Converting Audio 1000 TIMES To Its Spectrum And Back | Crazy Effect

Marques Brownlee's video:
• This Is What Happens W...
For those of you reading, yes, I did try tweaking the parameters fed to FFmpeg's 'showspectrum' and 'spectrumsynth' filters, and this doesn't help, i. e., the volume issue is still there.
Music credits:
Chronos by Alexander Nakarada (www.creatorchords.com)
Licensed under Creative Commons BY Attribution 4.0 License
creativecommon...
"Investigations" Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0 License
creativecommons...
"Sneaky Snitch" Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0 License
creativecommons...
#1000 #1000times #crazyeffect #crazyexperiment

[@Dark_Brandon_2024]

5:55 this is exactly what i've been waiting for the whole time, no kidding, feels like it's the ultimate punch-line on the horizon

[@astro607]

I gotta say the first few iterations could probably be used as some sort of sound sample for a sound track. I wonder if generation loss is somehow recreatable in an easy way inside of some music production software with out taking too much time of course.
🤔

[@CrazyRandomDot-ly4zp]

Good question! Perhaps it'll really depend on a specific case of generation loss, since generation loss from, let's say, converting audio from one lossy format to another isn't going to produce the same traits as that from converting audio to a spectrum and back. In theory, though, once the traits are more or less clear, it shouldn't be hard to "simulate" them (sort of), although I'm not aware of any software that offers this...

[@aliheja]

this channel is underrated also pls script

[@michaelchen2821]

nice!

[@CrazyRandomDot-ly4zp]

Isn't the generation loss phenomenon fascinating? ;)

[@michaelchen2821]

@@CrazyRandomDot-ly4zp Yes!

[@syntaxerorr]

The typewriter sound is a bit too much.