Just got mine, been playing with it a bit. So far so good.
The issue with the gap changing kits isn’t an issue really IMHO… I’m switching from the factory patterns and there is the tiniest (maybe 1/32 note?} bit of silence but I honestly didn’t notice it. Also the fact that you’re changing kits and completely new sounds are coming in (at least when changing the factory kits, they are very different from each other), the very short silence is the last thing you notice when they change really.
Of course it would be nice if it wasn’t there at all but I’ve played live for almost 20 years and with everything else going on in a set this wouldn’t even be noticed… If it’s the only instrument in the rig maaaaaybe someone would hear but I wouldn’t count on it. The clips that were posted sound worse than what mine is doing so not sure what’s going on there.
Positive notes are that the feel of the encoders and sliders are very solid and smooth. I wasn’t expecting them to feel cheap but they have much more resistance than I was expecting… buttery smooth. And most importantly it sounds amazing. Full stop.
One thing that’s giving me some heartburn is the way that p-locks work on this as opposed to the Elektron paradigm: If you adjust a parameter on only one trig, it doesn’t simply override the default value. So, if you modify a parameter on one trig, you need to modify that param on every trig. Or am I tripping and there’s some better way?
Maybe don’t use a plock at all if you’re trying to adjust all the sounds? Just adjust the sound of the gen/track, or am I misunderstanding your question?
All I am trying to do is adjust a parameter in the Elektron way. For example, I have 4 trigs down for a bass drum. I want to change the pitch of only one trig. By p-locking the trig I want the pitch change on, it affects the pitch of the other 3 trigs. The only way to reset the other 3 is p-locking them back do the default value.
Just did a timing check. The latency is insanely bad if it’s getting clocked externally, unless you choose layered gens. It still is about 5ms later than my other drum machine which is receiving sync from the TR1000.
Layered Gens on 64 buffer - 7ms. Layered Gens off - 51ms!! Crazy.
The thing is that it should be 2ms, that is where my SP1200 is landing, and that is the round trip latency of my RME. So it’s a small annoyance, but I will have to push the Roland 5ms ahead to have it’s audio be synced up in the DAW with my SP1200. It would be nice if I did not have to do this.
Where it gets real strange is when I turn off the MIDI coming into the TR1000 and use it as the master clock, as Roland says to do in the manual. In layered gens mode it is unusable, the drums sound completely wonky and out of time. Not even at the right tempo or anything. So I switch to the other mode and now the machines both play in time but the roland is WAYYY behind the SP1200.
So basically I am unable to send MIDI sync from the Roland clock to another drum machine if the Roland is the master clock.
The SP1200 clock is super reliable and accurate. It records dead on the grid, just 1-2 ms of latency which is nothing.
I am going to write to Roland Support and see if they can drop the latency even lower, and see what the heck is going on with my machine not being able to send a clean clock.
The 1 caveat here is I record in double time. So I record at 244 BPM instead of 122. I do not think that should be an issue though. The TR1000 goes up to 300 bpm.
The only good news is the TR1000 does not jitter when getting clocked from the DAW via a sync clock. So that is stable. But the clock still has some serious issues, especially if it’s supposed to drive an entire studio as a main clock. That literally can not happen right now, at least with my unit.
I am updated to firmware 1.1.1 . Let me know if any of you can recreate this.
Yes, if direct analog out without ever being ADed and DAed is important (which I understand) then this is not possible.
Theoretical technical ideas
This could be an option: Routing everything summed through a 192/32 buffer. That would introduce a delay. (5ms? more?)
I like the other suggestion of “lazy load” from above better. Load different tracks/instruments one by one, depending on if they are played in the “1” or later.
Given that pattern change happens some time after the command - there would even be time to pre-load some instruments that get played on “1” in next pattern.
Also, a check of how similar instruments are, could help. If same instruments on same tracks but slightly different settings - no gap but quick morph (the tech is in there - see slider), that could also be happening sequentially depending on where next trig.
FX tails could in theory be pre-rendered, buffered and played back for the time of 2 trigs to 4 trigs, same is for non-analog machines, while in background machines are switched already.
Not great because that doesn’t send start/stop. So it’s always on which means the drum machines will never start in sync with each other. I just tried it.
I am able to use the tr1000 as a master, and my mpc as a slave, with rock solid timing using din MIDI. but if you flip it and run the mpc as the master, there is flamming on the downbeats even with layered gens on. if i nudge the tr1000 ahead by 1-5 clicks (not sure what values they are using) it is in sync but as has been noted, this isn’t ideal to have to do this everytime you press play on your master. I sure hope they have a fix for this. The Tr8s, hell, even my boss 880 are all easily synced to my mpc. like, i’ve never even had to think about it. theyre just bang on.
How rock solid is the timing? Like if you record them both into the DAW are they hitting at the same time? Because mine do not. there is like a 5ms difference. Which doesn’t seem like much , but it really does make a difference. And that is when they are synced to the DAW. It’s worse timing if the TR1000 is the master, I do not get it.
I had the same issue, i tried to sync it to my SP1200 and it was flammy. I was really surprised though when I had it as the master and it was further off then when synced to the DAW. Just not good.
this hasn’t been my experience… just to be sure, are you choosing AUTO for sync source, or are you specifying USB or MIDI? On AUTO, everything is unusable for me it seems.
With the MPC (slave) clocked TO the TR1000 (master) via 5 pin MIDI, everything is super solid. No flammiing. Layered Gens on or off, it doesn’t matter. that works great for me. I only experience issues with external sync, and they are exactly what you describe. Not great with layered gens, and pretty mch unusable in the default state.
the problem is, for my workflow, you can’t record to the audio tracks of the MPC unless the MPC is the master.
Wit the SP1200 clocked to the TR1000 as master, that is the tightest I get. And yes it will sound fairly tight, but you will most likely see a delay between the MPC and the TR1000 in the DAW.
Can you try the TR1000 as master, and set to internal. Send clock to MPC that way. Leave the DAW out of it completely? When I do that, I can hear the TR1000 lagging behind the SP1200. You may be doing this and I misunderstood.
I am going to power mine down and let it sleep and see if anything changes when I fire it up again tomorrow.
i’m not saying to cache the analog audio, but cache things that take time to load into memory, or have a longer loading time like samples, i would guess that the analog routing would be very quick. 100ms is quite a long time. i would be confident that it could be much less, i do wonder if the timing is related to tempo, not sure what tempo that sample mp3 was at.
we do similar things by caching presets when preset chaining on metropolix, the eeprom is hella slow, and we only have 1mb of ram hehe
that scenario you describe is the one that works solidly for me. I haven’t done anything with a DAW at all, just the MPC. Thats literally the ONLY scenario where I don’t hear the TR lagging behind. all of the scenarios where the TR is the slave have varying degrees of delay/flamming, depending on the layered gens setting, or the auto vs MIDI settings. “AUTO” is the worst.
Ok I am wondering if it just needed to get shut down and reset. Ill see if things change and stabilize tomorrow. I put it through a bit of a workout trying different syncs, internal, external…etc.
You report things with the DAW in the mix, which of course adds extra buffers etc. Let me give you an alternative measure, by simply using the sync shift on my sample accurate clock to get it in time with a Eurorack drum module (I usually just use impulses or rim shots).
I found about 8-9 ms with the “layer gen” setting off, and about 50 ms with it switched on.
Like you, I am puzzled why the latency is that high. The TR8s has about 2.2 ms.
I do factor the daw latency in. I have it dialed in and have recorded a lot of gear in it. It’s 2ms that is added. Not a problem there. It’s the Roland that’s just too far from behind from my other gear that is currently a bit of an issue.
If you read what I wrote, it was not a criticism (apologies if it sounded that way!). It was a confirmation using a different approach to measure the latency of the machine (which, as far as I can see, Roland does not report in the manual)