Are both AU and VST2 affected?
yes.
I’ve just installed a plugin that I built yesterday before pulling the 5.3.0 commit on develop - probably a 5.2.1 develop from a week or so ago and the problem exists there too, so this problem is somewhere between them.
I’m wondering if this got broken a few weeks ago when all those parameter changes were made?
Any updates on this?
Thank you
I had the same issue in FL studio.
My workaround is to comment out the audioProcessorParameterChanged callbacks in AudioProcessorParameter::sendValueChangedMessageToListeners in file modules\juce_audio_processors\processors\juce_AudioProcessor.cpp (line 1441):
void AudioProcessorParameter::sendValueChangedMessageToListeners (float newValue)
{
ScopedLock lock (listenerLock);
for (int i = listeners.size(); --i >= 0;)
if (auto* l = listeners [i])
l->parameterValueChanged (getParameterIndex(), newValue);
// audioProcessorParameterChanged callbacks will shortly be deprecated and
// this code will be removed.
/*
for (int i = processor->listeners.size(); --i >= 0;)
if (auto* l = processor->listeners[i])
l->audioProcessorParameterChanged (processor, getParameterIndex(), newValue);
*/
}
I tested VST2 only
bump
Would be great to see a fix soon. The current release is unusable for plug-in developers because of this bug and i want to use the great new 5.3 features
It’s been 7 days since last update - v strange!
Fixed here:
Please give it a test. If all is well I’ll cherry pick the change to the master
branch too.
thx - will check in the next hour or so and report back
Live on Windows is looking good now, thx. Not around Mac currently so can’t test AU. prob in another hour or so
I’ve been waiting on this one as well; just built my plugin against the latest on develop, a632b57e7
, looks good here too. Tested VST & AU in Live 10.0.1 on OS X 10.13.
confirmed all is good on OSX too. thx
Just curious, has AAX the same problem?
The same issue has already been addressed in the AAX, AUv3 and VST3 wrappers
Amazing, thank you.
I’ve found a new way to resurrect this issue. In a pretty simple test the orange light of doom illuminates whenever I put in a call to updateHostDisplay, such as I may do when a parameter name is updated, which seems like a strange reason to knock out automation.