Prophet REV2, annoying by design ?

[AlainHubert]

I've had my REV2 16 voice for close to 2 years now, and there are still a few things that annoy me on it. One of the most annoying ones is the fact that the DCOs are not free running all the time. For some unknown reason, the programmers have decided to reset the timers controlling them depending on how wide is the frequency gap between played notes. The result is an irritating "oscillators all starting in phase" artifact that is clearly audible. So when I want to create a smooth detune, it is destroyed by an occasional unwanted restarting of the oscillators, similar to a sampler playing back a sample always starting it in phase.

You'd never hear this in a fully analog synth. And there's no reason why the oscillator timers cannot stay free-running, and keep the detuning between them constant, without being restarted in phase with each other when the space between played notes is greater than one octave.

This, and a few other annoying quirks, are making me considering selling it, more and more every day...
Luckily, I have the OB6 which doesn't suffer from this...

Sorry for the rant, but as much as I try, this REV2 irritates me a little more every time I try to design a new patch on it. Even my other DCO based synth, the DeepMind 12, doesn't behave like this...

[camillomuc]

I've had my REV2 16 voice for close to 2 years now, and there are still a few things that annoy me on it. One of the most annoying ones is the fact that the DCOs are not free running all the time. For some unknown reason, the programmers have decided to reset the timers controlling them depending on how wide is the frequency gap between played notes. The result is an irritating "oscillators all starting in phase" artifact that is clearly audible. So when I want to create a smooth detune, it is destroyed by an occasional unwanted restarting of the oscillators, similar to a sampler playing back a sample always starting it in phase.

You'd never hear this in a fully analog synth. And there's no reason why the oscillator timers cannot stay free-running, and keep the detuning between them constant, without being restarted in phase with each other when the space between played notes is greater than one octave.

This, and a few other annoying quirks, are making me considering selling it, more and more every day...
Luckily, I have the OB6 which doesn't suffer from this...

Sorry for the rant, but as much as I try, this REV2 irritates me a little more every time I try to design a new patch on it. Even my other DCO based synth, the DeepMind 12, doesn't behave like this...

I´m not sure, if I get you right, but I think, the "osc note reset off" function in the misc menu should provide free running oscs?

[maxter]

I second camillomuc, I can't see the problem here...

[S Y Z Y G Y X]

Alain....if you’re so unhappy with the Rev2, just sell it already ....I don’t see any negative problems with the Rev2, it’s a best, just not for you, just sell it already jeeze

[AlainHubert]

Even with Osc 1 Note Reset and Osc 2 Note Reset both set to Off, if you play a C2 and C4 (two octaves apart) successively you'll hear both oscillators being restarted in phase each time, no matter which voice is playing. Audible when setting both oscillators to the same C2 base reference, and Osc 1 detuned by -10 relative to Osc 2.

SYZYGYX
You're right. No use in whining about it. Sequential will probably never address this issue anyway in an updated firmware. I tried hard to like it, but even though it offers lots of voices, and lots of modulation possibilities, it lacks the rich sounds that my OB6 can produce. So it's going away as soon as the UB-Xa comes along.

I'll stop bothering people about it now, sorry.
 

[Tugdual]

Even with Osc 1 Note Reset and Osc 2 Note Reset both set to Off, if you play a C2 and C4 (two octaves apart) successively you'll hear both oscillators being restarted in phase each time, no matter which voice is playing. Audible when setting both oscillators to the same C2 base reference, and Osc 1 detuned by -10 relative to Osc 2.

SYZYGYX
You're right. No use in whining about it. Sequential will probably never address this issue anyway in an updated firmware. I tried hard to like it, but even though it offers lots of voices, and lots of modulation possibilities, it lacks the rich sounds that my OB6 can produce. So it's going away as soon as the UB-Xa comes along.

I'll stop bothering people about it now, sorry.

You’re not bothering me.
Ignoring problems is not what I consider a positive attitude.

[guyaguy]

Yeah he’s just asking if someone has found a solution for something. Maybe the thread title is a bit aggressive but I’ve seen lots of threads for various synths where the OP asks “why can’t my synth do X?” and gets advice on how it can do X.

[Tugdual]

Been tweaking my REV2 but couldn’t reproduce the issue. Wonder if this could be a 16 voices thing as I own the 8 version. However what I did notice is that the wave reset only locks the zero crossing hence you might end up with a 180° phase difference which can be very annoying.

[maxter]

Could be an OS bug. Some of these only affect particular units of Rev2, apparently.

Alain didn't mention he had osc reset set to OFF, which is why I didn't see the problem. No mention of OS version either, so I actually haven't bothered trying to reproduce it on my Rev2. I can't recall having any problem with osc reset OFF yet.

I'd actually consider selling my Matrix6 first, if I had one that I only used as a MIDI controller. My 5 cents.

[AlainHubert]

I do have the Osc Reset to Off, and my REV2 is on the very latest OS 1.1.5.9.

It's the way the timers that control the DCOs are programmed, if the frequency difference between the last payed note and the new note exceeds 2 octaves, then restart the timers (DCOs) for the newly assigned voice, hence the problem.
The Matrix 6 from Oberheim also exhibits the same behavior, although it does it every time a voice is recycled...
I guess it's too difficult to have 32 DCOs running freely and independently from each other continuously ?

[MPM]

Have you tried using glide?
May get you there.

[psionic11]

Can't you override hearing the OSCs reset by modding OSC waveshape?  Varying mod rate by using keytracking should mask any regularities, yes?

[Razmo]

I do have the Osc Reset to Off, and my REV2 is on the very latest OS 1.1.5.9.

It's the way the timers that control the DCOs are programmed, if the frequency difference between the last payed note and the new note exceeds 2 octaves, then restart the timers (DCOs) for the newly assigned voice, hence the problem.
The Matrix 6 from Oberheim also exhibits the same behavior, although it does it every time a voice is recycled...
I guess it's too difficult to have 32 DCOs running freely and independently from each other continuously ?

I cannot help but wonder if this is a limitation in the Curtis chip... if it resets the oscillators when a new note is more than one octave apart, then it may be because the change to the new pitch would somehow introduce an unwanted "pop" or "click" in the sound because something may need to stabilize in the CV department...

I say this becuase I remember quite clearly the dreaded pops and clicks associated with the same chip on the Evolver... if you played a low key and a high key alternating on a poly Evolver, you got some quite anyoing clicks and pops until all voices had been allocated with the same pitch... with glide on it was quite easy to hear...

I bet this is the reason... but it's of course guesswork.

[Razmo]

Tried to replicate it, but have not had any success.. maybe if you could post an audio example, so we could hear what you mean?

[AlainHubert]

Sorry for the late reply (been busy). I've recorded an audio demonstration of the DCO timers reset behavior when playing notes more than one octave apart.
It's a simple Unison with 16 voices, with a medium-slow detuning between Osc1 and Osc2.
At first you hear a continuous tone, with the expected beating between the two oscillators.
Then I start playing that same note in succession (staccato), with the beating still continuously free running, but when I play a note 4 octaves higher between the low tone, you can clearly hear the beating between the oscillators being restarted on each low note.
When I stop playing the high note between the low ones, the beating resumes its free running and is not restarted each time.

I've selected this method (Unison), which is the most audibly evident one. But the same thing happens in normal polyphonic mode.

It's a short 20 second recording (all in real time), in .wav format.

https://www.dropbox.com/s/pvfb93c0kpc8tr3/DCO%20Timers%20Reset%20behavior.wav?dl=0

[Razmo]

Sorry for the late reply (been busy). I've recorded an audio demonstration of the DCO timers reset behavior when playing notes more than one octave apart.
It's a simple Unison with 16 voices, with a medium-slow detuning between Osc1 and Osc2.
At first you hear a continuous tone, with the expected beating between the two oscillators.
Then I start playing that same note in succession (staccato), with the beating still continuously free running, but when I play a note 4 octaves higher between the low tone, you can clearly hear the beating between the oscillators being restarted on each low note.
When I stop playing the high note between the low ones, the beating resumes its free running and is not restarted each time.

I've selected this method (Unison), which is the most audibly evident one. But the same thing happens in normal polyphonic mode.

It's a short 20 second recording (all in real time), in .wav format.

https://www.dropbox.com/s/pvfb93c0kpc8tr3/DCO%20Timers%20Reset%20behavior.wav?dl=0

I hear what you mean... but I do not think it's because they actually reset the oscillator phases because if they did, it would sound much more pronounced (try switching oscillator reset to on with your preset, and you should know).

Even though I can hear what you are saying, I can actually also hear that something is moving on those low notes in between the high ones, it's just not sounding the way you would expect.

If i had to make a guess, it's something that happens when the pitch need to be changed on a voice from last time it was in use... in your example everything gets back to normal exactly after you stop playing the high note, but as you are using 16 voice unison, all voices would have been set to the last low note at once... so it sounds like it stops right away... what if you use a single poly voice where you're cycling thru the voices? ... on a 16 voice REV2, if I'm right, when you stop playing low/high notes, the effect should continue and happen on all the voices which previous pitch was the high note (as pitch is being reset to low from high)... I bet it has to do with changing pitch on a voice, and that the larger the difference, the more pronounced.