Bug - NS4 OSC ENV with ENV to PITCH

[ns4decode]

I've had my Stage 4 for about a week now and have been digging into the synth quite a bit. As I was putting together a synth preset to morph a pitch change to the control pedal using the OSC ENV to PITCH (posted here), I found that the pitch of the ENV ATTACK starting point changes when moving from an ENV DECAY of 43s to infinity; it's a dramatic change actually. Using a negative ENV AMT, one would think that if it takes 43s for the pitch to move from the negative setting to 0, that switching to a DECAY of infinity would cause the pitch to never get to 0 and stay forever at the negative ENV AMT. However when DECAY is set at infinity, the starting pitch of the ATTACK is changed so that the final state of the attack is 0, rather than the negative ENV AMT. The same problem is apparent when using a positive ENV AMT.

Playing around with the OSC ENV Decay related to a SYNC wave without ENV to PITCH selected, I found that the OSC ENV DECAY is implemented correctly on the SYNC wave when moving between the 43s setting and Infinity setting for both negative and positive ENV AMT. That is not the case for the ENV to PITCH implementation.

In the video I demonstrate the change between a sound with OSC ENV DECAY set at 43s and set at infinity. I also show the problem and how the sound of the current implementation of the infinity setting can be achieved so nothing is lost if Nord implements a true infinity setting for the ENV DECAY.

[ns4decode]

I posted a systematic description of the behaviors of the envelope-amount knobs (filter and oscillator) here: viewtopic.php?t=25972

It's consistent with your observations.

I wonder if the difference between finite and infinite decay times with env-to-pitch might be deliberate, not a bug. The way it is currently, it allows the final pitch to be controlled precisely (because it's the pitch of the note being pressed). If they didn't make the infinite-decay case behave differently than the finite-decay case, then setting the final pitch precisely in the infinite-decay case might be very difficult if not impossible.

[ns4decode]

Very nice analysis! Feedback from Nord Support in the U.S. is that the current implementation is by design to get the upward or downward move to Pitch, however as I put at the end of my video, that's achievable without the needing to implement the infinity setting as it is today.

[ns4decode]

Oh, you're right. I missed the significance of that last part of the video the first time I watched it. I withdraw my "very difficult if not impossible" statement.

I initially thought that since the minimum allowed attack time (half a millisecond) isn't quite zero, that very quick initial drop in pitch might have a perceptible effect on the sound in some cases, but after I saw your reply I did a few quick experiments with a saw waveform (rich in harmonics) and with a sine wave (no harmonics), and in both cases I couldn't hear any effect of that half-millisecond transient. If we can't hear it in a carefully controlled lab setting, then we can rest assured that nobody will hear it in a live performance, either.

The numbers are fun, too. If the nominal pitch is less than 2 kHz (roughly C7), then a half-millisecond transient occurs in less than the duration of a single oscillation of the waveform, and on that timescale concept of "pitch" isn't even meaningful.

[ns4decode]

On second thought (or third thought?), even though the initial transient is imperceptible, the shape of the subsequent slower rise is definitely perceptible.

Here's a setup that demonstrates the effect. Turn off all layers except synth layers A and B, and use these settings for both synth layers:

analog type (knob 1) = ANALOG
analog cat (knob 2) = PURE
analog wave/partial (knob 3) = Pure Saw
osc ctrl = 0.0
osc env-to-pitch on/off = on
osc env velocity on/off = off
LFO target = nothing
amp env attack = 3.0 ms (minimum)
amp env decay = inf
filter on/off = off

Use these settings for synth layer A:

osc env attack = 0.5 ms
osc env decay = 1.05 s
osc env amt = -5.5

Use these settings for synth layer B:

osc env attack = 1.02 s
osc env decay = inf
osc env amt = +5.5

(Notice the opposite signs of the osc env amt.) Now, solo one layer and play a note. The pitch starts at lower frequency and then rises to the nominal pitch of the note being played, but the shape of the rise is different depending on which layer is being solo'd:

* Layer A rises more rapidly at first and then carefully settles into the nominal pitch, like an airplane coming in for a smooth landing.

* Layer B rises at a constant rate and abruptly stops rising when it hits the nominal pitch. I wouldn't want to be a passenger on that airplane, but it could be musically useful.

The way Nord implemented it allows the user to choose either of these two riser-shapes. Don't know if that was their motive.

[WannitBBBad]

Thanks for looking into it further - that's a very good observation about the Attack versus Decay characteristics. I like the smooth landng on the Decay. The current functionality works fine for short-duration smooth pitch changes, but having a true infinity setting and a little more flexibility in the setting of the pitch (increment steps for Env Amt not exact to pitch) would be perfect. Thanks again.

[Mr. Marko]

IMO this OSC section of the NS4 appears to be still stuck in beta stage of refinement...

Timbre Mod of Digital Waves still cannot be modulated by OSC CTRL / ENV AMT or LFO OSC CTRL like the NS3 was able to...

ENV to Pitch works...but no Timbre Mod like the NS3.

All you can currently do on NS4 is FILTER a Digital Wave...introduced Feb of 2023 and things like this are still not working.

When?