fixes (possibly?) for AudioControl mapping #3063

joshpar · 2017-07-29T00:05:52Z

telephon

looks good - I'll test later. Could you fix the formatting please? You may have to set the preferences in xcode (if you are using that) to use 4 spaces instead of tabs, and then reformat.

telephon · 2017-07-29T08:24:17Z

include/plugin_interface/SC_Graph.h

@@ -35,6 +35,7 @@ struct Graph
 	uint32 mNumControls;
 	float *mControls;
 	float **mMapControls;
+    uint32 mAudioBusOffset;


formatting? maybe convert tabs to space?

joshpar · 2017-07-29T16:13:01Z

So - I have that setting already there, and it appears that my change is using 4 spaces instead of tabs. When I reformat the file, all the other lines are changed (and when I use show invisibles in Xcode, it looks like everything else is tabbed). So - I don’t feel live the white space of stuff I didn’t touch is appropriate to this issue, but I can still update the rest of the file if you think we should. Let me know what you want. Also - did you run the patched version to make sure it performed as expected? Best, Josh

…

On Jul 29, 2017, at 1:26 AM, Julian Rohrhuber ***@***.***> wrote: @telephon commented on this pull request. looks good - I'll test later. Could you fix the formatting please? You may have to set the preferences in xcode (if you are using that) to use 4 spaces instead of tabs, and then reformat. In include/plugin_interface/SC_Graph.h <#3063 (comment)>: > @@ -35,6 +35,7 @@ struct Graph uint32 mNumControls; float *mControls; float **mMapControls; + uint32 mAudioBusOffset; formatting? maybe convert tabs to space? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#3063 (review)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADTrLqobdidMn2Fdi0RAHCX37VQU4Zhks5sSuyYgaJpZM4OnMQZ>.

telephon · 2017-07-30T14:10:40Z

This definitely fixes the bug in #2839. Thank you very much for this.

The formatting seems to be tabs, not spaces, probably it was this way round. I can fix that.

the rest of the file uses tabs.

code formatting

telephon · 2017-07-30T14:39:56Z

OK, done.

joshpar · 2017-07-30T15:19:00Z

Thanks! Glad the fix seems to work. If other cases come up, please let me know. Josh /* Josh Parmenter www.realizedsound.net/josh */

…

On Jul 30, 2017, at 07:39, Julian Rohrhuber ***@***.***> wrote: OK, done. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

telephon · 2017-07-30T18:12:58Z

@joshpar wrote (#2839):

I think this can be closed. But I'll leave some notes here. Unlike In and other bus UGens, AudioControl does NOT look right at the bus. It looks at a chunk of memory in the Graph that is allocated to store audio mapped data (similar to the way Controls work). As a result, when the Graph has Audio being mapped to it, a reference to the audio bus that is mapped needs to be stored - I added a place in the Graph to store the index of the mapped audio control. This is then checked against to see if data has been written to it (touched) or not. However, I'm not positive this will work well with multiple audio mapped controls. If someone can test for that in their usual usage, and if there is a problem create an new issue, I can dig deeper for other use cases.

telephon · 2017-07-30T18:43:32Z

@joshpar

rereading the example above, I've found that the new version doesn't output any glitches, but also doesn't output any audio.

Here is a slightly easier to use version of the first one by @jamshark70:

(
SynthDef(\fxbus, { |out, in|
	var sig = In.ar(in, 2);
	ReplaceOut.ar(out, HPF.ar(sig, 800));
}).add;

SynthDef(\fx, { |out, a_in = #[0, 0]|
	ReplaceOut.ar(out, HPF.ar(a_in, 800));
}).add;

SynthDef(\boop, { |out, freq = 440, gate = 1, amp = 0.1|
	var sig = Mix(VarSaw.ar(freq * [1, 1.004])) * amp,
	eg = EnvGen.ar(Env.adsr(0.01, 0.1, 0.5, 0.001, curve: 0), gate, doneAction: 2);
	sig = LPF.ar(sig, 3000);
	Out.ar(out, (sig * eg).dup(2));
}).add;
)

s.boot;
s.scope;

(
b = Bus.audio(s, 2);

p = Pbind(
	\instrument, \boop,
	\degree, Pwhite(-7, 7, inf),
	\dur, 1,
	\legato, 0.25,
	\out, b
).play;
)

// OK:
f = Synth.tail(s.defaultGroup, \fxbus, [in: b, out: 0]);  

// end this one.
f.free;

// try the audio mapped version: currently silent
(
s.makeBundle(nil, {
	f = Synth.tail(s.defaultGroup, \fx, [\out: 0]);
	//s.sendMsg("/n_mapan", f.nodeID, \a_in, b.index, 2);
	f.map(\a_in, b);
});
)

p.stop;
f.free;

for me, the last bit outputs silence now (it used to outputs glitches).

telephon · 2017-07-30T18:48:36Z

Also, this one crashes the server now :
(to be sure I've checked with a clean build)



(
SynthDef(\fx, { |out|
	ReplaceOut.ar(out, HPF.ar(\in.ar(0 ! 2), 800));
}).add;
)


(
b = Bus.audio(s, 2);

s.makeBundle(nil, {
	f = Synth.tail(s.defaultGroup, \fx, [\out: 0]);
	f.map(\in, b);
});
)

The OSC messages are identical:

With an "a_in" control name:
[ "/n_mapan", 1002, "a_in", 10, 2 ]

With an "in" control name:
[ "/n_mapan", 1011, "in", 10, 2 ]

joshpar · 2017-07-30T20:49:07Z

Thanks for the example. I’ll try to work through this, but do you think you or someone else can make up an example of the version without sound that doesn’t use Pbind? I can’t always tell what is going on there with order of execution (and one thing I saw while debugging the other day had to do with this). Thanks - I’ll look at the crashing issue as well. That example works well for me to work on this.

…

On Jul 30, 2017, at 11:48 AM, Julian Rohrhuber ***@***.***> wrote: Also, this one crashes the server: ( SynthDef(\fx, { |out| ReplaceOut.ar(out, HPF.ar(\in.ar(0 ! 2), 800)); }).add; ) ( b = Bus.audio(s, 2); s.makeBundle(nil, { f = Synth.tail(s.defaultGroup, \fx, [\out: 0]); f.map(\in, b); }); ) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#3063 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADTrCE_wSSpKz_fgZbZElalhbcY5Nx0ks5sTNAEgaJpZM4OnMQZ>.

telephon · 2017-07-30T21:15:23Z

Sure, here you are:

(
SynthDef(\fxbus, { |out, in|
	var sig = In.ar(in, 2);
	ReplaceOut.ar(out, HPF.ar(sig, 800));
}).add;

SynthDef(\fx, { |out, a_in = #[0, 0]|
	ReplaceOut.ar(out, HPF.ar(a_in, 800));
}).add;

)

s.boot;
s.scope;

(
b = Bus.audio(s, 2);

{ Out.ar(b, SinOsc.ar(330) * 0.1 ! 2) }.play;
)

f = Synth.tail(s.defaultGroup, \fxbus, [in: b, out: 0]);  // OK

f.free;

(
s.makeBundle(nil, {
	f = Synth.tail(s.defaultGroup, \fx, [\out: 0]);
	//s.sendMsg("/n_mapan", f.nodeID, \a_in, b.index, 2);
	f.map(\a_in, b);
});
)

p.stop;
f.free;

joshpar · 2017-07-31T01:30:09Z

SO - I have it working for your solution below, and the numbers make sense to me in the code. But - running James’s example (copied below) it doesn’t. I think the Order of Execution is messed up though. The node plot shows this: ``` NODE TREE Group 0 1 group 1070 temp__3 1071 group 1077 temp__0out-353201172_0 1076 group 1084 boop 1073 group 1074 system_link_audio_1 1075 system_link_audio_1 ``` Shouldn't node 1077 be after group 1076? I’m thinking this is why the example below now outputs silence with me new changes. Please take a look at the node tree and tell me if I’m mistaken. I’ll push a change to the remote branch after I look at the crash. ```supercollider ( SynthDef(\boop, { |out, freq = 440, gate = 1, amp = 0.1| var sig = Mix(VarSaw.ar(freq * [1, 1.004])) * amp, eg = EnvGen.ar(Env.adsr(0.01, 0.1, 0.5, 0.001, curve: 0), gate, doneAction: 2); sig = LPF.ar(sig, 3000); Out.ar(out, (sig * eg).dup); }).add; ) ( p = ProxySpace.new.push; ~out = { \in.ar(0!2) }; ~out.play(vol: 0.2); ~player = Pbind( \instrument, \boop, \degree, Pwhite(-7, 7, inf), \dur, 1, \sustain, 0.2 ); ~player <>> ~out; ) p.clear; p.pop; ```

…

On Jul 30, 2017, at 2:15 PM, Julian Rohrhuber ***@***.***> wrote: Sure, here you are: ( SynthDef(\fxbus, { |out, in| var sig = In.ar(in, 2); ReplaceOut.ar(out, HPF.ar(sig, 800)); }).add; SynthDef(\fx, { |out, a_in = #[0, 0]| ReplaceOut.ar(out, HPF.ar(a_in, 800)); }).add; ) s.boot; s.scope; ( b = Bus.audio(s, 2); { Out.ar(b, SinOsc.ar(330) * 0.1 ! 2) }.play; ) f = Synth.tail(s.defaultGroup, \fxbus, [in: b, out: 0]); // OK f.free; ( s.makeBundle(nil, { f = Synth.tail(s.defaultGroup, \fx, [\out: 0]); //s.sendMsg("/n_mapan", f.nodeID, \a_in, b.index, 2); f.map(\a_in, b); }); ) p.stop; f.free; — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#3063 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADTrCutwKeKZaic2VmRlUqdbdKwZWhhks5sTPJrgaJpZM4OnMQZ>.

…sets

…percollider into AudioControlFix # Conflicts: # include/plugin_interface/SC_Graph.h # server/plugins/IOUGens.cpp # server/scsynth/SC_Graph.cpp

joshpar · 2017-07-31T02:47:04Z

new fixes are in for checking. @telephon - was the sample that crashed something that you would expect to hear audio with?

telephon · 2017-07-31T05:55:29Z

@joshpar no, I didn't expect audio to come out of that example. It was just there to crash the server.

joshpar · 2017-07-31T13:08:18Z

I pushed some new things to that branch - and some other files had to be touched. Does the pull request automatically get updated or do I need to do a new one? Ready for checks I think. Josh /* Josh Parmenter www.realizedsound.net/josh */

…

On Jul 30, 2017, at 22:55, Julian Rohrhuber ***@***.***> wrote: @joshpar no, I didn't expect audio to come out of that example. It was just there to crash the server. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

mossheim · 2017-07-31T13:09:15Z

Does the pull request automatically get updated or do I need to do a new one?

If you're pushing to the remote branch that you used to make the PR, the PR gets updated automatically.

joshpar · 2017-07-31T13:14:05Z

Great - that’s what I thought. Should be good for review then! /* Josh Parmenter www.realizedsound.net/josh */

…

On Jul 31, 2017, at 06:09, Brian Heim ***@***.***> wrote: Does the pull request automatically get updated or do I need to do a new one? If you're pushing to the remote branch that you used to make the PR, the PR gets updated automatically. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

telephon · 2017-07-31T15:57:29Z

Excellent. This works for the following test, and doesn't crash any of the above.

(
SynthDef(\fxbus, { |out, in|
	var sig = In.ar(in, 2);
	ReplaceOut.ar(out, HPF.ar(sig, 800));
}).add;

SynthDef(\fx, { |out|
	ReplaceOut.ar(out, HPF.ar(\in.ar(0 ! 2), 800));
}).add;
SynthDef(\source, { |out|
	Out.ar(out, SinOsc.ar(330) * 0.1 ! 2 * Line.kr(1, 0, 0.8, doneAction:2))
}).add;
)

s.boot;
s.scope;

(
b = Bus.audio(s, 2);

fork { 
	loop { 
		Synth(\source, [\out, b]);
		1.5.wait 
	} 
};
)

f = Synth.tail(s.defaultGroup, \fxbus, [in: b, out: 0]);  // OK

f.free;

(
s.makeBundle(nil, {
	f = Synth.tail(s.defaultGroup, \fx, [\out: 0]);
	f.map(\in, b);
});
)

p.stop;
f.free;

telephon · 2017-07-31T16:08:18Z

OK, so one more thing. So far, the audio rate controls worked as analogies to InFeedback, so that it would grab the previous buffer content if necessary. This allowed us to ignore node order if there is only one writer on that bus.

This is the reason why the ProxySpace example still doesn't work:


( 
SynthDef(\boop, { |out, freq = 440, gate = 1, amp = 0.1| 
	var sig = Mix(VarSaw.ar(freq * [1, 1.004])) * amp, 
	eg = EnvGen.ar(Env.adsr(0.01, 0.1, 0.5, 0.001, curve: 0), gate, doneAction: 2); 
	sig = LPF.ar(sig, 3000); 
	Out.ar(out, (sig * eg).dup); 
}).add
) 

(
p = ProxySpace.new.push; 
~out = { \in.ar(0 ! 2) };
~out.play(vol: 0.2);
~player = Pbind( \instrument, \boop, \degree, Pwhite(-7, 7, inf), \dur, 1, \sustain, 0.8); 
~player <>> ~out; 

)

While this one works:

(
p = ProxySpace.new.push; 
~player = Pbind( \instrument, \boop, \degree, Pwhite(-7, 7, inf), \dur, 1, \sustain, 0.8); 
~player <>> ~out; 
~out = { \in.ar(0 ! 2) };
~out.play(vol: 0.2); 
)

I'll post a simpler test case.

telephon · 2017-07-31T16:10:29Z

Here is a test case:


(
SynthDef(\fxbus, { |out, in|
	var sig = InFeedback.ar(in, 2);
	ReplaceOut.ar(out, HPF.ar(sig, 800));
}).add;

SynthDef(\fx, { |out|
	ReplaceOut.ar(out, HPF.ar(\in.ar(0 ! 2), 800));
}).add;
SynthDef(\source, { |out|
	Out.ar(out, SinOsc.ar(330) * 0.1 ! 2 * Line.kr(1, 0, 0.8, doneAction:2))
}).add;
)

s.boot;
s.scope;

(
b = Bus.audio(s, 2);

fork { 
	loop { 
		Synth.tail(s.defaultGroup, \source, [\out, b]);
		1.5.wait 
	} 
};
)

f = Synth.head(s.defaultGroup, \fxbus, [in: b, out: 0]);  // OK

f.free;

// add to head: no sound.
(
s.makeBundle(nil, {
	f = Synth.head(s.defaultGroup, \fx, [\out: 0]);
	f.map(\in, b);
});
)

// add after: yes sound.
(
s.makeBundle(nil, {
	f = Synth.after(s.defaultGroup, \fx, [\out: 0]);
	f.map(\in, b);
});
)

p.stop;
f.free;

joshpar · 2017-07-31T17:07:49Z

well - that worked because of the bug that we’re squishing. To get that back, we could possibly check if the bus was touched in the current OR last cycle, but that is really dodgy behavior in my opinion. More than happy to try it out, but in the end, we will always end up with a glitch of at least one control period at the end where a bus is no longer touched, but it WAS touched on the last cycle and that data gets repeated. Josh

…

On Jul 31, 2017, at 9:10 AM, Julian Rohrhuber ***@***.***> wrote: Here is a test case: ( SynthDef(\fxbus, { |out, in| var sig = InFeedback.ar(in, 2); ReplaceOut.ar(out, HPF.ar(sig, 800)); }).add; SynthDef(\fx, { |out| ReplaceOut.ar(out, HPF.ar(\in.ar(0 ! 2), 800)); }).add; SynthDef(\source, { |out| Out.ar(out, SinOsc.ar(330) * 0.1 ! 2 * Line.kr(1, 0, 0.8, doneAction:2)) }).add; ) s.boot; s.scope; ( b = Bus.audio(s, 2); fork { loop { Synth.tail(s.defaultGroup, \source, [\out, b]); 1.5.wait } }; ) f = Synth.head(s.defaultGroup, \fxbus, [in: b, out: 0]); // OK f.free; // add to head: no sound. ( s.makeBundle(nil, { f = Synth.head(s.defaultGroup, \fx, [\out: 0]); f.map(\in, b); }); ) // add after: yes sound. ( s.makeBundle(nil, { f = Synth.after(s.defaultGroup, \fx, [\out: 0]); f.map(\in, b); }); ) p.stop; f.free; — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#3063 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADTrJNfUEE9cnngVxqFRB7ngzSqJXl1ks5sTfx1gaJpZM4OnMQZ>.

telephon · 2017-07-31T17:53:09Z

Modulating the bus argument of an InFeedback doesn't cause glitches:

https://github.com/supercollider/supercollider/blob/master/server/plugins/IOUGens.cpp#L666


(
SynthDef(\fxbus, { |out, in|
	var sig = InFeedback.ar(in, 2);
	ReplaceOut.ar(out, HPF.ar(sig, 800));
}).add;
SynthDef(\source, { |out|
	Out.ar(out, SinOsc.ar(330) * 0.1 ! 2)
}).add;
)


(
b = Bus.audio(s, 2);
Synth(\source, [\out, b]);

f = Synth.head(s.defaultGroup, \fxbus, [in: b, out: 0]);  // OK

fork { loop { 0.5.wait; f.set(\in, 99); 0.5.wait; f.set(\in, b); } };
)

Why must the AudioControlbe different? (sorry, just asking.)

mossheim · 2017-08-04T03:24:01Z

glad @vivid-synth asked - using the original testing code, this now crashes supernova (before, it just had the same issue).

(
SynthDef(\fxbus, { |out, in|
	var sig = InFeedback.ar(in, 2);
	ReplaceOut.ar(out, HPF.ar(sig, 800));
}).add;

SynthDef(\fx, { |out, a_in = #[0, 0]|
	ReplaceOut.ar(out, HPF.ar(a_in, 800));
}).add;

SynthDef(\boop, { |out, freq = 440, gate = 1, amp = 0.1|
	var sig = Mix(VarSaw.ar(freq * [1, 1.004])) * amp,
	eg = EnvGen.ar(Env.adsr(0.01, 0.1, 0.5, 0.001, curve: 0), gate, doneAction: 2);
	sig = LPF.ar(sig, 3000);
	Out.ar(out, (sig * eg).dup);
}).add;
)

s.boot;
s.scope;

b = Bus.audio(s, 2);

(
p = Pbind(
	\instrument, \boop,
	\degree, Pwhite(-7, 7, inf),
	\dur, 1,
	\legato, 0.25,
	\out, b
).play;
)

f = Synth.tail(s.defaultGroup, \fxbus, [in: b, out: 0]);  // OK

f.free;

(
s.makeBundle(nil, { // SUPERNOVA CRASH
	f = Synth.tail(s.defaultGroup, \fx);
	f.map(\a_in, b);
});
)

p.stop;
f.free;

mossheim · 2017-08-04T03:27:46Z

i'm getting a bad access here:

supercollider/server/plugins/IOUGens.cpp

Line 232 in 5f7d811

int thisChannelOffset = channelOffsets[unit->mSpecialIndex + i];

this mirrors changes introduced to scsynth in fdf1fb0

mossheim · 2017-08-04T03:51:45Z

@joshpar , i took your changes from fdf1fb0 and applied them to similar spots in supernova. the original test code now works with supernova like it did with scsynth.

mossheim · 2017-08-04T03:53:48Z

include/plugin_interface/SC_Graph.h

@@ -35,6 +35,8 @@ struct Graph
 	uint32 mNumControls;
 	float *mControls;
 	float **mMapControls;
+	int32 *mAudioBusOffsets;
+	uint32 mAudioBusOffset;


this appears to have been accidentally reintroduced - could you please remove it?

or I can do it, I don't want this to get too confusing though :)

…ixes

joshpar · 2017-08-04T04:02:20Z

done

…

On Aug 3, 2017, at 8:54 PM, Brian Heim ***@***.***> wrote: @brianlheim commented on this pull request. In include/plugin_interface/SC_Graph.h <#3063 (comment)>: > @@ -35,6 +35,8 @@ struct Graph uint32 mNumControls; float *mControls; float **mMapControls; + int32 *mAudioBusOffsets; + uint32 mAudioBusOffset; or I can do it, I don't want this to get too confusing though :) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#3063 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADTrAxtAxtyi5F9SesayMC63gXMjVdkks5sUpXpgaJpZM4OnMQZ>.

telephon · 2017-08-04T09:00:21Z

great!

mossheim · 2017-08-04T17:04:37Z

all right. should any tests / other documentation be added for this? i am otherwise good to merge.

telephon · 2017-08-08T11:28:01Z

i think it is ok to merge. we do need some systematic tests for mapping, but this is another issue

vivid-synth · 2017-08-09T00:17:19Z

Thanks @joshpar and @brianlheim !

joshpar · 2017-08-12T05:03:21Z

@snappizz - pretty sure this is ready to go

timblechmann · 2017-08-12T08:19:46Z

include/plugin_interface/SC_Graph.h

@@ -35,6 +35,7 @@ struct Graph
 	uint32 mNumControls;
 	float *mControls;
 	float **mMapControls;
+	int32 *mAudioBusOffsets;


this looks like an ABI change. please increment https://github.com/supercollider/supercollider/blob/master/include/plugin_interface/SC_InterfaceTable.h#L24

@timblechmann changing an API from v1 to v2 gives me some pause -- will this potentially break existing code?

it doesn't break any code, but it does break binaries. you can't use server 3.9 with plugin binaries from 3.8, or vice versa and you have to recompile all your plugins.

This reverts commit 891fcbc.

dyfer · 2017-11-02T23:26:13Z

I know this is closed, but it's causing a rather likely crashes in supernova: #3196 I think the referenced issue needs to be addressed before this gets included in the 3.9 release...

nhthn · 2017-11-04T17:55:10Z

two issues that have since been identified as side effects of this PR: #3196 and #3266. i have marked them as required for 3.9.

In sc_synthdef::prepare(void), the memory requirement for a sc_synth structure was underreported because it did not change with supercollider#3063. As a result, there was memory corruption. This commit both fixes the issue and adds clarifying comments to the calculations. Fixes supercollider#3196, supercollider#3266.

fixes (possibly?) for AudioControl mapping

6262192

joshpar mentioned this pull request Jul 29, 2017

AudioControl outputs stale data #2839

Closed

telephon reviewed Jul 29, 2017

View reviewed changes

telephon added 4 commits July 30, 2017 16:32

code formatting

2eb23e3

the rest of the file uses tabs.

Merge pull request #3077 from telephon/AudioControlFix

fcf0b7b

code formatting

code formatting

61804a8

Merge pull request #3078 from telephon/AudioControlFix

dc025e7

code formatting

telephon added the Work In Progress (WIP) - don't merge yet label Jul 30, 2017

joshpar added 2 commits July 30, 2017 19:42

handle audio bus offsets for mapping better with a pointer to bus off…

fdf1fb0

…sets

Merge branch 'AudioControlFix' of https://github.com/supercollider/su…

b0536f1

…percollider into AudioControlFix # Conflicts: # include/plugin_interface/SC_Graph.h # server/plugins/IOUGens.cpp # server/scsynth/SC_Graph.cpp

supernova: handle init & set of audio bus offsets

b186e02

this mirrors changes introduced to scsynth in fdf1fb0

mossheim reviewed Aug 4, 2017

View reviewed changes

remove extra member declaration from SC_Graph from AudioControl bug f…

4a68135

…ixes

joshpar changed the base branch from master to 3.9 August 12, 2017 05:03

nhthn added this to the 3.9 milestone Aug 12, 2017

nhthn added comp: scsynth comp: server plugins and removed waiting for consensus labels Aug 12, 2017

nhthn self-assigned this Aug 12, 2017

telephon merged commit 891fcbc into 3.9 Aug 12, 2017

nhthn removed their assignment Aug 12, 2017

timblechmann reviewed Aug 12, 2017

View reviewed changes

joshpar mentioned this pull request Aug 12, 2017

Plugin api change #3129

Merged

nhthn deleted the AudioControlFix branch August 14, 2017 03:19

dyfer added a commit to dyfer/supercollider that referenced this pull request Sep 29, 2017

Revert "fixes (possibly?) for AudioControl mapping (supercollider#3063)"

bc11177

This reverts commit 891fcbc.

nhthn mentioned this pull request Oct 31, 2017

Issue triggering old pre-compiled synthdefs in v3.9 #3266

Closed

mossheim mentioned this pull request Nov 30, 2017

supernova: reserve enough size for synthdefs #3317

Merged

mossheim mentioned this pull request Jan 4, 2018

scsynth: port audio control fix to v1 synthdefs #3382

Merged

fixes (possibly?) for AudioControl mapping #3063

fixes (possibly?) for AudioControl mapping #3063

Conversation

joshpar commented Jul 29, 2017 • edited by mossheim Loading

telephon left a comment

Choose a reason for hiding this comment

telephon Jul 29, 2017

Choose a reason for hiding this comment

joshpar commented Jul 29, 2017 via email

telephon commented Jul 30, 2017 • edited Loading

telephon commented Jul 30, 2017

joshpar commented Jul 30, 2017 via email

telephon commented Jul 30, 2017

telephon commented Jul 30, 2017 • edited Loading

telephon commented Jul 30, 2017 • edited Loading

joshpar commented Jul 30, 2017 via email

telephon commented Jul 30, 2017

joshpar commented Jul 31, 2017 via email • edited by telephon Loading

joshpar commented Jul 31, 2017

telephon commented Jul 31, 2017

joshpar commented Jul 31, 2017 via email

mossheim commented Jul 31, 2017

joshpar commented Jul 31, 2017 via email

telephon commented Jul 31, 2017

telephon commented Jul 31, 2017

telephon commented Jul 31, 2017

joshpar commented Jul 31, 2017 via email

telephon commented Jul 31, 2017 • edited Loading

mossheim commented Aug 4, 2017 • edited Loading

mossheim commented Aug 4, 2017

mossheim commented Aug 4, 2017

mossheim Aug 4, 2017

Choose a reason for hiding this comment

mossheim Aug 4, 2017

Choose a reason for hiding this comment

joshpar commented Aug 4, 2017 via email

telephon commented Aug 4, 2017

mossheim commented Aug 4, 2017

telephon commented Aug 8, 2017

vivid-synth commented Aug 9, 2017

joshpar commented Aug 12, 2017

timblechmann Aug 12, 2017

Choose a reason for hiding this comment

vivid-synth Aug 17, 2017

Choose a reason for hiding this comment

nhthn Aug 17, 2017 • edited Loading

Choose a reason for hiding this comment

dyfer commented Nov 2, 2017

nhthn commented Nov 4, 2017

joshpar commented Jul 29, 2017 •

edited by mossheim

Loading

telephon commented Jul 30, 2017 •

edited

Loading

telephon commented Jul 30, 2017 •

edited

Loading

telephon commented Jul 30, 2017 •

edited

Loading

joshpar commented Jul 31, 2017 via email •

edited by telephon

Loading

telephon commented Jul 31, 2017 •

edited

Loading

mossheim commented Aug 4, 2017 •

edited

Loading

nhthn Aug 17, 2017 •

edited

Loading