Expressive E Osmose

Nope, Z (press) comes first, then Y (AT) on the Osmose. There are actually 5 levels of vertical travel: velocity (“tap”), press (Z), AT (Y), release velocity and, thanks to Osmose’s implementation of Y, there is a zone in between Z and Y where you can “shake” the key up and down, that is it sort of touches the beginning of Y, pretty cool.

Also, many original EM Continuum presets (available in the Continuum folder of the Editor) sound great out of the box on the Osmose, despite the completely different implementation of Y and the lack of a continuous X. Not all of those presets make use of all of the dimensions though and a few don’t appear to sound at all. However, there is literally years of discovery and exploration in there.

1 Like

Given the powerful DSP needed to run the EM, the category “softsynth in a box” doesn’t do it justice IMO :slight_smile:

1 Like

Thanks for the description! Now I am more stoked about getting my hands on one.

Oh, I agree 100%. I’ve just spent too long on GearSlutz where you need to proactively make statements to discourage the 5 trolls who want to tell you why you’re “stoopid”. :smile: Sure…. Every digital synth is essentially a “softsynth in a box”… but a lot of times, the box is worth having (Osmose, Hydrasynth, Montage, many others).

Thanks again.

1 Like

The 2020 video from UVI where they used Osmose to play their IRCAM solo instruments library is truly astonishing. Do we know if they’re sharing the setup for it now Osmose is shipping?

3 Likes

Yes but with mpe instruments it looks like a lot of the usability is down to specific configurations that are carefully built to play to the strengths of each hardware interface. There’s lots of chat above about this, and tweaking software. So it would be great to be able to access the specific config used in the video.

2 Likes

amazing how finished the Osmose already looks in this video

Played one today at perfect circuit. Really fun and easy to play! Sounded great! I’m not a keys player at all and I thought it was really cool.

3 Likes

I still hadn’t placed the preorder but that video right there pushed me over the edge. Pre order placed. And now the waiting game…

Mostly Percussive presets.
This thing sounds as organic as any synth I’ve ever heard.

1 Like

Yes, I am as impressed wit the EM engine as I am with the new key bed to be honest- but its the way they integrate that is the magic.

The MPE+ mode in the VCV Rack module moDllz should now be fixed in the latest version that is available, 2.1.2.

1 Like

Has anyone done a MIDI recording with the Osmose? I’m wondering how recording MIDI (and later editing MIDI tracks) works with what I presume to be a lot more data in the “MIDI Stream”… For example, if you record a C2 with some vibrato, then later transpose it to a G2, will it work as you’d expect? This may be a dumb question.

1 Like

FYI I updated the eaganmatrix firmware without a problem when I connected Osmose to the max editor. It asked to update firmware file 2.

Took a while though!

2 Likes

I think some of this depends upon which daw you are using… and if it supports MPE, and how well it’s implemented.

I’ve used Live 11 and Bitwig (and quite a time back Cubase) and didn’t really have any (technical) issues. (though as below, I dont use it that much !)

as (current versions of ) Bitwig and Live have MPE support you can indeed transpose notes, as they attach the modulations (X/Y/Z) to the note.

I personally would avoid using MPE+, not only is it more data, but Bitwig/Live has not been developed with these in mind… so Im pretty sure they will not handle it properly.
(and indeed, I know that you have to take out some dubious messages to make it work at all)

now this will reduce the ‘accuracy’ of the data, taking it from 14 bit to 7bit, but given form factor of Osmose, I personally don’t think this is an issue (due to relatively short travel, compared to continuum) - and Ive been fine with this in the past.


that said, as Ive stated many times here, and elsewhere…
personally I rarely bother with midi/mpe, I much prefer to record audio.

yes, I know WHY you might want to record midi, I dont deny the advantages, but FOR ME, currently the hassle outweighs the advantages.
so I only use MPE to control mpe enabled synths (like Madrona Labs Aalto), and even those, I still record the audio!

why don’t I do mpe recording?
even normal midi note editing is pretty tedious…
but, even with a decent MPE implementation in your daw,
there is just so much data from MPE its beyond tedious, so I never will edit the mpe.

also in my mind there is a fundamental issue…
the advantage of midi (over audio) is the ability to change the sound / preset later.
in my experience that does not work too well for expressive controllers.
why?
because your (expressive) playing is tightly linked to what you are hearing, you are ‘performing’ in a feedback loop - so you modulations wont transfer well to another preset.

put another way, if I played on a regular keyboard , and then changed the preset.
whilst the notes might work well on a different preset, Id almost certainly have to change the modulation lanes.

but , this is a very personal choice… its not that I cannot record MPE, its I choose not too.
(I say personal, however, I know a lot of players with different expressive controllers, and almost all of them, for the most part, record audio, rather than midi/mpe)

of course, if you are working in a pro studio with clients, and so need recall, re-playability.
you might not have the luxury of this choice… time is money :slight_smile:


anyway, the tech is fine, in my experience … dont worry too much about data rates etc.
it works, many of us have been using expressive controllers for quite a few years, both before and after mpe introduction…
and frankly, my Osmose doesnt spit any more data than my Eigenharp, likely much less :wink:

2 Likes

If you own an Apple device, you could try uwyn’s (aka Geert Bevin) MIDI Tape Recorder - it is a free AU(v3) and the coffee button is enabled. It records and plays back MPE with utmost precision, has no editing features whatsoever but allows up to 4-track recording, each one MPE or not.

Regarding the use case of MPE recording, editing is probably not what you want, too much data that are linked between each other in some way. Switching presets depends on the nature of those presets. Nevertheless, sometimes you want to decide about the preset’s own effects and sound design at a later stage because you don’t know yet what you want. MTR is perfect for this.

MPE is an agreed upon standard and MPE+ is Haken Audio’s own extension to it, given the huge resolution of its Continuum surface: MPE+ — Haken Audio . Not sure how the resolution of Osmose’s controller compares. The great thing about MPE+ is that it is backwards compatible with MPE, as explained in their article. So I’d assume that you could have it set to MPE+ all the time. To be tested …

Note that at the moment a patent issue means that if you record the MPE output from Osmose rather than the MPE+ output, any pressure-weighted portamento used wont be present in the midi data.

1 Like

Also watch out for tedious differences between MPE and MPE+ outputs in terms of what midi messages are sent when using the pitch & mod faders (and probably pedals), and also some differences in how many midi channels are used (Im not an expert with this bit but I think someone told me that EaganMatrix doesnt use channels 15 and 16 for notes). Sometimes these things could potentially trip you up if intending to use that midi data to control the internal Osmose synth. So I’d be tempted to try to use the MPE+ output for this purpose. Even if your DAW doesnt end up retaining the extra resolution from MPE+, which is done using CC87 messages, at least the rest of the data should remain compatible with the EaganMatrix synth when replaying it.

1 Like

Do you happen to know if those dubious messages that you had to remove also come out of the Osmose if you only set the Haken port in Osmose settings to notes, rather than notes + matrix?

can’t remember if I tried this…
but you can just filter them out of the input, which is what I would do

however, this doesn’t really address the main point.

something like bitwig only attaches cc74 to the note, it will not attach the other mpe+ messages to it
so that means that even if its recorded in other automation lanes (in bitwig it is not) , it still would not be editable using per note editing e.g. moving notes around.

Im also pretty sure (99%) this is true of Ableton…

however, when I was using Cubase, I think it recorded all cc per channel …
but Ive not used Cubase for a while now… so this might have changed for the better (or worst)
you’d have to try it.

also there are some midi recorders that could probably record all the MPE+ midi ‘as is’, but then no proper editing, and the extra data can only be used by the Haken engine…
(ok, the vcv module you mention will use it… but thats the exception that proves the rule)

btw: I know what the MPE standard says SHOULD be supported, Im talking purely from experience here… you can go talk to your daw provider of preference, about any shortcomings :wink:

but frankly thats not that bad… Ive been using ‘7bit’ MPE for years, it’s perfectly functional, synths have slew/step compensation anyway.

whilst of course we can look at MPE+ , and technically go ‘ooh more bits = better’, I do not currently think it adds much … and adds exactly zero unless you are targeting the Haken engine.
so sure, technically fun/interesting, but for most musicians , its not going to be useful.

reality for now, MPE is best option if you need it. better support/interoperability.
our best hope of higher resolution midi is going to be wider support/adoption of midi 2.0

(and I still stick by, limited travel range on Osmose, means this 14 bit midi is nowhere near as important as it was for the continuum)

Yes, at the end of the day context is king in terms of which version of the midi output is best to use. There may also be differences between the MPE and MPE+ output of the Osmose that are just down to the current implementation rather than anything that is a fixed consequence of differences between the MPE and MPE+ spec on paper. For example the pressure-weighted portamento stuff I already mentioned. But its also possible that there are differences in temporal resolution in practice, eg it could be the case that the MPE+ mode happens to output more expressive messages per second than the MPE mode does in current firmware (even when excluding all those extra CC87 messages that come with MPE+), although I need to do more testing to confirm or reject that possibility, its just something I maybe noticed yesterday when testing something else, but havent checked properly.