Nature and purpose of the Serge 'S/H source'

[windchill]

Please forgive the, probably stupid, Serge-noob question.... Can anyone enlighten me as to what the 'S/H source' on the Serge Noise Source and Random Source actually is? I know it's a sort of noisy saw wave, but what is the nature and purpose of this special 'S/H source' as opposed to, say, white noise? Is it just about obtaining a specific statistical distribution of values?

The reason I ask is because I was reading about the Random Voltage Generator patch (http://www.serge-fans.com/wiz_SSG3.htm) and wondered if I could do the same on a (R*S Eurorack) SSG if there is no access to the proper S/H source.

[cycad73]

It's probably getting a uniform rather than Gaussian distribution of values.

[syncretism]

Please forgive the, probably stupid, Serge-noob question.... Can anyone enlighten me as to what the 'S/H source' on the Serge Noise Source and Random Source actually is? I know it's a sort of noisy saw wave, but what is the nature and purpose of this special 'S/H source' as opposed to, say, white noise? Is it just about obtaining a specific statistical distribution of values?

I'm not an engineer, but Serge's noisy saw sounds similar to Mattson's "Slow Random" output; you can see screenshots of white, pink and slow noise here.

In both cases, the end result looks closer to red noise. I don't have the SSG in front of me, but I recall it sounding brighter than red noise. Random*Source's documentation describes S/H as "an ideal input for Sample & Hold functions to produce random voltages of equal probability" and "similar" to 1/f (pink noise). The only way I can reconcile that quote with what I see and hear is to interpret this as random fluctuations in voltage of equal probability in relation to the current value. In practical terms, these outputs seem to generate more "constrained" voltages than, say, sampling white noise. They sound more deterministic, even if they're not.

I wish I had a better grasp of this stuff.