RE: More ICE1712 hassles (was Re: [linux-audio-user] Minidisk question)

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: RE: More ICE1712 hassles (was Re: [linux-audio-user] Minidisk question)
From: mikko.a.helin_AT_nokia.com
Date: Mon May 20 2002 - 14:55:22 EEST


You can't mix at unity gain even two channels of let's say of 24 bit data without exceeding the 24 bit data range. So you will always need some negative gain setting for the mixer channel gains to avoid clipping. There may though be some hardware bug in design which just isn't documented (not the first one for envy24), maybe M-Audio have fixed this just by compensating the level settings (program the envy for gain -6 dB and show a 0 dB gain in mixer).
-Mikko

> -----Original Message-----
> From: ext Andrew Nesbit [mailto:alnesbit_AT_optushome.com.au]
> Sent: 19. May 2002 18:52
> To: linux-audio-user_AT_music.columbia.edu
> Subject: More ICE1712 hassles (was Re: [linux-audio-user] Minidisk
> question)
>
>
> Frank Barknecht <barknech_AT_ph-cip.uni-koeln.de> writes:
>
> > > and that using the monitor mixer with outputs set to "0dB" gives a
> > > harsh clipping-like effect.
> >
> > The monitor mixer is H/W In 1 and 2, isn't it?
>
> No. You have a choice of three signals to send to the output of the
> card, which can be selected in the patchbay:
>
> 1. PCM output from some "player" application;
> 2. the signal which arrives on the "H/W In" jacks on the card;
> 3. a mix of the above classes of signals, where the output levels are
> defined by the mixer levels in the monitor mixer panel of
> envy24control.
>
> > I don't hear a clipping
> > here, but PD shows me that the ADC seems to clip, but it does this
> > regardless of the settings in envy24control. I'm not sure,
> how PD reads
> > it's ADC information, though.
>
> Here is what I've experienced: i fire up envy24control. All the
> sliders in the monitor mixer are set to 0dB. I have PCM being output
> by aplay/XMMS/whatever. If I set the patchbay to "digital mix", and
> unmute the "PCM Out 1 & 2" channels (all other channels muted), then
> the output signal is (incorrectly, IMHO) hotter than if I select "PCM
> Out 1 & 2" for "HW Out 1 & 2" in the patchbay. This leads to
> clipping. Analogously for having "H/W In 1/2" signals pass through
> the monitor mixer instead of the "PCM Out 1 & 2" signals.
>
> I have not done any "real" scientific analyses of this; this is based
> on what I /hear/, and it is a very significant difference and is not
> hard to notice.
>
> My benchmark for how envy24control should operate is the M-Audio
> Control Panel under Windows because it behaves how /I/ expect, that
> is, no nasty clipping: whether I select "PCM Out" or "Monitor mixer"
> in the patchbay, I seem to get exactly the same effect (unitary
> gain).
>
> > > I also get a short burst of noise at the very start and end of
> > > playback (but I don't know if this is peculiar to this driver or
> > > card).
> > Now this is serious: I get the same, and I always thought it somehow
> > was my fault. But this is only a small and short clickish sound.
> > OTOH I remember, that my SBLive showed te same behaviour,
> but I can't
> > test this right now.
>
> These problems, including the lack of OSS Mixer emulation, and a few
> other problems, are things that I have resolved to try to rectify by
> the end of this year, at the latest. I am still working my way
> through Rubini & Corbet's "Linux Device Drivers" 2nd ed. and trying to
> understand the ALSA source code and the ICE1712 programming specs. On
> top of this, my main priority is all my university work. I am still
> trying to formulate my ideas for improvement in ICE1712 drivers and
> ALSA in general, and when I do, will post a request for comments on
> alsa-devel. Then I can get cracking. I feel bad about mentioning
> this before anything has yet come to fruition; all talk and no action
> is a bad thing in anybody's opinion, but this is probably the
> appropriate time to mention it.
>
> Andrew.
>


New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Mon May 20 2002 - 14:43:30 EEST