inconsistent strange voice behavior on new REV2

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
inconsistent strange voice behavior on new REV2
« on: May 26, 2018, 01:53:27 PM »
Hi, im new to this Forum aswel as to dsi synths,so first of all: nice to meet you all!

Yesterday I bought my prophet REV2 brand new from a shop here in the netherlands *europe.
Started out from an init patch and decided to recreate a simple patch i originally made on my juno60.
(as a way to learn where all the buttons are on the REV2).

This is when I noticed strange behavior when re-triggering the notes on de REV2.
each time i retrigger a key (cycling the voices): the sound changes
*it sounds a bit like Sample and hold LFO on the filter... but there is nothing like that active.


I figured this could be solved by calibrating the voices and filter.... but that didn't help!
then i found out i had an old operating system (came straight out of the box) ... so i updated and recallibrated
.... No result.

i mean... i understand its analog...and voices might sound slightly different each time you play ....but not this much right?

I made a short video... but cant get it to upload on here ....
so i will convert it to a sound file... *bad iphone recording ,... but the problem is so obvious that you'll hear it anyway :)
It's basically me hitting a key and sustaining the note.. then hitting the same key over and over on the keyboard... this is where you can hear the tonal differences between voices "problem" (there is no LFO active)


Maybe it's just something i havent figured out on the REV2 yet ? ... some setting orso? [/b]
Hope you guys can help me out.

Many thanks.
« Last Edit: May 26, 2018, 02:00:03 PM by elektrofunk »

Re: inconsistent strange voice behavior on new REV2
« Reply #1 on: May 26, 2018, 02:07:51 PM »
Sorry if this sounds overly basic but it's good to start there...What is your filter and amp Velocity set at? Based only on the clip it sounds like it could just be Velocity modulation. Do you still get that effect when Velocity is at 0?

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #2 on: May 26, 2018, 02:42:27 PM »
Thanks for the quick reply.

No velocity modulating the filter.
no LFO's active *there is no destination assigned to any LFO.
There is an envelope modulating the OSC 1 shape...and some filter action via envelope

i loaded the patch in the software... this way i could make a graphic overview of the patch....
am i missing something?
« Last Edit: May 26, 2018, 03:00:48 PM by elektrofunk »

Re: inconsistent strange voice behavior on new REV2
« Reply #3 on: May 26, 2018, 03:00:51 PM »
Hello and welcome.

From looking at the picture, I'd say that the filter envelope attack, decay, and release times are so high that some voices might not have time to reach the sustain stage, and keep rising the cutoff each time you press a key (but differently for each voices, since there's a separate envelope for each one).

If you turn those 3 values (attack and decay and release for the filter envelope) down to 0, does that removes the strange behavior between voices ?
« Last Edit: May 26, 2018, 03:02:34 PM by AlainHubert »
Oberheim OB-X8, Minimoog D (vintage), OB6 (Desktop), Oberheim Matrix-6 (MIDI Controller for OB6), VC340

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #4 on: May 26, 2018, 03:44:06 PM »
hi alain, nice to meet you and thank you for your advice.
It does make a major difference. Mostly the release time all the way to 0 does the trick!

But with the attack and decay almost all the way down i still get noticeable differences between the voices. (almost gone but still annoys me) Its the 8 voice version  ... and i can hear that its always the same voice (each 4th note when cycling voices... in split mode) sounds different.

oftopic ... i dont understand why my Juno 60 has no problem with a similar patch... even with the ADR far open?

Maybe it's a good idea to share the patch data so someone can load the patch in his / her REV2... to see if its a hardware related problem? or is this normal for a REV2? *i would say that my REV2 doesn't calibrate the voices perfectly?




Re: inconsistent strange voice behavior on new REV2
« Reply #5 on: May 26, 2018, 03:50:49 PM »
hi alain, nice to meet you and thank you for your advice.
It does make a major difference. Mostly the release time all the way to 0 does the trick!

But with the attack and decay almost all the way down i still get noticeable differences between the voices. (almost gone but still annoys me) Its the 8 voice version  ... and i can hear that its always the same voice (each 4th note when cycling voices... in split mode) sounds different.

oftopic ... i dont understand why my Juno 60 has no problem with a similar patch... even with the ADR far open?

Maybe it's a good idea to share the patch data so someone can load the patch in his / her REV2... to see if its a hardware related problem? or is this normal for a REV2? *i would say that my REV2 doesn't calibrate the voices perfectly?
The Juno 60 only has one envelope controlling the amp and filter, whereas the Rev 2 has one envelope for filter, one for amp, and one aux envelope. Your pic shows that the filter envelope settings are almost maxed whereas the amp envelope settings are 0 or minimal. So it's going to sound/behave quite differently than the Juno patch unless you match the envelopes. In fact you might want to set both to be about the same values and see if you get normal voice behavior.

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #6 on: May 26, 2018, 04:42:18 PM »
Thanks again guyaguy.

While i was doing that ... i noticed that every 4th time i hit a note, there was something sounding different...
i turned of osc2 and found that osc 1 was the problem. I removed the sub osc from osc1 ... and i found out that the 4th note played on this osc didnt make any sound at all (the sub osc did... but the raw pulse wave of osc1 sound didnt)

did a fresh calibration of everything again....started with a initialized patch... but still the same problem.

See video:
turned off osc 2 ... and dialed in a pulsewave on osc 1 with pwm far up so its almost turns the pulse into silence....
this is where its very noticable..... every 4th note sounds different. *and its not in split or stack mode... so really weird.
What else could i do wrong?

To show whats going on here the videos:
DSI_REV2 ODD VOICE: showing you the patch with the envelopes way high: but still hearable problem
4TH NOTE: testing after recalibrating with pulse wave.

link : https://drive.google.com/open?id=171cthNvPvO6REND8VeVvG_6vSoj9awb6

MPM

  • **
  • 100
Re: inconsistent strange voice behavior on new REV2
« Reply #7 on: May 26, 2018, 05:53:45 PM »
From the bottom of pg19 in the manual, there’s what I pasted below on OSC note sync., It’s not “hard sync” but it stopped that random 4th note kind of thing I noticed too on some patches.

You can force either or both of the oscillators to start its wave cycle from its beginning (its zero crossing) each time you press a note on the keyboard by setting note reset to on in the misc parameters menu. This is can be useful for creating sounds that have a very consistent sound each time you trigger them.
OB-6  Prophet-6  Prophet.Rev2/16  no kids

MPM

  • **
  • 100
Re: inconsistent strange voice behavior on new REV2
« Reply #8 on: May 26, 2018, 06:00:47 PM »
Oops.

It’s “Note reset” not note sync.
OB-6  Prophet-6  Prophet.Rev2/16  no kids

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #9 on: May 27, 2018, 05:19:15 AM »
Thanks for the tip MPM,
but unfortunately note reset was already activated ... doesn't solve it.
Is there anyone who could check if their REV2 acts the same as mine?

Gomjab

  • **
  • 110
Re: inconsistent strange voice behavior on new REV2
« Reply #10 on: May 27, 2018, 06:58:02 AM »
Thanks for the tip MPM,
but unfortunately note reset was already activated ... doesn't solve it.
Is there anyone who could check if their REV2 acts the same as mine?

You can find older threads here discussing similar issues.  The last several firmware updates have listed improvents on voicing.

I had similar issues with my 16 voice model but they were fixed with the latest updates however I did have to repeat the calibration procedure a few times.

There were a few cases where folks had bad voice boards which DSI handled with a replacement. If you are on the latest firmware try recalibrating a few more times and check an init patch.  If the problems remains email DSI support directly.

Good luck.   I love my Rev2!

Gomjab

  • **
  • 110
Re: inconsistent strange voice behavior on new REV2
« Reply #11 on: May 27, 2018, 07:09:09 AM »
One other thing I notice is on oscillator 1 you are modulating shape for a square wave near 90%.  Try modulating below 10% which should have similar sound as the square is collapsing to a pulse on both sides. 

Can you post the patch file you are using?  I could load it into my Rev2 and compare results.
« Last Edit: May 27, 2018, 07:11:17 AM by Gomjab »

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #12 on: May 27, 2018, 08:53:23 AM »
Hi Gomjab,

i did several calibrations yet again.
Can't get it solved.

I think setting the square modulation above 90% should still give me a consistent note each time i trigger the same key.

Calibration doen't fix it *did it around 5 times again today.
So I hope you guys can help me out by loading the same patch in your REV2 and trigger the keys ... i noticed the 2th A note on the lower part of the keyboard gave the most obvious result of the problem. I made a video again showing how it sounds on my REV2. and i added the sysex patch to load.

*video is named: sysex patch example.mov
*sysex patch is namde: U1 P2 4th note problem

Link: https://drive.google.com/open?id=171cthNvPvO6REND8VeVvG_6vSoj9awb6

Please let me know (video or audio recording would be awsome) what the patch does in your REV2 so i can compare it to mine.
Thanks so much for all the help so far.

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #13 on: May 27, 2018, 09:25:07 AM »
and i discovered another mystery ....
when i initialize a patch and play more then 4 notes...  notes gets killed....
meaning i only have 4 notes polyphony.... without any stacking / splitting or unison active.

Im about to mail to DSI technical service but i want to be sure im not overlooking something since the synth is new to me and could be some simple option somewhere that i have activated ? Or is my REV2 bad?

Gomjab

  • **
  • 110
Re: inconsistent strange voice behavior on new REV2
« Reply #14 on: May 27, 2018, 10:43:30 AM »
and i discovered another mystery ....
when i initialize a patch and play more then 4 notes...  notes gets killed....
meaning i only have 4 notes polyphony.... without any stacking / splitting or unison active.

Im about to mail to DSI technical service but i want to be sure im not overlooking something since the synth is new to me and could be some simple option somewhere that i have activated ? Or is my REV2 bad?

Make sure you are plugged in the main output and not the sub out.  I’m going to try your preset today if I can find time.


Re: inconsistent strange voice behavior on new REV2
« Reply #15 on: May 27, 2018, 11:01:45 AM »
Hello elektrofunk.

I've downloaded and sent your sysex file to my REV2 (16 voice), and it produces a similar behavior, though not as pronounced. All notes play on mine, but some are fainter then others, and of course it's always on the same voices (roundabout voice assignment).
So my conclusion is that it's because of pulse width calibration variations between voices. This is an analog synth, and although it has DCOs (digital frequency control of the oscillators), the waveshaper is analog and voltage controlled.

On my REV2, it came a time when the calibration procedure couldn't do a good enough job. I contacted DSI support, and they suggested clearing the calibration data table (which might have become corrupt).

You can do this by accessing the "debug menu" by holding "down" and "up" and pressing "global". The first item should be "1. Clear Calibration". To get out of that menu, simply press "global" again. Once it's done, power down the synth, wait 10 seconds, power it back, and call up the calibration. It might take longer to go through it since it will start from scratch on each voice. It might even lock-up at some point. If it does, simply wait at least 3 or 4 minutes to make sure it's really stuck, and power cycle the synth and restart another calibration. Everything should be fine after that.

It might, in the worst case, require more than one calibration to get things in order. But if after all of that the synth doesn't behave normally (within the bounds of acceptable minimal analog variations between voices), than DSI support should take over.

EDIT: after having a look at the patch itself, I noticed that you control the wave shape of OSC1 with ENV3 with maximum amount (in the mod matrix slot 1), even though the wave shape is already at 97%. Anytime you play with parameters on the edge like that, you're bound to encounter variations due to the analog nature of control voltage. After playing a little more with your patch, I discovered that when playing some lower keys on my REV2 there ARE a few notes that aren't producing sounds. Most probably because of calibration varations of my unit, too.
Not only that, but if you listen closely when holding a note down (one that produces a sound), you'll hear the timbre change abruptly at some point (especially on lower notes). This is probably due to the digital resolution limitation of the DAC that sends the converted control voltage to the waveshaper.
« Last Edit: May 27, 2018, 11:17:18 AM by AlainHubert »
Oberheim OB-X8, Minimoog D (vintage), OB6 (Desktop), Oberheim Matrix-6 (MIDI Controller for OB6), VC340

Gomjab

  • **
  • 110
Re: inconsistent strange voice behavior on new REV2
« Reply #16 on: May 27, 2018, 01:39:35 PM »
Here is my 16 voice Rev2 playing your patch.  It also has variations due to pulse width being on edge.  I remember someone posting that the pulse width is more consistent on the low limit vs high limit.

Also I set my levels based on factory patches.  This patch output was lower but I left the input levels where they were.


elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #17 on: May 27, 2018, 03:42:08 PM »
Thanks you so much Gomjab and Alain,

I believe i have fixed the problem for now....
Not sure what did the trick but i remember doing a Global settings reset...
and it seemed like hitting the sync osc button to on and off again... did something....
dont ask me why! maybe the button itself got stuck in between off and on?... if thats even possible.
or maybe something else solved it while i didnt notice.

The polyphony changed back to normal again ...
strangely was stuck in 4 polyphony without any split or stacking...
and now back to 8 again... and the inconsistent voices is to an acceptable level now.

Again thanks for all the help and time you guys put in helping me out!
Now i finally can use my prophet to work on some new songs for my band :-)
« Last Edit: May 27, 2018, 03:45:48 PM by elektrofunk »

Gomjab

  • **
  • 110
Re: inconsistent strange voice behavior on new REV2
« Reply #18 on: May 27, 2018, 04:18:23 PM »
Glad you got it worked out!

Now go make music!

 ;D

elektrofunk

  • *
  • 10
  • The Robots Are Coming
    • The Robots Are Coming
Re: inconsistent strange voice behavior on new REV2
« Reply #19 on: May 27, 2018, 04:22:22 PM »
haha thanks man!
music ,... not sure.
but noise! hell yeah! The Robots Are Coming
« Last Edit: May 27, 2018, 04:24:44 PM by elektrofunk »