Re: [ecasound] Re: Resampling in realtime

From: Aaron Heller <heller@email-addr-hidden>
Date: Tue Jan 25 2005 - 19:56:25 EET

--On Tuesday, January 25, 2005 15:46 +0100 Julien Claassen
<julien@email-addr-hidden-lab.de> wrote:

> Hi!
> Thes resampling clicking problem is still there. I have to use sox,
> which works very well. But still I'd like to be able to use ecasound for
> that task as well. What is the problem there: Ecasound or libsamplerate?

My tests indicate that the clicking problem was fixed in this snapshot:

 http://ecasound.seul.org/download/snapshots/ecasound-2.3.6-cvs-20041221.tar.
gz

Besides listening, I used od to compare the files produced by the two
programs (ecasound and sndfile-resample, the example program that is part
of libsamplerate) with my test sine-waves and found that at 16-bit word
length the results are identical and at 32-bit word length they differ at
most by 1 LSB.

I also observed that files produced by ecasound are epsilon shorter (like
50 to 200 bytes) than those produced by sndfile-resample. I haven't had
the time to figure out way, but this is pretty insignificant as far as I'm
concerned.

I'll post a pointer to my test files when I get a chance.

Thanks for fixing this Kai....

Aaron Heller <heller@email-addr-hidden>
Menlo Park, CA US

--
To unsubscribe send message 'unsubscribe' in the body of the
message to <ecasound-list-request@email-addr-hidden>.
Received on Tue Jan 25 20:15:09 2005

This archive was generated by hypermail 2.1.8 : Tue Jan 25 2005 - 20:15:09 EET