[sdiy] Anyone do a voltage controlled "Tilt" Filter?

[Richie Burnett]

Crossfading between the outputs of a 1st order low-pass filter and 1st order 
high-pass filter (both with fc=1kHz) gives the responses shown in TILT1.PNG 
attached.  Whilst it does approximate a "tilt" response, it's probably not 
quite what you want.  There are no phase cancellation or notches provided 
neither of the two filters inverts the signal, and they add to a perfectly 
flat "wire" when they're mixed in equal amounts, but the centre of the 
response drifts to the left and the right as the mixing ratio is changed. 
It doesn't always cross 0dB at the same place.

The correct tilt response is shown in TILT2.PNG.  Notice how the frequency 
response stays centred around the middle of the audio band and always 
crosses 0dB at 1kHz in this example.  In order to get this response the 
positions of the poles and zeros in the filters have to move.  (This is 
unlike the previous example where both filters had fixed 1kHz cutoff 
frequencies.)  When the pole and the zero are both at 1kHz they cancel each 
other out and the resulting response is flat.  Move the pole down below 
1kHz, and the zero up by the corresponding amount and you get a tilted 
response with LF boost followed by HF cut.  Conversely, if the zero comes 
first you get a response with LF cut followed by HF boost.  That is how I 
generated those response curves.  Mathematically.  With no thought to what 
hardware is needed to implement the transfer functions!

So, you can get a kind of tilt response just by mixing the outputs from 
1-pole LPF and HPF but it will move from side-to-side which might or might 
not be acceptable for your application.  To get a perfect balanced response 
will take something a bit more clever.

Hope this helps,

-Richie,


> Neil Harper wrote:
>
> my last guess was wondering if an equivalent circuit could be made by 
> combining a lowpassed and hipassed audio signal and then crossfading them. 
> I can imagine it would somewhat similar, but I'm not sure if you would get 
> the same totally flat frequency response in the middle setting.


> ColinMuirDorward wrote:
>
> As for x-fading between high and low shelves, you might have phase 
> problems and get a bit of notch filtering, no?


-- 
This email has been checked for viruses by AVG.
https://www.avg.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tilt1.png
Type: image/png
Size: 31227 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20210531/ec0e7761/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tilt2.png
Type: image/png
Size: 31248 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20210531/ec0e7761/attachment-0001.png>