I believe I made the AV report privately to Gab.

The primary Csound platform for my real-time pieces is still CsoundAV.
Csound5 remains beta, is limited in its driver choices (referring to Gogins'
edition), more difficult to install currently, etc.

My comment referred to the fact that AV 0.431 does not require a latency
setting for MME, whereas (with identical .csd) 0.432 does (50-100 ms of
delay). Why this is so remains a mystery to me (as many Csound things do),
but the impression I had is that Gab's changes were modifications to AV that
were moving the driver models closer to those of Csound5 (I may be
completely off here). (This, in turn, continues to make me somewhat
suspicious of the current Csound5 driver implementations for Windows -
again, especially with respect to the Gogins' version.)

So, indeed, I would currently recommend 0.431AV over 0.432. (Any other
differences are IIRC minimal, though I think a few old bugs may have been
dealt with as well.)

Art Hunkins

----- Original Message -----
From: "Istvan Varga" 
To: 
Sent: Thursday, September 15, 2005 1:00 PM
Subject: Re: [Csnd] about latency and buffer value


> The report was not about CsoundAV. Or do you suggest using
> that instead ?
>
> Art Hunkins wrote:
>
> > Have you shut down all background processes/programs possible? (On
Windows
> > ME and lower, this can be everything except Explorer and Systray.)
> >
> > Also, my experience was that on CsoundAV at least, better latency could
be
> > had with older versions of the audio drivers than with the new (old
being
> > 0.431 for AV, newer being 0.432). Why this is true, I do not know -
though I
> > did report this to Gab Maldonado. Both are still available at:
> > http://sourceforge.net/projects/csoundav
> > under "View ALL Project Files."
> >
> > I don't recall whether this observation applies to ASIO or just MME - it
> > could well be just the latter.
> --
> Send bugs reports to this list.
> To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk

-- 
Send bugs reports to this list.
To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk

Art Hunkins wrote:

> I believe I made the AV report privately to Gab.

No, it was sent to the Csound list.

> The primary Csound platform for my real-time pieces is still CsoundAV.
> Csound5 remains beta, is limited in its driver choices (referring to Gogins'
> edition), more difficult to install currently, etc.
> 
> My comment referred to the fact that AV 0.431 does not require a latency
> setting for MME, whereas (with identical .csd) 0.432 does (50-100 ms of
> delay). Why this is so remains a mystery to me (as many Csound things do),
> but the impression I had is that Gab's changes were modifications to AV that
> were moving the driver models closer to those of Csound5 (I may be
> completely off here). (This, in turn, continues to make me somewhat
> suspicious of the current Csound5 driver implementations for Windows -
> again, especially with respect to the Gogins' version.)
> 
> So, indeed, I would currently recommend 0.431AV over 0.432. (Any other
> differences are IIRC minimal, though I think a few old bugs may have been
> dealt with as well.)

Well, I think you have already explained more than enough times how
much better CsoundAV is than any other version. Knowing this, I would
most happily remove support for Windows from Csound 5, given that it has
no chance to be even comparable to CsoundAV, if it were not for the fact
that some people are still using it and did not make the switch as you
keep suggesting.

Still, this thread had nothing to do with CsoundAV. It was about using
real time audio with Csound 5, and I did already explain a few possible
solutions. Thus, a discussion about CsoundAV version 0.431 vs. 0.432 is
really off-topic.
-- 
Send bugs reports to this list.
To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk