Looking for additive synthesis plugins/vsts that let you basically draw in frequencies like in Afx's "Formula"

[Brisbot]

I only know of three of these vsts. The one from 20 years ago Aphex used for the above track. Izotope Iris 2. And to an extent Harmor.

HOWEVER I have forgotten the name of the Formula one, but I remember it was for Mac's Only so I never got it.

Izotope Iris 2 is really a combination of a typical synth with the additive synth stuff, however the workflow for it isn't as 'free' as what AFX used for "Formula" so Iris 2 is out. It feels restricted. I do own it btw but always disappointed when I try to use it purely for drawing the sound.

And for Harmor, only way to draw waveforms would be to do it outside of harmor. So again the workflow for doing something like that is inconvenient.

[TubularCorporation]

I used to like Coagula way back (from the guy who made Granulab - the original, 90s Granualb, not the VST that took the name a couple decades later).

 

Still works on Windows 10, and it's in the Wayback Machine.  https://web.archive.org/web/20130404110130/http://hem.passagen.se/rasmuse/Coagula.htm

 

If you're using something newer than Windows 2000 you'll probably want this, too, because it's not a very complex program but there's a bit more to it than it looks like at first:

https://archive.org/details/microsoft_help_viewer_hlp

Edited February 21, 2022 by TubularCorporation

[exitonly]

https://uisoftware.com/metasynth/

[TubularCorporation]

Reaper has spectral editing directly on clips in the timeline now, but it was still pretty new and limited last time I tried it a couple years ago.

Edited February 21, 2022 by TubularCorporation

[TubularCorporation]

Oh yeah,

 

Virtual ANS https://warmplace.ru/soft/ans/

And Phonopaper, which is a little like Virtual ANS except it's mainly a phone app (there's also a Linux version but I've never tried it) and it uses camera input: https://warmplace.ru/soft/phonopaper/

 

Edited February 21, 2022 by TubularCorporation

[kausto]

virsyn cube

[Brisbot]

23 hours ago, TubularCorporation said:

Oh yeah,

 

Virtual ANS https://warmplace.ru/soft/ans/

And Phonopaper, which is a little like Virtual ANS except it's mainly a phone app (there's also a Linux version but I've never tried it) and it uses camera input: https://warmplace.ru/soft/phonopaper/

 

woah that looks tight AF

[zkom]

I made a little Python script that converts PNG and JPEG files to WAV files using PIL and SciPy. Basically what it does is an inverse short-term Fourier transform on the image data after it's been vertically scaled to the FFT block size either linearly or logarithmically.

This is how it looks in a logarithmic scale. I had to add some black padding to the image so it doesn't screw up the bottom of the image totally.

The linear scale looks better because then you get 1-to-1 correspondence with vertical lines and frequencies:

If someone wants to try this script I can upload it to GitHub. It's run from the command line or shell with a few parameters.

[zkom]

I put the script in GitHub because why not? https://github.com/kzmo/img2sound

Should run pretty much on any platform that has Python >= 3.6, but I have tested it only on Python 3.9 and Windows 10.

[zkom]

I'm testing different patterns how they convert to sound and grids with horizontal lines with different kinds of distortion seem to work best. I basically just create something with GIMP and then use my script to turn it into a sound file.

Here's a grid with a kaleidoscope effect

 

[Wunderbar]

Syne: https://www.musicdevelopments.com/syne.html

[Freak of the week]

On 2/23/2022 at 2:27 PM, zkom said:

Basically what it does is an inverse short-term Fourier transform on the image data after it's been vertically scaled to the FFT block size either linearly or logarithmically.

this is what metasynth and coagula are doing, essentially.

anyway, xenakis killed it with upic

https://en.wikipedia.org/wiki/UPIC

[TubularCorporation]

28 minutes ago, Freak of the week said:

anyway, xenakis killed it with upic

And just in general.

[BaggerMcGuirk]

On 2/23/2022 at 12:35 PM, zkom said:

I put the script in GitHub because why not? https://github.com/kzmo/img2sound

Should run pretty much on any platform that has Python >= 3.6, but I have tested it only on Python 3.9 and Windows 10.

Got it running on my machine(Python 3.10), and it's excellent.  Really nice work!

[mcbpete]

A couple of years ago I bought Photosounder which is kinda a souped up Coagula:

 

 

[TubularCorporation]

Probably not in general, but with Mickey Mouse maybe since Disney has historically had more control over international copyright law than most governments do.

 

EDIT: lik the tracks, I could definitely hear some Coagula in there.

Edited March 14, 2022 by TubularCorporation