stability
60 Hz -8 kHz w.r.t. 1 kHz
or 2 dB
40 Hz -14 kHz w.r.t 1 kHz
(c)
125 Hz - 10 kHz w.r.t 1 kHz
(b)
6.1.17
(a)
6.1.18
(a)
(b)
6.1.19
(a)
(b)
(c)
(d)
(e)
6.1.20

1 kHz at +8 dBu
80 Hz at +8 dBu
Signal/Noise Ratio
M/c running, open gate, Weighted, Random, Peak
M/c running, open gate, Unweighted, Random, Peak
Buzz Track
Wow And Flutter
Weighted, Peak
DUAL CHANNEL SOUND
(a)
40 Hz - 63 Hz
(b)
125 Hz - 6.3 kHz
(c)
8 kHz - 10 kHz
40 Hz - 10 kHz
COMOPT
COMMAG/
SOUND
FOLLOWER
(Replay only)
0.75 dB
0.75 dB
2 dB
2 dB
+1 dB/-3 dB
-
+1.5 dB/2.5 dB
1.0 dB
3%
3%
3%
3%
50 dB
55 dB
30 dB
-
40 dB
45 dB
0.15%
0.15%
2 dB
1 dB
2 dB
2 dB
1 dB
2 dB
-20 dB
45
(a)
(b)
40 Hz
40 Hz - 125 Hz
(c)
(d)
125 Hz - 10 kHz
10 kHz - 15 kHz
-20 dB
Oblique
segment
-40 dB*
Oblique
segment
-30 dB
(e)
15 kHz
* Note: For 35mm Commag this figure is relaxed to -30 dB.
6.1.23
(a)
(b)

40 Hz
40 Hz - 200 Hz
(c)
(d)
200 Hz - 4 kHz
4 kHz - 15 kHz
(e)
15 kHz
A profile is shown in Reference Section, Ref. 19 (j)
6.2
25
Oblique
Segment
15
Oblique
Segment
25
25
Oblique
Segment
15
Oblique
Segment
25
VIDEO MEASUREMENTS
In the absence of suitable test films the measurement of the electronic circuits should give results as
good as the limits quoted.
6.2.3
Black Shading
The measurements are best carried out with the lamp switched off. 5% lift is introduced in the centre
of the picture so that the outputs are clear of black clipping effects.
The luminance shading is measured peak-to-peak on the encoder output (or Luminance (Y) output,
where the telecine is used in a component environment. Chrominance may be switched off for a
clearer display. The inner zone is identified using an active picture area generator which complies with
BS 5115.
The colour separation difference measurement is made on the telecine output signals; the peak change
from the centre, regardless of sign, is measured on the R-G, B-G and R-B signals. The peak (rather
than peak-to-peak) error is assessed since it relates more closely to the perceived effect.
All results are expressed as a percentage of the nominal peak output (700 mV) of the telecine.
6.2.4
White Shading and
6.2.5
Flicker (Flying Spot Types)
The measurements are made "open gate" with the mechanism running. The telecine is adjusted so that
the R, G and B signals are balanced in the centre. The output level controls should be adjusted so as to
achieve 100% (700 mV) on the encoder output (or Luminance (Y) output, where the telecine is used in
a component environment) but care should be taken to ensure that no part of the picture is affected by
white clipping circuits.
46
On Flying Spot telecines, the odd and even fields are examined separately. For any position, the
average of the two fields is taken as the shading, and the maximum difference at any point between the
two fields is taken as the flicker.
6.2.6
Resolution
It is intended to specify a sine wave test method when suitable test film is available; meanwhile tests
should be made using a square wave pattern such as Marconi Test Film Number 1.
Lift and gain are adjusted so that the central l.f. transition corresponds with a voltage excursion from
25% to 75% video level. The response from other frequency gratings is compared with that of the l.f.
transition.
Limits are quoted for a 'machine running' condition.
Correction for the shading performance may be necessary when measuring corner resolution.
On flying spot telecines the odd and even fields should be measured separately with the film running
and the worst figure taken as a result.
6.2.7
Waveform Response
It is intended to specify the use of a test film having a white pulse on a dark grey background and a
black pulse on a light grey background when such a test film is available. Meanwhile, the test should
be made using the Marconi Test Film Number 1 with the telecine running. Lift and gain are adjusted
so that large area blacks produce 5% video level and large area whites produce 55% video level. A
line just above the upper frequency wedges is strobed to measure the amplitude of the preshoot which
occurs just before the negative going pulse produced by a black vertical line. The amplitude of the
overshoot which occurs just after the negative going pulse and the amplitude of the negative going
pulse are also measured.
It should be noted that on camera type telecines misregistration may affect the results and the
measurements should therefore be made on the green channel.
6.2.8
Noise
Measurement is made using a test film or slide of neutral density 1.0. This should give an output of
approximately 57% of peak white and correspond to the point on the gamma correction curve at which
the gain is unity.
The red and blue gains are adjusted, if necessary, for optimum balance at this level. The following
measurements are made:(a)
Weighted Luminance
(b)
weighting filter shown in Reference Section, Ref. 7 using RMS detection.
(c)
Total LF Random and Periodic
The total LF noise should be measured peak-to-peak in the frequency band 40 Hz -10 kHz (7.5 kHz).
Care should be taken that the measurement is not affected by field shading signals.
47
(d)
Fixed Pattern Noise (Line array machines)
Impairment level 4 on the 5 point scale is defined as: "perceptible but not annoying". (Reference
Section, Ref. 18).
CCD telecines should incorporate correctors to limit the effect of CCD element variation to -45 dB
with respect to peak white luminance or better.
6.2.9
Geometry
A suitable grating test film having about 25 vertical and 25 horizontal intervals is run in the telecine.
The output of the telecine is combined with an electronically generated grid pattern having the same
number of horizontal and vertical lines and the combination is displayed on a picture monitor. An
estimate of the error is made as a proportion of one interval and the result is expressed as a percentage
of picture height.
Small adjustments to vertical scan amplitude are permissible during this test.
6.2.10
Registration (Camera Types)
A suitable grating test film having about 25 vertical and 25 horizontal intervals is run in the telecine.
The signal outputs are displayed differentially on a black and white picture monitor and the green
channel horizontal and vertical shift potentiometers are used to correct the largest registration errors in
the inner and outer zones in turn. Measurements are made of the change of shift potentiometer voltage
in each case and expressed as a fraction of the voltage change for one square. The result is expressed
as a percentage of picture height.
If both horizontal and vertical errors exist at the same point the results should be summed vectorially.
6.2.11
Interlace
With the SMPTE Alignment and Resolution Pattern (Reference Section, Ref. 15) in the gate and the
machine stationary, the structure of the near horizontal lines of one of the vertical resolution test
wedges (see Plate 6.1) is examined on a picture monitor. The interlace performance of the picture
monitor is unimportant.
On the picture monitor screen (or photograph of it such as Plate 6.1), the horizontal distance between
the left hand edge of one white element and left and edge of the next element to the right (i.e.. on the
other field) should be measured. This distance is referred to as A in Fig. 6.1.
Then measure the horizontal distance between the left hand edge of the second element mentioned
above and the left and edge of the next element to the right of that second element (i.e. on the same
field as the first element). This distance is referred to as B in Fig. 6.1
For telecines where none of the vertical scan is produced by the film motion, the interlace ratio is
given below:A
A+ B
B
A+ B
or the required standard of 40 : 60 may be expressed:1 .5 >
A
> 0.67
B
For telecines where approximately half of the vertical scan is produced by film motion (twin lens or
hopping patch machines) the required standard may be expressed:2.30 >
48
A
> 0.43
B
6.2.12
Interfield Registration (Flying Spot Types)
The SMPTE Alignment and Resolution Pattern (Reference Section, Ref. 15) should be running in the
telecine.
(a)
Vertical Interfield Registration
Provided that the horizontal component of error is small compared with the vertical component, the
method of measurement described in Para. 6.2.11 may be applied to the near horizontal lines of the
corners of the SMPTE test pattern.
The vertical interfield registration error =
0.174 x
A B
% of picture height
A+ B
The required standard of 0.05% of picture height is met if

1.810 >
b)
A
> 0.554
B
Horizontal Interfield Registration
On a photograph of a picture monitor screen (such as Plate 6.2) at 400 lines in each corner, the
horizontal component of the distance between the left hand edge of a white picture element on one
field and left hand edge of the corresponding element on the other field should be measured. This
distance is referred to as L in Fig. 6.2
The horizontal distance between the left hand edge of the first used element and the left hand edge of
the next element to the right on the same field is then measured. This distance is referred to as M in
Fig. 6.2
The frequency of 400 lines in the corners is 4.6 MHz, so horizontal misregistration =
L
x 0.5% of picture height
M
The required standard of 0.05% of picture height is met if

L
> 0. 1
M
49
Enlargements of Portions of SMPTE Test Film
6.2.13
Streaking
(a)(b) Streaking Short-Term

A 'de-streak' test film, consisting of white horizontal bars of different lengths on a black background,
(Reference Section, Ref. 15) is used. The telecine controls are adjusted so that the white bars
correspond with nominal peak output (700 mV) with the background clear of black clipping effects.
Short-term streaking is seen as either a positive or negative disturbance to the background following
the white horizontal bars. The luminance signal is displayed on a waveform monitor at line rate and the
change in level of the background after the white bar is expressed as a percentage of the bar amplitude.
Some care is needed to separate effects of streaking from flare. Using the waveform monitor on line
strobe the line selector is "rocked" up and down so as to compare a strobed line through the white bar
with one just below it. Each bar is assessed in turn and the worst result quoted. Overshoot components
on the luminance signal are ignored for the purpose of this measurement.
50
Colour streaking is measured using the telecine output signals. If colour streaking is apparent, separate
measurements of G-R, G-B and R-B should be made.
(c)(d) Streaking Long-Term
A test film consisting of a white patch on a dark surround is used (Reference Section, Ref. 12). The
telecine is adjusted to give peak signal output on the white patch with the background just above
blanking level.
The luminance signal is observed at frame rate. The level of the background may be affected by both
flare and streaking. The streaking is measured by comparing a line following the white patch with the
equivalent line before the white patch. The greatest difference in amplitude is expressed as a
percentage of the black to white transition. Any variations during the first two lines after the transition
due to vertical aperture correction should be ignored.
6.2.14
Flare
In addition to the following, alternative methods of measurement are also acceptable.

(a)
D.C. Flare
All flare correction circuitry should first be aligned according to normal station practice. The film gate
is closed and the lift should be set to 5% in order to avoid clipping or crushing.
A test film of 50% APL (Reference Section, Ref. 13a) is then placed in the gate. (In the case of certain
twin lens machines a running loop will be required to preserve the correct APL). The resulting video
signal amplitude is then adjusted to 700 mV.
The level produced at the centre of the black area of the film is compared with the original 5% lift and
the D.C. flare is equivalent to this difference expressed as a percentage of 700 mV.
(b)
A.C. Flare
A.C. Flare (the halo surrounding bright objects) is measured using a test film having a 99% APL.
(Reference Section, Ref. 13b).
The telecine gate should first be closed and the lift set to 5%. The test film (still frame or running loop
as appropriate) is placed in the gate and the video signal amplitude is adjusted to 700 mV.
The level produced at the centre of the black area of the film is compared with the original 5% lift and
the A.C. flare is equivalent to this difference expressed as a percentage of 700 mV.
6.2.15
Picture Steadiness
A test film similar to that in Reference Section, Ref. 14 can be used for this test; this film has a vertical
stripe and a rectangle with upper and lower edges inclined at 15 to the normal.
To measure horizontal frame steadiness (weave), the line direction positional variation of a point A
(Reference Section, Ref. 14a) on the vertical stripe is measured using external triggering of the delayed
time base of an oscilloscope with approximately 100 ns/cm magnification. The peak-to-peak weave of
the point HA is measured in nanoseconds and expressed as a percentage of picture height by:Weave p p =
H A 4
x
520 3
% of picture height
The weave limit of 0.2% of picture height corresponds to a peak-to-peak timing jitter of 78ns.
To measure the vertical frame steadiness (hop), a point B (Reference Section, Ref. 14c) on the inclined
edge of the rectangle is observed on an oscilloscope in the line rate mode with internal triggering of the
delayed time base by the narrow vertical stripe. The maximum line positional variation of that point on
51
the inclined edge (HB) is an indication of the vertical steadiness frame error. The hop referred to the
picture height is:Hop p p =
H B
1450
% (HB in ns)
The hop limit of 0.2% of picture height corresponds to a peak-to-peak timing jitter of 290 ns.
If the frame steadiness error of the test film is significant, then this must be arithmetically deduced
from the test measurements.
Measurements on 16:9 Telecines

Measurements on 16:9 telecines follow the same procedures as above. A 16:9 format alignment test
film is now available for this purpose and is shown in the Reference Section, Ref. 44. Additional
information also appears in the Reference Section, Ref. 45.
AUDIO MEASUREMENTS
6.2.16
Output Signal Level
(a)
Suitable test films are listed in Reference Section, Ref. 15.
(b)
The wow and flutter test film should be replayed and the peak-to-peak variation in output
during a one minute replay should be measured on a standard PPM (to BS 6840).
6.2.17
Suitable test films are listed in Reference Section, Ref. 15

Measurements should be made with any Academy filter out of circuit.
6.2.18

At the lower level, the presence of noise may necessitate the use of selective filters to isolate the
harmonic components.
6.2.19
Signal/Noise Ratio
The noise levels are measured using a test set incorporating a standard PPM (to BS6840), and a low
A line-up film, which produces an output of 0 dBu, should be replayed with the machine under test
adjusted to its normal gain setting. The output of the machine is connected to the test set and the gain
of the test set is adjusted so the PPM gives a scale reading of '4' (ie. 0 dBu). For each of the following
measurement conditions, the gain of the test set is re-adjusted so that the PPM again peaks to a scale
reading of '4'. The signal/noise ratio is the difference between the two settings of test set gain on each
occasion.
(a)(b)
Measurements should be made without film but otherwise under normal operating
conditions, i.e. with machines running, doors closed and operating area lights on.
(c)
Measurements should be made unweighted while playing a buzz track test film. (This film
is specified in BS550 - 5.4.2 and ISO6025).
(d)(e)
Measurements should be made while playing erased magnetic film.

52
6.2.20
Wow And Flutter
Wow and flutter amplitudes should be measured using an instrument complying with IEC Publication
386, the relevant details of which are given in Reference Section, Ref. 10.
The specified figures should be achieved immediately after the run-up time used in normal operation.
DUAL CHANNEL SOUND

6.2.21
The frequency response section of the test film shall be played and the output level between 40 Hz and
10 kHz of each channel measured. The level difference is the difference in magnitude between the A
and B channels.
6.2.22
The optical test film should have each test frequency (40 Hz to 10 kHz) at a modulating level of 100%
(5%) on alternate tracks with zero modulation on the other track during the same period.
Each channel should have the gain set so that the 100% modulation level produces a level of +8 dBu at
the signal channel output. The output level of the other channel, w.r.t. +8 dBu, is the crosstalk. Each
channel should have the same overall gain. The difference in channel output levels between the
modulated and the unmodulated tracks is the crosstalk.
The presence of noise may necessitate the use of a frequency selective measuring instrument or
spectrum analyzer.
6.2.23
The frequency response section of the test film shall be played and the difference in phase at each
frequency between 40 Hz and 10 kHz is measured. The measurement may be made by applying the
outputs of the equipment to a dual channel oscilloscope. Both channels should be viewed using one of
them as a trigger. A half cycle is displayed and the error assessed by comparing the zero crossing
points. The performance profile is given in Reference Section, Ref. 19 (j).
53
54
Section 7: Disc Reproducers
SECTION 7
DISC REPRODUCERS
7.1.
PERFORMANCE FIGURES
7.1.1
Explanatory Notes
The performance limits in this section refer to all forms of disc storage except hard disc recorders.
Those marked with an asterisk are not applicable to Compact Disc. The performance of Compact Disc
should be very much better than the minimum requirements given below.
Reproducers fitted with a variable speed control should have the measurements made with the control
set to the nominal speed.
The pick-up head (mounted on the normal tone arm) and equalising amplifier are considered to be part
of the signal chain.
7.1.2
40 Hz - 12.5 kHz w.r.t. 1 kHz
7.1.3 Signal/Noise Ratio
(a)
(b)
7.1.4 Rumble *
(a)
Weighted
(b)
Unweighted
7.1.5 Intermodulation Distortion *
Frequency Intermodulation Distortion, Unweighted, Peak
7.1.6 Speed And Wow And Flutter *
(a)
Absolute Speed
(b)
Wow and Flutter, Weighted, Peak
DUAL CHANNEL SOUND
40 Hz - 12.5 kHz
(a)
40 Hz
(b)
40 Hz - 315 Hz
(c)
315 Hz - 4 kHz
(d)
4 kHz - 12.5 kHz
(e)
12.5 kHz
A profile is shown in Reference Section, Ref. 19 (e)
(a)
40 Hz
(b)
40 Hz - 200 Hz
(c)
200 Hz - 4 kHz
(d)
4 kHz - 12.5 kHz
2.5 dB
50 dB
55 dB
55 dB
40 dB
5%
2.0%
0.12%
1.5 dB
-9 dB
oblique segment
-20 dB
oblique segment
-10 dB
40
oblique segment
20
oblique segment
55
(e)
12.5 kHz
A profile is shown in Reference Section, Ref. 19 (k)
7.2
36
7.2.1
Types of disc considered
The tests in this section apply to reproducers playing vinyl discs. Suitable test discs should be used for
testing Compact Disc players.
7.2.2
Measurements should be made using a standard PPM (to BS 6840) while playing the mono side of a
test record to DIN 45 541. (Note that the frequency range is covered in two bands, having different
recorded levels - the appropriate 1 kHz level should be used).
7.2.3
Signal/Noise Ratio
determined by the ITU network as defined in ITU-R BT.468-4 shown in Reference Section, Ref. 9(b)
The first band of the mono side of a test record to DIN 45 541 should be played (1 kHz at a recorded
velocity of 10cm/sec peak) and the test set should be connected to the output of the equipment. With
the equipment under test set to its normal gain, the test set is adjusted so that the PPM gives a scale
reading of '4' (i.e. O dBu). A second measurement should be made with the turntable running and the
reproducing stylus tip 2.5mm above the rotating turntable; the gain of the test set is re-adjusted so that
the PPM again peaks to a scale reading of '4'. The signal/noise ratio is the difference in the settings of
the test set gain. The measurements should be made both weighted and unweighted.
7.2.4
Rumble
The measuring instrument should comply with BS 7063 the relevant details of which are given in
Reference Section, Ref. 11.
The first track of side A of a test record to DIN 45 544 should be replayed and the measuring
instrument should be connected to the output of the equipment. With the equipment under test set to its
normal gain the gain of the measuring instrument should be adjusted to produce a scale reading of
'0 dB'. The unmodulated track is then played and the gain of the measuring instrument is readjusted so
that the meter again produces a scale reading of '0 dB'. The rumble is the difference in the settings of
the measuring instrument gain. The measurement should be made both weighted and unweighted.
7.2.5
Intermodulation Distortion
The measurement of frequency intermodulation distortion should be made in accordance with the
method described in DIN 45 411, i.e. the measuring instrument should be a wow and flutter meter
complying with DIN 45 507 (equivalent to IEC Publication 386/B4847) in which the weighting filter
has been replaced by a high-pass filter having a time constant of 1ms.
Parts III and IV of side B of a test record to DIN 45 542 should be played, and the larger value of the
distortion should be noted.
7.2.6
Speed And Wow And Flutter
Speed deviation and wow and flutter should be measured using a test record to DIN 45 545.
56
The measuring instrument for wow and flutter should comply with IEC Publication 386/B4847, the
relevant details of which are given in Reference Section, Ref. 10.
The record should first be centred with respect to the rotational axis of the turntable by playing the
concentric groove near the edge and adjusting the position of the record to minimise lateral movement
of the pick up. The 3.15 kHz band should be played, and the wow and flutter should be measured at
three portions across the band (largest radius, mid radius and smallest radius). The largest value of
wow and flutter should be note.
7.2.7
The monophonic side of a test disc to BS 1928: 1965 (or DIN 45 541) should be replayed, the A and B
channels being lined-up in the normal way. The outputs from the A an B channels should be measured
at frequencies between 40 Hz and 12.5 kHz and the differences calculated.
7.2.8
The stereophonic side of a test disc to BS 1928: 1965 (or DIN 45 541) should be replayed. For both the
condition of an output on the A channel only and also for one on the B channel only, the output levels
from the A and B channels should be measured on each occasion and the differences (i.e. Crosstalk)
calculated. Measurements should be made at frequencies between 40 Hz and 12.5 kHz.
A profile is shown in Reference Section, Ref. 19 (e).
7.2.9
The monophonic side of a test disc to BS 1928: 1965 (or DIN 45 541) should be replayed. The phase
difference between the output from the A and B channels should be measured and the test repeated at
frequencies between 40 Hz and 12.5 kHz.
A profile is shown in Reference Section, Ref. 19 (k).
57
58
Section 8: Waveforms
SECTION 8
WAVEFORMS
8.1
SOUND-IN-SYNCS
Pulse Amplitude:
The pulses will be inserted in the video waveform at an amplitude of 700 mV 30 mV and will be
inserted 300 mV 9 mV below video black level.
Pulse Shape:
The pulses representing the binary digits will be shaped so that they have a sine squared form with a
half amplitude duration of 182ns, nominal.
Pulse Group Duration:
The total duration of the pulse group will be 3.82ms 10 ns at the half amplitude points for Mono SIS
and will vary slightly from this figure for Dual Channel SIS depending upon the number of symbol
periods inserted into each sync pulse duration.
Synchronising Period Waveform:
The pulse group will be inserted into the video waveform in the position shown in Figure 8.1. In order
to accommodate the pulses during the field blanking interval, line rate equalising pulses will be
increased in width to approximately 4.5ms. These equalising pulses will be restored to the correct
width before radiation.
Fig 8.1 - Sound-in-Syncs Signal
8.2
VERTICAL BLANKING INTERVAL
8.2.1.
Synchronising and Blanking
The vertical synchronising and blanking waveform is shown in Reference Section, Ref. 25.
59
8.2.2.
Colour Burst Blanking
Blanking of the colour burst during the field interval is shown in Reference Section, Ref. 26.
The same subcarrier phase / sync relationship only occurs every eight fields because of the switching
of the V colour component and the quarter line offset of subcarrier which causes a phase shift of 90o
on successive lines in the PAL system. This is known as the PAL 8-field sequence and is important if
correct timing is to be maintained across edit points in programmes.
8.2.3.
VBI Line Allocations
The current allocation of lines in the Vertical Blanking interval is given in the table in Reference
Section, Ref. 27.
8.2.4.
Insertion Test Signals
The National Insertion Test Signals used by independent television in the UK are shown in Reference
Section, Ref. 28.
8.2.5.
Teletext Data Signals
A line in the field blanking interval is identified as a Teletext Data line by the presence of the clock
run-in followed by the framing code at an appropriate time as shown in Reference Section, Ref. 29.
8.2.5.1.
Data waveform
Each data-line contains binary elements (bits) as a two-level NRZ (Non-Return-to-Zero) signal
suitably shaped by a filter. The binary signalling levels are defined on a scale where television black
level is 0% and white level 100%. The binary 0 level is then 0(2)% and the binary 1 level is
66(6)% as shown in Reference Section, Ref. 29. The difference between these levels is the basic data
amplitude. The data waveform will contain overshoots so the peak-to-peak data amplitude will exceed
the basic data amplitude.
8.2.5.2.
Bit rate
The bit rate of the signal is 6.9375 Mbit/s (25 parts per million). It is 444 times the normal television
line frequency.
8.2.5.3.
Data timing
The data timing reference point is the peak of the penultimate 1 of the clock run-in sequence. This
point has been selected to reduce the effect of any transient distortions at the start of the data-line.
The line time reference is the half amplitude point of the sync pulse leading edge.
The data timing reference in the signal as transmitted shall be 12.0(+0.4/-1.0)s after the line-time
reference.
The data timing may vary from data-line to data-line.
8.2.5.4.
Data-line structure
Each data-line comprises 360 bits, which may be considered as 45 eight-bit Bytes.
The first three Bytes, which have even parity, serve to synchronise the bit and Byte recovery operation
in the receiver. The remaining 42 Bytes have odd parity and carry address and control information, and
the codes for a character row.
The use of odd parity during the variable part of the data-line ensures that there are never more than 14
bit periods between the data level transistions in the waveform. This simplifies the recovery of the bitrate clock directly from the data waveform.
60
All the address and page control information is transmitted using Hamming Code Bytes to reduce the
possibility of the wrong character rows being stored in the receiver.
8.2.5.5.
Clock run-in
The first two Bytes of every data-line comprise the clock run-in sequence of alternating bits, beginning
101010..., to indicate the presence of a data-line and to establish the timing of the bits on that line. In
some cases the first one or two binary 1s may be absent.
8.2.5.6.
Framing code
The third Byte of every data-line comprises the framing code 11100100. This code has been selected
to enable Byte synchronisation to be established even if one bit of the framing code has been
incorrectly received.
8.2.5.7.
Hamming codes
The fourth and fifth Bytes of every data-line, and a further eight Bytes of the page-header data-lines,
are Hamming codes containing four message bits interleaved with four protection bits dependent on
the message bits. Tests can be used to correct single errors in the received Byte and detect multiple
errors (when 2, 4 or 6 bits are in error). When there are 3, 5, 7 or 8 errors in the Byte this procedure
results in a false message being decoded.
8.2.5.8.
Character Bytes
The remaining Bytes of each data-line are seven-bit character codes with an added odd parity bit b8.
The bits are transmitted in numerical order from b1 to b8.
8.2.6.
Widescreen signalling
Information about a television programme can now be signalled so that a receiver can carry out
automatic adjustments of the displayed picture on the screen. The signalling waveform occupies the
first half of Line 23 immediately before the start of active picture in the first field as shown in
Reference Section, Ref. 30.
The table in Reference Section, Ref. 31 shows how bits in this waveform are set to convey information
about aspect ratio, enhanced services, subtitles and others like surround sound.
8.2.7.
Ghost Cancellation Reference
The Ghost Cancellation Reference signals are intended for use by appropriately equipped television
receivers to reduce or eliminate the effects of multi-path reception on displayed pictures and, in
addition, to improve the eye-height of received teletext data so as to reduce the likelihood of errors in
teletext decoding. The waveforms are shown in Reference Section, Ref. 32 and the parameter values
are given in the table in Reference Section, Ref. 33. For further information please refer to ETS 300
732.
61
62
Section 9: Peak Programme Meters
SECTION 9
PEAK PROGRAMME METERS
9.1
PERFORMANCE FIGURES
9.1.1
Definitions and Test Conditions
The British Standard BS6840: Part 10: 1991 (Methods of specifying and measuring the characteristics
of peak programme meters) is the reference for the information given in this Section. Only the most
important tests from that Standard are described and are sufficient to assess PPM performance using
test equipment currently available commercially. In certain circumstances it may be necessary to carry
out further tests from BS6840.
There are three types of peak programme level meter, type 1, type 11a and type 11b. The type 11a and
11b instruments differ only in respect of the scale divisions and marking. The type 11a scale is used by
United Kingdom broadcasting related organisations. The type 11b scale is particular to the EBU.
Reference Indication
The indication corresponding to the permitted maximum signal level (CMTT Recommendation 661)
on the circuit to which the meter is connected (usually mark "6").
Reference Input Voltage
The r.m.s. value of the steady state sinusoidal signal of 1 kHz which results in reference indication
(usually 1.94V).
Reference Input Level
The level of the reference voltage, in decibels, referred to a stated reference, preferably 0.775V.
9.1.2
Calibration
PPM Reading
0
1
2
3
4
5
6
7
9.1.3
(a)
(b)
(c)
(d)
Below 12 Hz
12 Hz to 31.5 Hz
31.5 Hz to 16 kHz
16 kHz to 40 kHz
(e)
40 kHz to 100 kHz
Relative Input Level

Negligible
-12 0.5 dB
-8 0.3 dB
-4 0.3 dB
0 dB
4 0.3 dB
8 0.3 dB
12 0.5 dB
Falling smoothly
+2 dB
0.3 dB
Level or falling smoothly to
0 dB to -6 dB at 40 kHz
Falling smoothly
63
9.1.4
Dynamic Response
Burst Duration
PPM Indication
100 ms
Mark "6"
0.5 dB
10 ms
"5.5"
0.5 dB
5ms
"5"
0.75 dB
1.5ms
"3.75"
1.0 dB
0.5ms (see note)
"1.75"
2.0 dB
NOTE:
Since it is necessary for the tone burst to include a minimum of 5 cycles, a
frequency of at least 10 kHz is required for this measurement.
9.1.5
Overswing
Application of 1 kHz tone
0.5 dB
9.1.6
Return Time
Time for indication to fall from "7" to
2.8s 0.3s
"1" after removal of tone
9.1.7
Distortion Introduced by PPM
Supply voltage on or off
<0.1%
9.2
9.2.2
Calibration
The relationship between input level and the meter indication.

A sinusoidal test signal of 1 kHz is applied to the input at a level of +12 dB. The signal level is
adjusted to give the correct indication of PPM mark "7" and this level is noted. The process is repeated
with the appropriate signal levels for PPM marks "6", "5", "4", "3", "2", "1" and minimum.
9.2.3
The difference, expressed in decibels, between the indicated level and the input level as a function of
frequency, referred to the value at 1 kHz. A sinusoidal test signal of 1 kHz at a level of +8 dB is fed to
the input and adjusted to give reference indication (mark "6"). The frequency is varied over the stated
ranges and the level is adjusted to give reference indication. The difference between these levels and
the level at 1 kHz give the frequency response.
9.2.4
Dynamic Response
The relationship between the duration of a burst of sinusoidal signal at a frequency of 5 kHz and the
corresponding indication.
The amplitude of a 5 kHz signal is adjusted to give reference indication (mark "6"). The steady state
signal is replaced by bursts of the same waveform, frequency and amplitude but of variable duration.
The maximum indication is noted for each of several values of tone-burst duration, which may be
measured using a counter or oscilloscope. Alternatively, a calibrated burst generator may be used. The
interval between the tone bursts should be long enough to allow the meter to reach its zero signal
condition between bursts. The burst should start and finish at zero crossings on the sinusoid.
Each tone burst should include a minimum of 5 cycles. Burst durations of below 1ms should therefore
be at a suitable frequency to fulfil this requirement. E.g. a frequency of at least 10 kHz is required for a
burst duration of 0.5ms.
64
9.2.5
Overswing
The transient excess of the indication after a 1 kHz input signal, which would result in reference
indication, has been applied. A sinusoidal signal of 1 kHz, previously adjusted to give reference
indication, is applied. The transient excess of the indication is noted.
9.2.6
Return Time
The time interval between the removal of a steady state signal (+12 dB) corresponding to a scale
indication of mark "7" and the moment the indicator passes mark "1".
9.2.7
Distortion introduced by PPM
The connection of a PPM to a sound circuit should not increase the level of Total Harmonic Distortion
by more than the stated limit. The tests for distortion are described in Sections 2.4 and 4.4 etc.
9.2.8
Additional Information
The performance requirements and test methods contained in this section are applicable to PPMs other
than the conventional analogue types. These include incremental types using lamps or LEDs, bar graph
displays and video displays. Additional requirements for these other types covering such aspects as
brightness, colour, decay time etc are given in BS6840 Sections 6.1 to 6.8.
65
66
Section 10: Satellite Links Paths
SECTION 10
SATELLITE LINKS PATHS
10.1
10.1.1
PERFORMANCE FIGURES (Vision)

This Section gives the performance limits for Satellite Links used for covering Outside Broadcasts and
for the purpose of News Gathering.
OB Link performance will be assured with an FM carrier in a full transponder and with a 3.7m dish
and modern receiver the minimum Carrier-to-Noise ratio given will usually be significantly improved.
Stereo Sound will usually be employed and the sections concerned with dual sound channels will be
relevant.
SNG Quality may be attained with an FM carrier in a full transponder and reduced dish size or with a
half transponder or 8.5Mbits/s digital carrier. The respective Carrier-to-Noise requirements are given
for these variations. If news material is linked with mono sound then the dual channel sound limits
should be disregarded.
Additional care must be taken if using digital transmission to avoid or reduce problems caused by
different compression ratios and motion algorithms in the system.
10.1.2
(a)
(b)
(c)
10.1.3
(a)
(b)
(c)
(d)
(e)
(f)
10.1.4
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
10.1.5
(a)
(b)
Signal Levels
Signal Level
Adjustment Error
Short period gain variation e.g. 1 second
Medium period gain variation e.g. 1 hour
2T Pulse Response
2T Bar Response
Non-Linear Distortion
Differential Phase
Burst/Chroma Phase
Differential Gain
Transient Gain Change - Luminance (see Note 1)
Transient Gain Change - Chrominance (see Note 1)
Transient Gain Change - Sync (see Note 1)
Chrominance/Luminance Crosstalk
Input/Output Impedance-Return Loss
Luminance
Chrominance
0.7 V
2%
1%
2%
2% K
2% K
2% K
2% K
4%
20 ns
5%
5
N/A
5%
5%
5%
5%
4%
-30 dB
-30 dB
67
(c)
Low Frequency
-30 dB
10.1.6
(a)
(b)
10.1.7
(a)
VLF Response
First Overshoot (see Note 1)
Second Overshoot (see Note 1)
Noise
Weighted Luminance
(b)
(c)
(d)
Total Low Frequency
Interchannel Crosstalk (see Note 2)
10.1.8
(a)
(b)
10.1.9
(a)
(b)
(c)

Sound Subcarrier Modulated (see Note 3)
Sound Subcarrier Unmodulated (see Note 3)
Carrier-to-Noise Ratio
FM Carrier (full transponder)
FM Carrier (half transponder)
8.5MBits/s digital carrier (see Note 4)
20%
8%
(i) OB Path
(ii) SNG Path
-52 dB
-49 dB
-44 dB
-40 dB
-50 dB
-52 dB
-50 dB
typically >17.0 dB
typically >15.5 dB
>16.0 dB
NOTE 1:
Transient changes are determined in the main by the effectiveness of the clamps in the
receiver and performance is expected to be well within the limits quoted.
NOTE 2:
Interchannel crosstalk depends on many factors peculiar to satellite transponders. There

can be same frequency cross polarized signals, interleaved adjacent channel, cross
polarized adjacent channel and factors due to power sharing near to transponder (TWT)
saturation from multiple in-channels signals. These factors are under control of the satellite
operator and/or the uplink power control systems. The EBU has set a target of 58 dB on the
new satellites, but great care needs to be taken when sharing transponders. -50 dB may be
more realistic in this situation.
NOTE 3:
If two or more subcarriers are used a product due to the difference in frequency of the
subcarriers can appear in the video baseband due to modulation derived distortion. The
amplitude of the added subcarriers is a critical factor and the effect is very non-linear.
The level of subcarriers is nominally 100 mV and must be specified before a meaningful
distortion value can be given. The resulting distortion is periodic and overlaid with the
combined modulation. Subcarriers of 6.65 and 7.2 MHz are commonplace with a resulting
product at (7.2 - 6.65) MHz = 650 kHz.
NOTE 4:
Digital carrier with QPSK modulation and 3/4 Forward Error Correction. The Carrier-toNoise ratio quoted will ensure the following Bit Error Rate criteria are met:
1 in 107 for more than 90% of any month
1 in 106 for more than 98% of any month
1 in 103 for more than 99.97% of any month
68
10.2
10.2.1
RECOMMENDED TEST METHODS (Vision)

Test Conditions
Before commencing a measurement, all test equipment should be checked for accuracy. Any
inaccuracies should be corrected if possible, or noted and allowed for in the measurement.
This section gives examples of test methods that use basic techniques. These examples do not preclude
the use of other valid methods. The use of ITS type test signals is also not precluded but the ITU
waveforms referred to in these notes are regarded as the primary standard.
10.2.2
(a)
Signal Levels
The test may be carried out using a calibrated television waveform monitor.
The signal level adjustment error may be measured by using a 75 ohm generator of the 2T Pulse and
Bar test signal as shown in Reference Section, Ref. 1. The generator should be adjusted so that the bar
amplitude is 700 mV and the synchronising pulse amplitude is 300 mV. The sine-squared pulse is
ignored in this application. The difference in amplitude of the bar centre at the output, expressed as a
(b)(c) Gain Variations
one hour. The measurements of 10.2.2 (a) should be repeated using the identical path and any changes
recorded as the results for these parameters.
10.2.3
(a)

The test signal should be the 2T Pulse and Bar Waveform as specified in Reference Section, Ref. 1.
The pulse-to-bar K-rating is defined as:
K=
BP
x 100%
4P
Therefore, in practice, to make the measurement, the pulse will be taken as reference.
measured at its mid-point, to obtain the K-rating.
the bar, the amplitude of the bar must be measured at its midpoint after first setting the blanking level
b)
2T Pulse Response
The test signal should be the 2T Pulse and Bar Waveform as specified in Reference Section, Ref. 1
Measurement may be made using a graticule such as that shown in Reference Section, Ref. 4.
bring the blanking level onto the base line at 30%. The horizontal gain is advanced and the horizontal
shift adjusted to make the waveform touch the H.A.D. markers on the 80% line. With normal gain the
graticule markers are 2% K and 4% K. For 1% K and 2% K the calibrated vertical gain is advanced by
69
2. For limits of % K and 1% K the pulse amplitude is first set to 80% and the calibrated vertical
gain then advanced by x 5.
worst pulse overshoot just touches the inner limits. The calibrated gain is then returned to normal and
K=
200
%
P x Calibrated Gain

Pulse Amplitude
100
80
67
57.5
50
(c)
5 Gain
0.4% K
0.5% K
0.6% K
0.7% K
0.8% K
2T Bar Response
The test signal should be the 2T Pulse and Bar waveform as specified in Reference Section, Ref. 1.
reach the outer limits marked on a graticule such as that shown in Reference Section, Ref.4.
(d)
The test signal should be the 50 Hz square wave test signal specified in Reference Section, Ref. 5.
With the horizontal scan at field rate the 50 Hz signal is adjusted as in 1.2.3. (c). For a stationary
display the signal must contain field synchronising pulses. Again, ignoring the first and last 2.5%
(250 ms) the percentage deviation of the worst half divided by 2 is the K rating of the bar. (It may be
noted that for the same deviation on the display a 4% K figure for the bar response looks the same as a
2% K for 50 Hz).
(e)
The measurement is best made using the 2Tc non-composite waveform (Reference Section, Ref. 6b).
The 50% luminance pedestal is used to calibrate the vertical gain of the oscilloscope. The chrominance
amplitude is then measured directly.
NOTE:
(f)
The use of the composite 2Tc waveform with a gain and delay test set will produce an
erroneous result in the presence of chrominance/luminance crosstalk.
The measurement is made using a 2Tc composite Pulse-and-Bar signal (Reference Section, Ref. 6a)
and a delay measuring test set where available.
The output of the test set is viewed on an oscilloscope and the test set adjusted to cancel any path
chrominance/luminance delay inequality. If a test set is not available then the level of distortion should
be estimated by examining the sinusoidal distortions at the bottom of the 2Tc composite pulse on a
waveform monitor of oscilloscope. The method is described in Part B, Guidelines.
70
10.2.4
(a)
The test signal consists of a 5-step staircase (Reference Section, Ref. 2 occupying one line in every
It should be noted that the staircase with added sub-carrier waveform is used to conform with C.C.I.R.
recommended practice.
Section, Ref. 3) and amplifier and displayed on an oscilloscope. The result is a train of five pulses.
Non-linearity is measured as the difference in amplitude between the largest and the smallest
expressed as a percentage of the largest.
i.e.
(b)
E max E min
x 100%
E max
Differential Phase
The test signal should be a 5 step staircase with added subcarrier (Reference Section, Ref. 2).
reference. The differential phase is defined as the largest departure in phase from that reference.
quoted.
(c)
Burst/Chroma Phase
Burst/Chroma Phase errors may be measured as follows. Display the output of a colour bar generator
directly on a vectorscope and after aligning the burst on the graticule, carefully measure the phase
displacement (if any) of the BLUE bar. Apply the colour bar signal to the equipment or path under test
and display the output signal on the vectorscope. After aligning the burst on its graticule, measure
again the BLUE bar phase displacement. Phase measurement minus the phase displacement of the
original signal indicates the burst/chroma distortion due to the equipment or path under test.
(d)
Differential Gain
The test signal should be a 5 step staircase with added subcarrier (Reference Section, Ref. 2.)
The differential gain may be measured by using a vectorscope in the line time mode. The six sections
of subcarrier are compared for their amplitude relationship and taking the blanking level section as a
reference, the differential gain is defined as the largest departure in amplitude from that reference.
quoted.
NOTE ON TRANSIENT DISTORTION APPLICABLE TO SUB-SECTIONS (e), (f) & (g)
BELOW
The transient gain change due to a change of APL is defined as the maximum transient departure in the
amplitude of each component from that which existed before the change in APL, expressed as a
percentage of the original amplitude. Separate measurements are made on the five step staircase with
added subcarrier (Section 9 Ref. 2), with the APL changed from low (intervening lines at blanking
level) to high (intervening lines at white level) and from high to low.
71
(e)
Transient Gain Change, Luminance (see note above)
Section, Ref. 2), amplified and displayed on an oscilloscope (some commercial filters with amplifiers
overload at normal signal level and require some 10 dB reduction of input signal level).
The oscilloscope should be synchronised by an external source and the black level clamp or dc restorer
should be switched off. Movement of the base line of the waveform when the APL is changed
indicates overload or some other non-standard measuring condition.
Set up the oscilloscope to make the amplitude of each of the spikes corresponding to the steps in turn
equal to 100% with intervening lines at black. Measure the maximum transient departure from 100%
of each of the spike amplitudes when the APL is switched from low to high and vice-versa.
The largest departure from 100% is taken as the result and it should be noted whether the change is
predominantly on only one spike and if so, on which spike.
(f)
Transient Gain Change, Chrominance (See note above)
Set up the oscilloscope using the chrominance filter and measure the maximum transient departure
from 100% of the peak-to-peak subcarrier amplitude on the third step, when the APL is switched from
low to high and vice-versa.
(g)
Transient Gain Change, Sync (See note above)
Using the differentiating network, amplifier and oscilloscope as in (e) above, set the oscilloscope so
that the amplitude of the positive spike corresponding to the trailing edge of sync equals 100% with
intervening lines at black.
Measure the maximum transient departure from 100% of the spike amplitude when the APL is
switched from low to high and vice-versa.
(h)
Chrominance to Luminance Crosstalk
The 2Tc pulse and bar waveform (Reference Section, Ref. 6b) should be used for the test. The
crosstalk, which manifests itself as a change in the mean level of the pedestal during transmission of
the chrominance component, should be expressed as a percentage of the picture level, as determined
by the measurement described in Para 10.2.2. (a), (nominally 700 mV).
10.2.5
(a)(b)(c)
Input/Output Impedance - Return Loss

Return Loss
The measuring points are the output to the uplink and the output of the station receiver.
The test is first carried out using a 2T pulse-and-bar waveform in conjunction with a return loss bridge
which should first be calibrated using two very closely matched, 75 0.1% ohm resistors.
In addition, the same leads should be used for calibration and measurement and the reference path lead
should be identical to the main path connection. With one return loss bridge presently available, a
calibration distortion of -40 dB is provided; if this bridge is used the output is displayed on an
oscilloscope and adjusted to give a reference display (5 divisions for example). The bridge is then
rearranged to include the circuit under test and the unbalance output measured. The return loss is then
calculated by linear interpolation. For large mismatches a 10 dB switch is
incorporated in the bridge to allow calibration at -30 dB. When measuring output impedance the input
signal should be removed and the input terminated. For very small return loss measurements an
external trigger to the oscilloscope is often necessary.
The test should be repeated using the 2Tc pulse-and-bar and 50 Hz waveforms. These results are
respectively recorded as the (a) Luminance, (b) Chrominance and (c) Low Frequency parameters.
72
10.2.6
V.L.F. Response
The signal used should switch all lines to black and white. The switching should occur at a sufficiently
slow rate to allow the waveform to settle before the following transition. The 1st and 2nd overshoots of
blanking level variation are measured (Fig. 10.1) and expressed as a percentage of standard picture
level (700 mV peak-to-peak).
It should be noted that the dc change ("c" in Fig. 10.1) is not measured since it is a function only of the
test signal.
Both the black to white and the white to black transitions are measured and the worst result quoted.
A dc coupled oscilloscope with a very slow timebase may be used for these measurements.
Alternatively, if only a television waveform monitor is available, a line rate display should be used
with the Y amplifier switched to dc coupled and the dc restorer switched off.
Fig. 10.1
10.2.7
Noise
(a)
Weighted Luminance
(b)
(c)
Measurement is made peak-to-peak in the band 40 Hz - 10 kHz (7.5 kHz).

(d)
The path to be measured is fed with a blanking and sync waveform. The output of the receiver is
monitored and the level of unwanted signals is noted.
Signal-to-crosstalk is defined as the ratio, expressed in decibels, of the normal peak-to-peak amplitude
of the picture signal to the peak-to-peak amplitude of the unwanted signals waveform.
73
10.2.8
a)
Modulation Derived Distortion (Sound to Vision Crosstalk)

Measured with whole-time 5-step staircase, without chrominance sub-carrier, into the vision channel
and +8 dBu at 1 kHz into the sound channel. The crosstalk should be measured unweighted, peak-topeak, in the frequency band 40 Hz to 10 kHz (7.5 kHz) using a noise measuring set. The result is
expressed with reference to standard picture level (700 mV p-p).
(b)
Measured with whole time 5-step staircase, with chrominance subcarrier, into the vision channel and
no sound modulation. The crosstalk should be measured luminance weighted, peak-to-peak in the
frequency band 40 Hz to 5.0 MHz, using a noise measuring set. The result is expressed with reference
to standard picture level (700 mV p-p).
N.B. Other methods of measurement using spectrum analysis are acceptable.
10.2.9 Carrier-to-Noise
The measurement uses the onset point for FM threshold as a repeatable technique in determining
Carrier-to-Noise ratio. It may be carried out using the station's satellite receiver or an L Band (0.95 1.7G Hz) spectrum analyser. The noise bandwidth of the spectrum analyser is not critical as long as
operation is at a level of about 10 dB above its noise floor.
A received picture/sync signal is examined on a spectrum analyser or picture monitor connected to the
output of a satellite receiver. An attenuator is then connected in the feeder cable from the LNB and is
adjusted until the signal peaks are taken down to the same level as the impulse noise in the 36 MHz
bandwidth as seen on the spectrum analyser or picture monitor. If the level of the reduced signal is
repeatable and within 1 dB of the peak impulse noise the setting of the attenuator can be read off to
give the value of the Carrier-to-Noise ratio.
A low value may be indicative of low satellite EIRP or poor alignment of the receiving dish and
appropriate action should be taken to improve the situation.
10.3
10.3.1
(a)
(b)
10.3.3
(a)
(b)
10.3.4
(a)
(b)
10.3.5
(a)
(b)
(c)
74
PERFORMANCE FIGURES (Sound)

Output Signal Level
Output Signal Level at receiver
Adjustment error
Gain stability over 1 hour
40 Hz - 15 kHz w.r.t. 1 kHz
125 Hz - 10 kHz w.r.t. 1 kHz
100 Hz 0 dBu
1 kHz 0 dBu
Signal/Noise Ratio
Weighted
Unweighted
0 dBm
0.25 dB
0.25 dB
1.0 dB
0.5 dB
1.0%
1.0%
53 dB
47 dB
-58 dB
10.3.6

Vision to Sound Crosstalk, Weighted
-45 dB

(a)
40 Hz - 15 kHz
(b)
125 Hz - 10 kHz
(a)
40 Hz
(b)
40 Hz - 315 Hz
(c)
315 Hz - 6.3 kHz
(d)
6.3 kHz - 15 kHz
(e)
15 kHz
A Profile is shown in the Reference Section, Ref. 19(a)
(a)
40 Hz
(b)
40 Hz - 200 Hz
(c)
200 Hz - 4 kHz
(d)
4 kHz - 15 kHz
(e)
15 kHz
A Profile is shown in the Reference Section, Ref. 19(f)
10.4
10.4.2
(a)
1.0 dB
0.5 dB
-35 dB
-6 dB/octave
-53 dB
+6 dB/octave
-45 dB
20
Oblique Segment
10
Oblique Segment
20
RECOMMENDED TEST METHODS (Sound)

Output Signal Level
Output Signal Level Adjustment Error
The Insertion Gain Adjustment Error is the difference between the input and output levels at 1 kHz.
The Measurement should be made with an input level of 0 dBu.
(b)
Gain Stability
The greatest change occurring in the output level when the path is measured again after 1 hour.
10.4.3
The input level to the uplink should be -10 dBu. As this test is a measurement of a variation of gain of
the equipment with frequency, corrections should be made for any variation in the input level with
frequency.
level at 1 kHz.
40 Hz,
60 Hz,
125 Hz,
250 Hz,
500 Hz,
1 kHz,
2 kHz,
4 kHz,
6 kHz,
8 kHz,
10 kHz,
12 kHz,
15 kHz,
frequency band.
75
10.4.4
The input level to the uplink should be 0 dBu and the test should be carried out at 100 Hz and 1 kHz.
10.4.5
Signal/Noise Ratio
1 kHz tone at the appropriate level is fed to the uplink and the gain of the test set is adjusted so that the
PPM gives a scale reading of '4' (i.e. O dBu). The input signal is then replaced by a 600 ohm
termination (as defined below) and the gain of the test set is re-adjusted so that the PPM again peaks to
the scale reading of '4'. The signal/noise ratio is the difference between settings of test set gain. The
measurements are made both weighted and unweighted.
(c)
Crosstalk
The path to be measured is lined up to zero level in the usual way and then terminated at the input. The
receiver output is monitored and the peak, weighted output level of any varying (programme) signal is
measured on a noise meter. The difference between this reading and that for the reference level tone is
the crosstalk.
10.4.6
Modulation Derived Distortion (Vision to Sound Crosstalk)
Measured as noise (Para 10.4.5), with vision channel modulated by 100% amplitude, 100% saturated
colour bars.
DUAL CHANNEL SOUND LINKS ONLY

10.4.7
injected into both channel inputs, the channels being lined up in the normal way. The output levels
from the A and B chains should be measured at frequencies between 40 Hz and 15 kHz and the
differences calculated.
10.4.8
A test signal at 1 kHz from a common source should be injected into both channel inputs, the channels
being lined up in the normal way.
The test signal should then be injected into the input of the A channel, the input of the B channel being
terminated in 600 ohms. The levels of the signals on the A and B outputs should be measured and the
difference (i.e. crosstalk) calculated. The inputs should then be reversed and the measurements taken
again to ascertain the crosstalk under this configuration.
Measurements should be made at frequencies between 40 Hz and 15 kHz. A Profile for crosstalk
performance is shown in the Reference Section, Ref. 19(a).
10.4.9
Test signals (initially at 1 kHz) from a common source should be injected into both channel inputs, the
channels being lined up in the normal way. The phase difference between the outputs of the A and B
chains should be measured and the tests repeated at frequencies between 40 Hz and 15 kHz.
A Profile for Phase performance is shown in the Reference Section, Ref. 19(f).
76
Section 11: Digital Video Circuits & Equipment
SECTION 11
DIGITAL VIDEO CIRCUITS AND EQUIPMENT
11.1
11.1.1
Introduction
The majority of equipment, currently being installed in television studios, has part or all of its signal
path in the digital domain. It may utilise internal digital processing of the signal as required or its
whole signal path may be digital with input and output digital interfaces.
The performance of these digital signal paths should be no worse than that of equivalent analogue
paths and traditional analogue tests can be carried out providing suitable converters, for going in and
out of the digital domain, are available. Digital technology introduces additional impairments and it
should be remembered that, whereas analogue systems fail gracefully with increasing distortion
producing a more noticeable subjective effect, digital systems are likely to fail catastrophically for any
increase beyond that which is just noticeable on the picture.
Performance figures can be, generally, extracted from relevant international standards such as ITU-R
BT.601-4 (Encoding parameters) and ITU-R BT.656-1 (Interfaces) and the EBU have published a
definitive document on the subject. This is EBU Tech. 3283, Measurements in digital component
television studios.
Tech. 3283 should be consulted if an in-depth study of digital measurements is required and it is not
intended to reproduce this work here although various performance figures have been extracted and are
reproduced below for convenience. It is recommended that these are met at the time of commissioning
any new system. In addition, this han dBook will define a number of measurement techniques that can
be used on a day-to-day basis to give confidence that digital systems are giving satisfactory
performance.
Composite Measurements:
The Recommended Criteria for Digital Composite Signals should be based on EBU Tech. 3280,
Specification of interfaces for 625-line digital PAL signals.
Parallel Measurements:
11.1.2
Signal Level and DC Shift
Line Driver (output terminated in 110

):
Maximum positive level on each of the symmetrical lines
-0.8V
Maximum negative level on each of the symmetrical lines
-2.0V
Common mode voltage between each symmetrical line and ground
-1.29V
11.1.3
Line Receiver:
Correct operation with minimum input signal of 100 mV p-p over a
duration of 22ms
General:
All 8 (10) data bits should be present and independent.
Rise and fall times

Line Driver:
Timing for 20% to 80% transition
<5ns
77
11.1.4
Difference between rise and fall times
<2ns
Timing
Line Driver:
Clock period
Pulse width
Positive edge of clock pulse should lie in middle of data signals
3ns
Line Receiver:
Deviation of clock signal from eye mid-point of each bit
<11ns
11.1.5
11.1.6
37ns
18.5 3ns
Jitter
The timing of rising edges of clock pulses shall be within 3ns of the
average timing over at least one field.
Impedance
Output of line driver
<110
Input of line receiver
110 10
11.1.7
EMC
The line receiver must receive the data correctly even if signal degradation occurs on both
lines due to interference. The maximum common mode signal between both terminals and
ground should be 0.5V p-p comprising an interference signal in the range 0 - 15 kHz.
Serial Measurements (270 MHz)

Measurements in the physical domain:
11.1.8
Output Level and DC Offset
Output Level
DC Offset
11.1.9
800 mV 10%
0V 0.5V
Rise and fall times
Timing for 20% to 80% transition
Difference between rise and fall times
0.75 to 1.5ns
<0.5ns
11.1.10 Overshoot
Overshoots on edges
<10%
11.1.11 SDI Jitter
Timing jitter
0.2UI p-p
Alignment jitter where UI(270 MHz) = 3.7ns and

the measurement time is 60s.
0.2UI p-p
11.1.12 Impedance
Input and output impedance
Return loss (10 - 270 MHz) (EBU Tech. 3267 and SMPTE 259M)
78
75
>15 dB
11.1.13 Crosstalk
There is no specification. Record worst value measured.
11.1.14 Input cable equalisation and amplifier dynamic range
Gain at 270 MHz
Headroom
up to 40 dB
6 dB (80m cable)
11.1.15 EMC
Emission limits to EN55 103-1
Immunity limits to EN55 103-2
Measurements in the data domain:

11.1.16 Signal Level
Quantising levels should agree with the following:

For 8 bits:
Y = 219 (EY) + 16
Cr = 160(ER - EY) + 128
Cb = 126(EB - EY) + 128
For 10 bits:
Y = 876 (EY) + 64
Cr = 639 (ER - EY) + 512
Cb = 505 (EB - EY) + 512
11.1.17 Rise and fall times
Luminance for 10% to 90% sin2 transition
>174ns
Cr and Cb for 10% to 90% sin transition
>364ns
11.1.18 Picture location relative to TRS
Refer to EBU Tech. 3283 Ref. 1 and Ref. 6. (See Reference Section,
Ref. 34 and Ref. 35.)
11.1.19 Luminance / Chrominance delay
Difference between Y and Cr or Cb
Difference between Cr and Cb
10 ns
=5ns
11.1.20 Forbidden data words
The following data words are reserved for TRS only:

00hex and FFhex
(8 bit systems)
00.xhex and FF.xhex
(10 bit systems)
11.1.21 Ancillary data
Structure should comply with EBU Tech. 3283 Ref. 9 and Ref. 17.
Audio sampling frequency, 48 kHz = 21s/sample.
79
Check channel status bits are set for intended operation in accordance
with Reference Section, Ref. 36 and that they are not modified
inadvertently as the signal passes along the chain.
11.1.22 TRS
The frame signalling should comply with EBU Tech. 3283 Ref. 6.
(See Reference Section, Ref. 35).
Lines
SAV
EAV
023-310
80.xhex
9D.xhex
311-312
AB.x
B6.x
001-022
AB.x
B6.x
313-335
EC.x
F1.x
336-623
C7.x
DA.x
624-625
EC.x
F1.x
11.1.23 Audio in data domain
Measurements specified in AES11-1991 and AES17-1991.
11.1.24 Timecode in data domain (transmitted during VBI)
Identifying signals in TT1 and TT2 fields should be 49.xhex and

15.xhex.
DTC length = 32 words as shown in EBU Tech. 3283 Ref . 13.
11.1.25 Transitions between 10 to 8 and 8 to 10 bits systems
A 10-bit system carrying only 8 bits of information should have the two
least significant bits set to 0 or 1. They should not be left floating.
An 8-bit system carrying data that was originally in 10 bits form should
have dither applied to reduce the contouring effects of truncation.
11.1.26 Relative timing between video and audio
+20 ms to -40 ms
The relative timing of sound to vision should normally lie in the range:
(sound leading by one field to sound delayed by two fields)
However, it may be a production requirement for sound and vision not
to be coincident (e.g. to convey the sense of distance in a large sporting
arena). In this case the intended relationship should be marked on
sound and vision and then maintained through playout, distribution and
transmission.
Measurements in the analogue domain:

11.1.27 A/D and D/A conversion
Signal/Noise
Static Non-Linearity
8 bits
10 bits
8 bits
57.6 dB Unweighted
69.7 dB Unweighted
0.23%
10 bits
0.06%
LSB
Ripple
80
11.1.28 Luminance / Chrominance delay
Timing difference between components
5ns
11.1.29 Switching point of video crossbar
To take place in line
6 or 319
At line-time
30 5s
11.1.30 Colour Gamut
Any signal component should have equivalent RGB values in the range:
0 to 102%
11.1.31 Synchronisation
Analogue synchronising signal specified in ITU Doc. 11/BL/39.
Jitter
Sync / blanking timing should be as shown in

EBU Tech.3283 Ref. 1. (See Reference Section, Ref. 34.).
<0.5ns
11.1.32 PAL 8 Field sequence interference in SDI
Signals originated in PAL and subsequently decoded and passed through a SDI can have a
6.25 Hz luminance flicker introduced when PAL encoded at the station output and
subsequently decoded in the receiver. The station output should be checked using a
monitoring decoder to ensure this impairment is not significant.
The EMC limits and other relevant documentation are available from the Institution of Electrical Engineers, Michael
Faraday House, Six Hills Way, Stevenage, Herts. SG1 2AY.
The address of the EBU is: Ancienne Route 17A, CH-1218 Grand-Saconnex, Geneva, Switzerland.
11.2
11.2.1

Introduction
This chapter of the handbook is designed to present a number of measurement methods that do not rely
on the use of very expensive test equipment and which can be used on a day to day basis to check that
satisfactory performance is achievable in a digital studio. It is slightly different in form from previous
recommended test methods chapters as it does not directly relate to the performance figures in the
previous chapter.
Suitable items of test equipment that may be found to be desirable are suggested in the following list:
AAVS DSA 309
AAVS S310
Pixel Power PC601 Framestore
Rohde & Schwarz VCA
Tektronix SDA
Tektronix WFM601 and WFM601i
Viewtronix line store with Digiview Software
Probel DIVAN.
81
11.2.2
Phase Locked Loop Performance
Reliable phase locked loop operation is, obviously, essential for virtually every item of equipment
handling serial digital signals. There are a number of simple tests that may be of use in establishing
confidence.
Breaking and restoring the input to a phase locked loop is not necessarily a good test of its locking
performance since the loss of input may well (should) be detected and used to widen the capture range
of the PLL. A subtler test involves shifting of the phase of the SDV signal with respect to a stable
reference. Any SDV signal generator with a genlock facility would suffice as shown in the diagram
below.
2 asynchronous
analogue sources
SDV generator
with genlock
PLL under test
Fig. 11.1
Switching between the two genlock sources is likely to upset a PLL with poor stability. N.B. switching
between sources must be accomplished with an electronic switch in order to defeat any capture range
changing built into the PLL. It is also desirable for the field phase of the two sources to be as far apart
as possible.
PLL operation may be checked by feeding a SDI Check field (pathological test signal) to the
equipment or path on test. The bottom half of the picture, which is a grey level, is observed on a
picture monitor and this should be completely free from zits or spots.
If a suitable generator is not available then the grey level could be provided as a full-field signal from
the colour background generator of a vision mixer or graphics device. The grey level is determined by
the values of 80.0hex (0 mV) and 44.0hex (166 mV) for Cr, Cb and Y respectively. Equivalent values are
R = G = B = 23.7%. It is important that these values are set accurately with the aid of a digital video
analyser as it is the particular run of nineteen 0s followed by a 1 from the serial scrambler that
stresses the system.
If a BER or EDH analyser is used for measuring, then the occurrence of zits in the picture can be
quoted as a result in terms of a mean time between errors or as a bit error rate. Ideally, there should not
be any zits to see.
Stress of a different kind can be imposed on the system by employing multiburst as a SDV signal. This
has a large number of frequency components and has in the past highlighted a poor return loss. The
observed fault was picture break up and the symptom could be cured with a reclocking DA. The
recommended EBU figure of >15 dB to 270 MHz is satisfactory.
Another method for checking stability of a signal path is to feed the equipment under test with a stable
signal generator and observe the CRCC for about thirty seconds. The CRCC should not change.
PLLs should recover from any signal disturbance or loss of input on their own without any help from a
human operator.
11.2.3
Equaliser performance
Equaliser performance may be checked by feeding a SDI Check field (pathological test signal) to the
equipment or path on test. The top half of the picture which is a shade of magenta is observed on a
picture monitor and this should be free from any scratches (about 50s duration) which indicate
errors.
If a suitable generator is not available then the signal could be provided as a full-field signal from the
colour background generator of a vision mixer or graphics device. The shade of magenta is determined
82
by the values of C0.0hex (200 mV) and 66.0hex (275 mV) for Cr, Cb and Y respectively. It is important
that these values are set accurately with the aid of a digital video analyser as it is the particular run of
eighteen 0s followed by two 1s from the serial scrambler that stresses the system. Equivalent
values are R = 79.4%, G = 9.1% and B = 89.9%.
If a BER or EDH analyser is used for measuring, then the occurrence of scratches in the picture can
be quoted as a result in terms of a mean time between errors or as a bit error rate. Ideally, no
scratches should be expected.
11.2.4
8-bit and 10-bit compatibility
Consideration should be given to this issue at the design phase by ascertaining what processing takes
place in any interface between equipment operating with different numbers of bits. Compatibility on
any signal path can be checked with the aid of a shallow ramp as shown in Reference Section, Ref. 37.
The level of the ramp is slowly adjusted, upwards and downwards, and the output of the path is
examined on a waveform monitor with attention being paid to any discontinuities, which are indicative
of problems occuring. When the output is digital it should be converted to an analogue signal for
display on the waveform monitor by means of a high quality DAC (i.e. one with 12-bits).
11.2.5
Error detection
The use of error detection and handling (EDH) is presently being considered as it is appreciated that
this technology, when more widely available, will give a very good indication of system performance.
A cyclic redundancy check (CRC) gives a good indication of bit error rate (BER) and generally, if the
CRC does not change then the system is performing correctly.
11.2.6
Colour Gamut
Colour gamut is the range of colours that can be displayed within the triangle defined by the specified
chromaticities for a given television system. In terms of R, G and B signal components, any colour can
be displayed as long as its R, G and B values each lie between 0 and 102% of the peak signal level.
When colours are originated or processed in a component system other than R, G and B e.g. Y, Cr and
Cb, errors can occur when the colours are coded back into the R, G and B domain. This situation is
shown in Reference Section, Ref. 38.
A test should be carried out to ensure that illegal colours are not generated in the component
equipment or signal path. This may be by means of the use of a waveform monitor fitted with warning
lamps which light up when R, G or B values go below 0% (NEG) or above 102% (POS). Test signals
containing appropriate colours can be derived from Reference Section, Ref. 38 where colours in the
inner (R,G,B) cube should give no error indication but colours outside this (e.g. point X) should give
an error indication. An easy test to carry out involves the use of colour bars which can be varied in
amplitude and should only give an indication of an error in the system if their equivalent R, G and B
values exceed 102%.
Files generated on computers using 24-bit file formats potentially exceed the range of values allowed
by the ITU-R BT.601 recommendation when converted. The following should be considered:
Values will correspond to levels in the range 0 to 255, which must be scaled to lie in the range 16 to
235 for the luminance channel, and 16 to 240 centred on 128 for the colour difference channels.
The slew rate of pixel levels must be limited so that the bandwidths of luminance and colour difference
signals do not exceed that specified in ITU-R BT.601.
11.2.7
Jitter
There is a comprehensive section on jitter in the EBU Tech. 3283 document and the important point to
be considered here is to be able to diagnose any degradations caused by jitter rather than attempting to
specify methods of measuring it. Careful observation at the output of a PAL encoder should be
83
sufficient to spot a jitter problem. Alternatively, any signal with a substantial high frequency content,
such as multiburst, should show the effects of jitter after D to A conversion if there is significant jitter
present at the DAC input.
11.2.8
Blanking and synchronisation
The relationship between digital and analogue blanking and syncs is shown in Reference Section,
Ref. 34. It should be noted that the active line time of digital pictures is greater than that for analogue
ones. Checks should be carried out to ensure correct timings occur in both domains and the final
analogue signal before display should be measured on a waveform monitor to ensure correct analogue
blanking has been inserted.
11.2.9
Vertical and horizontal shift
Problems can occur in multi-generation video recording employing digital VTRs with bit rate
reduction if pixel geometry (shift) or values are changed in-between generations. This is because
rounding and truncation errors are introduced each time a new set of coefficients are calculated. A test
pattern or picture with known vertical and horizontal detail should be fed repeatedly through the
equipment and paths concerned and the output observed at each pass to check that picture degradation
does not occur. Any shift will be apparent, subjectively, as spatial offsets of picture information and
corresponding loss of resolution, and an oscilloscope or waveform monitor can be used to investigate
where the problem arises in the system.
11.2.10 Sound / Vision timing

It is assumed that tests will be carried out at installation time to ensure there is correct lip sync on
programme material as it is routed around the station. Care then has to be taken to ensure timing errors
between sound and vision are not introduced when video effects are added or when recording is taking
place. On an all digital station there can be problems with AES/video timing and this needs to be
checked at input and output of equipment such as VTRs. Special care should always be taken when
sound and vision are processed separately and then brought back together again. The use of a clapperboard to check the coincidence of sound and vision is still one of the most reliable methods and some
special test tapes are now being created in international laboratories.
11.2.11 Subjective assessment of picture quality

Analogue production and distribution systems are tested at every technical line-up and are
continuously checked by the monitoring of picture quality. In digital production and distribution the
monitoring of pictures should detect early signs of spots which are the first indications of problems
somewhere in the system. This can be confirmed by BER measurement or the use of EDH.
Dedicated test signals exist for assessing some aspects of the performance of video equipment and
systems but there is an increasing reliance being placed on the use of real programme material in this
area. This is because many processes such as bit rate reduction (compression), format conversion and
movement portrayal etc. employ coding algorithms whose outputs are determined by the content of the
signals presented at the input, the mathematics of the algorithm itself and the transport rate. The
majority of scenes are likely to pass through such systems without any noticeable impairment unless
there is a fault condition but there will be a small percentage of scenes that cause significant
impairments. This percentage is likely to increase when several different types of system are cascaded
together or when compression using a bit rate less than 8Mbit/s is used.
Test sequences of real programme material should be used to assess the performance of such systems
and these should contain normal programme material and some critical material which will stress
the system and give an indication of a quality threshold. Suitable sequences can be obtained from the
EBU (mobile and calendar, Renata, flower garden and Diva with noise etc.) and locally made
sequences including material with areas of saturated colour, fine detail, movement, scrolling captions
84
etc. can be added to a test tape. This should be originated in a non-compressed, digital component
format.
11.2.12 MPEG-2
Systems are becoming available for testing MPEG-2 coders and decoders but are currently not
addressing the measurement of picture quality. Until such time as this position is changed the general
comments in the previous chapter should be observed in respect of subjective picture evaluation. The
new systems are concerned with measurements on the transport stream to ensure that no errors occur in
the coding and decoding processes:
Coder testing:
structure and content of packets.
check for time-stamps - programme clock reference (PCR), presentation time stamp (PTS),
decode time stamp (DTS).
evaluation of information tables - programme association table (PAT), programme map table
(PMT), conditional access table (CAT) and network information table (NIT).
Decoder testing:
check performance of clock recovery with jitter.
check performance of video decoders with stress signals.
Transmission links testing:
check for introduction of jitter.
Compatibility:
check for compatibility between coders and decoders and between products from different
manufacturers.
11.2.13 Cascaded Compression Systems

Various bit-rate reducing techniques are being used for the handling (non-linear editing), storage
(VTRs) and distribution (SDI) in the studio complex. They are also being used for distribution between
studios and to transmitters (34 Mbit/s and 140 Mbit/s). It is becoming more likely that a number of
these will be found in series and that an assessment of the technical quality of programmes passing
through them will have to be made. Work is being done in a number of laboratories and organisations
to devise new methods for assessing picture and sound quality and the extent of distortions likely with
various combinations of compression systems. A future edition of this handbook will deal with the
subject more fully as the results of this work become known, but until then the following notes may be
useful:
keep all television signals in digital form and avoid conversions back to analogue at equipment
interfaces.
when equipment employing the same compression algorithm is cascaded, the vast majority of
any accumulated distortion will be due to the first codec encountered. The final quality will
also be due to the scene being handled at the time.
avoid pixel value changes between passes through the same or different codecs.
transmission via inter-studio 34 Mbit/s is not a significant contributor to distortion unless

analogue interfaces are employed. The quality reduction caused by the use of analogue
interfaces is, however, sufficient to reduce the quality advantage of PALplus to that of
conventional PAL. Two passes through 34 Mbit/s compressed links with analogue interfaces is
85
sufficient to reduce component picture quality below that of studio PAL signals. Even the very
slight distortion introduced by a 34 Mbit/s link is sufficient to produce visible effects after
subsequent handling in a 4 Mbit/s MPEG-2 system.
some non-linear editing systems used in practical studio configurations degrade the signal to
the threshold at which distortion becomes visible.
noise before final distribution coding greatly reduces quality.
avoid slow motion after low bit-rate links.
86
Section 12: Digital Audio Circuits & Equipment
SECTION 12
DIGITAL AUDIO CIRCUITS AND EQUIPMENT
12.1
12.1.1
Introduction
The majority of equipment currently being installed in television studios has part, or all, of its signal
path in the digital domain. It may utilise internal digital processing of the signal as required or its
whole signal path may be digital with input and output digital interfaces.
The performance of these digital signal paths should be no worse than that of equivalent analogue
paths and traditional analogue tests can be carried out providing suitable converters, for going in and
out of the digital domain, are available. Please refer to Sections 2, 3 and 4 of this handbook. However,
the digital technology introduces additional impairments and these will be addressed in this Section.
Performance figures and measurement methods can be, generally, extracted from relevant international
standards such as AES 11-1991 and AES 17-1991. The EBU has also published useful information in
the documents, EBU Tech. 3250, Specification of the digital audio interface (AES/EBU interface)
and EBU Tech. 3283, Measurements in digital component television studios. Some figures from
various relevant standards and documents are reproduced below for convenience.
12.1.2
Signal Level and DC Shift

Line Driver (output terminated in 110
):
Signal amplitude
2V to 7V p-p
Line Receiver:
Correct operation with maximum line driver signal.
Correct operation with minimum input signal of 200 mV p-p over a

duration of a quarter of the biphase symbol period i.e.
For 48 kHz sampling
44.1 kHz sampling
32 kHz sampling
12.1.3
12.1.4
81.4 ns
88.6 ns
122.0 ns
Rise and fall times

Transition between 10% and 90% points
5 to 30 ns
Jitter
Data transitions relative to an ideal jitter-free clock
20 ns
Reference clock
<1 ns
A to D conversion (Ideally)
0.1 ns
12.1.5
Impedance (Measured over the frequency band 0.1 to 6.0 MHz)
Output of line driver
110 20%
Input of line receiver
110 20%
12.1.6
Output balance (Measured over the frequency band DC to 6.0 MHz)

Any common mode component with respect to signal
>30 dB
87
12.1.7
Common mode rejection

Check that no data errors are introduced by the presence of a common
mode signal of up to 7V p-p at frequencies from DC to 20 kHz.
12.1.8
12.1.9
A/D and D/A conversion

Signal/Noise:
16 bits
98 dB RMS
20 bits
24 bits
122 dB RMS
146 dB RMS
(see 12.1.4 above)
Jitter:
Synchronisation
Digital audio to analogue video timing as shown in
EBU Tech. 3283 Ref. 12a. (See Reference Section, Ref. 39).
Timing w.r.t Digital Audio Reference Signal (DARS):
1.0 s
5% of audio frame at transmitter
1.04 s
25% of audio frame at receiver
5.20 s
<2 ns
Jitter
12.1.10 Ancillary data
Structure should comply with EBU Tech. 3283 Ref. 11.
Audio sampling frequency, 48 kHz = 21s/sample.

Check channel status bits are set for intended operation in accordance with
Reference Section, Ref. 36 and that they are not modified inadvertently as
the signal passes along the chain.
12.1.11 Headroom
Minimum undistorted signal level
+12 dBu
Recording (line-up) level relative to FS
-18 dBu
12.1.12 Relative timing between video and audio
The relative timing of sound to vision should normally lie in the range:
(sound leading by one field to sound delayed by two fields).
However, it may be a production requirement for sound and vision not
to be coincident (e.g. to convey the sense of distance in a large sporting
arena). In this case the intended relationship should be marked on
sound and vision and then maintained through playout, distribution and
transmission.
+20 ms to -40 ms
The address of the EBU is: Ancienne Route 17A, CH-1218 Grand-Saconnex, Geneva, Switzerland.
88
12.2
12.2.1

Introduction
This chapter of the handbook is designed to present a number of measurement methods that do not rely
on the use of very expensive test equipment and which can be used on a day to day basis to check that
satisfactory performance is achievable in a digital studio. It is slightly different in form from previous
recommended test methods chapters as it does not directly relate to the performance figures in the
previous chapter.
Suitable items of test equipment that may be found to be desirable are suggested in the following list:
Audio Precision System One (Digital)
Prism Sound DAS-90 with Dscope software
Tektronix 764 Digital Audio Monitor
Tektronix AM70 Digital Audio Analyser
12.2.2
Phase Locked Loop Performance
PLLs should recover from any signal disturbance or loss of input on their own without any help from a
human operator.
12.2.3
Channel Noise
As well as the standard practice of measuring audio channel noise in the absence of a signal (idle
channel noise) it is important with digital systems to measure noise in the presence of signal to assess
the severity of any increase in the noise floor that might occur. With the equipment or system adjusted
to its normal operational control settings a 2 kHz tone is applied to its input. The output is applied to
an ITU weighting filter that is then applied to a 2 kHz comb filter which removes the test tone and its
harmonics. The output of the comb filter is measured using an ITU unweighted meter and the result is
recorded. The level of the 2 kHz tone should be varied between peak level and a level 50 dB down and
the result should not exceed the limit profile shown in Reference Section, Ref. 40.
12.2.4
(a)
Distortion
Difference Frequency Intermodulation Distortion (DFIM)
Two signals of equal amplitude are summed and applied to the equipment under test. The frequency
separation of the two signals should be chosen such that it is prime with respect to any sampling
frequency used within the equipment and that it avoids confusion with power supply related
harmonics. The mean frequency of the two signals, (f1 - f2)/2, shall lie in the range 320 Hz to 20 kHz.
The level of the signal applied to the equipment should follow the inverse of the ITU-T BT. J17 curve
(Geneva 1972) or 50s pre-emphasis characteristic as appropriate. The test should be carried out at
peak level and at a level 20 dB down.
For equipment with either 50s or no pre-emphasis the peak level at 320 Hz should be +8 dBu and for
equipment employing J17 this may need to be 0 dBu. The rms amplitudes of the second and third
intermodulation products are measured and should satisfy the performance profile given in Reference
Section, Ref. 41. A description of this technique and the associated arithmetic is given in IEC 268.3
(1969).
Frequency separations for the two tones which have proved to be useful are 100 Hz below 2500 Hz
and 200 Hz above 2500 Hz.
89
(b)
DFIM with respect to Level
The measurement procedure is similar to that in (a) above except that the level is varied between the
peak level and a level 50 dB below this. It will be found convenient to perform the measurement at a
mean frequency of around 2500 Hz although the performance profile given in Reference Section,
Ref. 42 should be met over the range of mean frequencies from 320 Hz to 20 kHz.
(c)
Harmonic Distortion with respect to Level
A signal at 40 Hz is applied to the equipment under test and the rms values of the second and third
harmonics at the output are measured and expressed as a ratio of the fundamental. The input level
should be varied between peak level and a level 50 dB below this and over this range the performance
profile given in Reference Section, Ref. 42 should be met. Note should be taken of the effect of any
pre-emphasis as described above.
12.2.5
(a)
Timing
Delay through a device or signal path
Test signals at a level of 20 dB below peak level are fed to the input of the equipment or path being
measured. The first method uses an impulse signal and the output timing relative to the input timing is
measured on an oscilloscope. The second method uses a random or pseudo-random noise sequence as
the test signal. The input and output are cross-correlated and the peak in the function is reported as the
timing delay.
(b)
Group Delay Error
The group delay error is a measure of the relative delay experienced by different frequency
components of the signal passing through the equipment or path. A group of sine-waves with known
phase relationship are applied to the input of the circuit being tested. The output signal is sampled and
processed using a discrete Fourier transform. The resulting imaginary component, which indicates the
phase relationship between the original signal components, is then processed to remove the phase
component that is due to simple time delay and sampling widow effects. The cumulative delay is
expressed relative to the delay at 1 kHz. Performance should be better than the profile in Reference
Section, Ref. 43.
12.2.6
Effects due to data compression
Considerable information on the performance of audio systems using data compression techniques can
be obtained by examining the output signal when conventional sinewave signals, usually at a level of
about 2 dB below peak level, are fed to the input. These signals can be either single frequency tones, a
single cycle or short burst, or a multi-burst. The resulting output waveform can be examined on an
oscilloscope or spectrum analyser and in general, any departure from a pure input signal is indicative
of a distortion or shortcoming in the processing algorithm.
Distortions that can be identified using sine wave test signals include:
Block rate distortion where signals are processed in frequency blocks (e.g. for FFT etc.) and
there are small discontinuities between blocks. These show up as a square wave modulation at
the block frequency and the output spectrum will contain additional frequency components.
Pre-Echo where there are additional signals at the output which appear before the input signal
which caused them. This can be caused when a signal change occurs within a block and the
whole block is then modified as a result. The effect is evident with percussive input signals so
can be seen if sinusoidal bursts are used as the input signal.
Noise intrusion where the error signal is not concealed under the critical band masking
threshold which usually occurs when the input signal is complex with several fundamental
90
frequencies, each with a number of harmonics. The effect may be visible if a sine wave multiburst is used as the test signal.
Spectrum shaping occurs in non-linear systems when the input contains several frequencies
each of which get amplified by a different amount. The multi-burst signal should again show
up such errors.
12.2.7
Subjective assessment of sound quality
The use of test signals for assessing the performance of audio systems is never a substitute for
subjective assessment using real programme material. Indeed, the strength of test signals has only been
made possible by the considerable work carried out many years ago to relate distortions on electrical
waveforms to those heard on real sounds by human ears. With modern digital systems, particularly
those that use some form of compression, it is very important to carry out subjective assessment as the
output invariably depends on the type of programme going through the system.
Test sequences using real programme material are available and one source for these is the EBU
SQAM compact disc (EBU Tech. 3253). This contains a wide selection of tracks ranging from full
orchestra and single instruments right through to choral work, speech and a pop group. There is also a
wide selection of solo percussive sounds with different characteristics, such as drums, triangle, gong,
glockenspiel and xylophone etc. Locally made programme sequences can also be used but these should
be mastered on a format which is not bit reduced or compressed.
Subjective assessment should be carried out in respectable acoustic conditions where the room will not
add its own coloration to the sounds being heard.
91
92

Handbook Tech Standar TV Production

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Handbook Tech Standar TV Production

Încărcat de

Drepturi de autor:

Formate disponibile

ITC

HANDBOOK OF TECHNICAL STANDARDS

ISSUE 2.0 - DECEMBER 1996

Standards and Technology, Engineering Division,

ITC Handbook of Technical Standards for Television Programme Production

Issue 2.0-December 1996

Contents & Introduction

On-line version source EBU website

Independent Television Commission, 1992, 1993 and 1996.

ITC Handbook of Technical Standards for Television Programme Production

Issue 2.0 December 1996

Contents & Introduction

On-line version source EBU website

HANDBOOK OF TECHNICAL STANDARDS

STUDIO CENTRES AND OUTSIDE BROADCAST FACILITIES

Technical Performance Working Party Membership

SECTION 1 VIDEO CIRCUITS AND EQUIPMENT WITHIN THE SIGNAL PATH

PERFORMANCE FIGURES (COMPOSITE PATHS)

RECOMMENDED TEST METHODS (COMPOSITE PATHS)

PERFORMANCE FIGURES (COMPONENT PATHS)

RECOMMENDED TEST METHODS (COMPONENT PATHS)

SECTION 2 AUDIO CIRCUITS AND EQUIPMENT WITHIN THE SIGNAL PATH

RECOMMENDED TEST METHODS

SECTION 3 VIDEO TAPE RECORDERS

PERFORMANCE FIGURES (COMPOSITE RECORDERS)

RECOMMENDED TEST METHODS (COMPOSITE RECORDERS)

PERFORMANCE FIGURES (COMPONENT RECORDERS)

RECOMMENDED TEST METHODS (COMPONENT RECORDERS)

SECTION 4 AUDIO RECORDERS

RECOMMENDED TEST METHODS

RECOMMENDED TEST METHODS

SECTION 6 TELECINES AND SOUND FOLLOWERS

RECOMMENDED TEST METHODS

SECTION 7 DISC REPRODUCERS

Issue 2.0 December 1996

Contents & Introduction

On-line version source EBU website

RECOMMENDED TEST METHODS

VERTICAL BLANKING INTERVAL

SECTION 9 PEAK PROGRAMME METERS

RECOMMENDED TEST METHODS

SECTION 10 SATELLITE LINKS PATHS

PERFORMANCE FIGURES (Vision)

RECOMMENDED TEST METHODS (Vision)

PERFORMANCE FIGURES (Sound)

RECOMMENDED TEST METHODS (Sound)

SECTION 11 DIGITAL VIDEO CIRCUITS AND EQUIPMENT

RECOMMENDED TEST METHODS

SECTION 12 DIGITAL AUDIO CIRCUITS AND EQUIPMENT

RECOMMENDED TEST METHODS

Ref. 1: Pulse and Bar test signals (ITU-R BT.451-2)

ITC Handbook of Technical Standards for Television Programme Production

Contents & Introduction

On-line version source EBU website

Ref. 22: Non-linearity Test Staircase Signals

ITC Handbook of Technical Standards for Television Programme Production

Issue 2.0 December 1996

Contents & Introduction

On-line version source EBU website

Technical Performance Working Party Membership

Channel Four Television

ITC Handbook of Technical Standards for Television Programme Production

Issue 2.0 December 1996