Re: [iiwusynth-devel] FluidSynth roadmap

[Antoine Schmitt]

Hello all,
Answering a few emails, including one that I realized Markus sent 
only to me. Sorry for answering late, my schedule is hecktic these 
days.

:::::::::6/11/02::::22:46 +0200::::address@hidden:::::::::
> -Chorus: Have you checked out a version from CVS recently? The 
> chorus was completely rewritten some months ago. The original sox 
> chorus did make some strange noises.

I'm not fond o fthe chorus sound, but I must say that I have little 
experience with chorus sound in general.
One thing is that the API accepts a 'number' of lines up to 99, which 
seems quite a lot, and the sound is really distorted after you pass 
10 lines...

> - Clicking noises: I'd guess, that hardware cards limit the smallest 
> possible attack time. The SF 2.01 standard says explicitly, that the 
> smallest attack time is instantaneous. I have that problem myself 
> occasionally with sound fonts from the web, it's quite annoying. 
> Could be fixed easy (some compatibility mode switch for the synth).

It seems that these clicking sounds only happen on a certain 
hardware/system/software configuration. Forget about it.
The only thing is that the default gain of the FluidSynth is 0.1, 
which makes a real difference with the volume of a soundfont-enabled 
soundcard. Maybe it should be changed to something like 0.5.

> - Reverb parameter query: exists, forgot to move it to the API :-(
> Chorus should be fine, though.

Right.

> - Stereo: The standard allows to explicitly lock the phases of two 
> voices, iiwusynth keeps them separate. If L/R voice are identical, 
> this does not matter. There might be a chance, that a random time 
> delay occurs between L/R voice, but I have never noticed anything by 
> ear (stereo piano with headphones).

Does that mean that stereo is implemented as two instrument zones in 
FluidSynth ?

> - Filter? Any problems?

I haven't mastered yet all the API of soundfont, and following my 
discussions with Peter, I was under the impression that the filter 
feature of the Fuidsynth was not accessible through the API. Maybe I 
misunderstood. Could you point me to the API to control it ?

> - Unloading sound fonts: Maybe stopping the voices using it before 
> freeing the sample mem helps, that would be easy to fix.

That would be really nice. I dont know where to look for to do this 
(how to get the voices using a certain soundfont).


:::::::::20/11/02::::20:42 +0000::::Josh Green:::::::::
> So thank you guys!
>
> Peter Hanappe
> Markus Nentwig
>
> 2 great persons helping to bring open music composition to the world at
> large :) Three cheers.. Hip hip hooray!

I agree ! hooray !

:::::::::6/11/02::::9:46 -0800::::Josh Green:::::::::
> On Wed, 2002-11-06 at 07:31, Antoine Schmitt wrote:
>  > Some other topics I'd like to submit to your sagacity:
> >  - have FluidSynth automatically take advantage of hardware devices 
> > if available
>
> How? Do you mean like a hardware SoundFont device like SB Live! etc?

Yes.

> Using one of these means you don't need FluidSynth. Of course there are
> many draw backs to hardware SoundFont devices (not well supported in
> Linux, not very portable between platforms, usually a pain to route
> synthesized audio to other applications).

Right, but on an end-user point of view, in certain contexts that are 
not rare, it would be interesting if a client program, written on top 
of FluidSynth, could take advantage of the CPU acceleration of an 
SBLive automatically if it is present and accessible.


>  >
> >  The only thing that is quite complex I think, and which is not caused
> >  by iiwu but by the opensource process, is that it is very difficult
> >  to rely on a stable version, as it changes all the time with bug
> >  fixes and new features. So for the production of a client of iiwu,
> >  one has to have the full development environment available, CVS, as
> >  well as all the other free software libraries (portaudio, MidiShare,
> >  now linsndfile for the Xtra...) + update regularely and recompile
> >  everything. I dont know how things are done usually with OpenSource
>  > software (I come more from the industry - NeXT, etc..), and I know
> >  that all this is done gracefully by all of us, but I think that it
> >  would be good to have stable releases from time to time, fully built
> >  and available in binary format, with documentation.
>
> What is your development platform? Are you using Linux or a win32
> environment?

I'm on Macintosh + Windows. Our project is targetted toward standard 
end-users, and our integration platform is Director. This explains 
the choice.
And this explains the difficulties of using CVS clients...
`

CVS changes ::

For your information, I will be working this week-end with Peter 
Hanappe to integrate my mods to the CVS tree, related to the 
integration of FluidSynth to the Xtra:

src/iiwusynth.h, src/iiwu_synth.c : ADDED synth_get_cc and synth_get_pitch_bend
src/iiwu_defsfont.c : FIXED unloadsoundfont memory leak
src/ iiwu_event.c : FIXED bug in heap allocation, that caused a crash 
during sequencer free.
winbuild/iiwusynth.dsw : ADDED a LIB and a DLL project
macbuild/iiwusynth.mcp : ADDED a shared lib and a lib project
src/iiwu_event.h, src/iiwu_seq.h, src/iiwu_synth.h, 
macbuild/iiwusynth.mcp : EXPORTS symbols using the __declspec(export) 
pragma on Macintosh (the iiwusynth.exp file is not used any more).
src/iiwu_sys.h : exported iiwu_curtime on Macintosh (for sequencer).
src/iiwu_seq.c : FIXED bug when changing the sequencer timeScale.
src/iiwu_ramsfont.h, iiwu_ramsfont.c : ADDED these two files for 
dynamic creation of soundfont in RAM, using API.
bindings/director : full implementation of the Director Xtra. 
Everything has changed.

Cheers !


++ as