SC-02-01 meeting, San Francisco, 2012-10

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 San Francisco, 2012-10-28

The meeting was convened by chair T. Kite.

The agenda and the report of the previous meeting, held on 2012-05-28 in Budapest, were accepted as written.

Projects assigned to this group but not mentioned here had no action requested or required - see for details.

Open Projects:

AES-X102: Liaison with IEC TC100 PT 61606
Discussion: The IEC TC100 had its meeting last week, with John Woodgate in attendance. Nothing has been reported yet. Secretariat will contact John Woodgate and report to the email reflector.

AES-X118: Liaison with ITU-R study group 6C
Discussion: The question of the response of the true peak meter is still outstanding. The group had reported its recommendation to the ITU-R group (namely, that 4x or higher interpolation should be used). However ITU-R have requested a more specific recommendation.

The purpose of the peak meter is to measure stored audio to make sure that no device later in the chain clips. The group agreed that it was better therefore for the meter to overread than to underread. We considered a recommendation that meter manufacturers choose the interpolation ratio and boost the peak reading by the worst-case underread for that ratio. This would ensure that the meter would never underread. However the prospect of different meters reading differently on the same source material (because of using different interpolation ratios) was deemed very undesirable.

The group once again looked at the table listing the worst-case underread of a peak meter with a given interpolation factor. We made two conclusions: (a) That the frequency of 0.45 fs rather than 0.5 fs is practically most significant, given the typical passband of the majority of converters on the market; and (b) that an interpolation factor of 8x should be used. The worst-case underread at 0.45 fs is 0.13 dB. (The worst-case underread for 4x is over 0.5 dB at 0.45fs. It was felt that the extra processing involved at 8x compared to 4x was worth this extra headroom.)

Mark Yonge will discuss the group's decision with A. Mason who had raised the issue.

AES-X187A: Revision of AES17
Discussion: Kite presented and the group discussed issues arising from the work that has been done since Budapest.

1. DC powered devices. In the existing standard it is stated that these should be powered by a supply whose peak-to-peak ripple is constrained. It was suggested that this be changed to specify that the supply be the supply that the manufacturer provides with the EUT. The primary motivation for this is that the EUT will typically be used with the provided supply, not with a laboratory supply, and therefore its performance should be measured with this supply. In particular the hum measurement is only valid with the provided supply. It was agreed that the provided supply should be used for all measurements. The group also considered EUTs which do not have a supply provided. There was no clear conclusion on this. Thomas Kite will propose something in the draft standard.

2. 10 Hz vs. 20 Hz minimum frequency. The standard is not consistent with regard to the lowest frequency of interest. Although the group was initially divided on which was better, it was noted that the AC generator is only specified to 20 Hz. This proved the deciding factor. The group decided that 20 Hz shall be the minimum frequency throughout the standard. There is nothing to stop users of the standard from extending measurements to 10 Hz if they so choose.

3. Distortion meter. It was decided to delete this from the list of measurement equipment. The existing standard notch filter and level meter are adequate.

4. Measuring the maximum input level if the EUT gain adjustment goes to zero. There was a lively discussion on this. The consensus was that this measurement should not be performed on such EUTs, and that the language should be modified to indicate that the 'minimum operable gain' should be used. Mark Yonge provided the new language in the document, and has uploaded it to the document site.

5. Allowing text report of an x-y graph ("+/- xxx dB from y Hz to z Hz"). Some measurements already allow this on an ad-hoc basis. The group agreed that it should be allowed for all x-y results. We agreed that the preamble to the measurement section should include a template for reporting different result types. Each measurement should specify the items that should be inserted into the template.

6. Sync paragraph in delay measurement. M. Poimboeuf recalled the reasons for this paragraph and asked that it be retained.

7. Measuring input-to-output phase. The group agreed that for EUTs with delay, the sine wave method is useless. The group approved replacing this method with one that uses (for example) the impulse response. Kite will propose something in the draft document.

Kite will post an update to the document as soon as possible. He requested that if possible the members of the group read through Annex 7 of the curent draft in particular. This Annex covers fast measurements for production, and will be important to the widespread adoption of AES17 in production settings.

AES-X187B: Companion ID to AES17
Discussion: Thomas Kite requested that this project be moved to the holding area. The rewrite of the 'A' document is more important and should be finished as soon as possible. The 'B' document can be tackled once the 'A' document is ready. The group agreed on this approach. Mark Yonge will so move this project.

AES-X206: Frequency Response Data file format
Discussion: Several members of the group had read the draft proposal by John Woodgate. John's work was praised as being a very good attempt to draft a very difficult standard. However, ultimately the group felt that it would not be possible to produce a standard worthy of AES under the constraints of existing files in the field. There is too much diversity and incompatibility, and the file format (such as it is) is very loose. Instead, the group agreed that we should publish an XML schema defining the structure of the .frd file. Users of the schema would then write XSLT translations as needed to generate a file that could be read by existing applications. Eventually, applications would read and write the XML-based file natively.

The group output would be the schema in electronic form (which would be normative), and a printed document listing the attributes defined in the schema. The first thing to do would be to propose a new scope for this project, with a rationale. We then agreed that we should notify SC-04-01, SC-04-03, and SC-04-04 to get their comments, since this approach will be of interest to those groups.

Thomas Kite agreed to write the scope and rationale. That will be posted to the email reflector for comments before passing this to these other groups.


There were no liaison issues.

New Projects:

There were no new projects.

New business:

There was no new business.

The next meeting will be scheduled in conjunction with the AES 134th Convention in Rome, Italy, in May 2013.

Many thanks to those who attended, and for the discussions on the reflector since the last meeting. Look forward to seeing you next time.

Thomas Kite
Chairman, SC-02-01

AES - Audio Engineering Society