Re: [ecasound] weird delay in real time processing

New Message Reply Date view Thread view Subject view Author view Other groups

Subject: Re: [ecasound] weird delay in real time processing
From: S. Massy (
Date: Fri May 19 2000 - 08:26:23 EEST

On Thu, 18 May 2000, Kai Vehmanen wrote:

> On Tue, 16 May 2000, S. Massy wrote:
> > # ecasound -i:file.wav -o:alsa,0,0
> > results: Output is grabbled and I get underruns report on exit.
> Hmm, what about "ecasound -i:file.wav -o:alsaplugin,0,0"? I tried
> 1.7.7r11 with my AWE64Gold(isa) without problems (ALSA 0.5.5).
When I do that, play a wave through alsa device, I get a minimal
number of underruns mostly at the begining of the processing. The big
issue is when recording, I get over a thousand overruns a minute
sometimes which makes quite a sad recording.

> I might have asked about this already, but anyway, how ecasound was
> installed... RPM, recompiled? If recompiled, did you have the
> correct ALSA header files installed (/usr/include/linux/asound.h and
> /usr/include/sys/asoundlib.h).
:) Good question, I recompiled alsa (0.5.7) and ecasound just to be
sure and I noticed little if no improvement. It's good to know though,
I had tried to install a prerelease version of alsa 0.6.0 but ecasound
wouldn't work at all with this one. I might reinstall that version and
compile ecasound just to see if I gain anything.

> > # ecasound -i:alsa,0,0 -o:alsa,0,0
> > results: Same except that since it's in realtime I get both underruns
> > and overruns reprots on exit.
> I got a few underruns (1 or 2 most of the time), but no sync problems.
Hum... for how long did you run the test? Most of the time the
overrun/underrun problem shows up after a minute or so. The sync
problem occurs when I use OSS emulation (/dev/dsp).

> > # ecasound -i:alsa,0,1,0 -o:alsa,0,0,0
> > result: "ERROR: [AUDIOIO-ALSA2] : "Sample rate 44100 is out of
> > range!"" (huh?)
> This is probably wrong... input card0, device1 (is this a typo?),
Well device 1 on card 0 (SBLive) is supposed to be the MIC according
to the output of "arecord -l". But it's bogus or I don't understand
something, I can bypass the "samplerate out of range" error by using
alsaplugin, then I have to take the channel parameter from 2 to 1,
then take the buffersize higher than 128 for ecasound to load with
this command line, but once I start the processing ecasound crashes and
freezes my shell. :(

> subdevice0. As subdevice is a low-level concept, you shouldn't need
> to specify it in normal use (default is -1, not 0).
Thank you, I didn't know that, I'm fairly new to ALSA :)

> > # ecasound -i:/dev/snd/pcmC0D0C -P:/dev/snd/pcmC0D0p
> This is an invalid command line.
I think you've made an error in the doc then.
From examples section in ecasound's documentation:
"The ALSA drivers have a somewhat different option syntax. You can
specify the actual device name (/dev/snd/pcm...) or specify "alsa"
and then the card number and the device number. Otherwise ALSA
inputs/outputs work just like OSS-devices."

Sebastien Massy
> --
> Kai Vehmanen <> ---------------- CS, University of Turku .
> . audio software for linux .. .
> . armchair-tunes mp3/wav/ra . .
> --
> To unsubscribe send message 'unsubscribe' in the body of the
> message to <>.

To unsubscribe send message 'unsubscribe' in the body of the
message to <>.

New Message Reply Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Fri May 19 2000 - 08:26:24 EEST