@d-healey I knew I was missing something! Thanks again!

Was wondering the same these days!

@Oli-Ullmann great, thank you!

yup, bug, should be fixed now.

@ustk genious, thank you!

@Christoph-Hart I checked the latest commit, that solves it! Thank you!

@d-healey ok, thank you!

@tomekslesicki Just took a look at this. Are you using a frame container with the full 12 channels?

I commented out the code that forwards the calls for channel counts > 2 because it blows up the code generation size, but I can add a macro that you can use to increase the number of available channels for frameprocessing.

@Christoph-Hart could you tell me more about the skew factor applied to the gain of the group xfades?

@Christoph-Hart fantastic, thank you!

@tomekslesicki There you go...

@d-healey yup, that was it! Thanks so much!

@Christoph-Hart yup, thank you!

Actually, no, it just gets stuck after a while when the parameters are linked to scriptnode controls at the top. @Christoph-Hart help!

@d-healey I’m using a global modulator in some, so I can’t compile everything.

@tomekslesicki Ah good catch, this slipped in when I was refactoring some of the complex data code (in particular the filter coefficient management across DLL boundaries). Should be fixed now.

@aaronventure I've added the gin stack blur at some point (it's already using this algorithm in the Graphics functions), however it's still ridiculously slow because it has to do graphics operations on a CPU and this is 1970 style hardware usage.

https://forum.juce.com/t/faster-blur-glassmorphism-ui/43086/7

I've went through the thread a few months ago and was close to port it over into HISE (I also sometimes get the urge of making blurry stuff), but I wasn't too convinced that the overhead justifies the performance gain (it's still being calculated on the CPU and this makes it easily a factor 100 slowdown compared to GPU based blur algorithms).

EDIT: this is the module I was talking about:

https://github.com/sudara/melatonin_blur

From revisiting the page, it might be actually worth it trying out how it performs...