@Christoph-Hart Thanks! That sounds interesting, too. I will have to look into this LAF thing. I thought I would not need it since I will always mess with pixel images. I have to watch David's video on that matter.

@DanH
A const array will always be an array whereas a reg could be something else later :

reg myArray = [] ; reg myArray = 42; // this will work. const myArray = [] ; const myArray = 42; // this won't

If I'm not mistaken, it has something to do with memory allocation. That's why it's always a good practice to reserve the arrays's slots number if you know it.

const myArray = [] ; myArray.reserve(4); // push 4 variables into the array later

Edit : ah, didn't see @Christoph-Hart 's answer

Thanks @d-healey ! I will have a deep dive into the server and file APIs then.

Should be fixed now.

@d-healey also working here... Weird
@problema your browser is too small so the pop-up area gets cut. IIRC you can change the font style with laf, but AFAIK centering the input field is not possible without getting into the source code. Either do that or expand your browser's horizontal size.

@johnmike Ok well you cant just copy and paste code out of the documentation and expect it to work...

FileSystem.browseForDirectory(var startFolder, var callback) you need to specify a value for startFolder, even an empty string will do you need to define a function for callback, that function will need a single param called result -this will be a file object, in this case a Directory - that you can manipulate and check in your function... // get the folder WeDontUSeThis = FileSystem.browseForDirectory("", function(result){ MyDownloadFolderString = result.toString(FullPath); Console.print( MyDownloadFolderString); });

@Christoph-Hart It has to do with the DM I sent

@gorangrooves Yeah me too for most of my instruments, but it depends on the plugin. Someone posted a good example above. When you have a mixer you want the default (double click) to be 0dB but you want the initial state of your plugin to have a good balanced mix so the faders probably won't all be at 0dB.

@d-healey It's called Deep Learning for a reason 😉

WaveNet: Feedforward network that learns to estimate a function from training data (one of the simplest networks), original paper from 2016

AutoEncoder: Takes a complex representation of data (usually a MelSpectrogram), simplifies (encodes) it, then learns to reconstruct the original complex representation from the encoded information instead. Useful for data compression, more useful when you make it variational

Variational AutoEncoder (VAE): Uses a fancy reparametrization trick to take each complex representation and "map" it to a latent dimension, the network can then decode from any arbitrary location inside that latent dimension. Think of it like burying treasure, you can bury one piece in the sandbox, one under the tree, one next to the car -- you then say "hey stupid neural network idiot, decode the stuff under the car" and it will give you (hopefully) exactly what was "mapped" to that location

Generative Adversarial Network (GAN): Uses one network (generator) to "trick" the other network (discriminator/critic), each network gets better at doing its job (tricking / not being tricked) until the generator gets so good at generating that it's indiscriminable, Deepfakes are a type of GAN iirc

Diffusion Model: Trains by taking a data sample and adding noise to it incrementally, it learns by understanding what the previous noise-step was, then you pass it a random noise value and it decodes something from that. Dalle and StableDiffusion are diffusion models, they aren't necessarily ideal for audio since they're quite slow at inferrence time

DDSP: differentiable digital signal processing, they basically train a neural network to control audio effects (or a synthesizer), by letting the network take the wheel it doesn't have to actually generate the raw audio data and instead just controls existng DSP tools

RAVE: realtime audio variational autoencoder, a fancy VAE that also includes a GAN stage at training. it's the current state-of-the-art for realtime synthesis and is embeddable on micro devices and such. I have no idea where to begin implementing it in hise as it's quite complex

Implementation: you basically need a fast library and some sort of time-series network, the former can be RTNeural (installing Tensorflow / torch is kinda annoying compared to the simple RTNeural library), the latter can be a LSTM, RNN, GRU or a convnet. It also must be trained (obviously)

all of these time-series models basically "predict" the next audio sample or buffer, based on the previous one(s). without temporal coherence it will just randomly predict each sample, which results in white noise (ie my problem in my VAE)

There's also guys like the dadabots dudes who generate full music instead of individual instruments/effects, they have a 24-hour metal livestream that is being generated in realtime which is really cool

you can find all sorts of tutorials on how to build simple neural nets on youtube using Python and a library like Keras. Be warned: you'll be looking at hand-drawn MNIST images for hours at a time

okay that's my last year and a half condensed into an ugly forum post 🙂

@Casmat If I understood right - you can use

const myCoolSampleMap = "{EXP::myCoolExpansion}myCoolSampleMap"; SamplerA.asSampler().loadSampleMap(myCoolSampleMap );

If you want this sample-loading handled automatically when an Expansion is loaded, you need to add it to the expansion's loading method, something like:

const expHandler = Engine.createExpansionHandler(); inline function loadSamples() { local sampleMap = "{EXP::myCoolExpansion}myCoolSampleMap"; //better ways to do this, note the EXP::expansion wildcard SamplerA.asSampler().loadSampleMap(sampleMap); } function expCallback() { currentExpansion = expHandler.getCurrentExpansion(); //useful to access subfolders ie "the better way" to do the above method :) loadSamples(); } expHandler.setExpansionCallback(expCallback); //calls our functions when we change expansions

@d-healey was a dinner patron for quite a while but finances took a hit. will be back soon as thats sorted. hopefully ill release my plugin one day 😧

@Christoph-Hart said in SNEX channel in processFrame:

If you never expect the function to be called in a mono context (which again should be the case if you assume that there is a data[1] element, you can leave it just empty:

How does this work with the "Support Mono FX" option? Fill both functions and just exclude data[1] in the first?

@d-healey thank you David..sigh ----Im an old man now...

@hisefilo thanks - I got the one david recommended to work.

@d-healey Thanks! Will look into it and see what it leads me to!

@d-healey said in CustomSettings - Use Open GL:

@Lindon
Floating tile's data property. Set useOpenGL to false

another dumb day for Lindon...thanks.

@Christoph-Hart I had no idea, that's a very cool feature! Wouldn't work in this case though as the sample map in each layer is selected from the UI each with separate controls.

@DanH yes indeed. You also can add a min Max node to keep it under control