I don't know who of you used ever Cool Edit aka Adobe Audition. There is a pretty nice freq filter, called FFT filter. It allows to setup spectral envelopes with very high accuracy (24000pts which means a fraction of Hz sharpness of filtering edge, "infinite" levels of damping/amplification). Minus of this plugin is - small window, poor preset support, updates can be really painful if you don't have a plan "how to".
What I need is a multiband resonant filter DX/vst plugin (similar to this in Cool Edit), which allows to set up accurate, small frequency bands. I mean really small bands (like between 1-30Hz width per band) and really resonant, and with amplification/damping difference like up to 200dB (not poor 18-30), to amplify the signal and remove unwanted components. For example, if you would filter a noise wave sample with that, you would get a very clean tone, irregular a little bit. It would be nice to have at least 3-4 such resonators per plugin (or even more; a matter of lenght of a chain). What I look for in such solution - is easy live update of freq ranges and amplification levels.
Other way of achieving similar tones is not wanted, what is needed is the filtering of existing sound sources to get the exact results.
I mean really small bands (like between 1-30Hz width per band) and really resonant, ... For example, if you would filter a noise wave sample with that, you would get a very clean tone, irregular a little bit.
With these resonant filters, the major difference between "similar" things is in the vibration. One irregular freq peak may seem to be unimportant, but when you begin to mix a structure made of multiple peaks like this one - a very audible complex and yet subtle pattern emerges. Even dithering / noise shaping methods and order of exporting steps matters here (not to mention the selection of sound sources for initial filtering); peak width accuracy should be to about 1-2Hz (otherwise it becomes "outside" the sonic movie). Forget about reverbs and similar postprocessing there too (-:
When you listen to the final result - even if you don't know what is the difference, you just feel it. You sense either that it's "right pattern" or "not right", no matter what sound comes out of the box. It was a week of blind experimentation to find "how". Now it's time to simplify the process before batch production.
Note to "brainwave entrainmenters" - it's not about binaural beats and altered states of consciousness. It's about auditory perception reshaping (something like in tomatis method I guess? just came to my mind); the way we hear depends on how our brain processes the auditory information. The way the brain processes the information depends on what auditory signal we receive. If we receive certain kind of auditory signal, then during the neural recognition - the "recognizer" in the brain - changes (neuroplasticity adjustment), and thus - we begin to hear everything else differently. If we repeat the process of reshaping oftentimes - then we become familiar with this particular way we hear, and later - we can recall it back just by remembering it. Or by repeating the listening session part (takes 2-4 days).
Reshaping matrixes and their principles - is a totally different story.
Timbre - something like this:
(APR program 1)
(use google translation from Polish)
What I need is a multiband resonant filter DX/vst plugin (similar to this in Cool Edit), which allows to set up accurate, small frequency bands. I mean really small bands (like between 1-30Hz width per band) and really resonant, and with amplification/damping difference like up to 200dB (not poor 18-30), to amplify the signal and remove unwanted components.
Try this VST plugin
It introduces a delay in processing the signal - the lower the freq you want your FFT window to cover, the highre the delay (i.e., latency). However, you seem to be concerned with sound editing rather than sound processing or live performance, so I am assuming the latency is not an issue.
Not sure it will be ressonant enough when boosting, but you can get -infinite filtering reduction (i.e., totally eliminate bands).
Does lots of other clever FFT processing (harmonics boosting/cutting, transpose, for example, which you may want to check ou too.)
Looks promising, I check it tomorrow, thanks!