[sdiy] Time-varying all-pass filters

ulfur hansson ulfurh at gmail.com
Tue Jan 18 13:38:41 CET 2022


super interesting! 

thanks so much for sharing, i for one am very excited to follow your progress and hope you figure out how to achieve deeper modulation/headroom with your apf!

best,
-ú

sent from outer space

> On 17 Jan 2022, at 20:55, Rutger Vlek via Synth-diy <synth-diy at synth-diy.org> wrote:
> 
> 
> Hi all,
> 
> Thanks for the helpful thoughts and suggestions. I wanted to share that I made some progress in understanding the circuit (an audio rate modulated, tracking all pass filter). Thanks to Ulfur I started thinking about how the APF approach relates to (sinusoidal) wavefolding done in the SineCore, and realized that the APF has no mechanism to deal with wrapping around of the phase (jumping to 0 after exceeding 360 degrees). While in the static, unmodulated case, the APF can achieve a 0 to 180 degrees shift without the need for wrapping, in the dynamic case the combination of a rapid change in modulation signal and a certain integrator state, could call for wrapping around, which... is not happening. Instead the signal is moving outside of the typical 10Vpp range. While in the example graph I sent it looks like a DC shift or LF component, this is not always the case. So DC blocking, or adding of modulation signal to the output is not really helping. Compared with normal PM of the same depth the APF approach preserves much more low end in the signal, while introducing a rather similar set of harmonics. So I will stick with it and try to work around the headroom issue.
> 
> Rutger
> 
> 
> 
> 
> 
> 
> Op vr 7 jan. 2022 15:06 schreef cheater cheater <cheater00social at gmail.com>:
>> Have you considered adding some of the CV to the output, or possibly subtracting? And then, as Lloyd suggested, follow that up with AC coupling.
>> 
>> On Fri, Jan 7, 2022 at 12:41 PM Rutger Vlek via Synth-diy <synth-diy at synth-diy.org> wrote:
>>> Thanks for the suggestion Ulfur! Do I assume correctly that it comes down to a mathematical equivalence of wave folding and phase modulation?
>>> 
>>> In reply to others: I see that my original question may not have been phrased clearly enough. It's not about making the frequency of an all-pass filter change in time, which I know how to do. The problem is when you change the frequency at audio rate. The figure from the paper I referred to in my original post describes the problem well. It shows how a rapid modulation of the APF frequency (top graph) leads to a low-frequency component in the output (middle graph), which could be avoided (bottom graph). The analog APF I am using acts identical to the middle graph, leading to (unnecessary?) headroom issues, while true PM via the oscillator directly would never cause this.
>>> 
>>> Despite headroom issues, the audio-rate modulated analog APF still sounds cool, by the way!
>>> 
>>> https://d3i71xaburhd42.cloudfront.net/dd2e6eff1c49d49479c0943615ebc36a19f4fa0d/3-Figure2-1.png
>>> 
>>> 
>>> Rutger
>>> 
>>> Op do 6 jan. 2022 om 22:58 schreef ulfur hansson <ulfurh at gmail.com>:
>>>> hey rutger!
>>>> 
>>>> i have had amazing results with openmusiclabs sinecore design. I've changed it a lot for my own personal needs, but the basic circuit is all there and presented in a very informative way :) recommend checking it out!!
>>>> 
>>>> very best,
>>>> -úlfur
>>>> 
>>>> fim., 6. jan. 2022 kl. 14:16 skrifaði Rutger Vlek via Synth-diy <synth-diy at synth-diy.org>:
>>>>> Hi Neil,
>>>>> 
>>>>>  Yes that schematic on page 22 describes very well what I'm doing, but with an audio rate modulation signal added to the tracking cv of the all pass filter. The problem occurs when input signal peaks and is added to twice the output signal of the low pass filter when IT peaks. That can happen when receiving a sudden change in the modulation CV.
>>>>> 
>>>>> Rutger
>>>>> 
>>>>> Op do 6 jan. 2022 13:08 schreef Neil Johnson <neil.johnson71 at gmail.com>:
>>>>>> Hi Rutger,
>>>>>> 
>>>>>> > Then it occurred to me that, assuming the input is a sine wave, a phase shifted version up to 180 degrees could be obtained from modulating an all-pass filter that is also tracking along with the pitch of the oscillator.
>>>>>> 
>>>>>> See page 22:
>>>>>> https://www.soundsemiconductor.com/downloads/ssi2130datasheet.pdf
>>>>>> 
>>>>>> Implementation here:
>>>>>> https://www.youtube.com/watch?v=bq-2UQ6zkSU
>>>>>> 
>>>>>> That was modulated at a slow (manual) rate, so no idea what it does at
>>>>>> faster rates.  Interesting question though....
>>>>>> 
>>>>>> > ps. Happy healthy new year!
>>>>>> 
>>>>>> Happy new year to you too!
>>>>>> 
>>>>>> Cheers,
>>>>>> Neil
>>>>> _______________________________________________
>>>>> Synth-diy mailing list
>>>>> Synth-diy at synth-diy.org
>>>>> http://synth-diy.org/mailman/listinfo/synth-diy
>>>>> Selling or trading? Use marketplace at synth-diy.org
>>>> 
>>>> 
>>>> -- 
>>>>>>>> ulfurhansson.com
>>>> instagram.com/ulfur_
>>>> .•°~ Úlfur - Arborescence on Spotify
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at synth-diy.org
>>> http://synth-diy.org/mailman/listinfo/synth-diy
>>> Selling or trading? Use marketplace at synth-diy.org
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> Selling or trading? Use marketplace at synth-diy.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20220118/4a599352/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 3-Figure2-1.png
Type: image/png
Size: 11840 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20220118/4a599352/attachment.png>


More information about the Synth-diy mailing list