I am using the individual upper and lower outputs into 4 discrete inputs of my interface.
Either in single/upper or single/lower with binaural off, I’m getting panning that alternates not to the left or right of the appropriate layer, but bouncing between layers. LR PHASE/spread slider is nulled at the bottom. If I repeatedly strike a single key, the note, even when centered bounces between both layers.
Please…why am I not getting discrete audio within the selected layer? Thanks in advance
In Single mode, the SG actually uses the two layers to get full polyphony. The patch is applied to the two layers, each note you play triggers either Upper or Lower, so the output bounces accordingly.
True separate output is achieved in Dual or Split. Why in Single it is not routed to the selected layer outputs? You can try asking UDO Support by mail.
I hope I answered your interrogation.
Thank you, Arnaude.
I did notice after posting, the behavior you quite accurately referenced while in dual and split modes. I can’t understand why the Gemini was designed to behave this way in single mode. Behavior like this is counter-intuitive.
I agree, counter-intuitive. And it makes it useless because of all the manipulation it implies when switching modes. I gave up using the separate outputs, I shouldn’t have to IMHO.
I can’t understand why either. This is such bad design, it’s either lazy firmware programming or not well-thought hardware. It sounds like MIDI notes alternatively sent to two sound engine in parallel, the finishing touch not being implemented - being the sum up of the voices on the selected layer output. Support does not see any issue there though.
I also agree, please change this!! It makes it so hard to use single mode.
You contacted support and they just said, ‘behaves as expected’? That’s a pity. I haven’t messed with the separate outputs in a while.
If you’re on Single, is the output of the unused layer silent? Just wondering if you can at least sum the two outputs?
No, not silent. Regardless if you’re in single lower or upper, Binaural off, every time you trigger a voice, the Gemini randomly directs it to each set of outputs. For example, if you’re playing a chord, notes will appear on each set of outputs. Fortunately, the stereo positioning is unaffected, however the entire purpose of a discrete output for each layer is functionally null if voices from each layer appear - willy nilly - from both sets of outputs.
For UDO to claim that this is normal behavior is troubling and frankly, illogical.
Totally agree it makes no sense, and makes the outputs kind of useless, which is why I never use them.
Was just wondering if you record both outputs, would it be a workaround for you to sum the two outputs in your DAW later. If that’s what you’re doing.
Did UDO suggest anything to help you out?
Actually, I would have preferred using the mix outs, but UDO decided that each layer must receive a separate midi channel (why can’t they both be the same like any other multi-timbral synth?!?!), which really complicates things when I’m using my S88 controller. I had hoped that having discrete control of each layer in my DAW would help, but the discrete outputs are not discrete.
Yeah their support is a disgrace.
Have we reported the "dedicated outputs / alternate switching when in single mode to support / George yet?
best
Pete
I haven’t. Based on feedback from this post, UDO (surprisingly) views this as nominal function. I understand if this is ‘baked into the hardware’, but if so, what a ridiculous design decision. If not, UDO should acknowledge that this behavior is anomalous and correct it.
The Gemini, by UDOs own admission, was conceived and designed to be 2 distinct instruments in one. Thus, separate stereo outputs for each. And, when using the Gemini in single mode, as a distinct instrument, to have voices bouncing around between each set of outputs is 100% a design flaw; a very serious one. No other instrument fielded with such a feature has ever behaved this way. It’s just wrong.
I tried using the mix outs, but could not get the Gemini to work properly with my S88 controller because UDO designed each layer to have a different midi channel. Can’t change this; a design flaw IMHO. Specifically, UDO claims that this is because the Gemini was designed to operate as 2 distinct synths. So I switched to the dedicated outputs - to operate as 2 distinct synths - expecting to control each layer via my daw only to be met with another design flaw.
When viewing the Gemini objectively, these obvious flaws seem to be viewed by UDO as intentional features. They have to; they have no choice because there’s no screen on the thing to provide their users the flexibility to choose. Had they designed the thing with a damn screen, they could acknowledge these things for what they are - flaws - and correct them in firmware.
I love the Gemini for what it is and it is absolutely a wonderful instrument - a modern classic. But, I also hate it for what it is not, and every reason thereof could have been addressed with a screen.
interesting, Jungo.
If UDO think this is normal behaviour then we must be looking at this from the wrong perspective, surely? what are we not seeing?
Would be really nice if UDO support could provide their thoughts on this here, so we all can try and understand what their vision is? if it does need to be like this, perhaps we can have a SHIFT option that turns it off.
All very confusing. Ill put in a support ticket anyway.
OK I did read the FAQ - i think reading between the lines, its hard coded in some way or at best extremely difficult to change it - and therefore they won’t. - which does seem a little off.
Maybe it really is ‘baked into the hardware’ as I previously surmised. If so, it’s a supremely poor design choice.
And yes, UDO should resume participation into these discussions. They were quite responsive in the past. It is their forum, after all. And I understand that they are a small company, but that should never be used as an excuse for scant participation in their own forum.
Arnaud, I just reread this explanation. If this is indeed the way the Gemini is structured - 2 independent 10 voice synths - 10 voices allocated in hardware, each to the lower and upper layers, then the behavior discussed herein would appear to be as designed.
The “Keyboard” section buttons would then determine how the 2 ‘synths’ interact - layered or split. “Single” allocates the voices from both synths to permit 20 voice polyphony.
Possibly UDO could confirm this? If so, it would put this entire discussion to rest.
To me they already have. It’s how it has be designed. It is dumb and makes dedicated outputs irrelevant in Single mode. It is usable only in Split or Dual. Both with their weird MIDI shortcomings. I don’t get how a screen could enhance the user experience. It just has not been thought through. They just glued two synths under a MIDI controller and shipped it, there is not even cross modulation to play with.
Sir, I couldn’t agree more.
I just installed v 1.4 for my Moog Muse last night. That incremental upgrade greatly increased the sonic capabilities of the synth…all because it has a screen that helps facilitate those capabilities; potentially even far more. And when it was released, most reviewers commented that the screen was ‘too small’!
For this sole reason, the Gemini will never be much more than what it already is, which is nonetheless impressive, but still a bloody shame as there’s so much unrealized potential in the synth. It’s difficult to comprehend why any designer would so intentionally limited themselves and their users.
I understand your point of view. I myself find it hard to see the point of a screen when the synth’s shortcomings are so low-level. For example, and I’m sorry for repeating myself, seeing the envelope without having modified the segment duration control curve in an update will not improve the synth’s capabilities.
As for the separate outputs, illustrating the path taken by the channels will not counterbalance the fact that the “separate outputs” are not usable in the Mode most suited to their use.
It could have been used without a screen if it had been designed as part of a complete modern installation, and not as a closed machine. It’s hard for us to understand because it’s mostly counter-intuitive. That’s my opinion, of course.
By “illustrating the path taken by the channels”, it becomes apparent that the entire discussion is rendered mute because it is impractical to correct a hardware signal path; no firmware upgrade can provide a solution, thus the problem is permanent. Time to move on from that particular issue.
My general problem with the Gemini, in use, is that there are frequent moments of confusion - interrupting the creative process - that could easily be addressed or avoided with even the most basic of screens. Obviously, this perspective is highly subjective. As an older person, I find myself increasingly unwilling to accept things as they are if they are counter- intuitive or lack common sense. I thought it would be the opposite as I grew older, but sadly, it’s not. 
I’ve enjoyed our interaction. Peace to you, sir.