I’ve seen all the reply’s and figured I would add Bomes Midi Translator to the discussion.
They now make a “BomeBox” piece of hardware.
I haven’t used the BomeBox, but I’ve used the software quite a bit.
It started out as a Keyboard midi translator.
So it excels at turning your computer keyboard into an insanely powerful midi controller.
These days you can send it all kinds of midi input messages and have that trigger a massive variety midi output messages.
It seems CH.LEN (or CHNG on my Rytm) doesn’t work exactly how I thought.
When changing patterns manually, on an AR, CHNG does work as described. It’s the count of steps from the start of the pattern until changing to the next one. However, in Chain Mode… it’s not. In Chain Mode, the sequencer automatically changes pattern after the LEN (or M.LEN) setting. Which is dumb because it means patterns behave differently depending which mode you’re in.
Ok… I’m gonna raise (another) bug with Elektron about this.
I think it’s ok for some compositions to use LEN to set pattern change times for Chains, if you’re gonna be performing. What bothers me most is the inconsistency and confusion this causes. Plus the manual describes it wrongly. It also means that you have to use polymeter reset rather than allowing constant polymeter shifting but with patterns that change after a fixed time (which is probably what most people that use polymeter want). The way it works now, you can’t have a pattern based on “pure” polymeter that doesn’t reset, and have it play for an arbitrary time (say for a long build-up), without turning off Chain Mode. You have to do the transition out of that manually. And if you want a pattern with all the same musical content that you can play “on chain”, as well, then you have to have two copies with different scale characteristics.
Elektron could fix this by respecting CHNG/CH.LEN in Chain Mode. Can anyone see any downsides to that? Any reason NOT to fix it?
It seems ridiculous that not only Elektron can’t/won’t fix it… they also perpetuated the misinformation in the manual for their newest instrument. The least they could do is correct the manual to match the behaviour. I’ve read somewhere here that there’s internal company politics around these aspects of the sequencer. I can imagine that happening around the implementation, due to historical code, long term customers etc… but I’m really struggling to understand how the politics could be so bad that the manuals are allowed to be wrong even for brand new devices (@elektron).
In my experience it works perfectly well for defining a 1, 2 or 4 bar change time when you give the command.
THE biggest issue is, and always seems to have been, that the Elektron devices can’t take real-time Prog Change messages on the very first beat of a bar.
It does work as described when manually changing patterns…
It doesn’t work as described in chain mode or when chained to other Elektron devices
Seems to have always been this way for the analogs and digis…
(The Octatrack can do it!)
There’s workarounds. As I said upthread, and you said, manual changes work as described. This means you can do most things you’d hope to do in a performance using a mixture of multiple copies of the pattern with different CH.LEN/M.LEN combinations, and using both Chain mode and Manual in a performance. I think this is OK… (but not ideal)… but the manual being wrong, across multiple devices, across multiple releases of the manuals, including the brand new instrument (@elektron) just boggles my mind.
Guys - Studio One is doing this for me, i can take a video and show you. But essentially i can start the song anywhere and the pattern will switch immediately to the correct pattern.
I have written the Program Changes as automation and even moving the song position pointer around the arrangement will change to the correct patterns ready to go…quite neat actually
I don’t think it’s easy, especially when connecting multiple devices.
Given that you can tell the sequencer to play the next pattern at any time, the sequencer needs to queue up the changes. This means it needs to collect them before changing to the next one. If you instead place a change at the start of a pattern, the sequencing goes like this:
press start
send MIDI Transport Start
start reading the pattern
send PRGM CH
you’d still have to play the rest of the pattern
if you didn’t do 5, the queuing mechanism would have to detect the difference between a PRGM CH issued automatically at the start of a pattern, and one played manually. A single device might be able to do this, but a receiving device couldn’t know the difference without the sending device sending extra data. Sending extra data a) takes you even further from the spec and b) adds the small timing overhead of processing the extra data, which you’d want to avoid in a realtime environment.
(I admit I’m attempting to fit a solution to the observed behaviour rather than designing for the expected behaviour)
I think there are a bunch of us who are happy achieving it externally.
I have a fantastic system of my own using Ableton and a Novation Launchpad Pro.
I control my Analog Rytm, Syntakt, Digitone and Analog Four, all in perfect sync and launching random patterns at will (i.e. I don’t need to keep patterns matched A1-A1).
So, that’s not really the problem, the external solutions are there for when you really want to multi-Elektron. You just can’t do it in an Elektron-only setup.
I got a reply from Patrik at Elektron. He tells me the Rytm (and by extension Syntakt, DN, DT) behaviour is working as designed. He’s going to nudge the technical writer(s).
I’ve replied that I think the design is wrong and doesn’t make enough sense. I’m pretty certain of this. I also think that if they prevented LEN=INF, and changed CHNG to RESET (giving us a “reset after N steps”, where N could be INF), we’d get all the behaviours and a consistent, coherent interface.
I’ve mentioned wanting a sequencer specifically designed for song mode. Where you can nudge pattern changes for a certain machine ahead of needed and controls to navigate the song. There is nothing on the market. At least I haven’t found it.
Maybe these are no longer “on the market,” (two steps forward, one step back) but:
Yamaha QY series and Roland MC series both can shift timing of sequencer tracks forward/backward. If I recall, the MC ‘forgets’ anything that gets moved before beat/measure 1.1.0. I don’t remember if the QY actually rotates it… it might depend on whether it’s a Pattern track or a Sequencer track.
Both of these sequencers have ‘edit’ menus that are a bit of a menu dive to perform these (and other) operations, which seems anathema to . Both of these sequencers also have intrinsic “song” features (and an extra ‘song performance’ mode, Super MRP, for MC). Both allow manual insertion & editing of PC messages.
Just saw this in the QY100 manual:
If you don’t read music or simply couldn’t be bothered, just record what- ever sounds good to you. Making music should never be a chore.
I remember seeing a video of a young-ish Tricky doing a music workshop for a school or community group. He talked about using a really small, portable QY-something, and loving how it didn’t force rigid timing. His older stuff was so wonky, but effective with it.