Back in the linked topic, I mentioned I’m struggling to find a comfortable workflow using a Rev2 and an Octatrack. I can get something going in various ways, but there’s annoying quirks and compromises I didn’t expect. I’ll try to describe my issues. I’ve temporarily got the Rev2 packed away (in favour of a Juno 106) so I might get a couple of details wrong.
In an ideal world
The Octatrack has two sets of stereo inputs; the Rev2 has two sets of stereo outputs, is bi-timbral and has a multi-mode allowing you to address the two layers on different MIDI channels (you pick one channel for layer A and the keyboard automatically uses ch+1 for Layer B). I imagined it would be easy to have some Octatrack MIDI tracks drive one layer and some other tracks drive the second layer, on different channels. I imagined I could set up the two devices so that I could play the Rev2 keyboard and address each channel separately, editing sounds (because I usually do sound design as part of composition), and for it all to be easy.
Rev2 MIDI Out → Octatrack MIDI In
Octatrack MIDI Out → Rev2 MIDI In
Rev2 audio Main Out → Octatrack In A B
Rev2 audio Alt Out → Octatrack In C D
Issues
It’s important to note that the knobs and other controls on the Rev2 all “send MIDI/NRPN” to control the synth. I infer this… if you engage Local Off, you can’t edit the sounds unless MIDI’s routed back to the Rev2.
if you set Rev2 MIDI Channel to 15 or 16, it uses Ch 1 or Ch 2 for layer B. So confusing; not in the manual; maybe a bug?
If you match the Octatrack Auto-Channel to the Rev2 output channel, you can play Layer B (by selecting a MIDI track on the OT that feeds the 2nd-ary channel). Great! But weird things happen if you try to do sound design. The controls don’t update, some of them don’t function. It gets even weirder if you try to edit the sounds for layer B (by picking an Octatrack MIDI Track that sends to the 2nd-ary MIDI channel).
Some of the knobs + switches on the Rev2 don’t take effect. My guess is that they send NRPN and the Octatrack filters that out, because it only knows about CCs (I’m aware NRPN is a kind of CC, so perhaps there’s a workaround using 2 or 3 CCs, but I haven’t figured that out yet)
The workaround for 4 is to turn Local On on the Rev2, but the Rev2 hasn’t been designed to trap alike notes so all the sounds double.
The Rev2 doesn’t let you run the seq and arp when it’s hooked into an external clock (unless the clock’s running), which makes it impossible to programme Arps and Sequences on the Rev2. I know the Octatrack could replace all that functionality and more, but I wanted to learn both devices (and potentially keep the sound design goodness of a sequence on the Rev2 for other uses such as playing live)
If you leave both sets of cables plugged into the Rev2 (e.g. because it makes sense to leave your cabling intact), but you turn off Multi-mode, and/or you have a sound that doesn’t use layer B, the synth will still connect half its voices to the Alt output. As it also round-robin’s the voices, you lose half your notes (if you don’t have both pairs up on your mixer or all of ABCD Octatrack inputs activated. Doubly annoying if you’re using Thrus rather than DIR). You HAVE to unplug the Alt cables if you’re not using that feature.
This is a bit ranty. Sorry. I just needed to get it off my chest. If anyone’s developed a smoother workflow that doesn’t involve toggling something every time you change activity (sequencing, sound designing, etc), I’d love to hear it.
Despite all the power and flexibility of the Rev2, I found it’s actually easier just to disconnect MIDI and sample it into the Octatrack, which is disappointing enough I’m considering selling the Rev2. (It’s a really, really, nice synth otherwise so I will give it a few more month’s experimenting first).
MIDI Param Send: Off, CC, NRPN—Changes to the values of front panel controls are transmitted via MIDI as Non-registered Parameter Number (NRPN) controllers or as Continuous Controllers (CC). Transmission of parameters can also be turned off.
Multi Mode: On, Off—Multi Mode allows separate MIDI control of voices 1-8 and voices 9-16 on 16-voice systems. (On 8-voice systems this is voices 1-4 and voices 5-8.) This enables independent, bi-timbral control of the Prophet Rev2.
When Multi Mode is on, the Prophet Rev2 responds to MIDI data on two consecutive MIDI channels — the base channel (selected in the Global menu) and the next highest channel. The base channel controls voices 1-8 (or 1-4 on 8-voice systems) and the other channel controls voices 9-16 (or 5-8 on 8-voice systems). Notes, program changes, and other MIDI data received on the specified channels provide independent control of
the two parts.
Note that a program change needs to be sent to each of the two channels when setting up Multi Mode. Only Layer A of the selected program is loaded. There are no stacks or splits in Multi Mode. Use CCs or Layer A NRPNs for automation of parameters. Program Layer B data is completely ignored in Multi Mode.
Edit : I would personally use the OT Sequencer and Arp. Much more convenient and powerful.
Hi mate, sorry to hear that your struggling with the workflow.
I can’t really suggest solutions for your case, as I didn’t have an OT at the same time as the Rev2.
My workflow is pretty simple, which seems to work well with the rev2 (for me). I only externally sequence it (no internal ARP or sequencers) using a Squid (soon to be the Hapax!). I don’t take midi from the rev2 at all, so all automation is done through CCs.
I find the layout of rev2 great for sound design, and so use its panel for all of that. It sounds amazing to me, it’s pretty much knob per function, and having the layers is crucial for my setup, even if it’s 4 voices per layer. (I do intend to expand to the 16 voices at some stage.)
How I use it is I generate sequences on both layers and also record loops and oneshot chords into my AR - something I imagine the OT would excell at. I imagine the OT’s LFO designer, plocks and amazing ARP could also open up the rev2 even more.
My only suggestion is maybe simplify your setup.
OT midi out to Rev2. Layers A and B audio to the OT’s inputs. Template a load of CCs. I like to have 2 midi tracks per layer set up, with different CCs mapped (squid only has 5 cc knobs), which means I can swap between or layer midi sequences.
With regards to the voice limits, it kind of doesn’t bother me too much as the rev2 sounds so big and dominating that large chords are lost a bit or can fill up too much space in the mix. Maybe it’s also cos I’m rubbish with music theory, and the styles of music I like are generally more sparse in this sense.
I’m sorry if that wasn’t much use mate. Good luck tho, I hope you keep the rev2 and upload some nice music on to the forum.
I gave up with multi mode on the Rev 2 because it’s rubbish. More flexible just to use splits or layers IMO. I also had a similar problem to the one you describe with the sequencer sync.
Regarding @Tchu’s manual excerpts: I think I have it set to send CCs. My comment about it maybe still using NRPN’s is guess-work. Some of of the controls seem to have no effect when you’re in Local Off, sending MIDI out and then back in again - regardless of Multi-Mode. E.G. LFO time (and perhaps even LFO assignment); ENV 3 Destination, and many others. You have to make the change, save the sound, deselect the sound, the reselect it again before the change becomes active.
I’ll try changing Param Send to off, to see what effect that has. My guess is that it will disable all panel editing (when in Local Off)… but I have hope that it redirects the controls straight to the synth
Regarding multi-mode: re-reading that passage makes me wonder if I have understood multi-mode wrong. I thought you could only select one patch, and the two MIDI channels would point to Layer A and Layer B. It seems to work that way some of the time. However, that paragraph makes it seem like you can assign the Layer A sound from any patch to either of the two channels. That’s actually more useful, and might make it easier to sound design whilst composing.
Ultimately, I suspect the best solution to all this is to stop trying to use the Rev2 as the master keyboard with the OT, and just hang it off the downstram MIDI chain (in Local On, for hands-on sound design). This is a shame as the Rev2’s the best keyboard I have.
Can confirm: Changing the MIDI Param Send or MIDI Param Recv to Off does indeed stop the knobs from sending CCs or NRPMs. Thus, if you also have the Rev2 with Local Off you can no longer edit the sounds from the panel.
This makes sense (insofar as it makes any sense at all to separate the knobs from the sound engine with Local Off ).
I believe you said in the linked thread that this is the module version. so it’s not about capturing notes from the Rev2 keys or having that be echoed back to it sort of thing; it’s just about the knobs. correct?
are you trying to have the OT capture/record your knob movements? is that why you are transmitting them and have local off turned on? if you only need to have OT sequence a CC change here or there I wouldn’t transmit them from the Rev2. just sequence those changes into parameter locks via OT knob movements. those will be transmitted to the Rev2. unless some of these are NRPN’s and that’s the heart of the issue here…? if not, and you’re not transmitting them/receiving them back, I wouldn’t think you’d have double-triggering or issues with sound design via the knobs.
I’d really like to be able to use the Rev2 as the master keyboard with my OT, whilst also being able to use all the front-panel controls for sound design. By “master keyboard” I mean (as in the OP):
Rev2 MIDI Out → OT MIDI In
OT MIDI Out → Rev2 In
Rev2 Thru → onwards to anything else (currently AR & Matriarch)
In theory I can then address either of the Rev2 voices (in multi-mode), or the downstream instruments, by matching the OT auto-channel to the Rev2 MIDI Send channel, and then picking an OT MIDI Track.
However, when you do this, the Rev2 doubles all the notes unless you set it to Local OFF. But then if you do that half the knobs on the Rev2 stop controlling their parameters in real time (they actually do set values, but you can’t hear the change until you save the sound, deslect it and then reselect it).
I’ve also tried various configurations without matching the auto-channel, by trying to make use of the OT’s “block MIDI flowing from audio channels to MIDI channels” behaviour. None of it works as smoothly as I hoped. It’s annoying compromises whichever way you configure it (unless I missed something. Which is likely).
oh I’m sorry. I didn’t see it in this thread and remember you clarifying in the other thread, I just remembered the wrong one…
have you tried any sort of midi filter between the Rev2 and OT? for example Midi Patchbay on Mac will let you filter notes. or with MidiPipe you could translate or filter the NRPN’s (or anything at all). not an ideal solution and not standalone, but there are standalone solutions to that (Blokas Midihub for example) if it works.
The problem is that the Tx and Rx MIDI Channel is the same on the Rev 2. My workaround for this is to use my Midihub (Blokas) to change the Tx MIDI Channel of my Master Keyboard (in my case, the UDO Super 6) to match the Autochannel of my OT.
I think I’ve run into this DSI behavior with my MoPho. It’s somewhere in the nexus of:
Some MIDI devices send MIDI clock pulses even when stopped (handy). Some don’t.
Some MIDI devices will not default to Internal clock, when their MIDI Clock Source is set to External but their sound engine is triggered one way or another (by starting their internal sequencer, triggering a note, etc.) (May be typical DSI behavior?)
Some MIDI devices won’t listen to MIDI clock until they see a “Start” message. Can be inconvenient.
Some MIDI devices will always send MIDI clock out of their MIDI Out port in certain modes. (This can be inconvenient/problematic in various setups, and can require filtering MIDI Realtime messages from this output.)
I generally like when these are the case (and when I get options to control them):
My MIDI clock source is sending MIDI Clock pulses even when its transport is stopped (does OT not do this?). (This is useful for tweaking the sounds of downstream synths, whose LFOs may be sync’ed, arps may be on, etc.)
My clockable synths listen for incoming MIDI clock, and use it when it’s available
My clockable synths will default to internal clock IF they’re set to External but no clock signal is coming in (again, I think DSI may not do this?)
I can turn off (completely) the transmission of MIDI Clock from the MIDI Out. This is usually the case, but not on everything. Especially inconvenient on controller keyboards which are not intended to be the clock source.
There. That’s the first draft of my MIDI Clock Manifesto.
Re: the NRPN/CC wobbliness — beyond the real possibility that Octatrack is filtering or stopping (some of) these messages — is it possible that they’re getting mixed in with other messages and thus jumbled somehow by the time they loop back to the Rev2? e.g. mixed in with clock pulses, notes, etc. A MIDI monitor could help here. Just thinking about how NRPNs are usually a sequence of messages which must appear in the correct order, whereas CCs are (I believe) only one message.
