@resonant

You can do DSP in hise script.
I imagine the method would be to load the ir into a hise buffer object (an array that holds each sample of your audio file), then do some kind of resample algorithm to fill a new buffer with a repitched version (making the file faster or slower by interpolating the samples) then load that processed audio data into the convolver.

Of course, if you use c++ you can do this all with high quality and efficiency .
But I'm guessing you're asking about doing it in hise script.

I've not done audio manipulation in hise script before. But I've seen that it's possible, looking at the docs

Link Preview Image HISE | Scripting | Buffer

A buffer of floating point data

favicon

(docs.hise.dev)

@oskarsh yes of course it's somehow possible to render stuff like this in realtime, it's just not what the waterfall display is made for.

This looks like 16-20 paths. The waterfall display can render 500 paths at once to make a proper 3D shape without that visible grid.

@resonant I really like vital_rev, you can find all the stock reverbs here:
https://github.com/grame-cncm/faustlibraries/blob/master/reverbs.lib

@Mighty23

Thank you for the explanation. I don't mean a graphic, but a modulation like the one in the image below. Not separate Right and Left, but a single one (like the scriptnode gate, comp...etc.).

alt text

@d-healey said in WDF Model is not working in Faust:

Interesting that it gave you the right answer,

because the 4o is guessing. You can ask it to run a python script that counts rs in a string and you'll get the correct answer. Pure inference in this case is mostly guesswork, because it works with tokens, not letters.

07eec81d-c908-44ee-b8a3-c2a703e8f569-image.png

The o1 has reflection so it can catch itself doing this. Plus, this test comes up so often that I wouldn't be surprised it already made its way into the training data.

@Orvillain Wow Thank you very much for this detailed explanation! I am grateful to you :folded_hands:

@Christoph-Hart said in The need for Spectral Processing and AI in HISE:

You should be able to use a custom C++ node with this code:

It would be great if there was some examples for custom C++ node usage

@resonant bump

@ustk Perfect. Thank you!!

Thank you @d-healey @orange

Both 2 solutions are really helpful.

@Christoph-Hart That's amazing, thank you!

@Soundavid I'll try this, thanks!

@ulrik Wow That's amazing! Thank you so much bro! Yes of course this will does the job for me.

@orange said in Calling multiple functions with delay on a Knob's Callback:

Especially watch out the Oversampling functions.

After doing tons of testing, I found that using the Engine.getMilliSecondsForSamples() function is one of the causes of the multi instance crashes.

I've used this after declaring the delay compensation with Engine.setLatencySamples() for compensating the delay with a simple gain's delay parameter when the Oversampling is off. So when the Oversampling is off, I am giving the exact sample delay (that we declared via delay compensation) with a Simple Gain unit like this: SimpleGain.setAttribute(SimpleGain.Delay, Engine.getMilliSecondsForSamples(4));

But it seems that method causes the crashes. Maybe getting the sample rate multiple times on a multi instance DAW project? I don't know.

Then I created a SN network like the below which converts the sample delay directly with converter, the crash is gone :) :

Screen Shot 2023-04-19 at 15.45.51.png

@Christoph-Hart Any progress about this?
I think it should be a easy one for you, since you've already did the mono version.

It's just for Labels that were set to saveInPreset=true, but I actually changed it so that all controls also restore the value.

@Christoph-Hart Ok, I'm sure it iwill be a very easy fix for a genious code guy like you ;)

@Win-Conway really? Will investigate how to do it. Thanks man!

Very nice and useful! :)