Re: About AudioPannerNode

Frederick makes a valid point. The HRTF data set used in WebKit does not incorporate distance, the impulse responses were measured at a single distance. There are in fact near-field acoustic effects, for example see the measurements made available by TU Berlin (https://dev.qu.tu-berlin.de/projects/measurements/wiki/2010-11-kemar-anechoic). However research in this area is still active and the implementation would become more complex by adding HRTFs at different distances. It should be considered for later versions of the spec but perhaps not version 1. Maybe though the description in the spec should be revised to indicate that distance-dependency in the HRTFs could be introduced later?
Likewise with directionality of sources, I feel that a sound cone approach is sufficient for now. Defining frequency-dependent polar patterns and implementing them will be complex, and would likely lead to them being used improperly. My opinion is that the initial specification should be quite simple and the complexity increased later as applications arise that demand it.
I agree with Robert on the HRTF files. The use of a standard HRTF set initially will make life easier for developers. The option for developer-supplied HRTFs will allow bespoke audiophile applications if necessary, but someone developing a simple computer game shouldn't have to compile an HRTF database themselves. Implementation complexity also increases with developer supplied HRTF sets, unless the measurement points are rigidly specified, which adds complexity for the developer.
Anechoic HRTF data (without room response, as used in WebKit) can be compressed quite effectively by decomposing the impulse response into minimum phase and all pass components. The minimum phase impulse has its energy concentrated at the start of the impulse, so can be truncated significantly, the all pass component can be approximated by a frequency-independent delay, which can be stored in a text file. Last time I did this I managed to compress the data to 25%. The perceptual effects of applying lossy codecs to HRIRs are not that well understood although I have seen a paper that shows degradation of basic inter aural and spectral cues. It should probably be avoided. I believe that MPEG Surround does HRIR processing in the psychoacoustic filter banks so it is very efficient.
Best regards
Chris

http://www.bbc.co.uk/
This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated.
If you have received it in error, please delete it from your system.
Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately.
Please note that the BBC monitors e-mails sent or received.
Further communication will signify your consent to this.

Received on Thursday, 24 January 2013 11:47:03 UTC