mkpcola wrote:I still believe in the concept that good music is the sequence of the right musical notes at the right time.
We made good music out of 16/32 notes polyphony machines...
you know...I don't see the point of having 500 notes polyphony..
we all use sessional recording...anyway.
On a polysynth or workstation, that would be 500 *voices* of polyphony, not 500 notes. One note can consume multiple voices. For example, on some synths, unison uses a full voice for each stacked oscillator -- so you could easily consume 6 voices for a single note. Then try even playing a basic triad... you can see how even something like 24-voice polyphony on the RADIAS gets used up pretty fast.
And not even block chords... Let's run with your "good music is the sequence of the right musical notes at the right time" concept. Say you're running an arpeggio, or a nice melody line. You want it lush, so you set nice, long release tails on each note. If another note sounds before the tail ends on the previous one, now you're using up two voices of polyphony, just as if you were playing a block dyad (though you don't tie up those two voices for quite as long). And, of course, the more notes overlap in this fashion, the more polyphony you consume.
And on something like the Kronos, where you have combis of up to 16 parts, each playing notes, and those notes use up varying amounts of voices. Stereo multisamples, for example, use twice as many voices per note as mono multisamples. Wave sequences use twice as many voices as a comparable multisample, so a stereo wavesequence would use *four* times as many voices per note as a mono multisample...
So say you're running a 16-part combi, with 16 HD-1 programs loaded. Max polyphony for the HD-1 engine is 140 voices, and that limit doesn't stack with multiple instances in a combi. Play a basic triad on each part, or an arpeggio with enough of a release tail on each note that they overlap, and you're using up ~48 voices of polyphony*, assuming each program uses only a single mono multisample. But maybe you want a more complex sound, so you add a second sample oscillator per program, again loaded with a single mono multisample. That's ~96 voices of polyphony used. If you've got a dense orchestral arrangement going, huge polyphony ceilings are quite helpful.
And yes, you can always bounce tracks to free up polyphony when recording, but then you run the risk of limiting your editing and mixing options further down the line. And if you're playing a multi-part combi live, then bouncing really isn't a viable option.
Back in the days of limited polyphony, you'd see much more MIDI chaining and track bouncing than you do today. High max polyphony -- along with all the other features of the modern workstation -- offers flexibility to the musician that wasn't possible in a single piece of older gear. A board like the Kronos can, all by itself, do things that used to take an entire room full of gear. It provides far more creative options to the musician than were ever formerly possible within a single physical instrument. It can be used as a mono-timbral performance instrument, playing piano or organ or synth patches, or what-have-you; or it can be used to facilitate complex multitimbral performances, whether simple two-patch splits or impossibly intricate combis. You could use it to sequence and record nearly anything, from a solo piano performance, to a 4-part pop or rock track, to a vast orchestral movement.
That flexibility is the point of having triple-digit polyphony. It's there if you need it, and some people definately do for the sort of music they make.
*This is all just for the sake of argument; I'm well aware that actual voice usage will vary.
