[sdiy] jitter analysis

[Fredrik Carlqvist]

> We're after some effect, the measurment data is only helping
> us to understand. But it is not the real target.
> The target is to understand.

Absolutely! We must find out the following: 1) there is a difference in
jitter between different synths, 2) The jitter can be perceived by the human
ear, 3) The jitter has positive effect on the listening experience, 4) which
jitter model gives the best performance. We are starting on 1).

>>  It is
>> not possible to measure the jitter with any precision after 
>> sampling it at
>> 44100Hz. 

> why?

Because the distribution of the jitter is essential and to measure it needs
all the resolution one can get. Anders Friberg at KTH has done simulations
of choirs using jitter to improve the sound quality and according to him,
the jitter distribution must be just right. (BTW, he has done several
interesting research programs). When synthesizing the human voice, jitter is
an essential parameter, according to him.

> So , as long as no profound error in my reasoning can be found,
> it is not clear to me why such obvious effects like livelyhood
> of a wave should not appear in the usual 22050 Hz sampling
> bandwidth.

That the ear is band limited to 20 kHz or there about does not mean the
resolution at lower frequencies is limited to 50us steps. Well, it could be,
but it doesn't follow from the band limitation. So even a small jitter could
be (at least subliminally) perceptible. In my understanding (which could be
completely wrong), band limitation says nothing about the resolution within
the pass band.

> But the effect must be somewhere in the samples, otherwise
> it is getting esoteric.

Let's say we sample at 44.1kHz. Are we certain we can reproduce the acoustic
wave with 1us precision with a 23us sampling period? I am not so sure. I am
certain a sampling system can be used to perfectly (with respect to the
human ear) reproduce physical sound, but I am not so certain the 44.1kHz
sampling frequency is enough. The sound may be band limited. But not to
20kHz. In some cases, the difference is striking.

I must say I am glad someone else is interested in this. I just found out
that someone on KTH did a masters thesis on the topic of "differences
between analog and digital synthesizers". He focused mainly on the filters
though. There is also some extensive research on jitter in the human voice.
I will dig it out.


Fredrik Carlqvist