Report of the meeting of the SC-02-01 Working Group on Digital audio measurement techniques of the SC-02 Subcommittee on Digital audio, held in held in Rome, Italy, 2013-05-05
The meeting was convened by chair T. Kite.
The agenda and the report of the previous meeting, held on 2012-10-28 in San Francisco, were accepted as written.
AES-X102: Liaison with IEC TC100 PT 61606
Discussion: J. Woodgate asked whether the AES wanted to propose any maintenance for this IEC standard. The secretary said that we probably did not, prior to completion of AES-X187A to update AES17. Woodgate said that it was undesirable that IEC 61606 and AES17 be incompatible. The secretary responded that we want AES17 to be a coherent set of measurements, that would hopefully be compatible with IEC 61606. IEC 61606-3 (professional use) has a 'stability date' of 2013, meaning that it is not expected to change before then.
Action: The secretary will draft a report to IEC TC100; we expect to offer proposals for updating IEC 61606-3 once the update to AES17 is complete.
AES-X118: Liaison with ITU-R study group 6C
Discussion: It was affirmed that 4x interpolation is our recommendation for true-peak measurement. This is based on:
(a) the maximum error being 0.436 dB at 0.4 fs, as reported by Kite;
(b) that for all reasonable program material the error will be much less;
(c) that an error of <0.5 dB in the worst case is acceptable for setting program levels.
Actions:
(1) MY will send TK an electronic copy of AES-R7, the original recommendation sent to the ITU, for amendment.
(2) TK will update AES-R7 with the new information about 0.4 fs and the observation that the error will typically be much lower on real program material.
(3) MY will convey to the ITU our confirmation of 4x as the recommendended interpolation ratio.
AES-X187A: Revision of AES17
Discussion: We began with a discussion of some of the latest issues raised by Kite during the drafting.
1. The Q of the standard notch filter is currently specified as 1.2 to 3, a wide range. Should we increase the minimum Q? Kite pointed out that a fourth-order notch with a Q of 1.2 is down by nearly 0.5 dB at the second harmonic. The advantage of a lower Q is faster settling time, especially at low frequencies. The meeting decided to leave the Q range as it is.
2. The standard bandpass filter is specified to conform to IEC 61260. Should we allow conformance to ANSI standards as well? Woodgate pointed out that IEC 61260 has changed since AES17 was originally written and may now have diverged from its ANSI equivalent. P. Treleaven asked how a filter could conform to two different standards.
Actions: MY will compare IEC 61260 with its ANSI equivalent. The text in AES-X187A will use 'shall' for conformance with IEC 61260.
3. The maximum input level method currently requires the user to monitor THD+N and compression simultaneously. Should we remove the compression requirement?
In an earlier discussion on the email reflector, it was generally agreed that devices which do not perform dynamic range compression will produce a THD+N of 1% before they hit a compression of 0.3 dB. Furthermore the standard already states that dynamic range processing is not covered by the standard and must be disengaged. Therefore the meeting agreed that the THD+N method is sufficient for establishing the maximum input level.
This discussion led to the suggestion from Woodgate, B. Olson, and M. Yonge that background discussions like this should be documented in the AES-X187B companion document, so that the reasons for these measurement methods are recorded. Much of the time spent re-working AES17 has been spent on trying to understand how existing measurement methods were evolved, and it would be good to avoid this in future. Olson described the content of AES-X187B as the answer to the "yes, but..." questions that users might have of AES-X187A.
4. The close-tone IMD measurement cannot be made with the standard equipment [Kite noted: an FFT analyzer, or something similar, is needed to get the selectivity to measure the 16 kHz tone in the presence of the 18 kHz and 20 kHz fundamentals]. The question was asked - should we remove this measurement, relegate it to an 'advanced' section, or update the standard equipment? The group felt that the FFT analysis has long been considered to be standard equipment, and that this should be recognized in the revised standard. It was felt that an informative note in the close-tone IMD measurement should explain that a high-selectivity analyzer is needed. Woodgate encouraged dropping the idea that the standard equipment is all that can be used to make the measurements in the standard.
[Kite notes: After the meeting it occurred to me that other measurements in the standard may have better performance if done with an FFT spectrum analyzer. An example is the power-line products measurement, which using the standard equipment requires the bandpass filter to be placed at successive harmonics of the power line frequency, but with an FFT could be done in a single measurement.]
Kite also mentioned that the draft of AES-X187A was taking a lot of time. In the interests of speeding things up, Yonge suggested that Kite pass technical ideas to him to do some of the writing and wordsmithing. Kite and Yonge plan conference calls for an 'ad-hoc editorial task group' from time to time.
AES-X206: Frequency Response Data file format
Discussion: In April, Kite posted an XML schema and an XSLT transform to the group document site to generate the legacy text format from the proposed XML interchange format. The group felt that this was adequate as a proposed format. Yonge and Olson pointed out that the AES has already published schemas for standards and that therefore no new infrastructure would be needed to publish this one. Yonge will look at the schema and suggest additions to it to include metadata relevant for more generalised interchange.
Liaisons:
A. Mason reported that there had been no change to ITU-R BS.1770-3 (loudness/true-peak metering). Our recommendation of the 4x interpolation ratio for peak meters must be reported before November. Since this is the current ITU-R recommendation, it should not prompt any changed action.
New Projects:
C. Lacinak had contacted the group to consider a proposed performance standard for analog-to-digital converters used in archiving. Lacinak has prepared documents listing the suggested performance metrics. Most of the measurements are derived from AES17.
The main question for the group was whether the AES should be involved with establishing performance standards. It was pointed out that there are examples of this already, for example the jitter specifications in AES3. However these concern the ability of equipment implementing AES3 to talk to each other, which is of critical importance to the standard being adopted. There was some concern that establishing a consensus for minimum requirements for converters could be difficult, and that measurement uncertainty would need to be carefully addressed. A benefit is that some of the proposed measurements could be incorporated into AES17.
The group agreed to accept the project. Mason pointed out that the draft submitted by Lacinak forms the proposal, and that it should be commented on by interested parties. Yonge and Woodgate expressed a desire to separate the measurement methods from the performance requirements. The structure of the document will need to change accordingly. Woodgate suggested a tabular format to save space.
Action: The secretariat will process the project initiation and start reworking the document to the new format in consultation with Lacinak.
New business:
There was no new business.
The next meeting will be scheduled in conjunction with the AES 135th Convention in New York, NY., US., 17 to 20 October 2013.
