synthforge-list — General discussion about SynthForge

Re: [Synthforge-list] SynthForge still active

[Damon C. <da...@ka...>]

On Fri, 2011-07-22 at 08:38 -0500, Matt Henley wrote:
> I really liked what I saw of your software but never got a chance to
> play with it much.
> 
> 
> Just out of curiosity,  did you stop work because of a limits on the
> audio subsystem in linux or ran out of time/interest?  I worked on a
> software project at one point and just couldn't manage to work all
> day, then come home and code..it felt too much like work.

Yes, it was quite hard work, and there wasn't much interest in it.

Plus I have loads of Windows audio software now, including Reaktor, so I
don't actually need it myself.

Damon

Re: [Synthforge-list] SynthForge still active

[Matt H. <nw...@gm...>]

I really liked what I saw of your software but never got a chance to play
with it much.

Just out of curiosity,  did you stop work because of a limits on the audio
subsystem in linux or ran out of time/interest?  I worked on a software
project at one point and just couldn't manage to work all day, then come
home and code..it felt too much like work.

Matt

On Fri, Jul 22, 2011 at 6:03 AM, Damon Chaplin
<da...@ka...>wrote:

> On Fri, 2011-07-22 at 13:02 +0300, Timo Westkämper wrote:
> > Hi.
> >
> > Is this project still active? Before I dig any deeper into this
> > project I'd like to know this.
>
> No, I haven't done any work on it for a year or two, I'm afraid.
>
> Damon
>
>
>
>
> ------------------------------------------------------------------------------
> 10 Tips for Better Web Security
> Learn 10 ways to better secure your business today. Topics covered include:
> Web security, SSL, hacker attacks & Denial of Service (DoS), private keys,
> security Microsoft Exchange, secure Instant Messaging, and much more.
> http://www.accelacomm.com/jaw/sfnl/114/51426210/
> _______________________________________________
> Synthforge-list mailing list
> Syn...@li...
> https://lists.sourceforge.net/lists/listinfo/synthforge-list
>

Re: [Synthforge-list] SynthForge still active

[Damon C. <da...@ka...>]

On Fri, 2011-07-22 at 13:02 +0300, Timo Westkämper wrote:
> Hi.
> 
> Is this project still active? Before I dig any deeper into this
> project I'd like to know this.

No, I haven't done any work on it for a year or two, I'm afraid.

Damon

[Synthforge-list] SynthForge still active

[Timo W. <tim...@gm...>]

Hi.

Is this project still active? Before I dig any deeper into this project I'd
like to know this.

Br,
Timo Westkämper

Re: [Synthforge-list] First thoughts

[Damon C. <da...@ka...>]

On Wed, 2009-12-16 at 22:05 +0100, Thorsten Wilms wrote:
> Hi!
> 
> I just took Synthforge for a short spin to see if it might fit my needs.
> Since my first steps with modular systems I'm never really satisfied
> with "closed" synthesizers.

I don't think SynthForge is quite ready for real use yet. I haven't done
much work on it for a while. I might get back to it next month, though
I'm not sure.


> The Idea of building and wrapping up custom synthesizers is not quite
> what I want, but comes close. I rather want to develop a new modular
> setup for each new song (though using a library of little assemblies
> like they were plugins is fine). It seems the current concept is about
> building _one_ synth at a time?

Yes, that's the main idea. Like Reaktor or SynthEdit.

Though you can of course still create a new project for each song if you
want to.


> What I would like to be able to do is have one patch (as in graph of
> nodes) that contains sub-patches for the different parts like lead, bass
> and so on. Since CPU power is never sufficient for me, at some point,
> it's crucially to be able to disable/enable different sub-patches.

You can't mute sub-patches in SynthForge yet, but it may be added.


> Damon, are you familiar with Ingen? I'm not quite sure I understand your
> approach to polyphony yet. In Ingen, each patch has a polyphony value.
> Each node can be a plugin or a sub-patch, trying to make them the least
> different. Each node within a patch has a boolean polyphony value. So
> you can have nodes that will be per voice or will stay unique without
> any special nodes or whatever other mechanism.

Yes, I have looked at Ingen.

In SynthForge polyphony is determined by the "Midi In" modules, where
you can select the number of polyphonic voices. Anything connected to
the "Midi In" outputs uses the same polyphony value (except the
"Modulation" output which is monophonic). So hopefully it is all taken
care of automatically.


> Consider to allow the addition of automatically connected nodes via
> right click on a port.

You mean things like adding a knob/slider input?


> I think knobs and sliders shouldn't be distinct node types, as they are
> 2 representations of the same thing, a range.

They are very similar from an internal viewpoint, but the user interface
controls are different and may have different control options at some
point.

Damon

[Synthforge-list] First thoughts

[Thorsten W. <t_...@fr...>]

Hi!

I just took Synthforge for a short spin to see if it might fit my needs.
Since my first steps with modular systems I'm never really satisfied
with "closed" synthesizers.

The Idea of building and wrapping up custom synthesizers is not quite
what I want, but comes close. I rather want to develop a new modular
setup for each new song (though using a library of little assemblies
like they were plugins is fine). It seems the current concept is about
building _one_ synth at a time?

What I would like to be able to do is have one patch (as in graph of
nodes) that contains sub-patches for the different parts like lead, bass
and so on. Since CPU power is never sufficient for me, at some point,
it's crucially to be able to disable/enable different sub-patches.


Damon, are you familiar with Ingen? I'm not quite sure I understand your
approach to polyphony yet. In Ingen, each patch has a polyphony value.
Each node can be a plugin or a sub-patch, trying to make them the least
different. Each node within a patch has a boolean polyphony value. So
you can have nodes that will be per voice or will stay unique without
any special nodes or whatever other mechanism.


Consider to allow the addition of automatically connected nodes via
right click on a port.

I think knobs and sliders shouldn't be distinct node types, as they are
2 representations of the same thing, a range.
 

-- 
Thorsten Wilms

thorwil's design for free software:
http://thorwil.wordpress.com/

[Synthforge-list] SynthForge 0.3 released

[Damon C. <da...@ka...>]

I've put SynthForge 0.3 up at:

   https://sourceforge.net/projects/synthforge/

Let me know of any problems.


SynthForge 0.3 (Aug 17 2009)
==============

 o Added lots of simple Maths and Logic modules and a Constant module.
 o Added an LFO module, with a choice of waveforms and modes.
 o Added an Input Switcher module for routing signals. (The Output
   Switcher isn't finished yet.)
 o Updated voice sleeping code to support up to 8 envelopes, in case
   multiple oscillators are used per voice, with separate envelopes.
 o Added support for "scale points", where specific values of ports have
   corresponding labels. When a Choice module is connected to one of
   these ports it uses these labels automatically. Used by the LFO.
 o Added a "Note" output port to the MIDI In module, to make it easier
   to modulate the note frequencies using a linear scale (this is used
   in the LFO demo). There are modules to convert from the MIDI note
   scale to frequencies and vice versa.
 o Added 2 new demos - a Filter demo and an LFO demo.
 o Fixed several bugs, including one that meant many LADSPA plugins were
   not visible in the "Add Module" dialog.


Damon

[Synthforge-list] SynthForge 0.2 released

[Damon C. <da...@ka...>]

Hi,

I've put SynthForge 0.2 up at:

   https://sourceforge.net/projects/synthforge/

There have been quite a few changes, and a few bug fixes. It's still
alpha quality though, so don't rely on it too much.

Let me know of any problems (or if that FM plugin crash is fixed!)


SynthForge 0.2 (Jul 25 2009)
==============

 o Can now play 2 octaves on the computer keyboard, q-i and z-,
 o Added support for putting voices to sleep when the envelope ends.
   Hopefully this will cut down on CPU usage a lot.
 o Moved unison functionality into separate modules, for better
   performance. It allows you to use unison for as few modules as
   possible, and also to use unison for some oscillators and not others.
 o Changed the MIDI In and Envelope module ports to use Gate ports.
   A special Gate value of -1 means that the note is about to be stolen,
   and the envelope should ramp down the amplitude to avoid a click.
   The MIDI In also has an amplitude output which is the note on
   velocity converted to decibels, from -20dBFS to 0dBFS.
 o Better indication of selected modules & controls, and code to stop
   accidentally moving modules slightly when selecting them.
 o Fixed problems selecting the preset manager control.
 o Made it easier to use frames, by keeping track of their children.
 o Added call to g_thread_init() to try to avoid odd occasional crashes.


Damon

Re: [Synthforge-list] crash when loading fm_osc_1415.so

[Damon C. <da...@ka...>]

On Sat, 2009-07-25 at 11:01 -0400, Paul Coccoli wrote:

> Does installing the debuginfo suffice?  The fm_osc_1415 plugin loads
> fine in ams and jack-rack, by the way.

Yes, I think this bug might be in SynthForge rather than the plugin.
I do get the occasional crash when loading different plugins. I'm
looking into that now.


> #7  <signal handler called>
> No symbol table info available.
> #8  0x00382e45 in blo_hd_run_cub () at util/blo.h:209
> 	low = 0
> 	frac = 0
> 	idx = 0
> 	th = (const float *) 0xb4480000
> #9  runFmOsc (instance=0x8b7ad30, sample_count=64) at fm_osc_1415.xml:32
> 	wave = 0
> 	fm = (const LADSPA_Data * const) 0x872eea0
> 	output = (LADSPA_Data * const) 0x8b25c80
> 	osc = (blo_h_osc *) 0x8b803b0
> 	pos = 0
> #10 0x00c56c3b in sfo_real_time_engine_run (engine=0x879e800, sample_count=128)
>     at sfo-real-time-engine.c:259
> 	process_data = (SfoProcessData *) 0x88c2aa0
> 	process_entries = (GArray *) 0x8b7f2a8
> 	entry = (SfoProcessEntry *) 0x419
> 	samples_remaining = 128
> 	n_samples = 64
> 	pos = <value optimized out>
> 	new_pos = 3
> 	i = 0
> 	sample_num = <value optimized out>
> 	output_buffer = <value optimized out>

I can't see anything immediately wrong with this. It may just be memory
corruption somewhere. I'll try valgrind.

Damon

Re: [Synthforge-list] crash when loading fm_osc_1415.so

[Paul C. <pco...@gm...>]

On Sat, Jul 25, 2009 at 5:23 AM, Damon
Chaplin<da...@ka...> wrote:
> This looks like the crash:
>
> #7  <signal handler called>
> No symbol table info available.
> #8  0x00185e45 in ?? () from /usr/lib/ladspa/fm_osc_1415.so
> No symbol table info available.
> #9  0x008c2c3b in sfo_real_time_engine_run (engine=0xa058800,
> sample_count=128)
>    at sfo-real-time-engine.c:259
>
> So it does seem to crash in the fm_osc_1415 code. Though you don't have
> debugging symbols installed so we can't see exactly where it crashed.
>
> I've just tried it again here and it still won't crash.
>
> If you really want to track it down you could remove the
> ladspa-swh-plugins package and build it from the source code, with
> debugging symbols. But you'd need to know what you're doing.

Does installing the debuginfo suffice?  The fm_osc_1415 plugin loads
fine in ams and jack-rack, by the way.

#7  <signal handler called>
No symbol table info available.
#8  0x00382e45 in blo_hd_run_cub () at util/blo.h:209
	low = 0
	frac = 0
	idx = 0
	th = (const float *) 0xb4480000
#9  runFmOsc (instance=0x8b7ad30, sample_count=64) at fm_osc_1415.xml:32
	wave = 0
	fm = (const LADSPA_Data * const) 0x872eea0
	output = (LADSPA_Data * const) 0x8b25c80
	osc = (blo_h_osc *) 0x8b803b0
	pos = 0
#10 0x00c56c3b in sfo_real_time_engine_run (engine=0x879e800, sample_count=128)
    at sfo-real-time-engine.c:259
	process_data = (SfoProcessData *) 0x88c2aa0
	process_entries = (GArray *) 0x8b7f2a8
	entry = (SfoProcessEntry *) 0x419
	samples_remaining = 128
	n_samples = 64
	pos = <value optimized out>
	new_pos = 3
	i = 0
	sample_num = <value optimized out>
	output_buffer = <value optimized out>
#11 0x08051958 in sfo_jack_client_process (nframes=128, data=0x8711600)
    at sfo-jack-client.c:271
	engine = (SfoRealTimeEngine *) 0x879e800
	jack_midi_buffer = (gpointer) 0xb5b6e240
	jack_midi_event = {time = 142295320, size = 3046666240,
  buffer = 0x87a7960 "/dev/shm/jack_fifo.500_default_system"}
	port_num = <value optimized out>
	num_events = 0
	i = 141628928
	event = <value optimized out>

Re: [Synthforge-list] crash when loading fm_osc_1415.so

[Damon C. <da...@ka...>]

On Fri, 2009-07-24 at 23:25 -0400, Paul Coccoli wrote:
> Bug report attached.  Let me know if I should rebuild with different
> configure options.
> 
> $ rpm -qf /usr/lib/ladspa/fm_osc_1415.so
> ladspa-swh-plugins-0.4.15-15.fc10.i386
> 
> I could look at it in gdb if necessary (and if I can remember how to
> disable that bug buddy thing).

This looks like the crash:

#7  <signal handler called>
No symbol table info available.
#8  0x00185e45 in ?? () from /usr/lib/ladspa/fm_osc_1415.so
No symbol table info available.
#9  0x008c2c3b in sfo_real_time_engine_run (engine=0xa058800,
sample_count=128)
    at sfo-real-time-engine.c:259

So it does seem to crash in the fm_osc_1415 code. Though you don't have
debugging symbols installed so we can't see exactly where it crashed.

I've just tried it again here and it still won't crash.

If you really want to track it down you could remove the
ladspa-swh-plugins package and build it from the source code, with
debugging symbols. But you'd need to know what you're doing.

Damon

Re: [Synthforge-list] crash when loading fm_osc_1415.so

[Paul C. <pco...@gm...>]

Bug report attached.  Let me know if I should rebuild with different
configure options.

$ rpm -qf /usr/lib/ladspa/fm_osc_1415.so
ladspa-swh-plugins-0.4.15-15.fc10.i386

I could look at it in gdb if necessary (and if I can remember how to
disable that bug buddy thing).

Re: [Synthforge-list] crash when loading fm_osc_1415.so

[Damon C. <da...@ka...>]

On Thu, 2009-07-23 at 23:15 -0400, Paul Coccoli wrote:
> Subject says it all.  To reproduce, I just need to load that plugin in
> the demo project. This is version 0.1 on Fedora 10 + PlanetCCRMA. Can
> anyone else reproduce?

It doesn't crash here. If you could get a stack trace that would be
useful.

This is one of the main things I'm worried about - buggy plugins causing
crashes. There's not much I can do about them, except maybe not show
them in the module list.


> Overrall, really impressive for a first release.  I'm just happy it
> builds on my system (unlike Ingen).

Thanks. There should be a new release soon, with quite a few
improvements. Though it does change a few modules, so don't build any
complex synths yet.

Damon

[Synthforge-list] crash when loading fm_osc_1415.so

[Paul C. <pco...@gm...>]

Subject says it all.  To reproduce, I just need to load that plugin in
the demo project. This is version 0.1 on Fedora 10 + PlanetCCRMA. Can
anyone else reproduce?

Overrall, really impressive for a first release.  I'm just happy it
builds on my system (unlike Ingen).

[Synthforge-list] SynthForge 0.1 released

[Damon C. <da...@ka...>]

Hi and welcome to the list,

I've just uploaded the first release of SynthForge, version 0.1, and
added a few screenshots.

SynthForge is now capable of creating very simple synths, though there
is still a lot of work to be done. And probably quite a few bugs to iron
out!

Damon