OK, Dennis, after you insisted so much that it can do this, I’ve downloaded the current version from Izotope’s website. Put it in a Cantabile project, wired a Mic input and a MIDI input into it and started trying.
Now I put it in MIDI mode, played around on a keyboard and sang with it. Observed the following:
I play a quick G chord and release it - no more harmonies, because no more MIDI input. On the VoiceLive, once a chord is recognized, it stays locked in.
I play the same chord in various octaves while singing - in Nectar, the generated voices move up and down with the MIDI input. So when I play high piano comping, suddenly all vocals become chipmunks. And when I do quick staccato chords, they even become chopped-up chipmunks . On the VoiceLive, wherever I play a chord, it doesn’t matter, because VoiceLive will add the generated voices where I’ve set them to be, RELATIVE TO MY LEAD VOCALS and in tune with the chord I play; one close interval above and one below the lead, or two above, or whatever. They will not move up or down when I change the octave on my piano. And of course, they won’t stop when I release the keys - they will sound as long as I sing the lead part.
So in summary, Nectar has no chord recognition in its MIDI input - it simply uses the MIDI input as fixed tone references for additional voices. The only modification it makes to the notes is shift them up or down an octave as my vocal line progresses up or down. An when notes get released, the respective note will not be generated anymore - because Nectar doesn’t “know” that I’m currently in G minor
There is no audio sidechain in Nectar for audio chord recognition, so it has no way to track my guitar playing and recognize the chords. The key detection only works on the voice input - that’s why it needs to listen to a full verse and a chorus before it can propose a scale. A VoiceLive has an audio input that you can feed with your instrument (guitar or other) or even a full mix, and it nearly instantly detects the correct chord and generates harmonies to fit it.
The Tracking Mode you mention only means that any lookahead functionality that creates latency is deactivated - has nothing to do with chord tracking.
So, the result of my quick test is very much like I expected: if you use Nectar in a production environment it is a great tool, also you CAN use it live if you focus your keyboard playing specifically on controlling the additional voices. You need to hold long chords to keep the voices sounding, and you need to play them in the correct octave to avoid “chipmunkcy”.
But as a live tool that analyzes my live piano and guitar playing and generates upper or lower harmonies that fit with my song, it’s not even close to practical for me.
@pax-eterna - please jump in if I have misrepresented anything here or if I have missed some hidden function.