EXS24 voices affect CPU?

Since the EXS24 is just playing samples, does the number of voices affect the CPU like when we use more voices on a synth?

Is there any extra processing or not really? My guess is that in this case it's only a feature to limit (or not) the number of simultaneous notes...

I assume that the more you're using the synthesizer engine to playback those voices (transpose, pitch bend, filters, modulations, glide etc...) the more CPU you'll use. Meaning that if you're using some filtering and transposition for example, then the more voices you're using, the more CPU resources you'll need.

I assume that the more you're using the synthesizer engine to playback those voices (transpose, pitch bend, filters, modulations, glide etc...) the more CPU you'll use. Meaning that if you're using some filtering and transposition for example, then the more voices you're using, the more CPU resources you'll need.

 

My understanding is that for example if you just increase the number of voices, you are just limiting the number of notes you can play at the same time, and the only time you are doubling the actual voices and the whole engine itself is when you use unison, because now when you are playing for example C#, is not just playing 1 sample, but multiples. Right?

 

So if Unison is off on the EXS24, the number of voices is just the number of notes at the same time. I would assume that in that situation, the pitch, bend, etc, would be applied to everything at the same time, almost like grouping tracks and sending them to an aux and using only one compressor.

That’s what I would assume, but I have no idea how things work under the hood...

The EXS24 is pretty light on CPU, always has been - it was designed to be quite CPU light even on machines in the Logic 4.x days in the early 2000s when it was introduced. It's a bit heavier if the filter is turned on, as you now have a more CPU-heavy component that needs to be calculated for each voice, but even so, on modern machines it's still very light. Obviously, the more voices you use, the more resources are used to generate them.

I would assume that in that situation, the pitch, bend, etc, would be applied to everything at the same time, almost like grouping tracks and sending them to an aux and using only one compressor.

That’s what I would assume, but I have no idea how things work under the hood...

 

No, voices always need to be calculated individually, each voice is a complete sample playback (oscillator) + synthesizer engine (envelopes, filters, modulation etc) in it's own right. If pitch bend is raising the pitch by +6 semitones, it just means when each sounding voice is being calculated, the modulations applied to each voice's pitch (in this case) get 6 semitones added on.

My understanding is that for example if you just increase the number of voices, you are just limiting the number of notes you can play at the same time

Ok I thought "increasing the number of voices" meant that you're using more voices ("Used Voices" in Sampler). Do you mean that you're raising the "Polyphony" parameter?

 

 

Oh scratch that you said EXS24....

My understanding is that for example if you just increase the number of voices, you are just limiting the number of notes you can play at the same time

Ok I thought "increasing the number of voices" meant that you're using more voices ("Used Voices" in Sampler). Do you mean that you're raising the "Polyphony" parameter?

 

voices-sampler.png

 

Oh scratch that you said EXS24....

 

Yes I mean the Voices parameter in between the poly switch and that thing that says how many voices are being used. I don’t have Sampler because I can’t upgrade to Mojave

I would assume that in that situation, the pitch, bend, etc, would be applied to everything at the same time, almost like grouping tracks and sending them to an aux and using only one compressor.

That’s what I would assume, but I have no idea how things work under the hood...

 

No, voices always need to be calculated individually, each voice is a complete sample playback (oscillator) + synthesizer engine (envelopes, filters, modulation etc) in it's own right. If pitch bend is raising the pitch by +6 semitones, it just means when each sounding voice is being calculated, the modulations applied to each voice's pitch (in this case) get 6 semitones added on.

 

Thanks for clarifying that. David was correct then as well.

And yes as you said the EXS24 seems in fact very CPU friendly. I’ve been experiencing big loading times every time I insert Ultrabeat and I like things to NOT slow down my workflow, so I’m transferring some samples to EXS24 and creating new channel strip settings.

 

Appreciate the clarification

Yes I mean the Voices parameter in between the poly switch and that thing that says how many voices are being used. I don’t have Sampler because I can’t upgrade to Mojave

Ok then indeed the parameter called "Voices" in the EXS24 is the same as "Polyphony" in Sampler.