Real-time stereo viewing in a 3D content-creation app

[Shokan]

Hi,
As the topic title says, I want to edit/model in real-time stereo in a 3D content-creation program such as Cinema 4D r10, ZBrush 3, Modo 301.
Of these, there is a plugin for Cinema 4D by SpatialView called SVI plugin. They make plugins for a few others as well, but only the one for C4D is able to let the user actually edit an object in stereo 3D.
My solution, though, needs to be across the board for 3D creation apps, since I don't know what I'll be using. DirectX is not popular as an API for 3D creation apps (of the few that do, 3DSMax displays its viewports in Direct3D, for example). So, my initial enthusiasm for IZ3D was dampened. It can display Direct3D content only; it's mainly targeted at gamers, as it should be. There are probably content-creation programs for game development that are mainly Direct3D-based, but since game dev is not my pursuit, the IZ3D monitor was ruled out early on. Unfortunately. Maybe there will be some support for OpenGL on this product down the line.
My next plan of attack was the dual viewport idea. That is, simply create two instances of the active editing viewport in a 3D content-creation program and use a simple viewer like the Pokescope. Could work, I reasoned.
Wrong! Programs such as ZBrush generate mega-density meshes (the opposite of game dev), and having the software and PC to draw two such instances of a two million-vertex model would shut everything down, or at least slow everything to a crawl. Unacceptable! Some ingenius solutions were devised however by some forum-goers such as using a second camera as a child of a parent camera for the stereo view and viewport pair. This idea also ran up against the huge demand on the system ina addition to a problem with one of the pair lagging in its update editing. That would ruin the stereo effect. The other solution would be to set the cameras up as sub-objects in a "null" object. This way, there is no parent/child setup resulting in unequal edit updates. This, too, was unfeasible because of the unavoidable system and software bomb.
The solution for me at this point is to use two whole systems and two monitors. I have not tried this yet because I do not have the wherewithal to swing the purchase. But, anyway, these two PCs would each be running the same content-creation program and the two monitors would be left eye/right eye. I could view my work in stereo using the Pokescope. By altering the viewport view of the object being edited to form a stereo pair and by also configuring the control panels in the two monitors so that there is a stereo version of them w/o doubling, it should all work. Each time I edit the object (move, zoom, sculpt, etc), I would still be seeing in continuous stereo due to the fact that I began with the correct alteration in view between the two views for stereo.
Oh, I forgot about the simultaneous pointer device setup... obviously, for this two system setup, I need one pointer device to control both screens with the cursor/pointer onscreen moving simultaneously in exactly the same way on both screens. This I think I will be able to achieve using a wireless USB tablet or pen-mouse that has a receiver on each PC tuned to the same frequency. I have not completely established that this is possible, but I'll know for sure soon.
As for alternatives to the Pokescope viewing of two monitors, I may also be able to use the Z800 goggles using the 3DTV.com Stereo viewer and Multiplexor for simple display of right eye-screen displaying one PC video and the other displaying the left.
So, this is how it stands for me and this project. I never knew it would take so much asking around... like two months of some pretty heavy emailing and forum-ing. I believe 3D content-creation apps SHOULD ALL be viewed in real-time stereo. It's a natural. Plenty of apps can produce a stereo pair of stills for viewing and even rendered output animations in stereo, but WORKING in stereo has resulted in me coming up with all the contortions above... AND, it'll be dang expensive!
I'll post a still from the Blender app of a stereo model pair sometime so you can see how exciting it would be. To me, at least. I'm open to comments and/or suggestions, for sure.

[LukePC1]

That's an interesting project. I wish you sucess, but I'm not shure, if you'd get good/acceptable 3D options. It might be a little difficult to adjust, but afterwards it might run quite good. And maybe it wouldn't even need high end systems. Just 2 similar/identically ones

Everytime you'd want to change 'seperatin' or 'convergence' you would have to manipulate only one computer. Would that work with a normal PS2 mouse/keyboard connected in addition to the other (Wireless) devices?

I don't know, if there was a question in there somewhere, but the iZ3D doesn't use 2 normal input immages. One is just black and white for polarization. So you shoudn't be able to use it in your plans...

[crim3]

With custom programming everything is possible: distributed systems, balance of load, synchronization... a stereo mouse cursor with mouse (or other input device) attached to one of the computers. Maybe if the guys of the project Blender, for example, would find working in stereo 3D essential, they will half redone their project to fit to whatever solution they find appropriated.
For example, I think that a distributed system is a great solution for stereo 3D, although it needs expensive hardware for visualization (though there is people over there feeding a z800 with two video sources http://www.3dvisor.com/forum/viewtopic.php?t=1038 ).

I think distributed systems are great in general for any balance of load solution, and should be considered seriously in this days, when any private person at their home has several computers connected in a LAN (who could imagine that 20 years ago when a pc would cost the earns of half a year).
For example, Flight Gear, a freeware flight simulator, allows for the flight model being calculated in a dedicated computer meanwhile another one is dedicated to 3d rendering and user interaction. Imagine how powerful can be a modern computer dedicated only to render graphics, because the physics are done in another computer. And vice versa.