Notices by andrew (vavassor@mastodon.social), page 7

@kibi Oh my god, what if coyote, though.

… Two degrees of freedom …

rotating coyote

@timj It was in Designing Audio Effect Plug-Ins in C++ by Will Pirkle, chapter 6 section 3.2. https://books.google.com/books?id=QddcxHLavrMC&lpg=PA168&ots=NqpAuhx2na&dq=resonator%20smith%20angell&pg=PP1#v=onepage&q=resonator%20smith%20angell&f=false

A lot of it has been opaque to me. But especially the sections on tapped delay lines, allpass filters, and "Time-Varying Delay Effects" have given me places to look for other articles to fill in the gaps.

Feel like with all the stuff it has there it's a good index of things to look up more about.

This site has been a really good reference for the synthesis and audio effects stuff I've been doing lately. https://ccrma.stanford.edu/~jos/pasp/pasp.html

I don't really know what I'm doing, y'all. I'm pretty limited on reference material. Feel like either info is limited or I don't know the right words to search for, today.

It looks like these are super focused on one note by design. It sings out the note but most other sounds come through only very quietly.

So if I want a more complex resonator effect I guess I'd want to make a bank of these filters and tune them to different notes. Then feed each filter separately and mix their outputs?

@timj The one I'm using is the Smith-Angell Resonator.

I didn't really pick it for any reason though. I just found a snippet from a DSP book that explained it and it seemed easy enough to code.

If you jack up that parameter it "sings" the note of the center frequency of the passband. It's really neat?

Otherwise, it just sounds like a normal bandpass.

"What's so resonant about this filter? It just sounds like a bandpass. That's a misleading name …"

*turns up the quality factor*

"o—Oh!"

@ervanalb NOW we're talking!!!

@lycaon I feel like this would just lead to the issue of folks who have few mutuals are just never recommended, which is unfortunate.

@dartigen @wxl It should be a little floppy disk "save" button on the compose page. And to load a saved draft there's a menu option in the side drawer menu that leads to it.

I haven't done FFT yet, but what I have done is make a spectrogram from the Discrete Fourier Transform.

The audio being visualised is white noise that is being put through a band pass filter, with the passband being sweeped up and down the frequency range to "pitch" the noise.

The x-axis is logarithmic because otherwise note frequencies got scrunched up on the far left. It also has the nice effect of showing the sine clearly, though! https://mastodon.social/media/tKafhn8piHPKK2OjAjw

When you want to subscribe to a japanese music artist's youtube channel but This video contains content from Sony Music Entertainment (Japan) Inc.. It is not available in your country.

FFT was something I wanted to code myself and understand, but it's been kinda outside my mathematical grasp?

I've been reading up and watching stuff and implemented the basic bit. I think I mostly get it, but it's weird

(It was running once every iteration of the audio thread's frame and quadrupled-ish the frame duration by itself.)

"Hmm, I know the naive Fourier transform is slow compared to the Fast Fourier Transform. Let's run it and see, though. For fun."

"OH WOW. Yeah, no, that's slow."

@sydneyfalk I guess! Although it's been easier to make low-pass in the problem domains I've worked with. But I don't see why not!

So … you can make a lazy high-pass filter by taking a low-pass filter and subtracting its output from the original signal.

And then you can make a band-pass filter by just applying a low-pass and high-pass filter.

So you can just make everything out of low-pass filters.