Check out my new API for instruments(free demo synth)

[jeffh]

Hello everybody,

I'm working on a new GPL'd instruments plugin API that aims to make writing high quality plugins on Linux much easier. I've only been working on it for about 3 weeks, but I have a reasonable demo synth together, and plan on making the first alpha release soon. It's based on DSSI, and I've included a jack-dssi-host in the attachment so it can be run standalone.

I'm trying to promote it early in hopes of attracting developers to use it, or possibly even contribute to it. The API is based on the official DSSI example synths, with my library of audio DSP code, and C++ Qt GUI widgets.

The library is a little rough at the moment, but I currently have:

*A small but growing collection of audio DSP modules designed to work together, with a focus on sound quality, CPU efficiency, and interoperability, written in C.

*A library of widgets designed to be easily theme-able with CSS style sheets, written in Qt/C++.

The demo plugin I've uploaded is a 2 osciilator analog-style synthesizer. It can make some pretty tasty acid sounds if you crank up the filter resonance, filter envelope and distortion. However, it does NOT yet have the following features implemented:

Pitchbend
Unison
Presets/Banks (although the instruments state does seem to be saved in Qtractor after closing the project)

Once the API is finished, I plan on releasing quite a few plugins of my own with it, and I'm hoping others will follow suit, so that we Linux users can finally have the same kind of plugin variety than Windows/VST users enjoy.

The project homepage is here:

http://sourceforge.net/projects/libmodsynth/

[slowpick]

Hexter is a nice dssi synth, this modular synth
might be interesting if more widely available in dssi form

http://drobilla.net/software/ingen/

used to be OM

[autostatic]

Hexter is a nice dssi synth, this modular synth
might be interesting if more widely available in dssi form

http://drobilla.net/software/ingen/

used to be OM

Ingen is not a modular synth. It's a plug-in host.

Best,

Jeremy

[autostatic]

Hello jeffh,

Basically all plug-in developers are focusing on LV2 in favor of DSSI. LV2 is the future. But great initiative anyways as there is not much wrong with DSSI and a lot of hosts support it. And more softsynths are always welcome!

Best,

Jeremy

[studio32]

Thanks, but a much better place to discuss this is the LinuxAudio Developer mailinglist from Linuxaudio.org (LAD)

[jeffh]

Basically all plug-in developers are focusing on LV2 in favor of DSSI. LV2 is the future.

Hi AutoStatic,

I evaluated LV2 before I decided on DSSI. The thing is, you can find posts of the internet back in 2007 where people were saying "LV2 is the future", and yet there is still almost no LV2 synthesizers. LV2 still has way too many experimental features for being something like 7 years in the making, DSSI is mature, stable, and works now.

[jeffh]

Thanks, but a much better place to discuss this is the LinuxAudio Developer mailinglist from Linuxaudio.org (LAD)

Thanks for the heads up, but it looks like joining that group requires me to attend their yearly conference and sign my life away by agreeing to all sorts of rules and board decisions, I'm not sure I'm ready for that kind of commitment yet

http://linuxaudio.org/policy

[autostatic]

The linux-audio-dev mailinglist doesn't acquire any membership. The consortium is something different, it consists of the entities listed here: http://linuxaudio.org/members

Best,

Jeremy
(webadmin linuxaudio.org)

[autostatic]

Hi AutoStatic,

I evaluated LV2 before I decided on DSSI. The thing is, you can find posts of the internet back in 2007 where people were saying "LV2 is the future", and yet there is still almost no LV2 synthesizers.

Hello jeffh,

True, but the number is steadily growing, and if the LV2-Juce wrapper counts too than the number of LV2 synths is even bigger.

LV2 still has way too many experimental features for being something like 7 years in the making, DSSI is mature, stable, and works now.

Agreed, but then LV2 tries to offer a complete API, so one single framework and no need for extra API's (like lv2-c++-tools for instance). And yes, it's been in he making for a long time but with the release of the new LV2 stack last year LV2 has made quite a progress and it's gaining a growing level of adoption. Which doesn't deprecate or obsolete DSSI, not at all, there are quite some great DSSI plug-ins available.
Also I just reread my initial post and it does look a bit presumptuous regarding the work you've achieved. This was in no way my intention, my apologies. If it becomes easier to create new sofsynths for Linux, in what form that may be, it most certainly has my support!

Best,

Jeremy

[StudioDave]

Hello everybody...

Hi Jeff,

http://linuxaudio.org/resources

Best,

dp

[studio32]

I agree with Autostatic. LV2 has progressed a lot in the last couple of months and there are some nice plugins atm, also synths:

http://calf.sourceforge.net/
http://code.google.com/p/foo-yc20/
http://deliriumdecrypted.blogspot.com/
http://blog.loomer.co.uk/2011/10/linux-lv2-preview.html

All though LV2 has good documentation and more and more developers are taking it up, DSSI seems to be more simple to code, but has also less possibilities.

[autostatic]

https://github.com/jeremysalwen/So-synth-LV2
http://distrho.sourceforge.net/
http://ll-plugins.nongnu.org/instruments.html
http://gna.org/projects/zyn/
https://github.com/pgiblox/lmms-lv2

But, to be honest yes, there are more usable DSSI synths available at the moment (whysynth, xsynth, hexter, amsynth, calf, zynaddsubfx-dssi, fluidsynth-dssi etc.)

Best,

Jeremy

[jeffh]

I agree with Autostatic. LV2 has progressed a lot in the last couple of months and there are some nice plugins atm, also synths:

...

All though LV2 has good documentation and more and more developers are taking it up, DSSI seems to be more simple to code, but has also less possibilities.

A plugin API only has to be able to accept MIDI, audio, and automation data in, allow code written by the plugin developer to accept and process the data, and then accept audio data back out. DSSI does all of the above, and has for a long time, and is very stable and mature. I don't need any features of LV2, because I can just program the features myself the way I want them to be.

I downloaded the LV2 API, and even talked to it's creator Dave before deciding on DSSI. If I had to guess, the reason that there aren't more LV2 plugins is because others like me had the same experience, and chose not to use it for the same reasons I did.

At the end of the day, if I put out a collection of the greatest synthesizers, samplers and effects that Linux has ever seen, users aren't going to care if it's DSSI or LV2 under the hood. Conversely, if I put out some of the worst plugins Linux has ever seen, being LV2 won't give anybody a reason to use them.

[slowpick]

Hexter is a nice dssi synth, this modular synth
might be interesting if more widely available in dssi form

http://drobilla.net/software/ingen/

used to be OM

Ingen is not a modular synth. It's a plug-in host.

Best,

Jeremy

This pic says its a patch. A modular synth jargon. The old version site label is
http://savannah.nongnu.org/cvs/?group=om-synth. The elements seem to be more fundamental than
exclusively a plugin host. I would consider lv2-rack, jackrack, calfjackhost to be plugin hosts, and
not modular synths, but I doubt they can load or even access many of the low level components shown. I am not a modular synth expert, but naming conventions would point to ingen being one. If it also can host lv2 plugins, so much the better
http://drobilla.net/blog/wp-content/upl ... g_vert.png

[slowpick]

I downloaded the LV2 API, and even talked to it's creator Dave before deciding on DSSI. If I had to guess, the reason that there aren't more LV2 plugins is because others like me had the same experience, and chose not to use it for the same reasons I did.

Good to see someone interested in creating new instruments, and researching the most productive
path forward. I think the lack of LV2, or any other format, is due to the education needed to succeed,
the desire to create such a product, and the money needed to do the work. It is much much easier to
have any 2 out of the 3, than to have the necessary trifecta.
Good luck in your projects!