@Felix-W said in Sampler.setSoundPropertyForSelection(); seems to have a bug.:

@CatABC said in Sampler.setSoundPropertyForSelection(); seems to have a bug.:

@ulrik

@d-healey said in Sampler.setSoundPropertyForSelection(); seems to have a bug.:

@CatABC Use constants instead of magic numbers

Sadly, I tried using constants, but the problem persists. Also, if I swap the positions of HIGH_KEY and LOW_KEY, and need to move the smaller value to the larger value, the contradiction still occurs.

const var ComboBox1 = Content.getComponent("ComboBox1"); const var ROOT = 2; const var LOW_KEY = 3; const var HIGH_KEY = 4; const var C1 = 36; const var D1 = 38; const var E1 = 40; const var B3 = 71; inline function onComboBox1Control(component, value) { Sampler.selectSounds("C3"); if (value == 1) { Sampler.setSoundPropertyForSelection(ROOT, C1); Sampler.setSoundPropertyForSelection(HIGH_KEY, C1); Sampler.setSoundPropertyForSelection(LOW_KEY, C1); } if (value == 2) { Sampler.setSoundPropertyForSelection(ROOT, D1); Sampler.setSoundPropertyForSelection(HIGH_KEY,D1); Sampler.setSoundPropertyForSelection(LOW_KEY, D1); } if (value == 3) { Sampler.setSoundPropertyForSelection(ROOT, E1); Sampler.setSoundPropertyForSelection(HIGH_KEY,E1); Sampler.setSoundPropertyForSelection(LOW_KEY, E1); } if (value == 4) { Sampler.setSoundPropertyForSelection(ROOT, B3); Sampler.setSoundPropertyForSelection(LOW_KEY, B3); Sampler.setSoundPropertyForSelection(HIGH_KEY, B3); } }; Content.getComponent("ComboBox1").setControlCallback(onComboBox1Control);

Ultimately, I set up two methods to call different methods depending on whether the value in the combobox increases or decreases.😹

Wow, this method is so cool! Thank you for the code comments; they really enlightened me.

@Oli-Ullmann Yes I think you're correct. I checked the docs and there is a line about it not meant to be a dynamic state.

@Dan-Korneff I haven't ran into that one. Do you have a minimal test project?

@Orvillain said in [bug] SlotFX.setBypassed no worky:

What is the benefit of using const, in my situation?

I think the best answer is from Christoph

it yields a huge performance boost (because it can resolve the function call on compile time).
There is absolutely no reason to not declare UI widgets, references to modules (via Synth.getModulator() etc.) not as const

reg is good for accessing script level variables in midi callbacks and anywhere else where you would have had to use a script level var. But the more reg you have (and you only get 32 per namespace) the worse the performance gets. Internally reg is like a predefined array that HISE is keeping track of. So the more values you add the more data it needs to go through each time to you access it.

Another addition to Javascript: Use reg instead of var when declaring temporary variables which are accessed in the MIDI (or audio) callbacks. It tells the interpreter to store this in a fixed size container with faster access times:
If you have a script with lots of variables, the interpreter must search the entire array for every variable access (so the 23 - 40 ms are depending on how many other variables are defined in the script while the access time to reg slots stay the same).

It's also possible since you're storing them in an object rather than as direct references that const gives no performance benefit. But it's still good practice to use a const here. It makes it clear to other developers who might see your code (or your future self) that this variable is not meant to be reassigned, and it also prevents it being reassigned accidentality.

So it appears I should use a stack with the insert method instead. Mistake in the doc...

@Lindon Worked for me, but im not using the latest version of HISE, maybe a bug?

hey did you ever figure out why this is happening i have ran into this issue too

@d-healey yeah

@Christoph-Hart oh awesome! Thanks!

AGI my ass…

Lol very true 😆

@13murderer в лс отправил всё

@dannytaurus Here's the original thread
https://forum.hise.audio/topic/10665/warning-snippet-browser-overwriting-the-current-project

@iamlamprey Fixed with this PR: https://github.com/christophhart/HISE/pull/793

Changes repaint(), which I believe uses a cached path, to refreshAsync(), which forces a proper repaint.

@Taihongcraft start here https://audiodevschool.com/courses/hise-bootcamp/

@Christoph-Hart Thought this was fixed, but apparently not 🤷♂

Screenshot 2025-10-24 at 23.25.38.png

can you confirm this is a vaiable solution?

Sounds like a hack, but then using HISE to create an installer app is already a bit weird.

Speaking of hacks, the "cleanest" hack would be to just deactivate the entire audio processing:

https://github.com/christophhart/HISE/blob/3ff4bbd9eb5596b084c7653e2d09fae2b4799c1e/hi_frontend/frontend/FrontEndProcessor.cpp#L96

This is the main callback for a compiled plugin (or standalone app). So if you add this code

#if IS_STANDALONE_APP return; #endif

at the beginning of this function, it will never call into the HISE rendering. Note that there are a few things that might be glitchy, some things in HISE rely on the audio thread to be working and initialised correctly.