i can never seem to hear any change when i play with "width"
-
- KVRAF
- 4707 posts since 16 Mar, 2004 from Columbia, MD
Yeah, if you are using any standard osc it will simply increase stereo width on dual or quad. Width 0 = mono. Max width = one voice in each ear (though I suspect there's some phase/delay stuff going on, since normally if you pan two identical voices exactly the same way it just sounds like mono.) Also works on FM/Comb in the same way.
Shreddage 3 Stratus: Next generation Kontakt Player guitar, now available!
Impact Soundworks - Cinematic sounds, world instruments, electric guitars, synths, percussion, plugins + more!
Impact Soundworks - Cinematic sounds, world instruments, electric guitars, synths, percussion, plugins + more!
-
- KVRAF
- 4141 posts since 11 Aug, 2006 from Texas
Urs,
Zircon's post above and some playing around have me curious; what exactly do you do to a dual or quad osc when the width knob is increased? I can hear the sound getting "wider" but I was wondering technically you are doing to the sound. Is it phase and/or delay as Zircon speculated?
I guess the nerd in me just loves to understand.
Zircon's post above and some playing around have me curious; what exactly do you do to a dual or quad osc when the width knob is increased? I can hear the sound getting "wider" but I was wondering technically you are doing to the sound. Is it phase and/or delay as Zircon speculated?
I guess the nerd in me just loves to understand.
-
- KVRAF
- 4197 posts since 23 May, 2004 from Bad Vilbel, Germany
Of course it's panning... but what turns width into an audible effect is the phase relationship (delay would cause cancellation at certain frequencies) between the two/four oscillators...bmrzycki wrote:what exactly do you do to a dual or quad osc when the width knob is increased? I can hear the sound getting "wider" but I was wondering technically you are doing to the sound. Is it phase and/or delay as Zircon speculated?
A little experiment: Set an oscillator to Dual, Width to zero, Voice to "Few" and repeatedly play the same note. You can hear obvious "round-robin" voice allocation (each voice has a particular phase), the phase relationship repeats every 4 notes.
-
- KVRAF
- 4141 posts since 11 Aug, 2006 from Texas
So regarding the delay, I'm guessing it has to be quite small of a delay to avoid the echo effect but long enough to fool the brain?Howard wrote:Of course it's panning... but what turns width into an audible effect is the phase relationship (delay would cause cancellation at certain frequencies) between the two/four oscillators...
A little experiment: Set an oscillator to Dual, Width to zero, Voice to "Few" and repeatedly play the same note. You can hear obvious "round-robin" voice allocation (each voice has a particular phase), the phase relationship repeats every 4 notes.
I've noticed what you're mentioning in your experiment. However, I don't understand why the phase relationship is 4 and not 2. The picture below shows 2 saw waves out of phase by a bit and I only count 2 unique points of intersection:
What am I missing?
Sorry to turn this into a trig thread.
-
- KVRian
- 795 posts since 2 Apr, 2006
It round robins 4 waveforms. Each is the same saw with a different phaseofset. It doesn't matter how many voices you have. Example
First time you press a note, voice one starts at phase 0 and voice 2 at phase Pi (or 90 degrees), second time you press a note voice one starts at phase Pi/2 and voice 2 at phase Pi/3. Third time you press voice one starts at Pi/3 and voice 2 at zero and the fourth time you press voice one starts at Pi and voice 2 at Pi/2
The result is a round robin of 4 phasedifferences with 2 voices. As the phase of a waveform is continuous (but cyclic) you can get a round robin with as many phasedifferences as you want using only 2 voices.
First time you press a note, voice one starts at phase 0 and voice 2 at phase Pi (or 90 degrees), second time you press a note voice one starts at phase Pi/2 and voice 2 at phase Pi/3. Third time you press voice one starts at Pi/3 and voice 2 at zero and the fourth time you press voice one starts at Pi and voice 2 at Pi/2
The result is a round robin of 4 phasedifferences with 2 voices. As the phase of a waveform is continuous (but cyclic) you can get a round robin with as many phasedifferences as you want using only 2 voices.
-
- u-he
- 30213 posts since 8 Aug, 2002 from Berlin
There's no magic here...
Zeb has got 16 voices. Each voice has got 4 oscillators times 4 phase counters for quad mode ("unison oscillator"). At the beginning each of the 256 phase counters are initialized with a pseudo random number based on a simple algorithm.
If one doesn't detune, the pseudo-random phase offset of the 2/4 phase counters in dual/quad keeps constant, i.e. the frequency cancellations stay constant. They are perceived as different timbre among different voices.
Regardless of polyphony(few/medium/many) and mode(poly/retrigger/...), the 16 voices are always iterated. Thus you get a repetitive pattern of 16 timbres under above condition.
The width parameter distributes the up to 4 "unison oscillators" in the stereo field. In dual mode and at extreme width with centered pan, there are two identical signals at spatial phase offset. This isn't necessarily perceived as difference in timbre but as variation of panning.
It's actually quite easy... dual/quad is meant to be used with detune
Urs
Zeb has got 16 voices. Each voice has got 4 oscillators times 4 phase counters for quad mode ("unison oscillator"). At the beginning each of the 256 phase counters are initialized with a pseudo random number based on a simple algorithm.
If one doesn't detune, the pseudo-random phase offset of the 2/4 phase counters in dual/quad keeps constant, i.e. the frequency cancellations stay constant. They are perceived as different timbre among different voices.
Regardless of polyphony(few/medium/many) and mode(poly/retrigger/...), the 16 voices are always iterated. Thus you get a repetitive pattern of 16 timbres under above condition.
The width parameter distributes the up to 4 "unison oscillators" in the stereo field. In dual mode and at extreme width with centered pan, there are two identical signals at spatial phase offset. This isn't necessarily perceived as difference in timbre but as variation of panning.
It's actually quite easy... dual/quad is meant to be used with detune
Urs-
- KVRAF
- 4197 posts since 23 May, 2004 from Bad Vilbel, Germany
This is true when triggering from the arpeggiator. In all other modes, few/medium/many changes the number of notes before the timbre repeats. Maybe this wasn't intentional, but that's how it isUrs wrote:Regardless of polyphony(few/medium/many)...the 16 voices are always iterated. Thus you get a repetitive pattern of 16 timbres under above condition.