Nord User Forum

[aureliopenna]

Bizarre...

[ajstan]

Ok, I think I’ve figured it out. It seems like you need to play enough notes within the release time on the AMP ENV to exceed the polyphony of the synth engine. Then instead of releasing the least recently played note to play the attack of the newest played note, the synth acts like it’s out of polyphony.

If I reduce the release time on a synth program to 0 or some other really low value, the polyphony issue seems to go away. Is that how it works for everyone else?

I think I wasn’t seeing it as I play a lot of strings, horns and pads that don’t require a lot of notes within a short period of time that would trigger the issue.

[anotherscott]

instead of releasing the least recently played note to play the attack of the newest played note, the synth acts like it’s out of polyphony.

I don't understand. To me, "acting like it's out of polyphony" would be the same as "releasing the least recently played note" - that's typically what a board does when it runs out of polyphony. So can you elaborate on exactly what it is doing that makes you say it's running out of polyphony, if releasing a previously played note is not what's happening?

[ajstan]

I don't understand. To me, "acting like it's out of polyphony" would be the same as "releasing the least recently played note" - that's typically what a board does when it runs out of polyphony. So can you elaborate on exactly what it is doing that makes you say it's running out of polyphony, if releasing a previously played note is not what's happening?

I could have phrased things better. Let me try again to state my understanding of the issue (unexpected behavior when the polyphony limit is reached, not a reduction in polyphony):

When newly played notes exceed the polyphony, the least recently played notes (according to some algorithm) are cut off to accommodate the new notes. This is most noticeable when notes are held (by key or pedal) as it affects sounds that are in the Decay stage of the AMP ENV where the sound is likely still prominent. I am not referring to this behavior in my post, although held notes may be an issue too, I just haven't checked.

I think that the issue in v1.46 (and it looks like some other previous OS versions tool) is that when the number of notes in the Release stage of the AMP ENV (notes that aren't held, but are still sounding and likely fading quickly) exceed the NS3 polyphony, the synth is not cutting off enough of the "old" notes to accommodate new notes being played.

Test 1: Set the AMP ENV Release of a Synth program to something like 2 seconds. Play a 4 note chord repeatedly. If you play the 4-note chord 8 times within 2 seconds the issue will appear and fewer than 4 notes will sound until the rate of playing is less than 8 times in 2 seconds.

Test 2: Turn the AMP ENV Release of that same program to 3.0ms. Again, play the 4 note chord repeatedly. The issue doesn't appear.

My thought is that it's something in the programming where notes in the Release stage are taking priority over new notes, not a reduction in polyphony. I wanted to post my findings to acknowledge the issue as I had stated that I had never experienced it, but it was only because I never was in a situation where I had 30+ synth notes from the same panel ringing out at one time. I heard about "voice stealing" and the "polyphony issue" but never saw anyone post about how to duplicate the issue. Hope that clears things up.

[analogika]

I wonder:

People were suggesting a while back that holding notes should be prioritised by something other than last-played — either lowest-note or loudest-note priority for sustain as a more musically useful algorithm.

Question: is it possible that they actually implemented this, and that it is now wreaking havoc as notes played within a repeated chord (as per description above) are played with lower velocity than notes played before and still decaying?

That would explain why they are killed immediately as other notes in the chord follow...

[Mr_-G-]

Have a look in this message
nord-stage-forum-f3/polyphony-lie-t14278-10.html#p93782
and this one
nord-stage-forum-f3/polyphony-lie-t14278-30.html#p94398
That way you can demonstrate how the polyphony is being affected.

[analogika]

Have a look in this message
nord-stage-forum-f3/polyphony-lie-t14278-10.html#p93782
and this one
nord-stage-forum-f3/polyphony-lie-t14278-30.html#p94398
That way you can demonstrate how the polyphony is being affected.

From the comments over recent updates, it would seem that something has changed since then in how the Stage handles polyphony.

The description in the post just above doesn't seem to gibe with the complaints in the old thread you linked to — or are they describing the same issue?

[Mr_-G-]

I can't tell if they are the same issue or not because I do not have a NS3 to try. Did you try it?

[AlexNagel]

I wonder:

People were suggesting a while back that holding notes should be prioritised by something other than last-played — either lowest-note or loudest-note priority for sustain as a more musically useful algorithm.

Question: is it possible that they actually implemented this, and that it is now wreaking havoc as notes played within a repeated chord (as per description above) are played with lower velocity than notes played before and still decaying?

That would explain why they are killed immediately as other notes in the chord follow...

Maybe my issue described here https://www.norduserforum.com/nord-stage-forum-f3/random-voice-stealing-polyphony-t16059.html has something to do with this?