[sdiy] VCO Jitter, Slop ...

Fast TriggerFish fasttriggerfish at gmail.com
Tue Jun 11 11:40:12 CEST 2013

What if you started on the assumption that the sound of classic VCOs
can be closely replicated by feeding a perfect VCO into one saturation
stage followed by a linear filter and a third saturation stage ?
Then parameterise each section ( say input and ouput levels for 2*tanh
sections, and a 4 band parametric eq for the filtering stage ).
Feed a few periods of various perfect waveforms through it ( with
adequate oversampling ) at various key frequencies and regress the
parameters against classic VCO waveforms so that the output of the
final stage matches the classic waveform in a well chosen measure ( I
have to admit I'm not sure which one, maybe the least squares of the
ampitudes and phases of an FFT ? )
Maybe also regress some form of bleed of 50/60hz and harmonics into
the oscillators.
Then once you have regressed all your parameters for a set of key
frequencies, decide on a way to interpolate them.
Finally, check if the ears agree with your resulting VCO, and maybe
tweak things by ear a little.
If it works you may be able to store the final "regressed VCO" output
into tables so that you don't have to actually compute these stages
each time.

Just thinking aloud here, If I was a DSP developer with access to
hardware and wanted fat oscillators that's probably thing that I would
try first.
Does that sound reasonable ?


On 11/06/2013, Terry Shultz <thx1138 at earthlink.net> wrote:
> Hi Folks,
> I agree with the premise of setting up a test case and proving or
> disproving a hypothesis.
> Plotting results and commenting on what jitter or Phase distortion
> issues sound best.
> Listen to a piano and monitor with a strobe tuner. You will see that
> there is modulation between strings. Some is okay but at what point does
> it interfere with performance on an instrument.
> I tuned pianos on occasion for extra money and voicing is a bit of art
> of listening.
> If you tune a piano by way of strobe tuner, the piano sounds terrible.
> Use a tuning fork and use your ears and the tuning will usually sound
> better as all pianos are not exact in nature. I am sure this is well
> known by our list of folks. Determining if a Digital Osc. vs. Analog VCO
> is better or worse is subjective to the listener.
> I find it interesting that many on this list are preferring Analog VCO
> design. I can easily build reliable and stable and perfectly in tune on
> my DSP or ARM products. I found that they sounded sterile in most cases
> and still enjoy going back to my analog synths.
> I found synths having a lower frequency modulation made a much fuller
> sound and fatter sound .
> I was playing with my friends synclavier a few weeks back and it never
> gave me the same lower end phase modulation as my E-mu/Moog/Oberheim
> vco's do. You can try all types of experiments and jitter redux, but in
> the end your ears will tell you what you need for a particular effect.
> I saw a note earlier about jitter and such. I tend to think jitter may
> be so small that I doubt you can notice it. I may be wrong, It might be
> a good experiment to try with a blind series of tests with folks and
> perhaps Dr. Diana Deutsch at UCSD can comment on this the next time I
> see her whilst I am in San Diego next.
> (I am in Germany again) interviewing for a new position and will do some
> testing when I return back to my office later next week.
> Phase modulation and Psychoacoustic illusion's are Dr. Deutsch's work
> area . Shepard tones etc.
> http://deutsch.ucsd.edu/psychology/pages.php?i=101
> Hopefully I am not too off topic as I cannot sleep well yet do to time
> change and so forth.
> best regards to you all,
> Terry Shultz
> On 6/10/2013 3:00 PM, Richard Wentk wrote:
>> I've experimented with most of these features using digital synthesis.
>> It's actually easy to add individual frequency drifts to individual
>> overtones with a controllable mod rate/depth using (say) SuperCollider.
>> I haven't found any reliable way to make a digital oscillator sound fat
>> yet. I can make some awesome stacked oscillator sounds, especially if each
>> note has a different detuning with an exponential distribution around the
>> nominal pitch. But 'fat' - not so much.
>> There are a few things I haven't tried, but I'm fairly sure that bandpass
>> filtered noise modulating the frequency won't transform a digital sound
>> into an analog-like one.
>> My suspicion is that in addition to the tanh distortion generated in a
>> filter, analog op-amps added a lot of colour with intermodulation
>> distortion and limited bandwidth. A lot of early analog circuits use poor
>> (by modern standards) op amps, and the sound isn't at all clean or simple.
>> It's also not very bright - often there's not much happening at the top
>> end.
>> Simulating these effects requires some fairly scary Volterra expansions,
>> so I haven't gotten very far with modelling it yet.
>> It's also worth remembering that analog waveshapes are often non-textbook.
>> A cubic sine approximation has a lot more character than a perfect clean
>> function generator sine. The sawtooth in the early Moog 901s was nothing
>> like a good saw. The rounded shape varied with frequency and added extra
>> emphasis on the low partials, which made the sound bigger and warmer.
>> And so on. Basically perfect VCOs sound boring. And a lot of early designs
>> were well short of perfect - in interesting ways.
>> Richard
>> On 10 Jun 2013, at 22:07, Tom Wiltshire <tom at electricdruid.net> wrote:
>>> I totally agree with Bruce. "Fatness" is a concept related to multiple
>>> oscillators. I don't think the human ear  (most people's anyway) have
>>> anything like the accuracy required to hear a little drift in a single
>>> oscillator.
>>> I also totally agree with Brian. We *really* need to test some of these
>>> hypotheses to get this nailed down once and for all. There is a
>>> widespread perception that analog oscillators (and some more than others)
>>> have some level of drift or noise inherent in their design that improves
>>> their sound, particularly with other oscillators. Digital technology has
>>> reached a point where we could/can model this property *if we can
>>> identify it*. Plenty of effort has already been made in this direction.
>>> My own experiments in this area have shown me that a little linear phase
>>> noise in a waveform isn't a bad thing, and that a little linear frequency
>>> wobble (if that's a different thing, which it only just is) helps too.
>>> Whether these effects are what's going on in "proper" analogue
>>> oscillators I've never been able to show. I only have CEM3340 oscillators
>>> around, and they don't show these effects, as far as I can determine.
>>> One important thing to remember is that even quite minor shifts in phase
>>> have a significant audible effect on the higher harmonics if you mix a
>>> pair of oscillators. Even if the main pitches remain very close, the
>>> highest harmonics will drift in and out of phase. We can easily hear
>>> this, especially on the incredibly harmonically rich synth waveforms
>>> we're typically talking about. Thus you can put in a little bit of
>>> "motion" whilst barely affecting the pitch.
>>> T.
>>> On 10 Jun 2013, at 21:25, BRUCE DUNCAN <modcan at sympatico.ca> wrote:
>>>> I don't think VCOs on their own can can increase their levels of obesity
>>>> by drift or unstable tuning or whatever. Probably only when multiple
>>>> VCOs
>>>> are played together that these attributes become interesting.
>>>> One of JHs main points was how Moog VCOs are more detuned in the low
>>>> ranges
>>>> adding animation in the bass notes while keeping higher frequencies more
>>>> or less in tune.
>>>> This does seem to work as I have tried it on a digital triple VCO with
>>>> good results.
>>>> It does sound fat to my ears anyway.
>>>>> From: tom at electricdruid.net
>>>>> Date: Mon, 10 Jun 2013 21:10:09 +0100
>>>>> To: aym-htnys at teaser.fr
>>>>> CC: Synth-diy at synth-diy.org
>>>>> Subject: Re: [sdiy] VCO Jitter, Slop ... (was: Re: Smith's Evolver
>>>>> Desktop's spikey VCO waveforms)
>>>>> JH argued that linear drift/offset was a significant part of "fatness"
>>>>> in oscillators. A detune of a few 1/10ths of hertz gives motion,
>>>>> without sounding out of tune. Doing this with a V/Oct CV leads to
>>>>> errors becoming increasing large as you go up the octaves, rapidly
>>>>> becoming unpleasant.
>>>>> http://www.jhaible.com/living_vcos/jh_living_vcos.html
>>>>> So however much drift you apply, perhaps it should be linear FM, rather
>>>>> than into an exponential FM input.
>>>>> T.
>>>>> On 10 Jun 2013, at 17:57, Andre Majorel <aym-htnys at teaser.fr> wrote:
>>>>>> On 2013-06-09 03:00 +0200, Michael Zacherl wrote:
>>>>>>> On 9.6.2013, at 00:11 , Andre Majorel wrote:
>>>>>>>>> VCOs suck. That's why they're interesting.
>>>>>>>> I can't prove it but I'm pretty sure it's their jitter that
>>>>>>>> makes VCOs interesting, more than their waveforms. For human
>>>>>>>> ears, all waveforms are equally boring if perfectly periodic.
>>>>>>> if you had to put it into numbers, what would be the amount of
>>>>>>> jitter, you think?
>>>>>> Hi Michael. Don't know. I experimented with modulating the
>>>>>> frequency of an MS-20 VCO with white and pink noise. As the FM
>>>>>> sensitivity knob was turned clockwise, it went from no audible
>>>>>> effect to unpleasantly warbly. I couldn't find a setting that
>>>>>> made it sound "even fatter" or anything like that.
>>>>>> My guess is that there is a rather narrow range of jitter
>>>>>> between "sterile and brittle" and "blurry and weak" and by a
>>>>>> happy accident, the state of electronics in the sixties and
>>>>>> seventies led to VCOs which fell right in the middle.
>>>>>> Recently read a very interesting bit about how much mains
>>>>>> frequency ripple there is on power rails inside synths. Wish I
>>>>>> could remember where and who.
>>>>>> Some years ago, Scott Gravenhorst posted here an MP3 of one of
>>>>>> his digital synth projects. I thought it sounded surprisingly
>>>>>> good (i.e. not DCO-like). He said he used a small amount of pink
>>>>>> noise FM on the oscillators. Maybe he could tell us more.
>>>>>>> et al ... and, in general, something like a slop parameter on
>>>>>>> DSI synths for their oscillators, is that just for the
>>>>>>> frequency and how much? Never had a play with it.
>>>>>> Since you can hear the difference between VCO and DCO on a
>>>>>> single note, I don't see how temperature drift & scaling errors
>>>>>> could be the source of "fatness". They may contribute to it, but
>>>>>> they're not the essential ingredient.
>>>>>> --
>>>>>> André Majorel http://www.teaser.fr/~amajorel/
>>>>>> _______________________________________________
>>>>>> Synth-diy mailing list
>>>>>> Synth-diy at synth-diy.org
>>>>>> http://synth-diy.org/mailman/listinfo/synth-diy
>>>>> _______________________________________________
>>>>> Synth-diy mailing list
>>>>> Synth-diy at synth-diy.org
>>>>> http://synth-diy.org/mailman/listinfo/synth-diy
>>>> _______________________________________________
>>>> Synth-diy mailing list
>>>> Synth-diy at synth-diy.org
>>>> http://synth-diy.org/mailman/listinfo/synth-diy
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at synth-diy.org
>>> http://synth-diy.org/mailman/listinfo/synth-diy
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at synth-diy.org
>> http://synth-diy.org/mailman/listinfo/synth-diy
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy

More information about the Synth-diy mailing list