[ecasound] Miscellaneous bugs and ECI related points

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

Subject: [ecasound] Miscellaneous bugs and ECI related points
From: S. Massy (theanaloguekid@tak.net.dhis.org)
Date: Mon Jul 09 2001 - 23:43:27 EEST

Ladies(?) and gentlemen,

Here are a few points I have gathered along the past few days, mostly
concerning the ECI.

- mp3 weirdnesses: moving around mp3 files has curious results sometimes,
especially moving backwards without previously moving
forward. Unfortunately this bug doesn't appear to be reproducible
every time, it seems to depend on a factor that I have not yet
discovered. But generally it goes like this:
- -i:mp3file
- -o:rtdevice
- start
- rw secs (no)
- fw secs (works fine)
- rw secs (now works alright)
Also, setpos 0 won't work.

I also noticed that you can have NEGATIVE positions if you rw beyond
the beginning (seems to be true for all formats), is it a bug of

- ECI and forked streams: Something really weird is going on here, it
seems impossible to use any format requiring forked streams from within a C
program. For instance, and so that we all start from the same thing,
try to use some_file.mp3 instead of some_file.wav in the eci example
and see what happens. Perhaps it is a problem peculiar to my
system. As it doesn't work with ogg either I assume the problem
extends to all forked streams. (tested with mp3 and ogg)

- ECI tip: If you are in a position where you execute a command that
moves in a file (such as rw/fw/setpos) and check for the engine-status
just after this command has been executed, you should be aware that
the engine status is switched from running to stopped while the
seek is completing. It is therefore a good idea to usleep() for at
least a tenth of a second before looking for the status, or you might
have surprises has I had.

- batch mode: Finally, I had an idea for a shell scripting friendly
interface. Basically it would be the interactive mode as we get it
with -c but without readline and its prompt. This, of course, is not
so useful now that we have interfaces to very powerful scripting
languages, but I thought it might be nice. What do you think?

S. Massy

To unsubscribe send message 'unsubscribe' in the body of the
message to <ecasound-list-request@wakkanet.fi>.

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

This archive was generated by hypermail 2b28 : Mon Jul 09 2001 - 23:44:13 EEST