Hi Eric,
Congratulations first for your work on SA and MP4. If I understand correctly
Youngmoo's post in the archive, you did this as a PhD student, I'm amazed :-)
To get back to the thread, thanks for the info. I'm no PCI or DMA expert
either, let's hope the experts will speak up.
Concerning latency, it seems that the OS and drivers may be largely
responsible for that. I was told (haven't checked) that on the same PC with
the same soundcard which gives you 30 ms or worse AudioIn/AudioOut latency
under Win98, you can get down to 3 ms under BeOS.
Eric Scheirer wrote:
>
> Hi Michel,
>
> >- it would be extremely useful if harware-synthesized audio (digital domain
> of
> >course) could be got back in real time from the card for further use by
> >software (an obvious use being recording it to a file), as well as going to
> >the card's audio out connectors.
>
> At least for some implementations of the whole MPEG-4 standard, this
> will have to be the case. In a full MPEG-4 implementation that includes
> AudioBIFS post-production as well as SAOL support, the synthesis is
> not the last step in the processing chain. The result of the synthesis
> is further manipulated (ie by mixing or delay) by the AudioBIFS
> instructions. Thus, for hardware/software implementations, it is
> necessary that the SA synthesizer, if implemented in hardware,
> return the samples back to the AudioBIFS processor.
>
> (Some implementations may integrate the AudioBIFS part and the SA part
> of MPEG-4 more deeply than is required; there's some good reasons for
> doing so. In such a case, the two parts could move in conjunction
> to the hardware side and there wouldn't necessarily be a reason to
> return the samples to the host).
>
> >- even better, allow the use of the card as an SAOL processing farm : an
> API
> >could be provided so that software can request processing of successive
> sample
> >frame blocks from the card and get them back as fast as the card can
> compute
> >them. Ideally you would implement a vst2 "proxy" dll for that.
> >
> >Is point 2 technically feasible ? Anything going on along these lines ?
> >(Forgive me if the questions have been raised before).
>
> It seems to me that latency issues are problematic if you want to
> do real-time processing in such a scenario. It requires two DMAs for
> each roundtrip cycle to and from the host, and so if you are trying
> to do several stages of processing, the DMA latency will add up
> quickly.
>
> But I'm not an expert in such matters; there are people here who
> understand these things much better than I.
>
> Best,
>
> -- Eric
-- Greetings, Michel ......................................................................... Michel Jullian Directeur General email mj@exbang.com Exbang Industries S.A. Mas Chauvain route de Villeneuve tel +33(0) 499 529 878 Maurin 34970 Lattes France fax +33(0) 499 529 879 .........................................................................
This archive was generated by hypermail 2b29 : Wed May 10 2000 - 12:15:25 EDT