Trouble with BT & XT classic - MK1 vs MK2 different?

Hi community.

I am used to produce bass lines with the rytm MK1 in our studio.

I bought a rytm MK2 not long ago and I got real trouble producing a reliably looping BT or XT pattern, especially if I use it as a bass line.

Both BT and XT classic engines (or the OSC of course) seem to be very unstable - beyond their “analogue nature” (see MK1 vs MK2 below).

The main and most obvious trouble is some notes getting “ducked” randomly. (They do not seem to be played, since they are lower in volume and darker in timbre.)

Here you can here what you see:
BT_Classic.zip (676.1 KB)

Here is a 1:1 comparison MK1 vs MK2

You can see that the MK1 does not produce this random/occasional ducking.

Here you can here what you see:
rytm mk1 vs mk2.zip (246.3 KB)

The BT in this examples is simple:

  • cleared Project, Kit and Sound
  • used BT classic standard/initial settings
  • BT OSC tuned different
  • triggered by different patterns (faster notes cause trouble more often)

If you want to test around here is a kit and a pattern I sent to support also:
BT & XT TEST - PATTERN & KIT .syx (17.7 KB)
BT 2nd call - PATTERN & KIT.syx (17.7 KB)

Please report back how your MK2 sounds. Is it having the same issue? Or is it more stable, just as the MK1?

2 Likes

AR2 here.

BT & XT TEST sysex appears to have an empty pattern A1. Is that correct?

BT 2nd call sysex comes with a pattern B1 that I was able to test. I do get the occasional dropouts like in your audio sample. Sounds like an envelope not triggering. It reminds me of some MD machine algo’s that have a somewhat similar behavior. It has to do with analog modelling there. Whether it 's intended by Elektron here is unclear to me. It’s definitely weird that the mk1 doesn’t do it.

Hi dtr, thanks for testing.

I always heard/read “there is no real difference in sound between the mk1 and the mk2” just some performance improvements.

Support says: “This definitely sounds odd.” and “I compared to an MKI unit here and the MKII is actually more steady/consistent sounding.”

Not here though on my end. Exactly the opposite seems to be true. I am not sure if its just my device or all mk2-s.

I can’t program any consistent BT-pattern that sounds at least mostly the same without having occasional low-volume notes, which seem to just get “dropped” in a complex beat.

Not good for any kind of BT/Bass loop and also not an expected behavior after “…and the MKII is actually more steady/consistent sounding.”

Right now I don’t see how I could use that for a live performance at all. In the music I produce I need reliable BT/bass lines. Analog differences like the mk1 does are more than fine with me.

About the kits/sounds/patterns for testing:

BT & XT TEST sysex appears to have an empty pattern A1. Is that correct?

If you go to SYSEX RECEIVE and select an empty pattern and press “YES”, “WAITING DATA” starts to flash on that pattern and you can load the one and only pattern in the syx file to that particular pattern of your choice.

Load the Kit to the same pattern and it should work.

Here is one even more obvious test and example:

BT_supp2_loop.zip (1.3 MB)

Here is the pattern and the sound for TRACK 5 - BT, please load it to a CLEARED KIT.

BT supp2 sysex.zip (1.2 KB)

1 Like

This also happened to me with the mk1, that’s why I didn’t pay much attention to it, maybe you just need to calibrate

2 Likes

indeed. try recalibrating your mk II to see if things improve.

1 Like

Intrigued I did my own examination to rule out stuff, clear kit/pattern etc etc

Simple repeating constant pitch notes, tuned lowish to present the most obvious difference between the two devices (bi-wired by midi so changes on one are reflected in the other)

It’s clear from the get go that for certain synth settings there is a remarkable difference between Mk1 and Mk2 - it’s fair to say Mk1 seems more ‘consistent’ , but it’s not so straightforward to leave it at that

I wouldn’t say the Mk2 is less consistent, but its behaviour at catering for the successive notes is clearly different (so will be pattern and settings sensitive)

Here are some screen grabs, the first two are from the same recording, just rendered differently, the third is what happens if you add a bit of noise or a transient to the same pattern - they both behave the same

When there’s no transient or noise, they seem to stack/build differently when they are run into one another in the first example with the absence of transient/noise - highlighted on the fourth screen grab (same image as image 1 but notated) - a very very different vibe is heard !! Just look at the trajectory of those rough red lines scribbled on !

I now wonder if there’s an element of there being subtle differences that can be tuned out to give comparable response or whether there are differences under the hood that have been undisclosed

In my experience there’s a lot more low rumbling on Mk2 whereas Mk1 is generally tighter

They appear to be (in some cases) quite different (which doesn’t really bother me tbh) but it’s good to know or to examine further

PS - there are two rim clicks on my Mk2 recordings to distinguish them whilst I cut/pasted these up in an editor - it was a looping 10 step pattern (pattern below) and the audio was cut from the middle of its longer looped recording (not from the start) so it’s in a steady state so to speak - tuning in the -20 > -30ish range iirc

:black_large_square: :black_large_square: :black_large_square: :black_large_square: :white_large_square: :black_large_square: :black_large_square: :white_large_square: :large_blue_circle: :white_large_square:

every figure : Mk1 on top, same pattern and settings so it’s a direct comparison

This has echoes of the A4 bass drum patch (built on a resonant filter) it has to build up steam to get the filter pinging - I wonder if the introduction of a transient/noise to the synth settings is the key to consistency (as per 3rd figure)

But I can agree that there’s a clear difference when the parameters are mapped 1:1 between Mk1 and Mk2

I would also love to know if there are indeed tiny differences to the circuits

2 Likes

Hi, I did calibrate a couple of times to exclude that. It does not help.

Thanks for your feedback and interest.

In your examples 1 and 3 I can see, that the top mk1-loops look rather consistently repeating volume-wise, whereas the bottom mk2 loops are inconsistent, the two loops differ.

I layed loop A (first half) on top of loop B (second half) to show exactly what I mean:
Your Example 1

Example 2 is not enough to show my point here, but I can see the difference between mk1 and mk2 of course as you say

Example 3 is how it should always look like for the mk2 as well.

Your Example 4

As you can see the loops of the mk1 are pretty much the same (its just black if layered inverted) whereas the mk2 has big differences in the volume of various notes between the two loops.

That is exactly why it is inpredictible to use. You never know when and which notes get dropped randomly.

I don’t mind analog variation and behavior (like the mk1). But the loops should sound consistent of course on a drum machine (like on the mk1, unlike mk2).

If this is a major difference (not just a bug on some devices), then I think it is actually very valid to say
"… the Mk2 is less consistent" (than the mk1)

2 Likes

The Mk2 acquires full tilt a bit quicker but it also seems to lose it quicker - in this use case (as the signal is by definition low frequency, it won’t harm to add a bit of noise and filter it as this will excite the circuit quicker - for whatever reason, in this use case the Mk2 is different, not wrong, it just behaves differently - the way it behaves will either be better, worse or neutral for various users - i think the Mk1 feels more true to the pattern, but maybe it’s like this because the circuit is noisier or otherwise different and the circuit gets going easier

Add a small dash of inaudible noise to the Mk2 to see if it becomes more like the Mk1

I don’t understand why the Mk2 loses its oomph on the third hit though and i don’t understand why the Mk2 is not repeating its own thing as you have indicated on the overlay - this seems like a legitimate question to ask for an explanation on - i can see that this could lead to a different playthrough even on the same looping track

The black overlay comparison within the same repeating passage is a very compelling observation

3 Likes

Noise actually isn’t making enough of a difference to help here and it’s fair to say that the Mk1 is still generally more consistent/right loop after loop - no doubt there’s something different about them that’s not too obvious to rationalise from this side of the fence (theories on noise may be very misplaced and transients only mask it)
whilst it’s not how i’d likely use it, i’d like to understand why it’s happening

when i get a moment, i may sequence both externally simultaneously and capture one channel from each to see what that reveals - i’m not sure how timing might come into this (or whether it was actually the cause at all) but i noted that passages sounded slightly different when the Mk2 clock was driving the Mk2 and when the Mk1 was driving the Mk2, but that may have just been this observed general fluctuation more evident on the Mk2 - either way something is amiss

I took the original upload for figure 1 (tweaked below) and shifted the second loop underside to underneath the first pass - it more clearly shows what you had identified in the overlay diagram above (in case you point support to this thread) -the red arrows are the North peaks of pass 1 mirrored, but the volume profile difference is already quite obvious - so much so I am wondering if the test was even robust !

Not only is it different and less predictable per pass, but it’s not even the same on the next pass and this was all chopped from the middle of a longer repeating passage of 8 repeats

The Mk1 is handling this far better - whereas of the six Mk2 beats only one is close to the first pass - this really doesn’t look at all right (I may need to redo this)

4 Likes

I hope that soon the tom row receives some love from elektron.
26

1 Like

I don’t understand why the Mk2 loses its oomph on the third hit though and i don’t understand why the Mk2 is not repeating its own thing as you have indicated on the overlay - this seems like a legitimate question to ask for an explanation on - i can see that this could lead to a different playthrough even on the same looping track

Not only is it different and less predictable per pass, but it’s not even the same on the next pass and this was all chopped from the middle of a longer repeating passage of 8 repeats

Yepp, that’s exactly my point. And it is not just the BT but also the XT engine that behaves unreliably. So really the whole Tom Row.

Hi guys, bad news :slight_smile:

I started to use different engines for some new productions. Now I faced this while using the BD Plastic Engine:

While I was running the loop, I noticed some standing out (irritating :smiley: ) mistakes occasionally. I recorded just a couple of minute and it happens every 2-3 minutes or so that one hit is just different than all others in volume envelope shape.

Here you can see clearly the difference to the other hits:

here you can listen:

Rytm 9 recording: 0:49 s

Rytm 7 recording: 1:33 s

The BD Sound:
BD Plastic_Crunch.syx (207 Bytes)

The pattern is simply 16 long with plain hits on 10 and 12, repeating. No trigger lock or vel difference/mod or anything. No LFO applied either.

Does this happen on your devices at all if you record say 10 mins?

Just a random thought due to how the “wrong” trigs sounds:

Can you rule out electro-magnetic interference with the electronics of the AR? Like a smartphone which tries to register itself with the nearest cell every now and then?

yes, I can absolutely.

Is the Track that is having issues part of a choke group? Maybe it’s choke-track is triggered at that step? Sometimes with choke tracks you will have some audio bleed / glitches if they hit at the same time.

You are right, but neither the BT/XT Classic nor the BD Plastic engine are used on tracks that choked.

When I face issues like that, I save the sound and load it to a clear project and kit to test.
I remove all modulation, if applied within the sound step by step, to come to the cause. In all 3 cases (BT/XT + BD plastic) it is seemingly the raw engine causing this with some basic settings.

Not all engines do that though. Did not have any problems with the HH, CY or snare engines so far. But I did not test everything excessively. I try not just to test for weeks by now, but also have some fun and produce. When I face an issue, I post it though of course.

My concern is that I can basically give up playing live. The beats do many rather experimental glitches, the device is lagging and freezing. I will swap my device and hope it get’s better.

Thanks for your interest in this case.

Good news: PROBLEM SOLVED by swapping the device.

It was an issue of my device after all with a bunch of other bugs and errors that do not occure anymore on the new unit after excessive testing. BT, XT and BD Plastic engines are running steady and consistent between the loops, as expected.

If you got similar trouble, insist on swapping. In my case support did not tell me for sure, if it is only a unique problem.

Here you can see the proof :wink: The top one is the new one with the settings of my example BT_supp2_loop.zip

The bottom one is the same on the old device, as posted already. Very different.

4 Likes

hmm, that’s interesting !

i think i need to create a new test case myself to be doubly sure first - presumably looking at more voices too

I told you …

1 Like