AR Direct Jump "concept"

SIM80 · 2015-01-05T19:32:50.000Z

I submitted a ticket on this back when the original threads on the topic popped up. At that time they said they were aware of the issue and that they were hopeful that they would be able to fix it. My guess is that they found a bug that was deep within the structures of how sequencing works and can’t fix it without reworking a lot of other code.

With me, it’s an issue even without using INF and using the Rytm standalone. Like panelist has described, it gets hung up from time to time and doesn’t jump properly.

xidnpnlss · 2015-01-05T21:22:09.000Z

""icosa"":

My guess is that they found a bug that was deep within the structures of how sequencing works and can’t fix it without reworking a lot of other code.

I am not a programmer so this all purely conjecture. But I imagine Olle’s response is verbatim what the programmers told him: that according to the concept of the code they wrote, it was intended for stand-alone only. Maybe left over from their old machines?

But, what I don’t get it, is how Overbridge plays into all this. I mean, they should have known it was on the horizon when coding the A4/AR. Right? So they should have taken it into consideration. Please, programmers, shed some light…

That email paints a grim picture for what I had hoped to be resolved issues with Overbridge and synch. He never says anything about it being cleared up with the release of Overbridge. And I’§m sorry but isn’t the entire concept of Overbridge to synch “seamlessly” with a DAW? How in Gods name are they going to be able to claim that without working out these snych bugs? I change a meter once and my synch is over. Maybe they should put an astrix somewhere “not intended to be used with complicated rhythms.”

secretmusic · 2015-01-06T12:09:47.000Z

""panelist"":

I’d side with you on this one. It happens to me even on 4/4 patterns with LEN set - I can’t always recreate it but sometimes the sequence will slip a step, sometimes the pattern I’m switching from will play a 2nd time before switching. Direct Jump is definitely buggy.

Well, to be really honest, I haven’t even tried switching patterns in real time since I found out about this bug. I’m very surprised to see that sync problems appear with “regular” master length.
This makes the answer even more bullsht in my eyes. It’s not a conceptual decision, it’s a bug. And disguising it as a “concept” is insulting my intelligence, I’m sorry. Elektron did NOT design a drum machine that can’t switch patterns while synced on purpose. They messed up somewhere, and are now telling us that is was by design.
I call bullsht when I see it, and this is it.
And stating that sync and direct jump don’t go “hand in hand” is even worse, when other drum machines do it just fine. Seriously ?
Please, fix this issue before Overbridge. If the AR can’t sync properly, should that not be a priority, so that Overbridge is build on SOLID FOUNDATIONS. You’re building Overbridge on top of something that does NOT work. This is nonsensical.

ps : i’m really annoyed by this.

/rant

Cheers !

panelist · 2015-01-06T17:08:42.000Z

To be fair, they may have meant that the problem is conceptual from a programmers view, meaning that they can’t just go in and say 'ah ha, here we go - arithmetic error on line 4862". It could very well be an issue with the model they set up for the structure of the sequencer, but that’s not really a consolation either because that means it’s a deeper problem that could take a longer time to resolve. Let’s hope that it does not become an irreparable problem due the added layer of Overbridge.

mercurio · 2015-01-06T18:40:30.000Z

I use Direct Jump mode a lot with AR and A4, both slaved to Ableton Live throught Innerclock’s Sync Gen II.

I usually work on 4/4 patterns. As some of you commented before, pattern changes doesn’t work as expected 100 % of the times. Rarely, they run to the end of the pattern before the actual pattern change occurs. It could be a bug.

Regarding the timing it works 100 % flawlesly though.

I think there must be a problem with MIDI timing through USB. May be Live’s MIDI clock output isn’t enough accurate to sync correctly the AR. Although IMO this could be a common problem with any DAW because of innacurate MIDI timing output of computers in general…

Anyway I think getting a rock solid MIDI timing through USB should be a priority for the upcoming Overbridge utility. Remember that Innerclock’s interface outputs sample accurate MIDI timing. I would like to see something similar in Overbridge.

So at least Direct Jump mode works perfectly syncing to physical MIDI with 4/4 patterns. My AR stays in perfect sync with my DAW no matter how many direct jumps are involved.

secretmusic · 2015-01-06T21:28:19.000Z

""mercurio"":

I use Direct Jump mode a lot with AR and A4, both slaved to Ableton Live throught Innerclock’s Sync Gen II.

I usually work on 4/4 patterns. As some of you commented before, pattern changes doesn’t work as expected 100 % of the times. Rarely, they run to the end of the pattern before the actual pattern change occurs. It could be a bug.

Regarding the timing it works 100 % flawlesly though.

I think there must be a problem with MIDI timing through USB. May be Live’s MIDI clock output isn’t enough accurate to sync correctly the AR. Although IMO this could be a common problem with any DAW because of innacurate MIDI timing output of computers in general…

Anyway I think getting a rock solid MIDI timing through USB should be a priority for the upcoming Overbridge utility. Remember that Innerclock’s interface outputs sample accurate MIDI timing. I would like to see something similar in Overbridge.

So at least Direct Jump mode works perfectly syncing to physical MIDI with 4/4 patterns. My AR stays in perfect sync with my DAW no matter how many direct jumps are involved.

It’s cool if you work in 4/4. Not so much when you want to do any kind of odd time signature or polyrhythms. I use MIDI not USB as I would like to play out live with the AR, and the problem still occurs. The bug has been confirmed by Elektron, and they are able to reproduce it reliably with the project I sent them. It doesn’t make a difference if it’s getting clock from USB or MIDI. It’s always broken.
I’m glad to hear that it works for you in a “best situation” scenario, and with the help of a sync gen. But there are some for whom it does not work, and it doesn’t look like it’s getting fixed anytime soon, or even at all.

Dataline · 2015-01-07T00:16:59.000Z

I too find this bug getting in the way, if using the tracks polyrythmicaly. As mentioned, it works most of the time when not using the sequencer polyrythmicaly.

Fingers crossed for a fix in the future.

I can not say much about how it can be fixed but i can see that it probably is not like a quick fix. The sequencer side of things must be very complicated in the OS.

benbin · 2015-01-07T06:17:12.000Z

Have an OT and just got my AR … this bug is really disappointing.
I appreciate the feedback from Elektron, Good communication = respect.

Its a drum machine or percussion synth based on the legend of THE Elektron sequencer.

Its reasonable to be suprised that there is a problem with this feature.

Hmmph

mkdsl · 2015-01-07T06:45:53.000Z

Yeah, it is more stable if the AR is master than when slaved, but even on 4/4 stuff it can skip a beat if I get too funky with pattern changing. My solution: stop the master and start it again. Feels like a mini break, and is rarely perceived as error. In fact it can bring release in a set, when there is so much energy involved in things changing you forget that ears need rest sometimes… Other people’s, that is.

secretmusic · 2015-01-07T11:10:44.000Z

""mkdsl"":

My solution: stop the master and start it again. Feels like a mini break, and is rarely perceived as error.

That can work once or twice in a set at most, but hardly a fix. And before you stop it, you do get a few beats of just wrong stuff. It’s not a solution that’s acceptable to me, I’d like my drum machine to switch patterns and stay in time. Which is why I haven’t performed with it yet. You see, it’s not just this major bug that gets in the way. It’s also the fact that, BECAUSE of this bug, I do NOT trust the AR to behave correctly while performing. And that broken trust with this expensive piece of kit makes me leave it at home. I’m sure it would work fine in some cases or for some people, but I choose to play on hardware for the reliability factor, and the AR is, in my case, very unreliable.

11 · 2015-01-07T14:48:33.000Z

Maybe we can get a tutorial video on what to do when performing live and your gear doesn’t maintain sync? lol

Olle · 2015-01-07T15:20:43.000Z

Can you test to set the LEN parameter in the Adv. Scale setup to a number instead of INF? I know 1024 steps before reseting is limiting the variations of rythms you are able to create but it solves this issue right now when using Direct Jump and external sync. Use the CHG parameter to set when the pattern shall change in Sequential mode.

I hope 1024 steps still shall give you complex patterns since its 64 Bars long and it helps you out in this case until the issue has been solved.

I’m very sorry if you got the impression that we will not work on this issue, since that is absolutely not the case. I was honest about the situation and the timeframe.

Again sorry.

/Olle

secretmusic · 2015-01-09T18:24:51.000Z

""Olle"":

Can you test to set the LEN parameter in the Adv. Scale setup to a number instead of INF? I know 1024 steps before reseting is limiting the variations of rythms you are able to create but it solves this issue right now when using Direct Jump and external sync. Use the CHG parameter to set when the pattern shall change in Sequential mode.

I hope 1024 steps still shall give you complex patterns since its 64 Bars long and it helps you out in this case until the issue has been solved.

I’m very sorry if you got the impression that we will not work on this issue, since that is absolutely not the case. I was honest about the situation and the timeframe.

Again sorry.

/Olle

The issue is not about what ways can be used to avoid the problem. You said yourself that you could not promise when or even if it could be fixed. THAT is what is worrisome. And stating that sync and direct change don’t go hand in hand is puzzling, when other drum machines do it just fine.

The real problem, is that you’re building Overbridge on something that doesn’t work. I don’t understand how this is not a glaring mistake. I would like the discussion to stay on this point, as it is the heart of the problem. Why are you building something so complex as Overbridge, but basing it on a sequencer that doesn’t behave properly ? Not only for polyrhythms, but also for plain 4/4, as confirmed by some users ? What will happen when you finally decide it’s time to try and fix the sequencer ? How will that affect Overbridge users ? Those are the questions that bug me, and clearly, no one with a real answer is communicating with us on these forums.

Snatches · 2015-01-09T19:28:06.000Z

1024 steps is a reasonable workaround for this issue. Even Steve Reich isn’t going to notice a polyrhythm repeating on a 64 bar pattern.

secretmusic · 2015-01-09T19:45:16.000Z

""Snatches"":

1024 steps is a reasonable workaround for this issue. Even Steve Reich isn’t going to notice a polyrhythm repeating on a 64 bar pattern.

Again, this is not the point of this thread. The point is that Elektron are building a feature around a sequencer that does not work as planned. I thought this was sufficiently clear from my previous post ! What will happen to Overbridge when / if they decide to fix the sequencer ? Is it not more logical to fix glaring issues before building upon them ?

Besides, this bug has been confirmed by users with 4/4 and a set length, so this post has NOTHING to do with how we could, in theory, circumvent the bug.

jamrod · 2015-01-09T20:33:16.000Z

""secretmusic"":

[quote=“Snatches”]1024 steps is a reasonable workaround for this issue. Even Steve Reich isn’t going to notice a polyrhythm repeating on a 64 bar pattern.

Again, this is not the point of this thread. The point is that Elektron are building a feature around a sequencer that does not work as planned. I thought this was sufficiently clear from my previous post ! What will happen to Overbridge when / if they decide to fix the sequencer ? Is it not more logical to fix glaring issues before building upon them ?

Besides, this bug has been confirmed by users with 4/4 and a set length, so this post has NOTHING to do with how we could, in theory, circumvent the bug.[/quote]
I think you’re making some assumptions about how the sequencer and Overbridge interact. It’s very possible that they’re two separate systems within the AR’s OS, and that changing the sequencer later won’t affect what Overbridge is adding (audio over USB).

The bug is annoying, but Elektron has responded and seems to understand the issue. It doesn’t necessarily seem like they’re building Overbridge on shaky ground. They’re not going to halt work on account of this relatively minor bug.

Correct me if I’m wrong – I’m just not seeing the connection.

djadonis206 · 2015-01-09T20:44:36.000Z

Can someone explain and point me in the direction of how INF works? Wont my individual tracks just loop / start over when they reach the end of their respective bars?

Why would you set the master length to INF unless you’re using really long one-shot trigs. Which are, or are not, available on the AR?

bluewolfse7en · 2015-01-09T22:14:10.000Z

It means that all Tracks/Patterns won’t reset at the same time at the end of a 'Master" cycle.

If you had Master setting of 32 with 4 Tracks running thus

15 steps
12
7
6

Then on Master step 32

would be on its own step 2
would be 8
3…4
4…2

Master will now reset all 4 tracks back step 1.

A setting of INF would allow them all to carry on on their own merry way.
Track 1 would reach step 3, Track 2 step 9 & so on.

djadonis206 · 2015-01-09T23:37:29.000Z

Ok, that makes sense. I was assuming different tracks would all be set to some even numbered length (16, 32, 64, etc)

But that makes sense if you set a track to an odd number of steps - it would loop infinitely at that odd number rather than restart on the 1 when the master length looped back around…

Gotcha! I learn something new everyday

Olle · 2015-01-10T00:50:48.000Z

Besides, this bug has been confirmed by users with 4/4 and a set length, so this post has NOTHING to do with how we could, in theory, circumvent the bug.

I’m more than happy to receive sysex exports of project with this issue since I have no data on this.

Your support case with the project files both for Abelton and the Rytm, including instructions were excellent and very helpful. This helped me identify and confirm the issue and narrow it down to occur when following conditions together are met:

Using Direct Jump.
and
Using External Sync.
and
Pattern has LEN set to INF(any other value is fine).
and
Pattern has individual track length.
and
Changing Pattern.

In your case, setting the LEN from INF to 1024 might help and affect the pattern minimally. It would help if you could test this to confirm. We will continue to work on what is causing this issue and I have given you an honest timeframe.

If you find any other conditions when this is occurring please let us know, but right now these are the only by us confirmed conditions.