@dustbro @d-healey Quick update, I stumbled upon this bug while archiving my app too. Just tried with Xcode 14 beta 2 and it works. It took a LONG time to link though, but at least it doesn't SIGTRAP!
Cheers.
E.

@Christoph-Hart Makes sense, I investigated more about having multiple MainController's and it was definitely the wrong route. Pretty satisfied with FRONTEND_IS_PLUGIN for my use-case.
Thanks!

Moreover, right now updating some sample properties (like Sample Start/End), is done via the KillStateHandler, which kill all voices on the main synth chain, which momentarily suspend all sampler instruments, app-wise.
I can avoid this by setting the define FRONTEND_IS_PLUGIN to 1, but that might not be the right approach for my use case.

Any help appreciated!

Thanks!
E.

Hi!

Quick question: is it currently possible to run multiple MainController's (and SynthChain's) simultaneously ?
I'd like to host each instance of a ModulatorSampler in a separate track plugin to allow rendering each of them separately (and possibly concurrently). The only concern about this is that the MainController has its own SampleManager, so if I create multiple instances, I get mulitple SampleManager's as well. It would be quite trivial to use a unique SampleManager, but I'm just wondering if that's the right approach to get a more fine-grain access to processing different chains separately.

TIA!
E.

Note that I had to disable the pool "search mode" when loading the same sample over a range of MIDI notes, otherwise updating the start/end positions would be mirrored on every sample.

auto pool = sampler->getSampleMap()->getCurrentSamplePool();
jassert(pool);
pool->setDeactivatePoolSearch(true);

I believe this is why this pool mode exists, @Christoph-Hart, i.e.: to allow each duplicate sample entry to have its own properties?

E.

Alright for those interested, the most straightforward solution (in my case) was to map one ModulatorSamplerSound per MIDI note and set the Sample start/end accordingly.

It would be really cool to be able to synchronously set the sample start/end positions without killing the voices when the sample is loaded entirely in memory (no streaming).

E.

Hi!

I am currently implementing a sample slicer (aka sample chop) and I want to let the user preview each slice individually.

What would be the most straightforward approach for this purpose:

• Map a ModulatorSamplerSound referencing the same sample over a range of MIDI notes (1 per slice), each sound having a specific SampleStart / SampleEnd
• Simply set the SampleStart and SampleEnd right before triggering the portion of the sample
• Use SamplePreview from the SampleEditorHandler

I believe all methods would work but some might have advantages over others.

Thanks in advance for any help,
E.

@Christoph-Hart Hi! Thanks for the quick reply. Yeah, that's what I came down to, I'll base64 encode my properties to avoid adding a child ValueTree. That'll work even though it's not really optimal.

Hi there! Quick question.

What would be the best approach to store extra properties (using a child ValueTree) within ModulatorSamplerSound? I tried using the data ValueTree (via ModulatorSamplerSound::getData()) and appending a new child ValueTree to store my own properties, but upon reloading it's getting interpreted has multi-mic data.

The alternative would be to store my properties outside but I thought I'd ask first, as there might be a way to do this already.

Thank you in advance!
E.