I'd be interested to read if anyone here is using this sort of tool for live recording mixes/mastering ?
Given that we generally select mics (and the rest of the signal chain !) to preserve the harmonic structure and integrity of the instruments we're recording...is there a place in the post-recording chain for this sort of manipulation: is it helpful when used with discretion, does it "mess with the DNA of the sound", is it a tool easily misused, helpful for spot mics/main pairs etc ? Are 'spectral profiles' of any use, or just glorified dynamic EQ engines ?
Here's a recent example (a version update), but there are surely others of similar ilk that might have found their way into your DAW's by now...useful, or not ?
Voxengo TEOTE 1.5 automatic spectral balancer plugin released
The following description might help to decrypt what this particular example is intended to achieve:
"TEOTE is an automatic spectral balancer AudioUnit, AAX, and VST plugin for professional music production applications. It was designed to be a very useful tool for both mixing and mastering. It automatically performs such tasks like gentle resonances taming, de-essing, tilt equalizing, usually performed during mixing and mastering. In mixing, TEOTE sounds good on pretty much any material.
While by definition TEOTE is a dynamic equalizer, its technology is solely based on multi-band dynamics processing. This allows TEOTE to have only minor phase issues, and to produce a subtle transient-emphasis effect associated with dynamics processing. TEOTE tries to make the program material follow the specified spectral profile, tuned to the contemporary mastering standards by default. It can be said that TEOTE “straightens” the frequency response, making further adjustments a lot easier; it removes a lot of repeating work.
Is TEOTE an AI plugin? In a sense that AI usually boils down to a “curve-fitting task”, TEOTE is an AI plugin that performs gain adjustment decisions in a quantity equal to “SampleRate multiplied by BandCount” per second. However, TEOTE does not use neural networks; it is based on an extremely-refined, completely predictable, curve-fitting function"