[Scummvm-devel] Linux music progress report (with Timidity)
lionel.ulmer at free.fr
Mon Nov 19 03:14:03 CET 2001
> > However I don't think Timidity is the right solution - at least not using
> > it externally. For a start, it doesn't quite sync right with the iMUSE code.
> > Watch the "Woodstock" sequence in day of the tentacle, and you'll see how
> > far behind it can fall. We need something that offers far more flexability
> > then Timidity does, at least as a remote server...
Where is this woodstock sequence so that I can check it out ?
> The FreeSCI sound server can be a good alternative. Any volunteers to
> check it?
Well, before anyone complains, Timidity is only a fallback solution for
people like me without a wave-table aware sound card. Otherwise, one should
write directly to /dev/sequencer or whatever (as it is done in the Windoze
code). BTW, does the synchro even work in Windows before anyone complains
that my code is bad ?
And integrating Timidity in the ScummVM code in NOT a good idea... Timidity
is rather big and I do not like these 'let's copy the code instead of
working out an interface' ways of coding...
Moreover, another solution is to use ALSA (and not OSS) so as to be able to
plug-in Timidity as a software sequencer and hope that ALSA manages the
synchro better than the Timidity external server. But in this case, I get
another set of people complaining :-)
If you look at how FreeSCI works (and yes, I did check some of that for my
code), it writes directly to /dev/sequencer. And if you do not have a
wave-table card, you can plug-in a hack to redirect these calls to Timidity
Anyhow, I will continue working in that direction. If you do not agree feel
free to send another patch with another solution :-)
Lionel Ulmer - http://www.bbrox.org/
More information about the Scummvm-devel