I never knew this existed - just released on a label called Soul Tribe...

https://shop.recordshack.org/sa/sounds/172491_B_01.mp3

If you mean the final mixed tape by two track mix downs, that's very close to being the record. I don't think you'll be able to seperate all the instruments and vocals as many have the same frequency and cut over each other, particularly the vocals and mid range instruments. I'm sure we'll be able to fanny about with different aspects of a recording but if it's a mixed tape or the actual record it won't be remotely as good as getting a multi track master.

Hello,

Yes, I mean working from a final mixdown tape (stereo or mono) when it's the only option because the multi-track isn't available.

A few companies have been trying to develop ways of isolating individual instruments (and voices) with some sort of multi-dimensional spectral process (best way I can describe it) from final mixdowns but without real success so far.

The real aim of the technology is to enable people to change wrong notes (or chords) to correct mistakes in old recordings but if and when they're able to do it really well the same technology should in theory also enable the isolation (at least to a degree) of individual instruments for a degree of remixing.

They have a long way to go but I think the technology will work one day, although obviously not on all recordings.

As you said, so many instruments (and voices) occupy the same or similar parts of the frequency range so you can't focus on one "component" part without effecting the sound of another. That's why multiband EQing and harmonics filtering can't always solve problems with a finished mix and engineers often reach a point of frustration where they wish they had the multi-track.

I suppose their biggest problem is controlling the wider harmonics of sounds, such as a kick drum which might have an obvious sweet spot of 65 Hz, for example, but it also has presence (to lesser degrees) in other parts of the spectrum.

Paul

...from what I understand, or assume, the technology and algorithms may be similar in principal to things such as multi-dimensional x-ray tools which can see and control previously unseen elements of an object or an image at different depths - viewed from a multitude of different angles.

It's all maths and science, there's no reason why they can't do similar things with audio becuase audio can now be digitally controlled by the wave forms of its component parts.

As with most major developments in audio technology, the advances may come as a result of military research etc.

Paul