The Official Sequential/DSI Forum

Rev 2 firmware is it just me?

Re: Rev 2 firmware is it just me?
« Reply #140 on: July 19, 2021, 09:41:40 PM »
Thanks, Invisible Homes.  That's considered a feature, though, rather than a bug.

Pym

  • *****
  • 181
Re: Rev 2 firmware is it just me?
« Reply #141 on: July 19, 2021, 09:53:58 PM »
Itís tuned so you can do karplus strong and other things. Some presets will use more of the range, some wonít. This happens on lots of knobs depending on the preset. Canít please everybody all the time.

Thanks, Invisible Homes.  That's considered a feature, though, rather than a bug.

Yes, of course you are right. That's the correct term. It's not a "bug" exactly. Since, however, this addresses a nagging and widely-held criticism of the REV2, it is hard to just call it a "feature request." Nomenclature aside, I hope that this is something that can be addressed in the last update.

I'm actually surprised that this issue and fix hasn't come up earlier in all the REV2 chatter. I've read so many threads on all the pages, but perhaps I missed something? I find it strange that the knob bottoms out at 200 Hz and that so much of its range is dedicated to 15 KHz+. As noted, that leaves a whole octave of range unaddressed on the low end and limits the resolution of the middle range where so much of the interesting stuff happens. From a sound design perspective I can actually understand the value in certain contexts (i.e. for less muddy pads etc), but it also limits filter response range in a way that isn't reflected on any of the other synths/filters I have. For bass oriented sounds, that range is essential and it warms up filter "wubs" considerably when using a slower attack and decay. I think it also makes the 12 dB filter so much more useful (and I already liked it).

The DC offset I used was -60, but YMMV. I suspect somewhere in the -50s is the sweet spot for the maximum range. [Edit/update: this seems to be the case for the balance with negative envelopes]. I'm curious to hear others experience and perspective on this.

Having a range option would be so nice and it would probably quiet some of the lingering criticism of the low end on the synth. It is a great compliment to Creative Spiral's amazing VCM work.
Sequential

Re: Rev 2 firmware is it just me?
« Reply #142 on: July 20, 2021, 05:12:12 AM »
Problem: One of the most common complaints about the Rev2 is the filter. This may in part be about the cutoff knob range. It is set up in an odd way so that the peak of the curve is at 20 KHz when the knob is around 12 o'clock.* Most of the usable range is in only half of the knob range. Further, the knob bottoms out just below 200 Hz!. This leaves an octave of filter range out.

Not sure if I'm doing something wrong, but I don't seem to be experiencing this on my Rev2.

At 12 o'clock, the peak is just above 200 Hz. Fully closed, there's no signal at all coming through.
With the filter opened so the screen reads "50", the peak is around 35 Hz.

It's the same in 12 dB mode, but the resonant peak is of course much less pronounced since the Rev2 12dB filter is barely resonant at all.

This is from a blank patch with just Noise, no modulation, no keyboard tracking, no envelope or velocity amount.

I've attached a shot of my spectrum analyzer with the filter at Noon and with the screen reading 50 (about 10 o'clock). No point in showing what it looks like at 0, since it's just silence. I've also attached an audio file of the resonant filter sweep (watch your ears!).

dsetto

  • ***
  • 387
Re: Rev 2 firmware is it just me?
« Reply #143 on: July 20, 2021, 10:16:30 AM »
Itís tuned so you can do karplus strong and other things. Some presets will use more of the range, some wonít. This happens on lots of knobs depending on the preset. Canít please everybody all the time.
...

I'm glad to be catching a discussion like this because it gets me more familiar with my Rev2. Invisible Homes, the subject you're raising sheds better light on something I've observed, but never considered. Moreso, Iím glad to learn thereís a way to adjust the frequency range to get greater throw in the lower and mid range. Pym, Iím glad to learn the intention and uses behind the design. Ö Out of curioisity, is another usefulness of the frequencyís higher resolution in the high range for tonal tuning and playing of the sine waves arising from high filter resonance?

Re: Rev 2 firmware is it just me?
« Reply #144 on: July 20, 2021, 11:48:25 AM »
Thanks, Pym, AdvanceFollower, SacredSynthesis, and Dsetto for responses and thoughts.

I get that one might want to have a higher filter range for certain uses, but it is unusual (if it isn't an error). This is not the case with my Prodigy, MS20, Electrix filters and (on the digital side) not how the U-He Diva emulates the P5's range. I'm getting some sense from AdvancedFollower's post and a parallel Reddit discussion that the filter ranges vary and/or that my unit may simply have a problem that calibration isn't addressing.

Edit/Update:

I've actually figured out the issue. This is not by design. My experience seems to be a result of the Global>24. CV footpedal setting. When it is set to "LPF Half" it bottoms out at 200 Hz and the results will stay in place even when the pedal is not plugged in, i.e. it alters the knob's subsequent behavior (even after shut down and restart). I'm not sure if that half range setting was by default.

I suspect that I received the unit with this setting on, since I have always found the knob range to be odd (esp. in the 12 dB mode). If it is, or if it is set that way on some units but not others, this might explain why some users experiences of the filter are variable.

This does start to lead to something more like a bug issue. There is no "off" option for the footpedal in the global menu. Given the above, that would be sensible.
« Last Edit: July 20, 2021, 12:30:22 PM by Invisible Homes »

Re: Rev 2 firmware is it just me?
« Reply #145 on: July 20, 2021, 12:58:36 PM »
That's a useful discovery.  At least it explains a strange behavior. 
"Simplicity, simplicity, simplicity!"

- Henry David Thoreau

The Musical Synthesizer YouTube Channel:
https://www.youtube.com/channel/UChLGwGiRVs7rlZXnOG9_mUw

The Musical Synthesizer Blog: https://themusicalsynthesizer.wordpress.com

Re: Rev 2 firmware is it just me?
« Reply #146 on: July 20, 2021, 02:23:41 PM »
Quote
I suspect that I received the unit with this setting on, since I have always found the knob range to be odd (esp. in the 12 dB mode). If it is, or if it is set that way on some units but not others, this might explain why some users experiences of the filter are variable.

That's very interesting. I wonder if certain batches were sent out with this setting at "LPF Half" by default. It's of course also possible some users accidentally changed this setting without being aware of what it does. That would explain why some complain about their Rev2's being overly bright, and also why some users report oscillator "bleed" with the filter fully closed (because the filter wouldn't actually be fully closed but rather at ~200 Hz, letting some of the signal through).

Given that even Sequential reps in this thread seem to be unaware of this setting and its effect on the filter cutoff range, instead saying the filter response is "by design", I could see how someone with their synth set to "LPF Half" (either from the factory or unintentionally by the user) would dismiss their Rev2 as extremely bright and harsh sounding. I know I certainly wouldn't have figured out that I needed to change the Foot Pedal setting, had my Rev2 arrived with this setting active, and writing to Sequential support, I would likely have been told the bright filter is "by design".

Re: Rev 2 firmware is it just me?
« Reply #147 on: July 20, 2021, 03:57:31 PM »
Quote
I suspect that I received the unit with this setting on, since I have always found the knob range to be odd (esp. in the 12 dB mode). If it is, or if it is set that way on some units but not others, this might explain why some users experiences of the filter are variable.

That's very interesting. I wonder if certain batches were sent out with this setting at "LPF Half" by default. It's of course also possible some users accidentally changed this setting without being aware of what it does. That would explain why some complain about their Rev2's being overly bright, and also why some users report oscillator "bleed" with the filter fully closed (because the filter wouldn't actually be fully closed but rather at ~200 Hz, letting some of the signal through).

Given that even Sequential reps in this thread seem to be unaware of this setting and its effect on the filter cutoff range, instead saying the filter response is "by design", I could see how someone with their synth set to "LPF Half" (either from the factory or unintentionally by the user) would dismiss their Rev2 as extremely bright and harsh sounding. I know I certainly wouldn't have figured out that I needed to change the Foot Pedal setting, had my Rev2 arrived with this setting active, and writing to Sequential support, I would likely have been told the bright filter is "by design".

Even if you don't use a foot pedal or CV input, would this setting affect the filter cutoff range? (What should it be set to to mitigate the "bright/harsh sound?"

Re: Rev 2 firmware is it just me?
« Reply #148 on: July 20, 2021, 04:43:33 PM »
In Globals, set Foot Pedal to either:

a. Breath CC2
b. Foot CC4
c. Expression CC11

Those three are essentially "Off" unless you specifically map them per-patch in the mod matrix.  I usually keep my pedal mapped to CC11, and then if I want it to map to cutoff in a given patch, I'll just use the dedicated Matrix Slot for it with an appropriate amount.     Using the Full LPF / Half LPF is sort of a blunt force method to make your Exp Pedal always do something... but it rarely produces results you would want in a real world performance scenario... and of course, if you forget about it and the pedal is in a forward position, it will offset the filter cutoff.

I just tested on my Rev2 keyboard to see whether that setting (Full LPF or Half LPF) changes the cutoff when a pedal is not connected... it doesn't seem to make any difference when you don't have a pedal connected...  that would potentially be a bug if it was the case, but I was unable to reproduce it.   When you tested it, did you power on with a foot pedal attached, and then remove the pedal after power was on? 

As far as I can tell, this is not a bug... operating as expected. 
« Last Edit: July 20, 2021, 04:45:57 PM by creativespiral »

Sequential Pro 3 and Rev 2, Deepmind, PolyBrute - Sound Sets, Patch Banks - Available on Sellfy: https://sellfy.com/sounddesign/

Re: Rev 2 firmware is it just me?
« Reply #149 on: July 20, 2021, 06:05:27 PM »
I just tested on my Rev2 keyboard to see whether that setting (Full LPF or Half LPF) changes the cutoff when a pedal is not connected... it doesn't seem to make any difference when you don't have a pedal connected...  that would potentially be a bug if it was the case, but I was unable to reproduce it.   When you tested it, did you power on with a foot pedal attached, and then remove the pedal after power was on? 

As far as I can tell, this is not a bug... operating as expected.

I did some investigation and reflection. Follow me, if you will:

1) The previous setting issue was persistent (i.e. almost 1.5 years of regular use).
2) The cutoff knob behavior was clear, i.e. I hadn't heard the 12 dB filter close all the way until this discovery.
3) When I tried to reproduce the issue (as per Creative Spiral), it was not persistent in the same way. That being said, when I unplugged the pedal while powered on, I encountered a situation where the cutoff knob didn't work at all and the filter stayed 100% open. This persisted across power up/down and the only way to make it go away is by changing Global settings to something other than the LPF Full or Half settings and power cycling.

Given anecdotal evidence re: filter tone and bleed through complaints, I still could not shake the suspicion that I'm not the only one to experience this issue.

...So, I tried to imagine what could possibly cause the Synth to think that there was a pedal plugged in. It occurred to me to check the difference in behavior with balanced TRS vs unbalanced TS cables. To experiment I plugged in an unbalanced TS 1/4 cable, set the global to LPF Half, and voila! If I unplugged the TS cable the setting persisted exactly as before with the cutoff falling at just below 200 Hz. This is reproducible and persists across power cycles. I can't even get it to go away until I plug in a pedal on a TRS cable, set Global 24 to LPF Full, and the pedal to reset the range.

To reproduce/test:
1. Power on.
2. Create a default Basic Program.
3. Change Global 24 to LPF Half
4. Plug and unplug TS 1/4" cable in Pedal/CV jack.
5. Test frequency range of filter cutoff. [I get just below 200 Hz] 
6. Power cycle and test knob on other settings.

I can reproduce this every time. The key is the TS 1/4" cable or...a balanced TRS that is either not plugged in all the way or slowly unplugged. This last bit leads me to go out on a limb with a big guess.

Hypothesis:
All REV2s were duly tested and shipped. The global was set to LPF Half on all or some units(?). During the testing on some REV2s either the expression pedal was not plugged in all the way or it was slowly unplugged in a manner that caused this error. This would be fairly easy to overlook even with diligent testing, and since (as above) it persists across power cycles it would seem to be a "feature." It would not be readily noticeable quickly checking most presets. This could plausible produce a situation in which some units shipped with this limited bandwidth in place and others shipped working normally.

The result: wildly varied opinions on the tone of the synth and the reports of bleed through. I think we may have solved a mystery. What say you ladies and gentlemen of the jury?

Pym

  • *****
  • 181
Re: Rev 2 firmware is it just me?
« Reply #150 on: July 21, 2021, 01:26:39 AM »
Glad you figured it out!

Sometimes itís really difficult to know whatís going on through text. Sending a quick video to support is worth a thousand words

Sorry we couldnít get that figured out sooner, I can only imagine how frustrating that has been!!

I just tested on my Rev2 keyboard to see whether that setting (Full LPF or Half LPF) changes the cutoff when a pedal is not connected... it doesn't seem to make any difference when you don't have a pedal connected...  that would potentially be a bug if it was the case, but I was unable to reproduce it.   When you tested it, did you power on with a foot pedal attached, and then remove the pedal after power was on? 

As far as I can tell, this is not a bug... operating as expected.

I did some investigation and reflection. Follow me, if you will:

1) The previous setting issue was persistent (i.e. almost 1.5 years of regular use).
2) The cutoff knob behavior was clear, i.e. I hadn't heard the 12 dB filter close all the way until this discovery.
3) When I tried to reproduce the issue (as per Creative Spiral), it was not persistent in the same way. That being said, when I unplugged the pedal while powered on, I encountered a situation where the cutoff knob didn't work at all and the filter stayed 100% open. This persisted across power up/down and the only way to make it go away is by changing Global settings to something other than the LPF Full or Half settings and power cycling.

Given anecdotal evidence re: filter tone and bleed through complaints, I still could not shake the suspicion that I'm not the only one to experience this issue.

...So, I tried to imagine what could possibly cause the Synth to think that there was a pedal plugged in. It occurred to me to check the difference in behavior with balanced TRS vs unbalanced TS cables. To experiment I plugged in an unbalanced TS 1/4 cable, set the global to LPF Half, and voila! If I unplugged the TS cable the setting persisted exactly as before with the cutoff falling at just below 200 Hz. This is reproducible and persists across power cycles. I can't even get it to go away until I plug in a pedal on a TRS cable, set Global 24 to LPF Full, and the pedal to reset the range.

To reproduce/test:
1. Power on.
2. Create a default Basic Program.
3. Change Global 24 to LPF Half
4. Plug and unplug TS 1/4" cable in Pedal/CV jack.
5. Test frequency range of filter cutoff. [I get just below 200 Hz] 
6. Power cycle and test knob on other settings.

I can reproduce this every time. The key is the TS 1/4" cable or...a balanced TRS that is either not plugged in all the way or slowly unplugged. This last bit leads me to go out on a limb with a big guess.

Hypothesis:
All REV2s were duly tested and shipped. The global was set to LPF Half on all or some units(?). During the testing on some REV2s either the expression pedal was not plugged in all the way or it was slowly unplugged in a manner that caused this error. This would be fairly easy to overlook even with diligent testing, and since (as above) it persists across power cycles it would seem to be a "feature." It would not be readily noticeable quickly checking most presets. This could plausible produce a situation in which some units shipped with this limited bandwidth in place and others shipped working normally.

The result: wildly varied opinions on the tone of the synth and the reports of bleed through. I think we may have solved a mystery. What say you ladies and gentlemen of the jury?
Sequential

Re: Rev 2 firmware is it just me?
« Reply #151 on: July 21, 2021, 07:55:48 AM »
Glad you figured it out!...
Sorry we couldnít get that figured out sooner, I can only imagine how frustrating that has been!!


No worries and thanks for your work. I love the REV2. As my previous post suggests, this is an odd discovery and one that has a certain element of chaos/unpredictability to it (a certain physical interaction with a certain software setting under the right conditions).

Again, I would not be surprised if this has affected perceptions of the filter on this platform and potentially other Sequential/DSI's using the Curtis PA397. I hope this insight helps in the future with other users and/or customer support. I imagine it might also prevent some returns on the retail end, if (and I grant that this could be still be a discrete case) it is the cause of other user issues.


Re: Rev 2 firmware is it just me?
« Reply #152 on: July 21, 2021, 12:55:40 PM »

To reproduce/test:
1. Power on.
2. Create a default Basic Program.
3. Change Global 24 to LPF Half
4. Plug and unplug TS 1/4" cable in Pedal/CV jack.
5. Test frequency range of filter cutoff. [I get just below 200 Hz] 
6. Power cycle and test knob on other settings.

I can reproduce this every time. The key is the TS 1/4" cable or...a balanced TRS that is either not plugged in all the way or slowly unplugged. This last bit leads me to go out on a limb with a big guess.

Hypothesis:
All REV2s were duly tested and shipped. The global was set to LPF Half on all or some units(?). During the testing on some REV2s either the expression pedal was not plugged in all the way or it was slowly unplugged in a manner that caused this error. This would be fairly easy to overlook even with diligent testing, and since (as above) it persists across power cycles it would seem to be a "feature." It would not be readily noticeable quickly checking most presets. This could plausible produce a situation in which some units shipped with this limited bandwidth in place and others shipped working normally.

The result: wildly varied opinions on the tone of the synth and the reports of bleed through. I think we may have solved a mystery. What say you ladies and gentlemen of the jury?

This is some good sleuth work, Detective @Homes.  I was able to exactly reproduce this. 

@Pym - This should maybe be looked into...    I can see how this may cause issues for a small handful of people - producing confusing behavior / making patches sound incorrect -- Either if the QC hypothesis is correct, or in the random case where someone unplugs their expression pedal while the unit is powered on... which causes the Exp Pedal value to max out and saves that value across subsequent power cycles.

The solution:  In firmware, upon startup/init routine, always reset Exp Pedal source value to 0 (then sample the current position if the pedal is present)    As is, it seems to remember the last stored expression value... even across power cycles, and even if the Exp Pedal is not present.   
« Last Edit: July 21, 2021, 01:07:03 PM by creativespiral »

Sequential Pro 3 and Rev 2, Deepmind, PolyBrute - Sound Sets, Patch Banks - Available on Sellfy: https://sellfy.com/sounddesign/

Re: Rev 2 firmware is it just me?
« Reply #153 on: July 21, 2021, 06:48:19 PM »
Creativespiral and Invisible Homes, I have to say that you've both done exceptional detective work for the Rev2 and for all Rev2 owners.  There is invaluable information on this thread from the two of you.  Great job, and thank you very much! 
« Last Edit: July 21, 2021, 06:53:10 PM by Sacred Synthesis »
"Simplicity, simplicity, simplicity!"

- Henry David Thoreau

The Musical Synthesizer YouTube Channel:
https://www.youtube.com/channel/UChLGwGiRVs7rlZXnOG9_mUw

The Musical Synthesizer Blog: https://themusicalsynthesizer.wordpress.com

Re: Rev 2 firmware is it just me?
« Reply #154 on: July 21, 2021, 06:50:04 PM »
Is anyone aware of a Rev2 ring rodulation bug?   Razmo had described it in this way:

"I just discovered that the "ringmod" bug is still present...just made a preset that introduce it...again a problem with layer B...this time, after loading this bugged preset, I also experienced strange behavior when switching to a preset that only used layer A after having loaded that buggy preset...at one time it started to ping pong the sound from left to right for eight keystrokes, then the sound changed to center for the next eight keystrokes...another time it played for eight keystrokes, and then went silent for another 8 keystrokes....

It seems that the bug changes character as I switch the Layer B effect on/off...."


Sequential needs a more detailed description in order to reproduce it.  Or perhaps it's been resolved in an update?
« Last Edit: July 22, 2021, 07:10:29 AM by Sacred Synthesis »
"Simplicity, simplicity, simplicity!"

- Henry David Thoreau

The Musical Synthesizer YouTube Channel:
https://www.youtube.com/channel/UChLGwGiRVs7rlZXnOG9_mUw

The Musical Synthesizer Blog: https://themusicalsynthesizer.wordpress.com

Re: Rev 2 firmware is it just me?
« Reply #155 on: July 21, 2021, 09:37:18 PM »
This is some good sleuth work, Detective @Homes.  I was able to exactly reproduce this. 

@Pym - This should maybe be looked into...    I can see how this may cause issues for a small handful of people - producing confusing behavior / making patches sound incorrect -- Either if the QC hypothesis is correct, or in the random case where someone unplugs their expression pedal while the unit is powered on... which causes the Exp Pedal value to max out and saves that value across subsequent power cycles.

The solution:  In firmware, upon startup/init routine, always reset Exp Pedal source value to 0 (then sample the current position if the pedal is present)    As is, it seems to remember the last stored expression value... even across power cycles, and even if the Exp Pedal is not present.   

Actually, I just had another thought about this, and tested this further... its more of an issue... it even happens if you remove the expression pedal when the unit is powered down.   

If you happen to end a session with expression pedal in foot forward position (which is not that uncommon) -  then power down the synth and remove the expression pedal to transfer it to another instrument or for some other reason (planning to play a gig, but not bringing the exp pedal), the Rev2 remembers that last stored exp pedal value when you power on next time... and if you were in Half LPF or Full LPF mode it will add an invisible/unchangeable offset to filter cutoff to all your patches.   And even changing the Foot Pedal mode at this point won't fix it, which is strange - it doesn't immediately detach itself from the cutoff offset... you have to change it in globals, then power cycle again to remove the invisible cutoff offset.   I can definitely see how this may have confounded users, and should probably be addressed. 

Fortunately, it seems like it would be a simple fix ... as mentioned, just add a Reset to 0 for Exp Pedal Source amount during Init/Startup, and that should do it.

Sequential Pro 3 and Rev 2, Deepmind, PolyBrute - Sound Sets, Patch Banks - Available on Sellfy: https://sellfy.com/sounddesign/

Re: Rev 2 firmware is it just me?
« Reply #156 on: July 21, 2021, 09:46:51 PM »
Is anyone aware of a Rev2 ring rodulation bug?   Razmo had described it in this way:

"I just discovered that the "ringmod" bug is still present...just made a preset that introduce it...again a problem with layer B...this time, after loading this bugged preset, I also experienced strange behavior when switching to a preset that only used layer A after having loaded that buggy preset...at one time it started to ping pong the sound from left to right for eight keystrokes, then the sound changed to center for the next eight keystrokes...another time it played for eight keystrokes, and then went silent for another 8 keystrokes....

It seems that the bug changes character as I switch the Layer B effect on/off...."


Sequential needs a more detailed description in order to reproduce it.  Anybody?

I did try and research this purported bug a few weeks ago, but could not find enough info, or any steps to reproduce...  And I personally have never run into it... also, I think after Razmo first reported it there was an OS update to 1.1.5.9, so perhaps that fixed whatever the issue was?    I did see this msg:  https://forum.sequential.com/index.php/topic,3263.msg35676.html#msg35676

But yeah, if anyone else has run into this since 1.1.5.9, please chime in...

Sequential Pro 3 and Rev 2, Deepmind, PolyBrute - Sound Sets, Patch Banks - Available on Sellfy: https://sellfy.com/sounddesign/

jg666

  • ***
  • 472
Re: Rev 2 firmware is it just me?
« Reply #157 on: July 22, 2021, 01:49:46 AM »
I have posted this before a while ago but worth mentioning again because there are more people viewing this thread and able to test on theirs...

This is regarding F1 P24

Using this patch, the octave from middle C seems to play differently to all the other keys. They seem quieter for want of a better description. Does anyone else notice this? so just play single keys from below middle C and through the middle C octave and beyond, it's just the one octave on mine that sounds different.

I haven't noticed this on any other patches but it happens every time on F1 P24. I could try and record a video with my camera if required (I don't record stuff or play on my computer)

DSI Prophet Rev2, DSI Pro 2, Moog Sub37, Korg Minilogue, Yamaha MOXF6, Yamaha MODX6, Yamaha Montage6

jg666

  • ***
  • 472
Re: Rev 2 firmware is it just me?
« Reply #158 on: July 22, 2021, 08:40:50 AM »
Further to my post above (it wonít let me edit my own post for some reason) if I change Octave using Octave +/- then the issue moves along the keyboard so this issue is nothing to do with the keys.
DSI Prophet Rev2, DSI Pro 2, Moog Sub37, Korg Minilogue, Yamaha MOXF6, Yamaha MODX6, Yamaha Montage6

timborťale

  • **
  • 168
  • MIDI nerd
Re: Rev 2 firmware is it just me?
« Reply #159 on: July 22, 2021, 11:45:05 AM »
Fortunately, it seems like it would be a simple fix ... as mentioned, just add a Reset to 0 for Exp Pedal Source amount during Init/Startup, and that should do it.

I'd be concerned that perhaps the filter amount as adjusted by the pedal may not be sourced from the pedal but cached separately, that would need to be zeroed as well (for instance, set pedal to LPF Half/Full, put a value on that, unplug it, change the pedal to CC #11 or something, plug it back in and change it again -> does the persisted LPF offset get reset to 0?).

However, my understanding is that the synth has no memory that is updated frequently during operation that persists across boots - you need to enter and exit globals to trigger a persistence of settings - so the fact that this is getting stored at all is a little surprising... but it may be that that's in a static allocation and they may skip zeroing that memory segment on boot... SRAM can persist values between power cycles.
Rev2-16, Prophet 12 Keyboard, Nords, etc...