Hi

I'm trying out the new (to me) fluid opcodes. They sound great! However 
I get some strange behaviour when using the sustain pedal. Here's what I do:

1) pres the sustain pedal
2) play note (and hold it down)
3) release the pedal very quickly after playing the note

This results in the note being cut off :-( It makes it impossible to use 
the pedal for me.

Any ideas what could be wrong? Is it a bug in the opcodes, if so is it 
fixed in later csound versions, or am I doing something wrong? Anyone 
else experienced this strange behaviour?

Here's my csound code for using the fluid opcodes:

; somewhere in , before instruments defs...
gienginenum1 fluidEngine
isfnum1      fluidLoad 
"/home/atte/music/soundfonts/Steinway_IMIS2.2.sf2", gienginenum1, 1
              fluidProgramSelect gienginenum1, 1, isfnum1, 0, 0

         instr mis_steinway2
inotenum	notnum
ivel		ampmidi		1

fluidNote       gienginenum1, 1, inotenum, ivel * 127
	endin

	instr 10000
afluidl, afluidr fluidOut   gienginenum1
	outs		afluidl, afluidr
	endin

I'm using csound 5.05.0 build from source under debian/linux.

I tried building 5.07.0, but that crashes on me, and unfortunately I'm a 
bit pressed for time, so I didn't go any further with that, sorry...

Atte André Jensen wrote:

> I tried building 5.07.0, but that crashes on me, and unfortunately I'm a 
> bit pressed for time, so I didn't go any further with that, sorry...

I found some time to dig into this. Now I'm trying out current cvs, 
which also segfaults. It seems that csound works fine on some of my 
smaller csd's. However my complex, partly programatically generated .csd 
segfaults. I tried removing all instrument definitions but it still 
segfaults. I get no error on the commandline, just:

...
displays suppressed
0dBFS level = 32768.0
Csound tidy up: Segmentation fault
inactive allocs returned to freespace
end of score.              overall amps:      0.0      0.0
            overall samples out of range:        0        0
0 errors in performance
Segmentation fault
...

How do I get closer to the problem? Is there a way to make csound be 
more specific than just "Segmentation fault". As mentioned a build of 
5.05 works just fine on the same machine, build and used the same way...

I tried looking at the changelog for any hints about changes related to 
the stuff I use, but I can't make a connection.

Any input greatly appreciated!

-- 
peace, love & harmony
Atte

http://atte.dk       | http://myspace.com/attejensen
http://anagrammer.dk | http://modlys.dk

Without the actual orcestra it is very hard to determine what problem you 
are seeing.  I would run the program under gdb or equivalent and see where 
it stopped.  Otherwise we are as much in the dark as you!

==John ff

jpff@cs.bath.ac.uk wrote:

> Without the actual orcestra it is very hard to determine what problem you 
> are seeing.  I would run the program under gdb or equivalent and see where 
> it stopped.  Otherwise we are as much in the dark as you!

Sorry for not posting the orchestra, but it's +2000 lines and uses a 
bunch of huge soundfonts and samples.

Anyways I found the problem: My old use of soundfonts with the sfplay3, 
sfload and sfpassign opcodes. Are those depreciated to the point of 
segfault, or have something changed from 5.05 to 5.07 that means that 
they should be used differently?

-- 
peace, love & harmony
Atte

http://atte.dk       | http://myspace.com/attejensen
http://anagrammer.dk | http://modlys.dk

Atte André Jensen wrote:

> I'm using csound 5.05.0 build from source under debian/linux.

An update: I got rid of the segfault with 5.07 and the sustain-pedal 
weirdness remains. Can anyone reproduce this behaviour (as described in 
the original posting)? Am I using the fluid opcodes wrong?

I can post a ready-to-roll csd with simple soundfont example for easy 
testing/reproduction-of-problem, but I don't think it'll be before 
sometime sunday :-(

-- 
peace, love & harmony
Atte

http://atte.dk       | http://myspace.com/attejensen
http://anagrammer.dk | http://modlys.dk