Trying to merge back changes from SuperDough fork #303

telephon · 2024-09-05T09:26:20Z

There are lots of changes, so it is a little hard to tell what would impact what. But given this merge request, we can see the diff better.

geikha · 2024-09-07T19:26:15Z

let's also consider that we could extend further the first handshake between Tidal/Strudel to SuperDirt, so that SuperDirt knows which one is connected and maybe interpret differently the messages (majorly on breaking changes)

telephon · 2024-09-10T10:58:49Z

library/strudel-synths.scd

@@ -250,20 +250,22 @@
    //emulation of juno 60 chorus
    SynthDef("strudel_chorus" ++ numChannels, { |out, mix, depth, speed|
 		var input = In.ar(out, numChannels);
-        var l;
-        var r;
+        var l = input[0];


input may have any number of channels, and if someone chooses to run it mono, it will not be an array. You could use the technique we use in dirt_leslie.

did not know that it was not always an array, Ill fix that!

daslyfe · 2024-09-13T18:23:17Z

let's also consider that we could extend further the first handshake between Tidal/Strudel to SuperDirt, so that SuperDirt knows which one is connected and maybe interpret differently the messages (majorly on breaking changes)

good callout, hopefully overtime tidal and strudel can be brought closer together in terms of things like note scaling but this is possibly a good short term solution

telephon · 2024-09-18T13:49:02Z

let's also consider that we could extend further the first handshake between Tidal/Strudel to SuperDirt, so that SuperDirt knows which one is connected and maybe interpret differently the messages (majorly on breaking changes)

What differences would need to be handled? Wouldn't it be much easier to just keep things the same? We are a really small community!

telephon · 2024-09-18T13:51:18Z

classes/DirtEvent.sc

@@ -67,6 +83,7 @@ DirtEvent {
 		var accelerate = ~accelerate.value;
 		var avgSpeed, endSpeed;
 		var useUnit;
+		~release = ~release.value ? 0.01;


default values are usually given in the default event. Unless you need to caclulate something with ~release here, also the .value will be called automatically.

telephon · 2024-09-18T13:52:07Z

classes/DirtEvent.sc

@@ -102,24 +119,38 @@ DirtEvent {
 		};

 		sustain = ~sustain.value;
+
 		sustain = sustain ?? {


The changes below: are you sure strudel couldn't translate its logic to the way tidalcycles works? It will be very confusing to change all that …

telephon · 2024-09-18T13:52:36Z

classes/DirtEvent.sc

@@ -129,6 +160,7 @@ DirtEvent {
 		~channel !? { ~pan = ~pan.value + (~channel.value / ~numChannels) };
 		~pan = ~pan * 2 - 1; // convert unipolar (0..1) range into bipolar one (-1...1)
 		~delayAmp = ~delay ? 0.0; // for clarity
+		~z1 = ~z1 ? ~n;


can you explain what ~z1 is used for?

telephon · 2024-09-18T13:53:25Z

classes/DirtEventTypes.sc

@@ -108,15 +109,15 @@ DirtEventTypes {

 				if (hasNote) {
 					freq = ~freq.value;
-					~midinote = (freq.cpsmidi).round(1).asInteger;
+					~midinote = ((freq.cpsmidi).round(1) ? 0).asInteger;


the default is not necessary and can be dropped. freq.cpsmidi).round(1) will never be nil.

telephon · 2024-09-18T13:54:08Z

classes/DirtEventTypes.sc

 					// Assume aftertouch means no noteOn, for now..
 					if(~polyTouch.notNil) {
 						val = ~polyTouch;
 						note = ~midinote;
 						schedmidi.value({ midiout.polyTouch(chan, note, val) })
 					} {
 						// match dirt_gate SynthDef amplitude scaling
-						~amp =  ~amp.value * pow(~gain.min(2) + ~overgain, 4);
+						~amp =  ~amp.value * StrudelUtils.gainCurve(~gain + ~overgain, 4);


Let's introduce gainCurve to SuperDirt. Should it be global or orbit-wise?

telephon · 2024-09-18T13:54:58Z

classes/DirtOrbit.sc

-			~n = \none; // sample number or note
-			~octave = 5;
-			~midinote = #{ ~note ? ~n + (~octave * 12) };
+			~n = nil;


we should find a way around setting this to nil. Why can't it be \none?

Basically, you can just directly replace the default if you need, no need for StrudelUtils:

~dirt.set(\n, 42); // whatever you want

telephon · 2024-09-18T13:59:32Z

classes/DirtOrbit.sc

 			~fadeTime = 0.001;
+			~delaytime = 0.1875;


I am not 100% sure, but I think this means that there is always delay on. This can be set for each piece separately.

telephon · 2024-09-18T14:00:44Z

classes/DirtSoundLibrary.sc

@@ -179,7 +203,9 @@ DirtSoundLibrary {
 		};

 		files = pathMatch(folderPath.standardizePath +/+ "*"); // dependent on operating system
-		if(sortFiles) { files.sort };
+		if(sortFiles) { 
+			 files.sort;	


in sclang, there is no ; inside the last statement of a function. It is not wrong, but the change is not needed.

telephon · 2024-09-18T14:03:47Z

classes/DirtSoundLibrary.sc

@@ -298,8 +327,7 @@ DirtSoundLibrary {
 			^event

 		} {
-			// the index may be \none (a Symbol), but this converts it to 0
-			event = allEvents.wrapAt(index.asInteger);
+			event = allEvents.wrapAt((index ? 0).asInteger);


This can be revisited in SuperDirt.

telephon · 2024-09-18T14:05:43Z

classes/DirtSoundLibrary.sc

@@ -320,8 +348,9 @@ DirtSoundLibrary {
 	}

 	makeEventForBuffer { |buffer, metaData|
-		var baseFreq = 60.midicps;
+		var baseFreq = StrudelUtils.baseNote().midicps;


I intentionally kept separate DirtSoundLibrary and SuperDirt. This StrudelUtils.baseNote(). couples both to one system. MAybe we find a way around this?

telephon · 2024-09-18T14:07:06Z

classes/SuperDirt.sc

@@ -53,10 +54,14 @@ SuperDirt {

 	init {
 		soundLibrary = DirtSoundLibrary(server, numChannels);
+		outputvolume = server.volume;
+		outputvolume.setVolumeRange(-90, 0);


We want to be able to use SuperDirt with other libraries – we shouldn't mess with the server volume, I think.

I am sure there is another way to set volume? Normally we are in the range [min = -90, max = 6] so here you have found a reason why this should be less?

telephon · 2024-09-18T14:09:26Z

hacks/jitlib-hacks.scd

@@ -74,7 +75,7 @@ Pdef(\x,
 		\s, Pn(Pshuf([\bd, \hh, \cp, \imp], 5)),
 		\begin, Pn(Pseries(0, 0.002, 20)),
 		\dur, 0.1,
-		\legato, Pwhite(0.01, 0.4)
+		\clip, Pwhite(0.01, 0.4)


What is clip doing?

telephon · 2024-09-18T14:10:58Z

library/default-synths-extra.scd

@@ -220,6 +225,7 @@
 		sound = MoogFF.ar(sound, SinOsc.ar(basefreq/32*rate, 0).range(lfof1,lfof2), resonance*4);
 		sound = MoogFF.ar(sound, min(env2*lfof2*1.1, 22000), 3);
 		sound = sound.tanh*5;
+		sound = sound * StrudelUtils.synthGain;


No need to hardcode the volume adjustments, you can just do that by setting the ~amp parameter in the input. So all the below can easily be avoided.

telephon · 2024-09-18T14:12:14Z

synths/core-modules.scd

-				hresonance: ~hresonance,
-				out: ~out
-		])
+// ~dirt.addModule('hpf',


why do you remove hpf and lpf – are there problems with them?

telephon · 2024-09-18T14:12:51Z

synths/core-synths-global.scd

-	if(\SwitchDelay.asClass.notNil) {
-
-		SynthDef("dirt_delay" ++ numChannels, { |dryBus, effectBus, gate = 1, delaytime, delayfeedback, delaySend = 1, delayAmp = 1, lock = 0, cps = 1|
+	SynthDef("dirt_shimmer" ++ numChannels, { |dryBus, effectBus, gate = 1, delaytime, delayfeedback, shimmer = 0, lock = 1, cps = 1|


let's move shimmer to SuperDirt ...

telephon · 2024-09-18T14:14:37Z

synths/core-synths.scd

 		var signal = In.ar(out, numChannels);
-		var env = EnvGen.ar(Env([0, 1, 1, 0], [fadeInTime, sustain, fadeTime], \sin));
+		var env = EnvGen.ar(Env([0, 1, 1, 0], [fadeInTime, totalDuration - (fadeInTime + fadeTime), fadeTime], \sin));


yes, in supercollider generally, sustain means totalDuration – this is a bit unusual, but the standard …

telephon · 2024-09-18T14:18:24Z

@daslyfe I've left a lot of comments for you so it hopefully makes it easier to separate those things that are really needed and those that just happened due to a lack of documentation in superdirt.

In the meantime, since superdirt is live hackable, many of the changes could just be moved to a file that is evaluated at startup. Like all the SynthDefs etc.

Maybe you could check and see if you can separate things a little? That would be immensly helpful.

daslyfe · 2024-09-19T11:28:01Z

@daslyfe I've left a lot of comments for you so it hopefully makes it easier to separate those things that are really needed and those that just happened due to a lack of documentation in superdirt.

In the meantime, since superdirt is live hackable, many of the changes could just be moved to a file that is evaluated at startup. Like all the SynthDefs etc.

Maybe you could check and see if you can separate things a little? That would be immensly helpful.

Thanks for taking the time to comb through all of this! It helps to know how things can be better separated. I am still in a “hacking stuff together” phase of making superdirt strudel compatible. I intend to break changes into smaller much more manageable PRs

telephon · 2024-09-19T13:22:43Z

that is good! You could take my comments as things to mind to be able to do it with less effort / changes.

telephon · 2024-10-18T12:27:15Z

@daslyfe I just say that you changed the monitor and added wet and dry.

SuperDirt/classes/DirtOrbit.sc

Line 66 in ad53027

    
           GlobalDirtEffect(\dirt_monitor, [\limitertype, \outgain, \dry, \wet]).alwaysRun_(true),

and

SuperDirt/synths/core-synths-global.scd

Line 26 in ad53027

    
           SynthDef("dirt_monitor" ++ numChannels, { |dryBus, effectBus, outBus, gate = 1, limitertype = 1, outgain = 1, dry = 1, wet = 1|

Is there a reason why the local wet / dry controls of each effect are not sufficient?

daslyfe added 22 commits August 18, 2024 17:12

fixed amp

b0dce42

clip

fc4c24f

adjusted envelope behavior

e58583e

lp envelope

789eace

feat: filter envelope

42b282f

working on supersaw

d0c7fcb

simplifying supersaw

7b768d0

supersaw finally

c0e7366

fixed scale ration of samples

9b85220

need fix...

04bd40a

flix clip duration

88bfe27

add distortion

e6350ab

update doc

e60f514

fix stupid null handling bug with supersaw :)

7a0e601

fix sound order

cbbe8ba

fix release sample repeat bug and sample clipping behavior

cda96c6

add remaining filter envelopes

fd6c624

kinda working

36be99b

phaser and compressor

368613d

fix phaser

f32c1b9

phase_depth_min

561903a

add noise

7072d23

telephon mentioned this pull request Sep 5, 2024

Merging StrudelDirt ? #302

Open

daslyfe added 2 commits September 5, 2024 23:58

add pulse oscillator

2859c80

superpulse

287947f

daslyfe added 4 commits September 8, 2024 00:32

fix: gain staging

266b616

add non sc3 feedback delay with filtering

1becbcc

feat: shimmer delay

a47e9ff

feat: juno chorus

c96ef2a

fix chorus stereo imaging

bcf605e

telephon commented Sep 10, 2024

View reviewed changes

daslyfe added 2 commits September 11, 2024 00:13

fix reverb and delay send behavior

b114473

fix volume normalization

48d2bad

first synthkick

019c1ba

telephon commented Sep 18, 2024

View reviewed changes

daslyfe added 2 commits October 12, 2024 23:42

feat: limiter gain, wet gain, dry gain

ad53027

updated readme

e9ab9d9

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Trying to merge back changes from SuperDough fork #303

Trying to merge back changes from SuperDough fork #303

telephon commented Sep 5, 2024 •

edited

Loading

geikha commented Sep 7, 2024

telephon Sep 10, 2024

daslyfe Sep 13, 2024

daslyfe commented Sep 13, 2024

telephon commented Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon Sep 18, 2024

telephon commented Sep 18, 2024

daslyfe commented Sep 19, 2024

telephon commented Sep 19, 2024

telephon commented Oct 18, 2024

Trying to merge back changes from SuperDough fork #303

Are you sure you want to change the base?

Trying to merge back changes from SuperDough fork #303

Conversation

telephon commented Sep 5, 2024 • edited Loading

geikha commented Sep 7, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

daslyfe commented Sep 13, 2024

telephon commented Sep 18, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

telephon commented Sep 18, 2024

daslyfe commented Sep 19, 2024

telephon commented Sep 19, 2024

telephon commented Oct 18, 2024

telephon commented Sep 5, 2024 •

edited

Loading