Home pit preparation - is decoupling HMD from the main camera possible for large pit?
-
Hi Guys,
After years of solo flying, I am thinking of finally building a pit in my attic, but I have one technical grievance that I hope exists a solution for. I want to create a large beamer/projector based as visible in the picture below, but I am afraid of always breaking immersion in dogfights by using TrackIR as normal in this setup, but if I disable it, it is pretty much a handicap loosing targeting with HMD. Is there a way maybe to get a HMD decoupled from main game camera? E.g I would move my head, and only the HMD would move around on the screen and not move the main game/world camera with it ? Picture below maybe helps explain.
I do not want to sound petty, but this is kind of my last make or break issue for going for the pit as I love the idea, just hate actually becoming handicapped by having it
-
Theres a bit more to it, pretty complex
If you have a cylindric setup, whats the standard, with 270 FOV, you need also the ability to look up and that still needs to function like the normal headtracking does.
Unlocking the HMCS from the rest is (AFAIK) not doable for BMS atm, but it would be possible (maybe) to center the HUD in the screen at least for the x-axis and to make headtracking working the normal way for the y-axis.But TRACK IR doesnt work then anymore, because you need a 1:1 resolution of Input:Headtrackmovement for the x-axis, so you need a gyro-based tracking.
As far as it goes for HMCS, you would need to disable it from rendering in the screen-projection, build a HMCS-replika with a real projector onto a visor and then you have to hope for a alignment feature develepod by BMS.
At the moment not possible. Someday, maybe, but thats definitely something I can not judge or tell you anything about in generell.
-
As far as it goes for HMCS, you would need to disable it from rendering in the screen-projection, build a HMCS-replika with a real projector onto a visor and then you have to hope for a alignment feature develepod by BMS.
That is overthinking it, no need to get extra helmet projector, movement of the HMD on the screen if unlocked can be properly scaled on any background with the curves in TrackIR in both X-Y axis. Yes I will probably never have a big space to look UP, but better than nothing. In regards to visuals Having the HMD itself flying over the projector screen is good enough, the primary HUD will be this way anyway in most pits anyway and even on the picture I posted the HUD is part of the main projector, the one in the physical pit is a fake or mirror. Practically the only thing I want is making the HMD symbols to be a cursor-like thing flying around the screen and locking targets (literally like playing RTS and clicking on something locks it, just controlled by the fake TrackIR analog inputs).
PS/EDIT: With all those simpit guys out there, I wonder why I do not see much about this problem, or are all those giant beamer simpit setups out there changing camera angles with TrackIR as a normal small monitor just to get HMD working and accept killing the immersion? From a programming perspective should be simple enough if allowed access to the engine code, somewhere there needs to be camera and HMD angles tied together, and just decoupling it and experiment. Pity FreeFalcon didnāt survived, right now I would be deep in the public code trying to find this, with BMS donāt know where to begin.
-
Maybe you get something with future release. Taken from the F4DrawingSharedMemTester out of lighning tools
// - NOTE: Check DrawingAreaSize in FalconSharedMemoryArea2 for the actual size of this area! // - NOTE: Treat this shared memory area as a pure "char*", not as "DrawingData*", since the size is not fixed! // changelog: // 1: initial BMS 4.35 version: added export of 2D drawing commands for HUD, RWR, and HMS in the following string variables // HUD_commands (only populated if g_bExportDrawingCommandsForHUD is set to 1 in Falcon BMS config) // RWR_commands (only populated if g_bExportDrawingCommandsForRWR is set to 1 in Falcon BMS config) // HMS_commands (only populated if g_bExportDrawingCommandsForHMS is set to 1 in Falcon BMS config) // // Common data format for 2D drawing command strings: // Command strings consist of a symbol indicating the command type, followed by a colon, followed by a comma-delimited list of arguments for that commmand, all followed by a terminating semicolon. e.g; // COMMAND: arg1, arg2, ..., argN; // // The following table lists all the available commands and their arguments. // // Command Arguments Type Description
the HMCS Display itself is accessible throught the F4SharedTexMem
-
Personally, Iāve been thinking of this idea since I started my own cockpit project - and from the āoverthinkingā standpoint, seeing as Iāve had some limited opportunity to play with the real thing.
My first thought was to use something like Google Glass to provide my JHMCS reticleā¦I also do not want to use TIR or any head tracker to move my OTW display - other than to provide up/down look, so I intend to lock my head tracking for OTW to Z axis only. Then to blank the JHMCS reticle from the OTW and display extract it into a projector in my helmet - there are a few options for either making or buying such, some expensive and some really expensive. Making one using a small round LCD (like 1 inch) to produce a projection is probably the least expensive way to go - there are lots of these available on Ali Express.
I also am not a fan of TIR - especially if I use projectors for my OTW displayā¦donāt want it on my head, and I also donāt want to have to deal with light interference from the projection (my plan is to make a 270 degree projection screen so I can just move my head naturally) - so I have an inertial head tracker - an EDTracker - that works very well. Itās also small enough to fit inside of a flight helmet liner nicely.
Which brings me to the real hangup - a way to align the reticle to the OTWā¦BMS does seem to have some of the hooks for doing a JHMCS alignment in itās DED pages, but not the full implementation. And thereās also no way to decouple to get the LOS info from my helmet to affect targeting input alone and not shift the OTW.
Soā¦Iām pretty much out in the cold unless I can figure out how to write a routine that will do these two things and feed the inputs into BMS via SM. I know some parts of SM are write enabled now, but what Iād need may not be, or even be available in SM. Noodling continuesā¦
-
Disabling z-axis completely wouldnt work, you need a 1:1 fesolution, otherwise, the otw just looks up in the forward-facing direction. But if you want to look up at the aft left side then, it actually will look down, if otw doesnt know where to look up.
-
That is overthinking it, no need to get extra helmet projector, movement of the HMD on the screen if unlocked can be properly scaled on any background with the curves in TrackIR in both X-Y axis.
Thats what I wrote already, simply not possible at the moment. Statement of BMS coder, not mine. Already got in touch with him some time ago, for exactly these considerations. Making the HUD stay in the center of the screen would work, HMD floating around according to headmovement not. Wont puplicate his nickname here, I talked to him about that, because I know him personnaly, but I dont want to waste his very very little freetime with a need to make comments on the public forum about stuff like that.
-
HMD floating around according to headmovement not. Wont puplicate his nickname here, I talked to him about that, because I know him personnaly, but I dont want to waste his very very little freetime with a need to make comments on the public forum about stuff like that.
Oh, ok than thanks if that is going from a dev then that is solid info ā¦. guess it is simpiting the same way as a monitor with TrackIR, just larger or nothing ā¦ pity.
Btw if you have this level of access to the developer and the HMD is exportable using F4SharedTexMem, the simplest form to inquire this implemented for us would be something like splitting the FPS to :
- every even frame normal camera, ignore TrackIR input
- every odd frame blank/blackout main world camera (or if possible display the previous frame from buffer), take TrackIR input and just get HMD rendered / or export from the F4SharedTexMem so we take it to second screen or some other head mounted projection device. Also in this frame any attempt to locking enemy plane as event can be attempted.
This way might be the simplest for the devs to give us what we need without actually decoupling the game engine logic much, just block/unblock game inputs and renders in every other render cycles, we simple enable-disable trackIR every other frame to practically get what we need. Yes we kill half the FPS, but simplest prototype to aim for. Maybe, talking out of my semi-technical arse here.
-
I already talked with him about these exact topics.
He said, he will have a look into it, if he finds time for.
NO PROMISES or info about any release here. -
Disabling z-axis completely wouldnt work, you need a 1:1 fesolution, otherwise, the otw just looks up in the forward-facing direction. But if you want to look up at the aft left side then, it actually will look down, if otw doesnt know where to look up.
Youāve got me backwards - I want to disable everything BUT the Z axis. Iāll be able to turn my head to see everything else in 270 degrees, but I want up/down tracking only when I nod.
-
Oh sorry, replace the z with a x in my description and I think, you will get my point
-
I can see thatā¦I think. Need to get my coordinates caged!
-
I already talked with him about these exact topics.
He said, he will have a look into it, if he finds time for.
NO PROMISES or info about any release here.Of course mate, just make sure to let us others know if something like that gets done. For me I am putting my sim plans on ice for the moment, or will try taking old office beamer to check how does it feel using TrackIR on something large/close to you
-
I can tell you from experience, headtracking is okay to use the way it works at the moment, but far from nice immersion.
Thats my current setup and you can see, i messed up the warping a bit, as the desktop is a bit squeezed in the center, that maximises the effect of the stretched FOV in Falcon BMS. If i have time and the patience I will change that in the opposit direction and stretch the desktop in the center, squeeze it at the edges, which will work against that effect of stretching in BMS a bit.
I have the headtracker setup to have a 1:1 resolution for the horizontal headmovement until it reaches almost the edges of the 214Ā°screen, then I pitched up the curve to cover the rest of the rearview, which is not projected.
But even with a 1:1 resolution the HUD doesnt stay centered. Better than nothing but far from perfect. -
Very nice, something like that is my baseline scenario, since probably more easier to setup on BMS side (so far flown on ultrawide). Right now however I am still a bit researching the option of half-dome system like this. Negative side is heavy distortion and the need to somewhere get a perfect mirror ball. But on the other hand turn based combat would benefit.
What warping software do you use? And noticed half-dome options there ?
EDIT: Dome technical materials here : http://paulbourke.net/dome/mirrordome/
-
Dome is f*n expensive, because you need a lot more projectors. My setup is built and ready for 360Ā° FOV, but currently only 3 of 5 elements and 3 of 4 projectors are in use, because of spaceproblems and the current lack of ability of BMS to use more frustums, what would enable a sharp and clear 360 projection without any stretching of the OTW.
When BMS is able to render multi-frustums someday, which it will be, then I have a complete cylinder which is also very tall next to me, so its a pretty good imersion, only the vertical view must be moved a bit by a headtracker, if you want to look up very high. But thats okay and waaaaaayyy cheaper, than any type of dome.
To cover the dome you suggested properly youād need at least 4 projectors for much less āfeltā projection surface, than I cover. Also you need definitely a higher room than a normal living room, because you want the vertical center of the screen at your eye-level.For warping I use fly-elise calibration pro, it can do every type of projection warping you can imagine, you could even back-project a canopy, if you have the 3d modell of it.
Yes, it will take a while, until we get multi-frustums for BMS, but it will be there someday and as you know that, dont shut yourself down on building a cockpit.
Good research at the beginning of such a project is very important, but dont miss the point to acutally start building one, it will take longer - Iām pretty sure about that - to build a COMPLETE and FUNCTIONING cockpit, than it will take BMS to have multi-frustum rendering.
Do it and start building your ACES -
Not only do you need a lot of $$$ and projectors for a dome, you need a lot of ROOM for a dome! That certainly make a dome and non-starter for me, even though Iāve spent hundreds of hours flying domed Trainersā¦a 270 degree flat screen is fine for me.
And Iām going to get around the problem of the HUD moving around with head tracking by building a working HUD and removing the HUD from the OTW all together. I know of at least one utility that will do that, and I can think of another trick I might consider - just turn the graphic to transparent. I want to do that to get rid of the nose model as well - I donāt think Iāve ever had a seat in a real fighter where I could see the nose over the glare shield, and Iād think you canāt see it sitting tilted back 30 degrees.
-
Ehmā¦ did you looked at the picture I posted above and where the single projector is there located? THat is a single projector half-dome idea, just that the projection light has to bounce off a mirror-ball to get 180 horizontal and 0-90 vertical dome. That is what I want to try out, but currently have trouble finding a warping software that would be able to handle something like this. Want to run this as experiment with old office beamer, if working properly then a 4k beamer should be a good second attempt since this setup is heavily loosing resolution quality on the edges.
-
And where exactly do you wanna place your pit?
That would obstract the projection, as its shown in your picture -
And where exactly do you wanna place your pit?
That would obstract the projection, as its shown in your pictureI was hoping the F16 cockpit has the HUD projector quite in your face behind the selector numerals, might be candidate holding the mirror ball. And yes, the 180 FOV will not be directly achievable as my head will be some distance from the mirror, putting you off center of the half-dome, still, right now in my mind having vertical 80-ish degree space is worth having only 170-ish FOV.
Still, office beamer and I have relatively round and white tent Was hoping to beg some warping software sales guy for a trial and try it in it with some makeshift setup to validate the idea before committing to building anything.