@lala It works! The only problem in your patch is that you don't compare the phases of the processed and unprocessed signals.
MfreeformPhase works differently, it looks like they're using an FFT and do the phase shifts in the frequency domain (please read about FFT, real/imaginary to amplitude/phase conversion and inverse FFT)
Because you're using a trigger level on the scopes which defeats the phase difference. If the scope had 2 Y inputs and you'd trigger by one signal threshold, you would see the two sinewaves shifted against each other.You can build a DIY 2-channel scope using an X-fader and square LFO, then disable the scope trigger.
Allpass is like a number of delays(infinite) each working on different band and length. (Google for phase response) It wont affect sine wave shape, but other will be altered.
Each waveform is a sum (mix) of sinewaves, look here, it shows mixing more and more sinewaves with speciffic amp, frequency (partial) and phase (made it for you ! :) If you pass it to spectrum analyzer you will see a partial after partial appearing (each bar in fft analyzer is that particular sinewave)
Allpass filter adds different delay do each of these component sinewaves. It sounds the same, but waveform looks different.
So it's a filter that mangles with phases of component sinewaves. If you use sinewave, its just a delayed sinewave :) If you use mix of 2... each of them has different delay.
Comments
8 views no answer
looks like I am not the only one who doesn’t know how to do it ^^
If by "any signal" you mean a sinewave at a certain frequency then just use a tuned all pass filter.
Other than that, imagine two sinewaves mixed, one at 440Hz and one at 523 Hz, how would you define shifting that signal by 50 degrees for example?
I hope that demonstrates the point 😉
By any signal I mean what ever audio I feed it...
i was trying to build a dimension c/d myself
but
no phase shift no dimension d
...
shift the phase compared to original signal can’t be voodoo magic
im sure it’s something simple, it’s in a god damned stomp box
There are approaches to apply a halfway constant phase shift over an acceptably wide frequency range (with limitations), for example:
Like I wrote, you can use all pass filters to implement frequency dependent phase shifting.
That all seems very involved?
for simple 4 steps 90 degree phase shifts ...
so I’m Feature requesting that hilbert transformer @giku
dimension d
triple chorus (ensemble)
...
i have things I want to build
AllPass filter
Nothing happens
What is the point that I am not getting?
@lala It works! The only problem in your patch is that you don't compare the phases of the processed and unprocessed signals.
MfreeformPhase works differently, it looks like they're using an FFT and do the phase shifts in the frequency domain (please read about FFT, real/imaginary to amplitude/phase conversion and inverse FFT)
Scope 1 shows output of osc
scope 2 shows after all pass
nothing happens
Of course, phase shifting does not change the signal other than its position in time.
So why do I not see another part of the window time frame in 2?
like sine not starting at zero crossing ?
that’s what I expect
Because you're using a trigger level on the scopes which defeats the phase difference. If the scope had 2 Y inputs and you'd trigger by one signal threshold, you would see the two sinewaves shifted against each other.You can build a DIY 2-channel scope using an X-fader and square LFO, then disable the scope trigger.
Now I get why nobody except you is doing that
way to involved
i even need to build ish to see what I am doing, pff
im off for some plugin shopping I guess
Wait, my Dimension clone is almost finished!
BTW, here's a demo that may help other Drambo users too:
:)
Aha, sometimes a pic is more telling
ok if you do dimension I’ll do ensemble
Damn, Im impressed with the trick you made to show 2 waves in oscilloscope :)
Allpass is like a number of delays(infinite) each working on different band and length. (Google for phase response) It wont affect sine wave shape, but other will be altered.
I don’t understand why a delay would not affect a sine but a triangle, but ok
Because phase processed by allpass is frequency dependent and any waveform other than a sinewave consists of more than one frequency.
@giku Thanks 😊
So this process needs more than one frequency
hm, I still don’t get why a delay would not offset a sine , but ok
i don’t need to understand everything I guess. :)
So if it’s a delay of some sort
the cutoff represents time
and reso does what?
tell me what i need to know,
allpass filters are still somewhat a mystery to me. 😅
@lala Just remember that the delay time depends on the input frequency.
Here's an article for you:
Each waveform is a sum (mix) of sinewaves, look here, it shows mixing more and more sinewaves with speciffic amp, frequency (partial) and phase (made it for you ! :) If you pass it to spectrum analyzer you will see a partial after partial appearing (each bar in fft analyzer is that particular sinewave)
Allpass filter adds different delay do each of these component sinewaves. It sounds the same, but waveform looks different.
So it's a filter that mangles with phases of component sinewaves. If you use sinewave, its just a delayed sinewave :) If you use mix of 2... each of them has different delay.
Resonance in APF does nothing.
@lala
More: https://www.desmos.com/calculator/qye3ybae9k
Use slider P to mangle with sinewave phases... thats what APF does. This waveform looks completely different, but sounds the same :)
We don't preceive phase in harmonic serries.
Aha That clears my mind.
thanks very much 🙏🏽
Maybe the whole discussion could be moved over to ‘how to’ section...
its impossible to move Threads from beta to public forum.
and it wouldn’t be to clever, we always talk about this and that what’s up next and what we are building, in some subclause.
besides this makes me look stupid 🤣
Just moved..