Looks like VR workarounds are making *some* progress…
-
I’m seeing more and more BMS VR videos with functional 3d cockpits and MFDs being uploaded, each time with a few less steps to get it working, but obviously not wearing the device myself I can’t attest for latency.
This guy Ivan Vargek seems to have a pretty solid setup: (details in video description)
This uploader Callsign: Squeak is getting good framerates, although that’s just one part of the equation, but states his method is still a little “screwey” in the comments.
Still not worth dropping $400 on a rift, but it makes a DIY Rift much more tempting to attempt.
https://www.reddit.com/r/DIYRift/comments/2vur7o/build_a_working_diy_rift_tracker_based_on_the/DIY TrackIR was like $13 to make (including cam), 20 minutes of soldering and gluing and totally worth it…
This is obviously more involved, and probably ~$50 worth of components… but I might have to give this a shot. -
When I look at Vargek’s video it doesn’t appear to be in 3D (the two eyes are seeing the same thing - they should be offset to provide stereo 3D). BMS doesn’t generate a depth map with Reshade’s SuperDepth 3D, so getting true 3D injection remains a challenge. Once you have 3D injection, you need to handle the HUD collamation to infinity and the HMCS collamation as well. Using Opentrack and VR Toolbox I’ve been able to get a pretty good 2D VR experience. You need to have a zoom button bound to your HOTAS because the resolution’s not adequate to pick out small things on the TGP or manage a cluttered HSD or FCR, plus it’s helpful for visually ID’ing other aircraft.
-Rabbit
-
None of them would produce enough parallax that would fit same IPD to that of VRHMD we have. Thus they don’t have a good stereoscopic image. I could give good parallax that would meet IPD=70mm for my Rift, and set the HUD collimation to infinity and the HMCS collimation as well by adjusting parameters of Tridef 3D(video link below), but that made external wing model (external f-16 body model seen from cockpit view) separate from cockpit model.
BMS’s external wing model, scenery, and cockpit seems being rendered separately then combined together like a composite photograph, Wing model is a half build 1/72 model kit is in front of pilot’s helmet, the wing model picture is at behind the cockpit and above the scenery, which would look naturally fit when drawing the single 2D image. However, when TriDef3D made a stereoscopic image from their z-buffer, this trick collapses the view. It is like you are looking the diorama of the city from its side, so you will notice the buildings have smaller reduce scall as it comes close to the front edge of the table, that was making perspective trick when looking the whole set from the front. You will notice diorama doesn’t have actual Length.
-
Diorama?
I believe it’s obvious that BMS ain’t vr build.
It’s like trying to make a car go to Mars and u in alive all the way.
Well it lacks the fundamentals…Στάλθηκε από το MI 5 μου χρησιμοποιώντας Tapatalk
-
Diorama?
I believe it’s obvious that BMS ain’t vr build.
It’s like trying to make a car go to Mars and u in alive all the way.
Well it lacks the fundamentals…Okay, here is the context.
TriDef3D is a software that produces a stereoscopic image from Z-buffer information from DirectX.
So it is a software to try producing a stereoscopic image for games that ain’t VR build obviously.
Some game works just as good as VR build one, others are not.
BMS is one of those others because it seems there is some kind of tricks when rendering cockpit - wing model - scenery, and “Diorama” is just a parable to describe this situation.Actually, BMS is working in VR except for external wing model separation.
-
Probably s wrong use of the term diorama. Well the actual term - word isn’t the issue.
In ur vid (sorry my conclusions will not be that accurate as I show it on my phone) the wings and pit look ok in place and normal.
In the front things are messed up and moving where they shouldn’t.
The front antenna parts looks like they are one 3d object but not attached to the body. My idea - conclusion is that the software u use sees it as a different 3d object and that why it’s floating all around.
Another perhaps is that the antenna part is not in the same model with the nose but with the pit object.
Z buffering is iirc for zbias. Zbias is not that much used or setup in Falcon.
Still if it’s setup for what u display I don’t think it’s a z bias problem.Στάλθηκε από το MI 5 μου χρησιμοποιώντας Tapatalk
-
Another thing that just came to my attention is that the nose part was somehow patched afterwards… So things might got messy there.
Στάλθηκε από το MI 5 μου χρησιμοποιώντας Tapatalk
-
Yes, it is like that. Thanks for understanding and correcting my wrong used terms.
I forgot actual parent number but the nose and the wing is the same model in LE.
Pit and Scenery look Okay but the wing-nose model is floating.
You also can see the gap between strake and canopy frame while the other side of view they are penetrating into F-16’s body.
When I see them through HMD, the wing looks tiny scaled than the pit.
As long as Pit and Scenery could get a good parallax via TriDef, Z-bias is correctly setup for them and TriDef is successfully reading them. The wing model also has a good stereoscopic effect for itself, I can see them solid via HMD, it also has z-bias information, but only the positional relation between it and the pit is wrong.Here starts my conjecture. As because it has a larger parallax then the pit has, which is why they are floating, the external wing model is “closer to the camera than the pit is, while it is drawn beyond the pit frame.” Strange? but they are…
Well, think about Photoshop layers or any power point images you can bring them forward and backward. From the front to back, they are placed in order like “Pit - Body - Scenery”. BMS might be rendering scenery first, external body for second, and pit for the last. However, in terms of Z-bias, the whole external body model is closer than whole pit model.Perhaps making larger scaled external model might solve this. I haven’t tried yet.
Still also has problems with mouse cursor(it looks double while the center of two cursors is the correct position) and latency, but if this external model has been fixed,
VR view will be just as perfect in BMS through TriDef-Opentrack-VirtualDesktop-Reshade(Polynomial Barrel Distortion). -
Ok just saw the video on my monitor. I don’t think that his can work unless the pit and the external airframe are the same object. Now they are different thus floating as the software sees they are not connected thus acts like they are two different 3d objects and tries to apply the same 3d view - movement effect on 2 different objects and they float.
-
Thanks for the upload Wax … I agree it’s nice to see the early efforts by others with VR. Never say never
BTW … is it just me or was the video uploaded by Squeak really VR? The video is way to stable … movement seemed to be very linear (like was using the keyboard to change views)?