Analog synths and round robin

[SampleScience]

Is it worth it to include round robin (4 layers) for analog synth sample sets? I'm in the process of recording old synths and I'm wondering if people actually care?

[DJ Warmonger]

No idea about old analog synths, but definitelly round robin is good feature for samples.

[murnau]

Of course you need round robin even if you sample long notes and from analog synths anyway. If it's to static you lost any realism.

[DSmolken]

I say sample one, do a double blind test asking people which one has RRs. If 90% of the people get it right, it's probably worth it. If 55%, well...

[SampleScience]

That's a bit of a problem for me, because I can see the difference between the samples in my audio editor, but I don't really hear it. That is for dry synth samples, for recorded to tape samples I hear the difference.

For some people it will be perceived as added value, for other it could be seen as useless making the product look bloated.

[Cinebient]

I think it is really useful. Especially if you repeat notes often or arpeggiate them round robin gives it more live and movement. Even if it's subtle.
Otherwise it will be too static.

[AnX]

You're better off using modulation in the sample player to add movement to the static samples.

No matter how many RR samples you have, they are ALL static.... (forgeting multi vel layers of course)

[EvilDragon]

I think it is really useful. Especially if you repeat notes often or arpeggiate them round robin gives it more live and movement. Even if it's subtle.
Otherwise it will be too static.

This. Plus additional randomized sample startpoint modulation.

[tehlord]

It'll sound very different polyphonically as well.

[Aryaroman]

That's a bit of a problem for me, because I can see the difference between the samples in my audio editor, but I don't really hear it. That is for dry synth samples, for recorded to tape samples I hear the difference.

It's now always what you think you hear. The lack of RRs can subliminally affect the way you perceive the sound. You might not notice it until it's missing.

[Saukar30]

It'll sound very different polyphonically as well.

+1

[Zookes]

I am unsure of this analog synth sampling.
Synth patches have sometimes much movement from free running osc, hardware quirk, and modulation also. Maybe not so practical or useful to RR sample this.

Much easier for realism to use sample modeling and maybe use wavetables plus noise quirks for variety than to sample RR. But for this effort, it is maybe more productive to model circuitry and do not use sampling?

[Shabdahbriah]

I am unsure of this analog synth sampling.[...]
Much easier for realism to use sample modeling and maybe use wavetables plus noise quirks for variety than to sample RR. But for this effort, it is maybe more productive to model circuitry and do not use sampling?

It's for a sampler which doesn't have any of those features.

[Aryaroman]

I am unsure of this analog synth sampling.[...]
Much easier for realism to use sample modeling and maybe use wavetables plus noise quirks for variety than to sample RR. But for this effort, it is maybe more productive to model circuitry and do not use sampling?

It's for a sampler which doesn't have any of those features.

While I get the fascination with sampling actual hardware synths, versus "a computer-modelled program", I kind-of agree with Shabdahbriah here. The samples might have their charm, but a sampler instrument is not that practical for extensive editing and a wide variety of playing the way an analog soft synth is, example, arpeggios and quick/polyphonic playing without RRs.

I like the way you can mix analog synth samples with the wavetables in Omnisphere. That way you can have the best of both worlds.

[SampleScience]

I am unsure of this analog synth sampling.
Synth patches have sometimes much movement from free running osc, hardware quirk, and modulation also. Maybe not so practical or useful to RR sample this.

Much easier for realism to use sample modeling and maybe use wavetables plus noise quirks for variety than to sample RR. But for this effort, it is maybe more productive to model circuitry and do not use sampling?

Yes, but I might as well code a new synth plugin while were at it. Funnily, I'm currently building a wavetable synth in Flowstone (Synthmaker).

It's for a sampler which doesn't have any of those features.

It's for various samplers; Reason NN-XT, Sforzando, Maize, Kontakt, EXS24, etc. I even make soundfonts! Since I'm an independent developer I need to cover a wider range of formats.