SYNTAKT Bugs Thread

[touchy]

Not sure if a bug, but at least unexpected behavior / felt unintuitive to me:

In live playing / recording mode trigging a track selects that track, so far so good.
When the FX track is selected, you can still trig the tracks with the buttons, but when exiting the FX track, the selection hasn’t been updated to the most recently trigged track, but is still on whatever track was last selected before entering the FX track.

Hope I explained that ok, should become clear when trying it out. Thanks for bearing with me

[deadinside]

Similar to what others have reported - having an older version of the project on boot up.
Usually solved by reloading the project, thankfully. (I always save project as I go, never rely on temp save if powering down)

However, a twist on this and the LFO assignment bug is that on reloading the project, the LFO assignment on a particular track isn’t showing BUT the LFO can be heard in operation on that track.

I have cycled power and loaded another project but it is still like this when I reload this particular project.

Anyone else?

[Watto]

12/SEPT/23: EDITING AGAIN A DAY LATER: Problems became worse, more strange issues popped up, I typed up a full report getting ready to send to support… last thing I tried was an initialised project… yep it all works fine now, I can only guess I had a corrupted project file… So! DONT READ MY ORIGINAL ADVICE BELOW and MODS PLEASE DELETE THE WHOLE POST IF ITS JUST CONFUSING AND ANNOYING… ITS BEEN CONFUSING AND ANNOYING FOR ME THATS FOR SURE!

THE TRUTH: If you want to sequence ST with OT and have the CC’s mirror in each machine… set PARAM OUTPUT to CC. It works a treat, you wont get fine resolution adjusting tune/cut off/lfo speed etc from OT, but you will be able to access that fine resolution by adjusting those parameters on the Syntakt itself. alls well that ends well, back to making music.

11/SEPT/23: SOLVED WHILE WRITING BUG REPORT. GUESS ITS A GOTCHA REPORT NOW INSTEAD.
(mods if theres a better spot for this please move it)

If you notice weird behaviour from ST’s fine resolution encoders(tune, filt cutoff/res, lfo speed/dep and more) while sequencing it from OT’s midi channels (I’m using OT with CC DIRECT CONNECT enabled) … Go to MIDI>PORT CONFIG>PARAM OUTPUT(on ST)… and set it to NRPN instead of CC.

NOT doing this will mean the fine res parameters will be limited to whole numbers, so if you dont make it from 0-1(for example), the value will bounce back to zero… super annoying.

Definitely do try out sequencing ST from OT if you haven’t already. It’s lovely to use OT’s arp in all its p-lockable glory with the stunning machines available on ST.

[kh]

I just lost a few hours of work while saving frequently. I ended up in some wonky state while copying a pattern and had to reboot- then the machine didn’t boot, just displayed an Elektron logo. Turned it off, unplugged it, tried again, and it booted up to a version of my project from like three hours ago. How is this possible? Was I in some terrible state where the machine was lying to me when telling me it was saving? There should be some process to verify that saves are actually successful, I should have seen an error message instead of being blissfully unaware all my work was being written into the void.

[Watto]

Well the saga continues. After trying out an initialised project the problem seemed to be solved, came back a day later (powered back on) and after 30 minutes of use, the problem returned!!! I’ve lodged a support ticket, but still wonder if anyone else has this issue when sequencing ST from OT’s midi tracks. See below for the details I gave to Support.

Syntakt Bug Report 13-September-2023

Equipment.

-Octatrack running firmware 1.40B (current)

-Syntakt running firmware 1.20A. (current)

My full setup consists of Octatrack, Digitakt, Syntakt, Digitone, Arturia Keystep. The machines are connected to each other via din midi ports(aside from keystep) with iconnectivity mio xm. I have successfully replicated the issue with just Octatrack and Syntakt connected via DIN midi in a closed loop(out-in, in-out) which rules out the possibility of the issues originating in the mio xm or from some conflict with the other machines.

Scenario and Replication of Issues
-I am using the Octatrack’s MIDI tracks (1-8) to sequence 8 of the Syntakt’s tracks(1-4 and 9-12).

-OT’s audio/sample midi channels and Auto Channel are switched off, so they should not be part of the problem.

-Syntakt Tracks 5-8 are midi tracks to control FX on DT and DN, I have tried using an initialised pattern on Syntakt (without the midi tracks) and replicated the issues.

-I have replicated the both issues with CC DIRECT CONNECT on and off, this isn’t surprising as on page 37 of the manual it clearly states:

“Messages received on the MIDI channel of the active MIDI track always behave as if CC DIRECT CONNECT is activated.”

-The issues persist whether CC knobs in CTRL1 and CTRL2 are enabled or not. It makes no difference what CC numbers are selected for control in the Octatrack.

• I have tested this with Digitone also in a closed loop and did not replicate the problem, therefore I believe the issue is with Syntakt.

• I have also tested on an initialised project on Syntakt. Initially the problems went away, and I thought something had happened and the project file had been corrupted… but the problems returned the next day within 30 mins after turning Syntakt on again (ie issue returns after powering off/on after it had ceased with an initialised project.)

Problems

While sequencing Syntakt with the Octatrack’s MIDI channels I have come to 2 sub-satisfactory options.

1. If Syntakt’s PARAM OUTPUT is set to NRPN:

• Adjusting Attack on Syntakt’s Amp page also adjusts Release.

• When adjusting Attack from Octatrack using CC 79, the value appears to move on Syntakt’s screen but then snaps back to zero.

• On Syntakt’s AMP page, adjusting SUSTAIN to 127 causes RELEASE value to jump to INF(occurs between 126-127), this does not happen when controlling SUSTAIN from Octatrack with CC 82.

• Value changes to parameters made from Octatrack’s midi channels are generally reflected on Syntakt, however the opposite is not true. (perhaps this is to be expected as Octatrack doesn’t receive NRPN messages?)

• Octatrack’s MIDI LEARN doesnt work and I have to check the manual to set the right CC (again perhaps this is expected because… NRPN not CC).

• Problems with adjusting fine resolution parameters mentioned below (2.) are not evident.

2. If Syntakt’s PARAM OUTPUT is set to CC

• Adjusting fine resolution parameters like TUNE, DETUNE, FILTER FREQUENCY, RESONANCE, FILTER ENV DEPTH, LFO SPEED, LFO DEPTH on the Syntakt is confined to whole numbers. Decimal points snap back down to the whole number below immediately (ie 1.45 snaps back to 1 as does 1.93) and aggressive turning of the encoder is required to get to the next whole number value, this doesn’t just take away fine resolution but also makes it very difficult to accurately set a whole number value. This behaviour is consistent unless:

The parameter in question is both selected as a CC on Octatracks CTRL1/2 AND the encoder is enabled for use. However the encoders now only give fine resolution when increasing values, decreasing values go back to the same problematic behaviour of snapping to whole numbers just as before. Also since these parameters need fine resolution to be useful, they are not suitable for control by the Octatrack and are thus taking up valuable CTRL knobs for no good reason… this is not a valid solution or workaround.

On Octatrack, if the channel is changed so that it is no longer able to control the track on Syntakt, the issue ceases immediately. This in particular suggests that this is a Per Track or Per Channel issue. Turning channel to off for one track will allow for fine resolution adjustment as usual on that track, while other Syntakt tracks that do have an Octatrack midi track channel assigned to their channel exhibit the problem in the same way.

• Value Changes made on either Syntakt or Octatrack ARE reflected as expected oneither machine’s screen.

-Octatrack’s MIDI LEARN feature works as it is supposed to.

[AM]

I had a similar problem with my Digitakt once. I think the temporary state gets erased (not displayed) when you unplug the machine and reboot it.

For the next time — try to reload the (lost) pattern from the last saved state. SETTINGS > PATTERN > RELOAD WHOLE PATTERN

At least, it worked for me.

(only works, if you haven’t re-saved the project after the reboot)

[avantronica]

I can confirm there is a BUG with AMP : NRPN allocations

@Watto

Adjusting Attack on Syntakt’s Amp page also adjusts Release.

I am not going to comment on OT control of the Syntakt - I am using Max (like a monitor) to sniff/send/debug the midi i/o

There is a problem with the NRPN allocation on the AMP Parameters

It’s also confirmed borked in the Documented allocation numbers for the LSB parameters

What is now obvious was only evident to me after i sniffed the NRPN strings from the ST

The table demonstrates the control issues

The three X in the table show the parameters which do not take control over NRPN, the 1,24 pair for Attack are indeed working as an NRPN control of ADSR Release, however, this Pair is assigned and is sending as the same pair from four Controls ADSR-Attack, ADSR-Sustain, ADSR-Release and AHD-Attack

Only ADSR-Release responds to the NRPN string, so there are 3 non working control assignments on that table, everything else works fine

=========

@Watto

perhaps this is to be expected as Octatrack doesn’t receive NRPN messages?

The ST will understand CC and NRPN when set to NRPN output

The Octatrack will see a string of four CC messages for every NRPN string you send on the ST outwards

CC99 CC98 CC6 CC38

The values of these CC control messages are respectively

NRPN MSB num / NRPN LSB num / MSB val / LSB val

The CC number for Volume is 7, the CC val is identical to sending only the MSB val (or with LSB val 0)

So the OT will not meaningfully process the NRPN string as there is no correlation with the relevant CC numbers

If the OT listens(learns) from an NRPN string it will be meaningless as it will think you are using one of the CCs from the four above, likely CC99, or CC38 depending on its learning algorithm

Everything else you discuss is just a consequence of using a coarse controller to talk to a device capable of finer resolution

When using Max, it’s an easy process to send NRPN and everything works fine - although ironically there’s no LSB resolution on any AMP Envelope parameters to access - so CC is much better for AHDSR as it does not needlessly HOG the serial comms with excess data (that also isn’t working properly)

So rather than awaiting a Bug Fix - just use CCs 79 through 83

[protrusion]

@kh

I ended up in some wonky state while copying a pattern and had to reboot

Did the pattern being copied include MIDI tracks?

[kh]

It did, yeah. Usually if it does that (shows a blank page) just switching tracks will bring it back.

[Watto]

thanks for the reply, I have read, re-read and tested further… I think theres more to it.

1. thank you for confirming NRPN bug with the amp page, thats a weird mistake to make and not catch for several firmware updates.

2. I’m still unclear on wether your advice is to set Syntakts PARAM OUTPUT to CC or NPRN?

a) I only get reciprocity of values at the moment if I set it to CC(by which I mean: I turn the knob on OT and it changes on ST AND I turn the knob on ST and the value changes alike on OT.

b) OT’s midi learn only seems to work when set to Syntakts PARAM OUTPUT is set to CC.

Both of these are important, not having reciprocity means sudden jumps when I adjust values from octatrack that are different to syntakt, big very noticeable problem when changing patterns and the sound on the track has a very different setting. And midi learn is just pure bliss when it works and saves me from looking at the manual…again.

3. [quote=“avantronica, post:396, topic:173511”]
Everything else you discuss is just a consequence of using a coarse controller to talk to a device capable of finer resolution
[/quote]

I’ve edited my post to include some numbers, bold text and italics that didn’t make it onto the post when I copy/pasted over from a google doc. so if you missed it, thats probably why…

@Watto

2. If Syntakt’s PARAM OUTPUT is set to CC

• Adjusting fine resolution parameters like TUNE, DETUNE, FILTER FREQUENCY, RESONANCE, FILTER ENV DEPTH, LFO SPEED, LFO DEPTH on the Syntakt is confined to whole numbers. Decimal points snap back down to the whole number below immediately (ie 1.45 snaps back to 1 as does 1.93) and aggressive turning of the encoder is required to get to the next whole number value, this doesn’t just take away fine resolution but also makes it very difficult to accurately set a whole number value.

I completely understand why I can’t expect fine resolution controls when sending CC values from OT to ST, that isn’t the issue I’m facing, the issue is that now: I cant adjust with fine resolution controls on the ST’s encoders either. that is, not unless I disable the OT’s midi track, then it goes back to normal, but then theres no more lovely ARP and quick mutes for ST from OT’s MIDI TRACKS… I’d like to do all my muting for this setup across OT and DT (DT midi tracks send mute dest to DN so it works out quite neatly when its working). Also, if I select a CC for one of the fine res controls on the OT, I can get fine resolution on ST’s encoders when increasing values, but it snaps back to whole numbers when I try to decrease the value. this is very strange behaviour if you ask me.

When I tested the issue on ST with an initialised project, it worked for about 15 minutes before I switched for the night (and slept well thinking I’d solved the issue) and then another 30 minutes of
beautiful arpeggiated detuned osccillators care of SY RAW before suddenly the detune just snapped back to 0 in an instant, as though that functionality had crashed.

I have done the exact same experiment with Digitone, with PARAM OUTPUT set to CC, there is reciprocity of values between OT and DN, MIDI Learn works flawlessly, can adjust fine resolution parameters like offsets(that only have NRPN options for external control) ON THE DIGITONE and I can adjust Detune, Filter Frequency, LFO DEP (which all have CC options for external control) in their full fine res glory ON THE DIGITONE… Selecting these CC’s and adjusting them from OT only results in whole number, 0-127 values being sent and yes: that is what I expected from the outset.

If it all works perfectly with DN then why not with ST? Its either a replicate-able bug, or my ST is to use your terminology: borked… in which case its still under warranty.

[avantronica]

That’s a lot to digest, you found one bug, perhaps you have a fair point, that will be something support may have looked into - i’m not really using other devices to do bi-directional so it’s not something i will likely try until i have a moment for curiosity

But i will eat my hat if this is an issue with your ST, that’s not remotely likely at all

[Watto]

If it wasnt for the AMP page NRPN numbers being borked(im adopting that btw) I could still send notes (hello arp!!!) and Mod Wheel, Breath Control, etc to ST giving me access to those sweet macros, that would be enough for a long while. BUT!!! At the moment with the ST sending out NRPN, its not just adjusting those AMP parameters from OT that is buggy, it happens when I adjust them on ST too, so I’d be sacrificing control of my amp envelope for an arp and the MW/BC/AT macros… maybe its still worth it, maybe I could work around not having amp envelope by using filter envelopes like I’m on a buchla setup and come up with some super original shit… but I think I’ll probably switch out ST for my A4 for this setup and use one of the A4 tracks as external inputs to make up for losing ST’s FX block, and focus on embracing some more minimalism in my techno.

[protrusion]

@kh

It did, yeah. Usually if it does that (shows a blank page) just switching tracks will bring it back.

Yep, I’ve seen that UI issue too. But I’ve also seen the issue where copying patterns makes things bug out somehow. Usually my patterns had some MIDI tracks, so I’m wondering if there’s a connection.

[djenzyme]

Having a minor issue with a little spike in the Sound at the end of a pattern (which switches to a silent pattern when it ends).

I’ve copied all the pattern settings across to the silent ‘end’ pattern, matched up the effect settings and fx track. No matter what I do, there is always the intended filter sweep down, that glitches slightly at the pattern change.

[Stlaub]

I’ve noticed a weird thing a couple of times but have yet to properly work out what/when.

I feel like both times I’ve noticed this the Syntakt has been hooked up to record directly into Overbridge.

So while a pattern is playing, I think that pressing the track button and a trig to select a track to edit the paramters/mix of makes everything jump in volume for a second, it’s almost as though the filter is being temporarily bypassed or something (the FX track filter is fully open though).

Has anyone noticed something like this? Will have a play and try and narrow it down later but I’ve only ever noticed it happening while recording into OB.

[Watto]

@avantronica in case youre interested…

The workaround I’ve found for just sending note(so arp), MW/BC etc to Syntakt from Octatrack without losing control of fine res values on Syntakt is to make Syntakt’s encoders INT only.

In this setup it doesnt matter if ST’s PARAM OUTPUT is CC/NPRN because syntakt isnt sending either anyway, and the loop that was causing buggy quantisation of the fine res values is broken.

Theres no reciprocity, but I can still use the modulation macros and arpeggiator, and I can mute my syntakt sequences from the quick mute page on octatrack. This allows for a lot of what i’d hoped for from sequencing ST with OT’s midi tracks… but not everything.

The bug still needs to be fixed, seems to me that the incoming values are taking priority over the internal values in a destructive manner. High Resolution values must always be available for adjustment on Syntakt wether that parameter is being controlled by external CC or not. If it works with digitone, it should work with syntakt too.

I’m yet to hear anything more than an automated acknowledgement email from support. Its been 6 days so it’d be nice to hear something soon.

[kh]

If you hit Clear with the step sequencer up, it clears the track. Cool. But if you hit Clear while recording, it clears the whole pattern!

[avantronica]

@kh

it clears the whole pattern!

That’s not a bug, that’s how all the current Elektron sequencers work by design - check out the pdf, there are lots of ‘expectation’ pitfalls

[kh]

Hm, you’re right that it’s in the manual, but I’d still consider a design flaw that bad a bug

[rtme]

I wouldn’t consider this a design flaw either.

Many buttons on the Syntakt and other elektron devices have different functions depending on what you’re doing / which menu you have open. This makes sense, as you can’t have dedicated buttons for all functions of this machine, you would end up with 20 clear buttons for instance.