Meeting of SC-02-01, Berlin 2004-05-09

Report of the SC-02-01 Working Group on Digital Audio Measurement of the SC-02 Subcommittee on Digital Audio, meeting held in Berlin, 2004-05-09

The meeting was convened by chair R. Cabot.

The agenda was accepted as written. The chair later changed the order of topics to place the IEC liaison activities at the end of the meeting so that they bridged into the IEC meeting which followed.

The report of the previous meeting were reviewed. T. Kite expressed confusion over a sentence in the third paragraph of "New Projects / Class D amplifier measurements". The report reads, "Ranada observes this to be less acceptable than a higher - but steady - background noise value." It should be changed to read, "Ranada observes noise pumping to be less acceptable than a higher - but steady - background noise value."

Maintenance Projects

AES-6id-R: Review of AES 6id-2000: AES information document for digital audio - Personal computer audio quality measurements.
Kite and W. Jones presented a document they had developed which suggested improvements and updates for AES-6id. They passed out paper copies and Jones promised to post a PDF version to the server that evening.

The details of their document will not be repeated here. In summary, the existing AES-6id document is analog oriented and needs to be revised to handle sound subsystems with digital I/O and support for compressed digital signals. A few comments made in the meeting should be noted:

Under section 3.4, though speaker level outputs may not be common on sound cards, since the information document is intended to be useful for characterizing complete PCs which might well (for example in laptops) include speakers, amplifier testing and appropriate loads must be considered.

There was considerable discussion on the determination of the EUT digital full scale. The tradeoffs between determining full scale through distortion measurements vs doing so by sensing full scale digital codes were discussed. Alternate methods of determining full scale using lower level test signals and calculating full scale from the observed digital codes were also discussed. An explanatory note should be included in the document which describes the inherent tradeoffs and the rationale for choosing a particular method under particular circumstances.

Considerable discussion occurred about high sample rate systems (above 48 kHz) and the effect on test procedures as well as the ability of the equipment under test to actually operate at the higher rate. It was suggested to add a test which verifies the bandwidth and, if possible, the operating sample rate of the equipment under test.

Kite and Jones agreed to edit the existing document to reflect the changes they were suggesting, rather than submitting documents describing the changes which should be made. This was felt to be easier for the committee to review and comment on. To keep the review task from becoming too daunting, they agreed to post the edited document in three stages. The first revisions (approximately half the expected scope of work) would be completed by the end of June and posted for review and comment. The second half would be finished by the end of August and posted for review and comment. The draft which takes into account suggestions from the committee would be posted by the end of September and that version could be reviewed and further comments considered at the next committee meeting in late October.

The secretary agreed to provide a Microsoft Word format version of AES-6id, with appropriately changed headers, to be the starting point for this editing effort.

AES17-R: Review of AES17-1998: AES standard method for digital audio equipment - Measurement of digital audio equipment
I. Dennis suggested that further work on enhancing AES17 with new measurement methods be diverted to efforts at revising AES17 into a format appropriate for IEC publication. Any new measurements which are to be considered should be dealt with in a form which would appear directly in the new format document. [See Liaison with IEC 61606-3, below].

Development Projects

AES-X102: Liaison with IEC MT61606
Since AES-X102 is associated with IEC MT61606 parts 1 & 2, the liaison for IEC MT61606 part 3 will be assigned a new project number. The chair observed that the IEC meeting would immediately follow this meeting. The agenda order was therefore rearranged to put all IEC related material at the end of the meeting.

AES-X118 ITU study group 6 liaison
M. Keyhl reviewed the progress in the ITU studies of loudness measurement. Formal testing had found that a two pole highpass filter followed by an rms meter yielded better correlation with subjective assesments of long term loudness than any of the complex models of hearing which had been submitted for consideration. Keyhle noted that G. Soulodre, who ran the comparative testing, was presenting a paper on this work later in the convention. The chair promised to upload a PDF copy to the FTP site by the next day.

AES-X141: Accurate measurement and indication of peak signal levels to avoid overload of digital media
Keyhl reviewed the document produced by C. Todd detailing the ITU request for peak level metering. Todd had answered some of the questions Keyhl had put to him clarifying the intent of their request. He had also created some new questions in the process which the group discussed. In summary, the group concluded that the goal was to monitor signals were currently being carried and use a reasonably short buffer memory of signals which had already passed by. This buffer could in effect become "predictive" information since a smal time lag displaying results to a human observer would be imperceptible. The display lag effectively displaces the buffer into the future as far as the peak calculation algorithm is concerned. The goal is to determine when a peak has exceeded full scale and by approximately how much it did so.

Keyhl expressed concern that the standard requested was trivially simple, merely requiring a sample rate conversion to an integer multiple of the original rate and a peak detection on that resulting stream. The chair noted that his previous conversations with Todd led him to believe that the ITU merely wanted to standardize the minimum processing required for a manufacturer to be able to indicate peaks within a maximum error, probably less than 1 dB. Manufacturers who supplied equipment in conformance with the standard could implement more complex algorithms which yielded more accurate results if they chose. The statement of conformance with the standard would then insure the purchaser (ITU member or otherwise) that the equipment metering would be acceptably accurate for practical use and would not merely display the sample values on a metering device. The committee's job was to determine the minimum upsampling required for a given maximum error (probably including this error function in the standard document), choose (and publish) a maximum error, document the metering technique and provide verification methods in the standard for evaluating implementations.

A task group to draft such a document was created consisting of R. Caine, Kite, Keyhl, B. Joergenson and Cabot. It was noted that the SC-02-01-D reflector existed and waould be the appropriate vehicle for communication within this task group.

Joergenson agreed to obtain any documents which his company could share on their peak metering implementation (which is one of the few which currently upsample before peak display) and post them to the reflector. Kite volunteered to help with simulations of the error function. The group goal is to produce a document for review by the next meeting in late October. The draft would be posted to the SC-02-01 reflector by late September for full committee review in advance of the meeting.

New Projects:

Jitter Measurement. C. Travis indicated that he is continuing development of measurement methods for characterizing sampling jitter on audio signals. He wants to present his ideas in a paper at the October 2004 meeting of the AES and will submit a document to the group with the intent to create an information document. A formal PIR will be submitted for this work.

Liaison with IEC PT61606 part 3 Dennis has contributed a draft conversion of AES17 to IEC standard format which he had recently completed. The chair noted that this immense effort this obviously had required and thanked him for his effort. The draft was quite well done and did not on first review elicit any negative comments. The group will review it in more detail and post any comments to the reflector. Prior to the meeting the Chair had offered to help with generation of figures for the draft, since illustrations of measurement setups is a common feature of similar IEC standards. When these are completed a new draft will be posted on the reflector.

New Business

There was no new business.

The next meeting will be during the next convention in San Francisco.

AES - Audio Engineering Society