Omnisphere 2.6 announced at NAMM [RELEASED 27-03-19]

[noiseboyuk]

Yeah I get what you’re saying but if I’m using a VA synth with a fun simple layout, I’d rather the authentic character of an old analog synth, like Repro, Odyssey, Tal Juno, a Roland cloud synth, etc. with Omni I’d like to take advantage of the hybrid nature of it and be able to use oscillators as well as samples and granular in particular. Don’t need access to everything but thenprophet x looks good because it lets you use a couple depending oscillators and a couple sample based sources plus two filters, some effects etc. D-50 profile might be cool too but I don’t think it controls granular.

I counted the controls of the Prophet X and I think there were around 300, which includes many buttons. I could probably make use of 25 to 30 per page on a Novation SL so 10 to 15 pages should do it. That’s not too bad really. And the nice thing with the SL is that you can hit a button to go right to a midi template page. So I’m thinking the Zero SL used in combination with something else with quick controls like Bitwig device controls on a push, or komplete kontrol, might give a great way to controlling a lot of Omnisphere entirely from hardware without owning an expensive hardware synth.

I am however very curious about how they mapped the Virus TI. Perhaps they assigned graintable functions to granular for example with could be really cool. I’d maybe buy a TI desktop if so, since I also wouldn’t mind having a virus.

Actually some of those new D-50-based presets did use granular, didn't they? Not 100% sure if that was extra to the programmer or if they remapped something, probably the former. That's the thing... any of these options like granular are only ever a click away in Omni itself anyway, and that's a really important aspect of the Hardware Integration for me. Its not just the voluntary restriction of following other synths' sonics and architecture, there's the knowledge that you can then go wild with it in totally different ways. And the D-50 is a special case in a way, there there's the big draw of the sampled presets, and the attraction of using those within full-fat Omni, regardless of whether or not you're using a controller.

15 pages on a generic remote.... sheesh, not for me. I've started work on the Moog Voyager, that one is really fun, there's the XY pad and idiosyncratic modulation options. It's pretty cool to me how just a relatively small number of controls (many of which are broadly common) can produce such different kinds of sounds. But the idea of a curated 2 page Prophet X or D-50 focused on big stuff is also quite fun though. Might take a look at the X next, while my enthusiasm for all this is still going strong. It's quite rewarding to solve problems and create a little bespoke UI.

[Echoes in the Attic]

Yeah you certainly don't need to use all the controls of something like the Prophet X, you could leave out many of the modulation source and destinations for example, and some lfo's etc. You can decide how many lfo's or envelopes are enough for a hardware control template. But I want as much control over the oscillators/samples as possible.

[noiseboyuk]

I did take a quick look at the selection of samples in the Prophet X profile last night. There’s one control for 16 soundsources categories, another for scrolling through the individual soundsources. In the hardware that’s limited to 100 files, but in Omni that could increase to many many hundreds. While it’s kinda fun in a lucky dip way, it just doesn’t make sense to me really - so much better, quicker and easier to use Omni itself, use the tags etc. And a further blow was that the category selection was pretty flaky, depending on the speed of the control change, it would skip to different categories, so you couldn’t name those 16 consistently.

So I dunno... it would be possible to have a drastically slimmed profile, I have a shark-jumping feeling about it. The whole point is to get a more tactile immediate experience, but I can’t help but see it as a retrograde step. Hardware beats software for basic knobs, faders and buttons, but software beats hardware on multi-function browsers and selectors. On the hardware X you have one LFO control section and a selector between LFOs, much like Omni itself, but in our emulation scenario you have separate controls for absolutely everything. The hadware experience on a synth like the X is already emulating the software one, and it uses simple computer power in the hardware itself to do this. Removing that with an emulation is a retrograde step I think.

So you could simplify it to a handful of options, use step changers for slower selection of categories, but it’s not a terribly appealing prospect. Now I’ve seen it in action, I’m more convinced that avoiding the synths with some level of internal computer control makes most sense for our Frankenstein hybrid of hardware and software.

[noiseboyuk]

Just an addendum to the above, and the overall philosophy of (mis)using Hardware Profiles. With a complex synth like Prophet X, here's the workflow using the hardware:

Physical Control -> Synth Brain -> CC/NPRN data -> Omni map CC/NPRN to internal control(s)

That's 4 steps. We can change the first one, and replace the physical control withsomething we can control, either on tablet or a generic controller - hooray. But all the other steps are our of our hands. We can't do that complex internal mapping that is the essence of Hardware Integration, so we can't change the CC/NPRN value and of course there's nothing we can do about a Synth Brain. So when it comes to this workflow, we're a bit stuck really.

One semi-interesting option would be to do some really tricksy programming to curate the sample set for example, so rather than having thousands of random options that aren't practical to select using just 2 knobs we intentionally limit to some useful options. Even that is a problem though, since I can't think of a way to assign two numbers at once - category and soundsource. So in practice you'd be best picking interesting options from the default. And one more problem - every time Omni is updated with new goodies, those values will change.

It's just a barrage of problems really, which leads to the whole jumping the shark thing. The immediacy of grabbing Osc 1 range, Osc 3 shape, mixer etc is the essence of the appeal for me of going down this road at all. This is the work in progress Moog Voyager:



Clearly labelled, familiar layouts that evoke the hardware, grab and go. Obviously there would be a 2nd page for filter and envelopes, but even so it feels a world away from blind menus and multi page configs. (It did occur to me in a flash of madness yesterday about whether or not 2 iPad minis side by side would be a nice thing, but I think that might be overkill!) And while I remember - many controls are stepped in specific increments (eg a 3 position switch, a 6 click knob etc) which could be a problem for a physical universal controller.

(this layout has presented a few challenges. It's the first one I've come across that uses HiRes Midi CCs which for a long time I thought was the same thing as NPRN. It isn't I eventually learned, you just set the MSB value like any midi CC control, and internally it steals a second value from a 2nd midi CC. All the variable pots here are using HiRes Midi, all the fixed position knobs use regular CCs. Then there's the XY pad which is mapped to Omni's Orb - couldn't have found those controls out without Omni making its Midi Learn report. Finally I rearranged the mixer to suit Omni's remapped function - on hardware there's an External In at the top, and Spectrsonics stole that control and mapped it to a nice FX chain, so that's something I put at the bottom of the mixer).

[noiseboyuk]

Rather than clutter up this thread with any more Midi Designer Pro templates, I'm posting on the old Lemur For Omni 2.5 thread and making it dual purpose Lemur / MDP - viewtopic.php?f=1&t=504702&p=7302573#p7302573 . Up and running now and confirmed working with Omni 2.5:

Mine
Moog Voyager
Roland SH-01a
DSI Oberheim OB-6
Roland VP-03
Roland JU-06

Others
Prophet 12
Roland SE-02

[noiseboyuk]

...and just to add that I've updated to the latest versions now on Windows, and its pretty responsive. In Cubase, reckon it's usually that 0.5-1s between patch changes on non-sampled patches - would be nice to be quicker, but it never seems to be more than a second.

[Echoes in the Attic]

Yeah I sometimes get around 3 seconds changing a patch without samples.

[aMUSEd]

They have added Virus Ti now which at least gives me the possibility of using this, but unless I can control all OSC types with it inc granular I will probably still prefer to use Kore and/or Komplete Kontrol as they are more flexible.

Wish they would add V-Synth though, that might be interesting.

[Examigan]

Wish they would add V-Synth though, that might be interesting.

They are probably working on more as we speak, or rather... type. It would be nice to know what’s in store for the future, or at least one or two that they may have in the works.

[mrsugmad]

All that work on Omnisphere...

...so little attention to Stylus RMX...

I wrote yesterday this very question to them, and here is their response:

""I have nothing to announce at this time, but we are constantly working on our instruments.
We haven’t forgotten about rhythm ""

Sounds encouraging....

[noiseboyuk]

I wrote yesterday this very question to them, and here is their response:

""I have nothing to announce at this time, but we are constantly working on our instruments.
We haven’t forgotten about rhythm ""

Sounds encouraging....

Welll.... only semi-encouraging at best. The "we haven''t forgotten about rhythm" line has been used for a few years now, it amounts to their only comment on it.

I've no doubt it will be incredible when it finally arrives, but it could still be years away for all any of us know. And still, really, nothing else comparable to RMX from any other developer imo after all this time.

[vitocorleone123]

Maybe I'm just reading the thread too quickly, but are you saying that something like MDP will NOT be able to serve as a controller that tells Omnisphere that it's a certain synth so you can use that "profile" within Omnisphere? That is, you still need the physical synth(s)?

[DocAtlas]

Maybe I'm just reading the thread too quickly, but are you saying that something like MDP will NOT be able to serve as a controller that tells Omnisphere that it's a certain synth so you can use that "profile" within Omnisphere? That is, you still need the physical synth(s)?

I would think that could work if your controller can send sysex for the hardware you're trying to emulate.

[cletiscake]

I wonder how far away omnisphere 3 is, and what kind of improvement it would bring

[noiseboyuk]

Maybe I'm just reading the thread too quickly, but are you saying that something like MDP will NOT be able to serve as a controller that tells Omnisphere that it's a certain synth so you can use that "profile" within Omnisphere? That is, you still need the physical synth(s)?

Either I’m mis-reading your post, or that’s not how it is at all. MDP can emulate every protocol and mapping used by hardware, and Omni doesn't know or care which is used.

Doc - no SysEx required for most synths, only the specific models that used it such as the Juno 106 and I believe the D-50. Pretty much anything post-1990 seems to use either Midi CCs (occasionally HiRes midi CC) and NRPN.

Cletiscake - a total stab in the dark, but I think Omni 3 is quite a way off.

Pretty much all we know about how Spectrasonics works is that internally they develop several products simultaneously, but it takes many years for them to reach a point where they get released - I think they said that one product is something like 17 years in development and counting. I think we’ll see other Keyscape / Trilian style products based on the Omni engine before Omni 3, whatever that would be.

I’d lay money they they too expected the RMX replacement to be out by now, but the development process has taken considerably longer than they expected. It’s kind of admirable that they won’t rush to market and it should be incredible at the launch party in 2056. I suspect that if they knew beforehand how long it was going to take, they’d have done at least more expansions for Stylus and at least a tag browser update, but decided to plough their resources into the new instead.