Slots A and B Polyphony

[maxpiano]

Looks like we have the same on pianos too, I tried the following test:

I am using the "Octave piano" factory program in NS2, which layers 2 pianos on A+B, 1 octave apart, so if theory holds true for pianos I should have a polyphony of 20

I am playing a fast sequence of 2 hands (4 fingers each) diminished chords, starting from a low octave and going up one octave per strike, while holding sustain pedal down.

In this way each chord strike triggers 8 notes and actually at 3d strike, i.e. when I go over the 20, I start hearing lower octave notes disappearing. If I have time I will try to do the test via MIDI file too.

[mon8169]

I've finally been able to test it in my classic, with the same result. Very, very smart voice allocation algorithm!!! It almost feels like having full polyphony!

[FunKey]

I just ran into the synth polyphony limit pretty badly today. I had a piano and strings with slow attack and release layered on slot A, and everything was fine. Then I turned on slot B as well, with the same strings shifted by an octave, and suddenly there were strange moments of silence at the beginning of every new chord. First I got confused as hell, but then I remembered this thread, so thanks to everyone for bringing up this topic.

BTW, has anyone combined such layered strings into a single sample? (If not, I guess I'll do it myself.)

[Mr_-G-]

Before resampling the strings, maybe you can try with the detuned saw waveform (+12 or -12) if that approaches the 2 bank synth sound. In combination with one of the Unison modes you might be able to create a thick string sound without the need of 2 banks.

[mon8169]

I just ran into the synth polyphony limit pretty badly today. I had a piano and strings with slow attack and release layered on slot A, and everything was fine. Then I turned on slot B as well, with the same strings shifted by an octave, and suddenly there were strange moments of silence at the beginning of every new chord. First I got confused as hell, but then I remembered this thread, so thanks to everyone for bringing up this topic.

BTW, has anyone combined such layered strings into a single sample? (If not, I guess I'll do it myself.)

Hi FunKey,

this is trange. Can you repeat it giving a little more (or less) attack time to the second shifted strings? It could be a problem trying to access to the start of the same (identical) synth program at the same time.

Ramon

[Mr_-G-]

Reading FunKey's post again... I also noticed this in some patches (i.e. not hearing the original sound, but the delayed echoes) when the delay is in action, and I put up the "Amount" dial to a high value near the end of the scale. Not sure if this is the case described above but it looks like a bug or a design "feature"

[metalfingers]

Yeah, limit appears to be 9 for the synth when both slots are on. My test patch used one FM slot and one sample based (if that makes a difference). For a soft, thick, multilayered pad with a long release time, those notes add up fast.

Really wish they could bump it up to 12 per slot or so. :/

[FunKey]

Before resampling the strings, maybe you can try with the detuned saw waveform (+12 or -12) if that approaches the 2 bank synth sound. In combination with one of the Unison modes you might be able to create a thick string sound without the need of 2 banks.

Good idea, thanks. I tried it, but in this case I think I need somewhat more realistic strings.

this is trange. Can you repeat it giving a little more (or less) attack time to the second shifted strings? It could be a problem trying to access to the start of the same (identical) synth program at the same time.

Changing the attack time just a bit in one slot doesn't stop this behavior, but in general, the longer the attack time is, the worse it gets. I think what happens is that when I hit a chord, a lot of old notes are suddenly cut off, but the new notes fade in slowly because of the long attack time.

Anyway, I went for the custom sample solution now. Works great. I'm actually using a three-layer version now, lol! (I'm a bit afraid to upload it because of copyright, but if anyone is interested, please PM me.)

Reading FunKey's post again... I also noticed this in some patches (i.e. not hearing the original sound, but the delayed echoes) when the delay is in action, and I put up the "Amount" dial to a high value near the end of the scale. Not sure if this is the case described above but it looks like a bug or a design "feature"

I think that's a feature. At an amount of 5, the first echo is equal to the original sound, and with higher settings, the echo stays the same, but the original sound becomes quieter. Too lazy right now to check whether that's documented in the manual.