Anyone On Linux Using Ams

General Discussion
1

Anyone else here using Linux? I’ve just had to reinstall Ubuntu 10.04 to get qjackctl and Alsa Modular Synth running again.
Anyone else think AMS is the BOMB… I know I do.
Also what’s the best way to render AMS if I’m routing it back in on midi return Line in? Anyone tried this?

2

I use Linux but I’ve never tried AMS. I use Zynaddsubfx all the time and it’s my favorite softsynth (EVAR, and I used quite a lot of commercial VSTs when I ran Windows). Of course it doesn’t come with gazillions of presets made by an army of sound designers, but if you like fiddling, you can do a LOT with it. It sounds great and the oscillator editor is just awesome. It comes with FM, PM, RM (all with custom oscils built with the editor) and some effects also. The latest 2.4.1 version supports unison mode for the voices. There are some odd limitations, though:

  • It has only one stereo output which means that you need to run multiple instances if you wanna route your different voices through different Jack clients. This sucks big time since I regularly have to work around it.

  • There are only 3 (or four?) effects per instrument. If you want to use more, you need to use external inserts (which means they are not a part of the instrument itself, so you can’t save them along with the instrument patch, only with the whole zyn patch which includes all instruments, inserts, sends and mixer params). But most of the time I don’t hit that limitation.

  • It doesn’t work properly if running Jack in realtime mode.

  • [edit]You can’t change parameters on the fly, they only take effect when the next note is played. Doesn’t matter for me because I can get around it quite comfortably by using evelopes (you can draw them by hand) and velocity-sensing[/edit]

  • [edit]oh and yeah, the UI is fugly and if you really use the synth (and multiple instances) you’ll end up with a sh*tload of open windows[/edit]

Don’t let that scare you, though. If you’ve never used it, fire up your local package manager, give it a go and get lost in it.

I think I’ll try AMS sometime. I already heard at several places that it’s supposed to “be da BOMB”

To answer your last question: AFAIK there’s currently no way to incorporate InDevices into Renoise’s rendering process. So you need to record your stuff with Audacity/Ardour/. This requires quite a bit of extra number crunching, though. I’m not able to record my stuff of my laptop because it’s a crappy old 1Ghz machine with 512 MB RAM and an Intel soundcard. XRUN hell!

You could also try the sample recording feature of Renoise (if you wanna bounce a whole track, which is not always an option since you lose some flexibility)

3

I’m gonna give it a go right now. I like the idea of custom oscilators for everything very much indeed. AMS is indeed “da bomb”, sounds more authentic than any soft synth I’ve ever heard and it’s just had a nice GUI update (Mine’s purple, with red throbbing wires on it!). You can also incorporate Ladspa plugins in the synth wherever you like… Great!

4

I just tried it and I really like it’s flexible FM capabilites!! Zyn only supports pairs of modulators/carriers. I think I’ll use it in my next DnB project to build a nasty bassline.

Oddly, “sometimes” it segfaults when adding LADSPA plugins. For example, when I open the example FM patch and add a GLAME Butterworth filter -> segfault. Do you have this problem, too? (btw, I’m running an all-32bit Ubuntu Karmic)

5

kickofighto wrote wrt to zynaddsubfx:

“- It doesn’t work properly if running Jack in realtime mode.”

if you run up against this you might want to try Yoshimi a proof-of-concept version of zyn designed to overcome these issues.
Ubuntu packages are available here Autostatic’s PPA if you don’t want to build from source.

some tutorial vids by the Zyn developer: zyn on youtube

best wishes, G.

6

It also supports on-the-fly parameter changes. But, AFAIK, the Yoshimi developer(s?) (is that you?) has not implemented the unison feature yet, which is kind of a show stopper for me. I think it also sounds somewhat different, can’t tell why or how though. Last time I checked it only supported Jack MIDI but Renoise uses plain old ALSA midi. You can use a2jmidi and the MIDI-Thru ports to route the midi to/from ALSA but that requires more wiring.

Thanks for posting the YouTube link, I totally forgot about these.

7

Hi, Noisebeuter. No, I am not the Yoshimi dev, I wish I was that clever!
You can choose ALSA or jack midi as the default option at compile time, but also set it on the command line, see: yoshimi --help

Some time ago, there was some discussion on the linux.audio.users mailing list about changes in the sound, but IRCC these were addressed to everyone’s satisfaction. The dev, Cal, can be contacted on that list: he is very receptive to user feedback.

I am not familiar with the unison feature, so I cannot comment on that.

BTW I discovered today that yoshimi is also available in the debian/testing repository.

best wishes, G.

8

I’ve been using AMS alot in the past, and just decided that I’m gonna get back into it. So I’d be interrested in how you go about integrating it into a good renoise workflow. There’s no way to have the AMS (or any other external noise) be part of a “render song to disk”, right?

Also I’m gonna try and hook PD up and see where that takes me.

Not the next week, though, since I’m on vacation now…

9

No, InDevices do not get rendered. Your best bet is recording the “noise” to a sample or use a Jack-enabled recording application like Ardour or Audacity.

It’s easy to do that with Jack! I use it as an effects processor sometimes and record that stuff with the sample editor. Be sure to use pd extended, though. It has far more patches built-in than the standard distribution. For example a feedback delay network which is good for creating the sound of a tsunami of snares that crashes into a wall of drumsticks!!

PS it’s called “pd” ;)

10

That’s good to know. I think I’m gonna give it another try.

I’ll check that list out.

It basically copies the same voice up to 32(? can’t remember the maximum) times and detunes every one of them by an amount you can specify. It’s good to get a phasing or Supersaw-like effect. Of course you could do this by hand but it’s tedious.