Filter envelope inconsistency across voices

[1000hz]

I'm observing lots of inconsistency on my P10 when playing the self-oscillating filter with the filter envelope knob turned up a bit. The envelope is applied unevenly across voices, resulting in the synth sounding pretty out of tune. This happens in both Rev 1/2 & Rev 3 filter modes, though is much more pronounced with mode 1/2. Running the tune procedure doesn't help tighten up the envelopes either. Is this normal? I should mention that the synth is otherwise in tune pretty well, including self-oscillating the filter with the filter envelope turned down to 0.

Here's how I'm testing:
- INIT patch
- turn off OSC A waveform
- set Filter to Rev 1/2
- turn Cutoff to 0
- turn Resonance to 10
- set Keyboard Tracking to Full
- turn Filter Envelope to ~4 (tune to A 440)
- play the same note repeatedly, cycling through the voices (round-robin mode)

Each voice plays a noticeable different pitch, up to a semitone sharp. I'm guessing this points to the filter sustain values having a bit of variance between the voice cards.

[Quatschmacher]

I'm observing lots of inconsistency on my P10 when playing the self-oscillating filter with the filter envelope knob turned up a bit. The envelope is applied unevenly across voices, resulting in the synth sounding pretty out of tune. This happens in both Rev 1/2 & Rev 3 filter modes, though is much more pronounced with mode 1/2. Running the tune procedure doesn't help tighten up the envelopes either. Is this normal? I should mention that the synth is otherwise in tune pretty well, including self-oscillating the filter with the filter envelope turned down to 0.

Here's how I'm testing:
- INIT patch
- turn off OSC A waveform
- set Filter to Rev 1/2
- turn Cutoff to 0
- turn Resonance to 10
- set Keyboard Tracking to Full
- turn Filter Envelope to ~4 (tune to A 440)
- play the same note repeatedly, cycling through the voices (round-robin mode)

Each voice plays a noticeable different pitch, up to a semitone sharp. I'm guessing this points to the filter sustain values having a bit of variance between the voice cards.

What is vintage mode set to? I think on an init patch it might not be fully on 4 (the most in-tune setting).

Have you also calibrated your pitch and mod wheel?

Have you populated your tuning tables with enough readings at different room temperatures over a few weeks?

When you hear inconsistency, do you notice 5 consecutive voices being in tune with each other then the next one differing but being consistent with the following 4? I found pulse width differs between voices on the two voice boards; the five on the mother card are consistent with each other and the five on the daughter cats are consistent with each other but voices between the two cards aren’t. Could be the same with filter tuning.

[Starkstrom]

I'm observing lots of inconsistency on my P10 when playing the self-oscillating filter with the filter envelope knob turned up a bit. The envelope is applied unevenly across voices, resulting in the synth sounding pretty out of tune. This happens in both Rev 1/2 & Rev 3 filter modes, though is much more pronounced with mode 1/2. Running the tune procedure doesn't help tighten up the envelopes either. Is this normal? I should mention that the synth is otherwise in tune pretty well, including self-oscillating the filter with the filter envelope turned down to 0.

Here's how I'm testing:
- INIT patch
- turn off OSC A waveform
- set Filter to Rev 1/2
- turn Cutoff to 0
- turn Resonance to 10
- set Keyboard Tracking to Full
- turn Filter Envelope to ~4 (tune to A 440)
- play the same note repeatedly, cycling through the voices (round-robin mode)

Each voice plays a noticeable different pitch, up to a semitone sharp. I'm guessing this points to the filter sustain values having a bit of variance between the voice cards.

In my experience this is normal with analog filters. I have dozens of analog filters in my modular system. They all track the keyboard differently and usually their intonation is far from perfect. After all this is not their primary function.

Using the filter at high resonance as a sine oscillator, something to experiment with:
Make sure the P5/10 is calibrated, filter tracking is set to “full” and the Vintage knob is set to ‘4’ - the most stable setting. Initially I would stay away from applying the filter envelope or at least set the filter envelope sustain to ‘0’. (A filter envelope can be added later if required.) It often helps to turn up the triangle wave of Oscillator B tuned to the same pitch as the filter (or an octave below). Then back down on the resonance until you get a usable mix of the osc triangle wave and the sine wave the filter is producing. In this case the resonance sweet spot is usually found between 12 and 3 o’clock. This way you get a stable fundamental while the sonic result is still reasonable close to a pure filter sine wave sound. It may also help to apply some resonance level compensation. There will alway be some “beating” as the filter sine wave and the osc triangle will never be in perfect tune. While counter intuitive, I often found that adding a touch of white noise in the mixer “stabilizes” the filter a bit for this type of sounds. The intonation of the resulting sound will not be perfect across the entire 5 octave keyboard but will hopefully be acceptable within 2 or 3 octaves. Often minute changes of the settings will produce noticeable changes in terms of intonation and timbre. Some experimenting and patience may be needed.

The previous settings are also a good starting point for filter FM sounds i.e. modulating the filter cutoff at audio range using the poly mod section and osc B as the mod source. In this application you may also want to try to switch to the rev 3 curtis filter as to my ears the rev 3 curtis filter is a bit more stable when doing filter FM.

I hope the above helps to get the sounds you are after. At least on my P10 I was able to create some in-tune sine, bell and gong type of sounds even for playing chords across the keyboard.

[dangerous]

Have you tried clearing the tuning completely, then running the tuning routine?

* Clear the calibration - hold "Record" and press "Tune"
* Turn it off and on again (not sure if necessary but can't hurt)
* Tune and cross fingers.

FWIW when I try your steps out, I get consistent notes from each voice but it doesn't track the keyboard very well, for instance I may get a minor third between the D-E.  Instead of adjusting the Envelope Amount, leaving it at 0 and adjusting the Cutoff itself gives much more usable results.  Basically as good and in tune across the octaves as the oscillators themselves.

[verstaerker]

i was wondering if the Filter envs are actually analog or digitally created? Do we have any verified info about it?

[Quatschmacher]

i was wondering if the Filter envs are actually analog or digitally created? Do we have any verified info about it?

The envelopes are generated digitally and converted to analogue control voltages.