Re: [ecasound] Help understanding -B -b and -z options

From: Kai Vehmanen <kvehmanen@email-addr-hidden>
Date: Sat Apr 23 2005 - 16:55:38 EEST

On Tue, 15 Feb 2005, Cassettes2CDs wrote:

> I do bunches and bunches of stereo 32-bit recording
> with ecasound and my M-Audio Delta 1010LT's. I
> currently do up to 8 stereo pairs at once, but don't
> need them to be -sychronized- with each other at all.

So are you using ALSA/asounrc and the multi-plugin, or the 4Front OSS
drivers, to provide the stereo pairs...?

> The recording process is very fragile, though, and if
> the computer does anything disk-y, I drop frames.

Hmm, have you compared recording invidual stereo pairs vs recording all
channels with single ecasound instance? The latter one should not
be fragile even on a not-so-well-tuned system.

> ecasound -z:db -r -b 8192 -f 32,2,44100,i -i /dev/dsp5
> -el valve,0.015,0.015 -f 32,2,44100,i -o filename.raw
> So my -z is "db," my -b is "8192," and I don't even
> use -B.

Basicly -B option selects the profile (i.e. which set of default
parameters to use for all buffering related options). The other options
(-z:db, -b, etc) then tune individual parameters.

A good intro to this topic can be found at:

A link to above message is mentioned on the tutorials page at:

Now as for reliability, -z:db protects against disk i/o scheduling
problems, while -b and -z:intbuf/-z:nointbuf help against kernel
process scheduling and interrupt latency related problems. For very
reliable recording, the following set of option should provide good
results in most cases:

ecasound -z:db,500000 -r -b:1024 -z:intbuf -f:32,2,44100,i -i /dev/dsp5 -o foo.raw

With -z:intbuf, the value of -b:xxx doesn't have a big impact anymore, as
then ecasound will in any case try to utilize maximum amount of sound
device buffering. With -z:intbuf, -b:xxx then just controls how often
interrupts are raised (1024-4096 are reasonable values, many soundcards do
not support higher values). I've also increased the disk i/o buffer to
50000 frames (default 100000). With 44100 srate, that's about +5secs of
buffering for "foo.raw" and other disk objects in the setup, which is
quite a lot.

But I also want to emphasize that normally you shouldn't need to specify
any of these options. Ecasound should be able to guess the right options
to use for most chainsetups.

  Audio software for Linux!
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
Ecasound-list mailing list
Received on Sat Apr 23 17:32:03 2005

This archive was generated by hypermail 2.1.8 : Sat Apr 23 2005 - 17:32:03 EEST