A filter

 

It makes me think of VSTs that generate sounds from images.
IF there was one to turn sound INTO images, then edit the image with graphics software, then import the edited image ...
Interesting idea, probably already thought and done....by someone, someplace...
The question is: Did they make a VST for it?!

 

FFT processing is not new. iZotope RX does this with ease now, where you can clean out sounds within a spectrogram.
iZotope RX7 is both VST and standalone.

 

Oh yeah - I have done some with RX7 - so neat to do things spectrally ...
But i was thinking of a VST that could output the music into an image, especially if by some user-chosen parameters.
For example, to map / transform the sound into a picture of a face, and then use graphics-oriented tools , or even (and this is the cool exploraratory aspect) change the hairstyle, or use aging software, or other crazy things. Who knows what might happen!
Even better, transform the sound into a 3D model, and have it do battle with another 3D model created from sound, and record the audio of the battle! Or to have sexual relations resulting in a baby,and then import the baby into the VST to create new music!
The ultimate mashup!

 

This could make perfect sense when developers continue to put more effort into visualising what a simple FFT/Wavelet/Gabor/Wigner transform cannot do: Find harmonics that relate to the same instrument.
Being able to modify "blobs" of instruments over time instead of unrelated spectral content would help a lot in modifying audio more deeply.
I expect a lot like that to come in the future because research and progress on the field of artificial intelligence has come quite far in identifying and classifying audio content. It's not that we wouldn't know what we're looking for, Mr. Neural ;)

 

Just throwing this out there but possibly a speakerphone degraded patch somehow faded into silence via wet knob. I just came up with that so please don't ask how to execute

 

Speakerphone (plugin ) my friend. An you can use the mix knob to execute this.

Also on the O.P question, if you want a plugin you can use all these mentioned before to achieve the sound you want.On the other hand if you want that "pencil" or "brush" action like photoshop for sound, there is Izotope RX

 

thanks everyone, i was actually beginning to think in terms of pictures and removing 'blobs' of spectrograph ...could that represent a literal scratching away of layers if i then combined that track with blobs removed, with the same track without any blobs removed?

 

You could use Izotope RX7's new Music Rebalance to extract the "Other" (or bass/vocals/precussion w/ wider settings) then copy the result to a new file and add/remove more from the processed result.

You could also use a dereverb plugin by companies like Acon set to "difference monitoring" and play w/ the dials to extract just a verb based print of the source.

Lots of options but RX's modules and spectral editor is the way I'd go.

 

a little shabby --> Roland R-Mix

 

It SHOULD represent a literal scratching away of layers, if you automate the volume between them.
That, and perhaps some other automations will give you the change over time.
What @BobbyMonfrado said about using some other tracks that have been "Music Rebalanced" can give you so many options. The problem with music rebalance is the artifacts, so I would go easy on settings if you use that.
You could have a few tracks, each with various spectral erasures and/or effects and automate volume between them, depending on the effect you want.
It may be that (seems to me, without testing) for something that sounds most like "erasing" spectral erasing with RX7 (or 6 or another product) would do just what you want. The frequencies you erase could take care of most filtering. I think erasing higher frequencies would not sound so much like an erasure...
I assume you will be duplicating the track a couple of times (at least) to do the erasing, which will make comparing pretty easy.

 

Virtual ANS
You can import images, wavs, draw & modify sonogram like photoshop.
http://www.warmplace.ru/soft/ans/#scr

 

Image-Line Harmor. Also BeepMap included in FL Studio.

In terms of blur-like effects, there's an option called Blur in Edison (built in audio editor in FL Studio) - also available as a VST plugin. There's a video from Image-Line where they show you how to create a single sample impulse, record it, process through Blur in Edison and use it in real-time with any convolution reverb. I believe they also attached the resulting IR in the description of the video.

 

I had not realized that in Harmor-I've just begun to play with it. Have never used "Blur" in Edison either - THANKS