Phase shift
- KVRist
- Topic Starter
- 87 posts since 19 Jan, 2019
Good Afternoon, Friends. Please tell me why when playing the same note sequentially on a virtual synthesizer, the phase of the instrument shifts on each subsequent note. And whether it is possible to fix the phase?
Synthesizer without any modulation and processing.
Synthesizer without any modulation and processing.
-
parallel loners parallel loners https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=444039
- KVRer
- 20 posts since 6 Jul, 2019
- KVRian
- 1361 posts since 17 Jul, 2007 from Riversland Valhalla
Most of the synths meant to mimic the vintage analog ones, where the oscillators drifts in phase indeed.
Although some synth plugins offers retrig phase, or putting a fixed rate.
Such as:
Sylenth1, Spire, Dune, Zebra, Hive to name a few.
Although some synth plugins offers retrig phase, or putting a fixed rate.
Such as:
Sylenth1, Spire, Dune, Zebra, Hive to name a few.
- KVRAF
- 15263 posts since 8 Mar, 2005 from Utrecht, Holland
If you play notes sequentially, one after the other, how do you know the phase has shifted (or not)
We are the KVR collective. Resistance is futile. You will be assimilated.
My MusicCalc is served over https!!
My MusicCalc is served over https!!
-
- KVRAF
- 2564 posts since 2 Jul, 2010
Many synths do offer control over this, some do not.
The origin is that in an analogue synthesiser the oscillator is usually running all the time in the background. When you press a key, the pitch is adjusted and the amplifier opened up so you can hear it. So the starting phase is a function of the oscillator frequency and the exact time you press a key.
Electronic producers who want an ultra-consistent phase behaviour generally achieve it with sampling.
The origin is that in an analogue synthesiser the oscillator is usually running all the time in the background. When you press a key, the pitch is adjusted and the amplifier opened up so you can hear it. So the starting phase is a function of the oscillator frequency and the exact time you press a key.
Electronic producers who want an ultra-consistent phase behaviour generally achieve it with sampling.
-
parallel loners parallel loners https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=444039
- KVRer
- 20 posts since 6 Jul, 2019
So a virtual synth shouldn’t generate phase shift unless it’s added on purpose to make it sound more analog? It seems like an odd thing to put in; is it even possible to hear the difference?imrae wrote: ↑Tue Jul 09, 2019 6:35 am Many synths do offer control over this, some do not.
The origin is that in an analogue synthesiser the oscillator is usually running all the time in the background. When you press a key, the pitch is adjusted and the amplifier opened up so you can hear it. So the starting phase is a function of the oscillator frequency and the exact time you press a key.
Electronic producers who want an ultra-consistent phase behaviour generally achieve it with sampling.
- KVRian
- 1361 posts since 17 Jul, 2007 from Riversland Valhalla
Yes, VA developers intend to add this as a part of emulating real analog synths
-
- KVRAF
- 2564 posts since 2 Jul, 2010
It isn't necessarily any more effort. Say you have a simple ramp generator used to move through a wavetable: that could be naively implemented as (t modulo period) where "period" is the time it should take to sweep through the waveform once.parallel loners wrote: ↑Thu Jul 11, 2019 10:44 amSo a virtual synth shouldn’t generate phase shift unless it’s added on purpose to make it sound more analog? It seems like an odd thing to put in; is it even possible to hear the difference?imrae wrote: ↑Tue Jul 09, 2019 6:35 am Many synths do offer control over this, some do not.
The origin is that in an analogue synthesiser the oscillator is usually running all the time in the background. When you press a key, the pitch is adjusted and the amplifier opened up so you can hear it. So the starting phase is a function of the oscillator frequency and the exact time you press a key.
Electronic producers who want an ultra-consistent phase behaviour generally achieve it with sampling.
To start the oscillator from a fixed phase, you need to work out the time (t) since the key was pressed. For free-running, you could just use the master/system clock.
-
- KVRian
- 629 posts since 15 Jun, 2017
In virtual synths the oscillator can be either always restarted or some form of free running (or random phase) is implemented. Synths may offer a choice between the two.parallel loners wrote: ↑Thu Jul 11, 2019 10:44 am So a virtual synth shouldn’t generate phase shift unless it’s added on purpose to make it sound more analog? It seems like an odd thing to put in; is it even possible to hear the difference?
Can you hear the difference? Well, if it results in a relative phase *difference*. E.g. oscillator 1 runs out of phase/sync with oscillator 2 *and* the difference in phase is not static (but "random"). Or the difference is "static", but can be modulated (or set as a "static" parameter).
As said phase difference can be the result of 2 freerunning oscillators (in analogue gear seldomly at the exact same and stable frequency/phase), where the gate (to amp) is opened whenever you press a key and you get whatever phase the oscillator is running at at that moment.
But you can also get phase issues if the oscillators are restarted at every keypress *AND* the notes overlaps. Newly started note will most likely not be in phase sync with the still running note. Easy to check: repetitively striking the same key (on a patch with some release time) will then result in phase issues.
A phase shift can be considered to be a delay effect, causing comb filter effects. Any detuning can be considered a form of (cyclic) phase shifting too.
The Delay Spectrum - comb filtering
https://www.youtube.com/watch?v=dQ5DEdO9ELg