The JS Modular Synth Thread

[schwa]

coolness!

screenshot

[tspring]

Originally Posted by ph69

I think it is nice to use the channels freely.

Agreed. Still, I think that it is important to avoid establishing approaches that are likely to make the synth chains difficult to use with rest of Reapers procedures and plugins.

Probably this is just temporary until you get further along, so I apologize if I am jumping ahead on a trivial issue. Anyway I noticed that the MoonSynthi_Controll was hard-coded to use channels 1 to 10. For example:

@sample
i=0;
loop(10,spl(i)=0;i+=1;

I don't think that swizzle could prevent use of these channels from zonking other audio on those channels. So I do think that we need to follow a convention of using only higher channels for controller modules, or at a minimum the controllers ought to be coded to allow use of the higher channels. Obviously, if the controllers use the higher channels, for simplicity, all of the other modules ought to be able to accept the higher channels as input.

I am really enjoying this, and learning a lot. Thanks for sharing your work.

T

EDIT: While I was typing this, Schwa posted. GOOD GRIEF THIS IS AWESOME. A SYNTH IS HAPPENING RIGHT BEFORE MY VERY EYES!

[ph69]

Originally Posted by tspring

Anyway I noticed that the MoonSynthi_Controll was hard-coded to use channels 1 to 10. For example:

@sample
i=0;
loop(10,spl(i)=0;i+=1;

Ooops, I added this code when I had troubles with spl()... I will remove them, they are useless.

Originally Posted by tspring

all of the other modules ought to be able to accept the higher channels as input.

I think it can be the best way to do it. But it will cost us sliders.
So, @gfx and @serialize may be usefull to add routing controls

[Alex Stone]

Ph, clever stuff, and i like the modular approach.

Thanks for this!

Alex.

[Consul]

I too was thinking that the arbitrary split between audio and control signals was pointless. I'll do away with that in the next standards draft, and make the other changes we've discussed.

[lowkey]

Love what your doing here guy's

Why dont you ask Benno to add JS import on LinuxSampler? They are making it modular. You'd get an optimised engine, they would get a leg up with their modules and users could contrbute easier .
What do you think?

[Consul]

Okay, here's the second draft, as far as I am able to get. How is frequency information getting passed around? Is it in radians per sample?

https://stash.reaper.fm/oldsb/72624/J...0Draft%202.pdf

EDIT: I just realized I made a big mistake with the spl(var) example. That's fixed in this latest version.

[Consul]

Okay, let me see if I understand this "radians per sample" idea.

A complete wave cycle is 2*pi radians...

So, to convert radians/sample into a frequency value, I would:

freq = srate/(radsPerSample*2*$pi);

Is that the basic idea?

[ph69]

Originally Posted by Consul

Okay, let me see if I understand this "radians per sample" idea.

A complete wave cycle is 2*pi radians...

So, to convert radians/sample into a frequency value, I would:

freq = srate/(radsPerSample*2*$pi);

Is that the basic idea?

freq (cycles per second) = srate * radPersample / (2*$pi);

[ph69]

I corrected a bunch of bugs (combo box indexes, ADSR retrigg...)
The modules are not exactly compatible with the former.

[Shan]

Killer thread!

How about a Multi Segment Envelope Generator. Including multi segments after the release. ADSR style is very limited from the geek viewpoint. If I could code, I'd contribute to this great community.

Oh, and I have 25 years of Synth programming behind me...and on every type of synthesis you could think of.(no joke) Time sure does fly.

Shane

[ph69]

We are still at the very beginning. So I implemented the easiest to implement modules and synthesis form (phase modulation).

Maybe, we may list "Ideas", "Wishes", "Todo"s and "ToTry"s
And then work together.

Cut here and add your own stuff
| |
V V
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Ideas:

Wishes:

Todo:
- Bandlimited saw generator (I m working on it)
- Filters (I will code a simple Chamberlin filter first)
- LPM operator (includes sine gen, ADSR and DCA in one module)
- Multisegment control function generator (a la eMu)
- Extend gfxscope and C° to listen to channels 2 and above

ToTry:
- Oversampled versions of existing modules.
- "Poly to mono MIDI dispatcher" : a midi controller that distributes mono midi controls to multiple mono channels. Polyphony in the way of the old Oberheim SEM synthis.
- Use of @gfx for channel selections.

[lowkey]

Originally Posted by lowkey

Love what your doing here guy's

Why dont you ask Benno to add JS import on LinuxSampler? They are making it modular. You'd get an optimised engine, they would get a leg up with their modules and users could contrbute easier .
What do you think?

I take it you dont think its a good idea lol

Keep up the good work anyhow.

[tspring]

Originally Posted by lowkey

I take it you dont think its a good idea lol

Keep up the good work anyhow.

Not a bad idea, just a little early to see how things would fit together... Baby steps now.

T

[tspring]

Hmmm...

Loading the project moonsynthi_LMP.RPP crashes Reaper here. Also loaded the JS modules into a new project, but made the mistake of hitting the play button before I had all of the connections sorted. It appears that JS doesn't like it much when you have unconnected channels. Out of time now. Will try again tomorrow.

T

[ph69]

Originally Posted by tspring

Hmmm...

Loading the project moonsynthi_LMP.RPP crashes Reaper here. Also loaded the JS modules into a new project, but made the mistake of hitting the play button before I had all of the connections sorted. It appears that JS doesn't like it much when you have unconnected channels. Out of time now. Will try again tomorrow.

T

Did you downloaded Justin's patch ? (see bug forum).

Here is another release, it is still alpha (versions may not be compatible with each others). Less bugs (hope so), and new modules : fully antialiased saw generator and x2 phase generator, PM sine and 9 tap decimator.
I extended the number of channels to 12 (it is necessary for x2 modules).

[ph69]

Originally Posted by Consul

Okay, here's the second draft, as far as I am able to get.

Hi,

It is very cool.
I enriched the gate signal so that it includes velocity information. It seems to be efficient and practical. One must be carefull with boolean signals : it seems that reaper adds an epsilon to avoid float denormals... See code for the ADSR

No filter yet, I will work on it this week end.

Is there anybody to work on the advanced function generator (extended enveloppe generator) ?

[schwa]

I'm quite interested in working on this, though I'm quite busy just now finishing a VST synth

Things I could knock out quickly would be
- bandlimited saw generator using JS fft and wavetable
- very, very CPU expensive moog ladder filter

Things that would be nice but not quick to make would be
- graphical ADSR envelope input
- other stuff

I also have an idea for how to do bandlimited arbitrary user-drawn waveforms, but I'm not sure how easily that will work in JS.

[tspring]

Originally Posted by ph69

Did you downloaded Justin's patch ? (see bug forum).

Yep. I downloaded and installed the patch (replaced the old jsfx.dll) before I did anything else. Justin posted two builds of of the .dll maybe I picked downloaded the preliminary version by accident. I'll check it out tonight.

Nice job on the docs, Consul.

ph69, I'm impressed with how quickly you are turning out modules. What kind of programming background do you have?

T

[ph69]

Originally Posted by tspring

What kind of programming background do you have?

Sinclair ZX81 (Timex 1000 in the US)

[tspring]

Hah! I remember it. Brainchild of Clive Sinclair... came out about the same time as the VIC 21 if I remember right. If you ever programmed on one of those you are no youngster. OK, you remain a man of mystery

T

[ph69]

born 1969.
JS reminds me Sinclair Basic. It is very limited, no functions, global variables. So when you have something to implement you know how you will do it.
After playing with the ZX81, I went to college and university, up to the phd degree. But ZX81 is definitely my programming background
Today, I work with the latest C# and other stuff. Terribly complicated and opaque, mystified, fullfilled with gadgets software.
So I love JS because it is limited to what is necessary for programming plug ins.

What is your background ?

[Consul]

I'll see if I can get the lowdown on what's happening so I can get it documented.

I started programming in BASIC on the Commodore 64, before moving to Assembler on the same platform. These days, I'm mostly Perl and PHP, but I'm taking a C++ class this semester.

[ph69]

The C64 was very nice. By the time of C64, I was tweaking my ZX Spectrum.


I will be alone this weekend. No ladies to distract/disturb me So I will focus on modules But on monday, I will get distracted/disturbed, so you will have time for documentation.

I coded all these modules so that we can evaluate how we can use Reaper's multiple channels.
So, comments, wishes and ideas are welcome.

Soon, we will need a multichannel spy/analysis tool so that the design of a synth from modules would not be a pain in the... fx chain.

So this WE, I will work on simple filters, LPM operator, use of gfx, additive synthesis and maybe this multichannel stuff.

Is there anybody that would work on an advanced enveloppe generator, LFO ?

[tspring]

Originally Posted by ph69

born 1969.

What is your background ?

Heh. I am a PhD ecotoxicologist, but I do a lot of statistical consulting too. I'm really a programming lightweight. Originally learned Basic on Apple II, and Fortran and SAS on mainframe. Programmed in Pascal quite a bit. Now I mostly program in SAS and VBA.

T

[argee]

Originally Posted by schwa

...Things that would be nice but not quick to make would be
- graphical ADSR envelope input...

I might work on this in the coming weekend.

cheers

[ph69]

Originally Posted by argee

I might work on this in the coming weekend.

cheers

That would be a great module.
I have coded a noise generator, a geiger noise generator and a static filter (Chamberlin). I will optimize the filter so that it can be modulated.

[ph69]

Originally Posted by tspring

Heh. I am a PhD ecotoxicologist,

Dave Rossum (Emu) has been trained as a biologist

[tspring]

Originally Posted by ph69

Dave Rossum (Emu) has been trained as a biologist

Well, that just prove that it is not absolutely necessary to be a dummy to be a biologist or ecotoxicologist

R

[tspring]

Originally Posted by tspring

Yep. I downloaded and installed the patch (replaced the old jsfx.dll) before I did anything else. Justin posted two builds of of the .dll maybe I picked downloaded the preliminary version by accident. I'll check it out tonight.

Nice job on the docs, Consul.

ph69, I'm impressed with how quickly you are turning out modules. What kind of programming background do you have?

T

Finally got a few minutes. Downloaded Justins second patch. Now the synth works! I dropped a couple of existing JS effects into the effects chain chain. Distortion and delay can do some interesting things. Give this a listen

https://stash.reaper.fm/oldsb/73776/tmoonsynth.RPP

Great fun.

T

[ph69]

I will try it once back home.

Here is a new release.
Noise generator, Geiger generator (with controllable density), 2pole multimode static filter and 2 pole multimode dynamic filter (code based on MusicDSP...).
I changed the saw generator so that we can change its level and choose between replacing and adding its output to a channel, this way, we can avoid the use of adder/mixer modules.

There's a lot of stuff to do and to try, so the interfaces will certainly evolve... still alpha

[ph69]

Originally Posted by schwa

I'm quite interested in working on this, though I'm quite busy just now finishing a VST synth

Things I could knock out quickly would be
- bandlimited saw generator using JS fft and wavetable

Once you have that one, you can do square and PWM.
Here is the recipe :

Take a sawtooth.
Take another sawtooth, from the same table but with a phase offset.
Substract the second to the first.
And you obtain a delicious square and the phase offset is your pulse width. Modulate it to match your taste.

[ph69]

Hi,
I am still working on the modules.
I will remove frequency control in rad per sample.
I think that using a pitch control in "float" midi units is more versatile, and the design of a look up table to translate it into whatever unit is needed quite easy.
I will post another release when it will be debugged.

[tspring]

Originally Posted by ph69

.
And you obtain a delicious square ...

Haha! Spoken as only a true engineer or bit twiddler would.

My shallow understanding of synths in general has made it a bit difficult to understand how this particular synth is supposed to work. I did notice that in one place you noted that phase modulation synthesis was the method of sound generation. The description of phase modulation synthesis in the following excerpt made things a bit clearer to me. I thought that it might be useful to other folks who would like to experiment with the modular synth.

-------------------------
Phase modulation synthesis—frequency modulation (FM), phase modulation and phase distortion

Phase modulation is a synthesis technique with a long history. The first forms of phase modulation can be found in, where else, radio technology. There it is commonly employed in FM radio. Phase modulation was also used with analog synthesizers, but the limited accuracy of analog oscillators and the difficulty of building oscillators with negative frequency support hindered the analog implementations. The true break‐through of FM technology came with John Chowning and the subsequent patenting of the method for sound synthesis by Yamaha. The result was DX7, probably the most successful single synthesizer in existence. More recent derivatives include OPL2‐4 synth chips (ADLIB etc.) and Yamaha’s more mature version of the DX7 synthesis principles, the SY series.

The idea behind FM synthesis is that quite rich and deliciously time‐variable timbres can be created by modulating the frequency of a carrier sine oscillator by another, the modulator. When the modulator frequency stays below 20Hz or so, only more or less rapid vibrato results. But when the modulation frequency rises to the audio band, the characteristic sidebands resulting from the modulation process can be heard. The sidebands are (generally) not harmonic, except in special cases. These come about when the frequencies of the carrier and the modulator form a simple ratio. The method has few variable parameters, these including the volumes of the two oscillators (the modulator volume affects timbre, not volume) and modulator frequencies. The basic configuration is two oscillators cascaded, as described above, plus envelope generators to control the amplitudes. Often more oscillators are used as well, since interesting (and complex) inharmonic spectra are thus easily produced, allowing for quite realistic bell and brass sounds to be generated. The most characteristic FM sound is a slow sweep of the modulator volume, while keeping the carrier‐modulator frequency ratio constant. This produces the well‐known ADLIB timbre. Common modifications include several two‐oscillator complexes in parallel (allowing for a form of additive synthesis), multiple oscillators in series (allowing for extremely inharmonic and noise‐like spectrum formation), non‐sine components (they produce a richer sound and, when added in parallel, result in modified group additive synthesis that complements the capabilities of the base FM system), layering, feedback (for noise and weird sounds and adding long term development to the sound), addition of filters (since FM can produce most of the basis waveforms for subtractive synthesis, this also complements the capabilities of the synthesizer) and several combinations of the preceding. Some specific modifications include a limited form of FM, called formant FM, which is capable of producing voice like timbres and formant peaks and has an associated analysis procedure which makes instrument design considerably easier, and a couple of other academic projects, with no presence in the commercial music business. For example, Yamaha’s FS1R uses a combination of formant FM and filtering to create its timbres.

The implementation of FM synthesis is very easy, the only problem being aliasing which results from high modulator frequencies. Specifically, the computational cost of the algorithm is very low due to the high simplicity of the algorithm—nothing more than a couple of table lookups are needed to produce a sample by FM. The cost increases when more oscillators, options and enhancements are added...

(Excerpt from http://www.helsinki.fi/~ssyreeni/dsound/dsound)

T

[ph69]

Hi,
No release today... But I work on new modules, lots of debug and optimisation to do before I go forward.

[ph69]

Hello everyone.

Reaper is very versatile, that's quite impressive.

I still work on these modules. I hope I will release some stuff this week.
Maybe, I have to make a catalog and some tutorials so that more people can play with them.

Maybe somebody else will release some monster module before me

[ph69]

Here is a new release.
As far as I remember what's new:
Pan : a controlled panner.
LFO : a four output LFO (with different offsets).

MidiControl has been debugged (it stops playing when Reaper stops playing). But modulation, pitch bend and expression are not yet implemented.

ScopeII : two channels, better synch capabilities, subsample synch.

[tspring]

Can't wait to have some time to work with this. But I am engulfed by totally consuming work for the next few days. Don't interpret my lack of action as lack of interest.

T

[ph69]

See the JS repository for the latest version

https://stash.reaper.fm/711/MoonSynthi.zip

[lemmy]

I hope you will keep going with this ph69, it's a great project. The latest examples are sounding good. Keep on!