A1 polyphony in unison modes?

[analogholic]

Hi everyone,

How is the A1´s polyphony affected by the various unison modes?

[Marlowes]

Hej, herr A!

26 voices to start with ... I would be really surprised if unison did not reduce that number ... as in previous models ...
But that aspect is not mentioned in The Book;

Unison

unison.png (45.49 KiB) Viewed 3859 times

(Kroffe should really know, but he is not on line very often. Ricard and Elektromin; you are pros - how does it work?
OK, I'm an engineer too ... but too old.)

/Amicalement
Michael in Scania

[analogholic]

Hej Herr Micke
Får la hugga Kroffe på 99:an då

Yeah, I wonder if unison 1 takes two voices and unison 2/3 takes four?!?
I think the NL3´s polyphony wasn´t affected by unison but I´m far from sure.

[AdamStage2]

Hej Herr Micke
Får la hugga Kroffe på 99:an då
I think the NL3´s polyphony wasn´t affected by unison but I´m far from sure.

No it doesn't affect the polyphony count on the NL3,you can add as much or as little as you desire via the Dial.

NL2 allowed you to switch between values of how many voices were used for Unison in globals but they were fixed,I suspect the A1 is pretty much a preset option again but having those Chorus and Ensemble effects onboard can help fatten up a sound rather than maybe using Unison if your bothered about polyphony/note sacrifice

I'm surprised Clavia haven't built the NL3 unison option into more recent NL models,its a great Unison feature,often overlooked,there's lots of negativity regarding the NL3 buts its a great synth by all standards and using the unison can really warm up pads allowing you to add your own amount rather than preset fixed values and it doesn't sacrifice Polyphony too.

[AdamStage2]

I am starting to like this A1 more and more,I'm torn between getting this or the NL4,I'm looking for either to finally retire my Nord Lead 1 from live performances,the on board efx on the A1/NL4 will also make another outboard device redundant also.

There are some great sounds coming from this synth in this video

Look beyond the Trance riffs and its quite a meaty synth...I'm starting to warm more and more to it

[ricard]

Unison mode on the NL3 results in 5 voices per note instead of just one, I think panned left-left-centre-right-right. The number of voices in the unison cannot be changed (unfortunately) however, the detune spread can. The polyphony does not go down, so the machine is still 24 voice polyphonic.This was a major selling point at the time. In contrast, on the NL2, detune really does consume twice the number of voices.

On the A1, if I remember correctly, Unison 1 (which doubles each voice) does not eat polyphony, you've still got 26 voices. Unison 2 (3 voices) and 3 (4 voices) drop the polyphony to half, so you get 13.

All this is not at all intuitive and does not mimic how a real analog would behave where unison eats physical voices; I'm not sure exactly what type of DSP trickery is used to acheive it, my vague guess after having bugged Clavia's support about it (i.e. Kroffe at the time) is that there are enough DSP resources to double up a voice, but not to have completely separate parameters for them. Hence it's possible to create two voices with essentially similar parameters, only the oscillator frequencies differing, but not to have completely separate voices, i.e. a polyphony of 2*26 voices. Or there could be something else going on, like the output from a single voice being split and pitch shifted in order to create the detune effect. My hunch is that the latter would create some artefacts that would be noticable in the sound though.

[Marlowes]

Hej Ricard!

Thanks! I really love being at the same forum as you.

By the way, when can I join your band?

Amicalement
Michael in Scania, too

[kroffe]

On the A1, if I remember correctly, Unison 1 (which doubles each voice) does not eat polyphony, you've still got 26 voices. Unison 2 (3 voices) and 3 (4 voices) drop the polyphony to half, so you get 13.

I belive this is correct, ricard.

[Mr_-G-]

I'm not sure exactly what type of DSP trickery is used to acheive it, my vague guess after having bugged Clavia's support about it (i.e. Kroffe at the time) is that there are enough DSP resources to double up a voice, but not to have completely separate parameters for them.

Maybe a number delay lines with fixed LFOs to generate the various detunings (like the chorus principle) so you process just the waveform, and not the rest of the synth paramters.

[S h a w]

On the A1, if I remember correctly, Unison 1 (which doubles each voice) does not eat polyphony, you've still got 26 voices. Unison 2 (3 voices) and 3 (4 voices) drop the polyphony to half, so you get 13.

Is this the definitive answer? I only ask because you qualify your answer with "if I remember correctly", and I have never seen this in print from Clavia... Which is REALLY a shame because a manual OUGHT to answer basic questions like this.

Thanks all,