Perfect blend from real to virtual, but how???

I'm hoping someone could shed some light onto how they have managed to achieve such a seamless join between their actual racing tub and the display screen show in the youtube link below.



Apart the camera/FOV settings to get it lined up perfectly with the tub, how do they get such a perfect blend? Are they running 3 TFT screens, with the centre screen right up against the front of the tub? If so, how come there isn't a slight shadow where they meet??? Alternatively, could one do the same by using a projection screen with a perfect cutout for the tub to slot in and project the image onto the screen that way? The reason I ask is because I might know a friend who has a damaged tub on his hands, and I'm toying with the idea!

Cheers,

Julian

There is no shadow because you are looking straight into the light source.

From those camera angles the tub doesn't need to touch the display. If you match the size of the real and virtual cars, it will give very good visual transition.

Projection wouldn't be any harder to do. Front projection would probably work better.

Interesting... So do you think that a different camera angle would expose the clear cut-off from the real to virtual car? The reason I ask is before normally at the front of the tub you would find your master cylinders etc etc and hence would not be able to get the screen closer than 150mm to the tub. If I do embark on this project, I plan on doing it right and with a healthy budget as I've been wanting (and saving!) to build a simulator for a while now.

In the back of my mind, I've always been interested in a curved screen display(1 or 3 PJ set-up) projecting a 1:1 image and really immersing the driver, but have always contemplated going down the TH2GO LCD/TFT screen route as its more familliar to me. I've also been following the dome/canopy experiments on here, which really impressed me with the immersion factor. The only issue with opting for a dome, is that I would not be able to align the real tub with the visual image as the dome would need to sit ontop of the tub, making me wonder if I'd lose that realistic aspect? Also I believe that you'd have to scale things down by some extent,losing the 1:1 scale effect.

Unfortunately, I'm pretty new when it comes to projectors, and am finding it hard to choose the right one (and configuration) as there's such a variety. The good news is I'm always keen to learn and do have an engineering background. The only thing I'm sure of is that I'd be willing to spend in this department for the right one, as I believe that since this would be a static simulator, the display aspect is crucial.

All advice is more than welcome, and once the build is underway (if it does go through), I'll be more than happy share step by step pictures and the end result :yippiee:

Julian

P.s. Here's a little clip of what I get up to in my spare time

In my opinion you should rather spend the money on motion first instead of visuals. You will get a much better immersion/realism value for your money. I would choose a motion sim with a single 42 display over a static plastic tub with huge screens any day!

Yes, I'm almost certain that different camera angles would expose the cut between real and virtual in that video. I can't see how a curved screen would be any different at the connection here since the tub doesn't have to touch the screen. As long as the two cars are scaled and aligned properly, the transition will be good from the cockpit.

The sim in the video is limited to very specific, open-wheel cars. Driving a touring car or similar would totally loose the smooth transition and look a bit silly IMO. If I was to build something similar, I would definitely make it work with a wide variety of cars.

Thanks for the advice Frakk. The thing is I really want to keep the tub idea as I've been in one and the experience is very good! Another plus is that the driving position would be spot on and I'd integrate the driver controls into the actual steering and pedal systems (btw I only intend on using it for rfactor formula racing). I do see where you're coming from and after a night to think about it I'm actually thinking of going down the 'motion' route, but here's the big question... Can I rig the actuators to the tub directly, thus inkeeping with my original idea? The tub weights 58.9kg and I would not let anyone heavier than 80kg get into it. I suppose 2DOF would be a sensible route, but then again I haven't researched motion platforms that much.

Thanks again,

Julian

I've had a look through a few 2DOF projects and it seems that everyone goes for the SCN5's, attached to just the seat (which I imagine is to keep weight as low as possible and hence keep momentum down? apart from the muscle tensioning affects...) I've also read up on the SCN6, which to my understanding is a heavier duty actuator, but less responsive? Digging deeper, both are available in different displacement ranges; 100 & 150mm seem to be the most popular. Furthermore, this displacement is not strictly to increase the maximum displacement of the simulator, but to increase fidelity by allowing one to position the actuators in a different manner?

To my understanding (again i could be wrong?) a motion platform simulator is balanced on its centre of gravity using a universal connector, and two actuators are connected to either side of the rear to simulate pitch up/down and roll, yet no yaw as this would damage the actuators in this configuration? I can see there is clearly a large amount of detailled measurements for this configuration, but what would happen if this 'configuration' was applied to a tub arrangement rather than just a seat; Would I still follow the same actuator angles, and more importantly, would it give good fidelity and realism? Would I get away with using SCN5's or would I need to opt for the slower SCN6's?

I've driven a few open wheelers in my time and I think I'd prefer the simulator to only pitch and roll to small angles, with the rate at which it is peformed being the crucial factor. Another reason why I'm slightly against large displacements is due to the fact that my display is going to be fixed. I do think that the end result could be quite good, but since I don't have a strong background in motion platforms, I'd appreciate some feedback?

Kind regards,

Julian

Hi Julian
I have tried all three LCD, Big screen (but only one beamer) and now the dome concept.

I have found watching Youtube films are deceiving as you never get the correct frame FOV; that is sit the camera in the driver’s seat and you get the camera view not what a human eye would see as the FOV for the camera is very much limited compared to the human eye. Case in point, I built my first screen after thinking 3 LCDs and a Fresnel lenses was the way to go, (I never found a Fresnel lens large enough).
The resulting 3 LCD set up was like looking at three separate screens, although the main focus its always on the center the peripheral view and the way my brain perceived the images was not one of being in a car but looking at at three screens I never felt that I was in a car.

The big screen maybe ideal if you try 3 beamers, I have spent time on the flight sim sites and some of those look awesome with huge wrap around screens an 3 PJs but again in my set up on a 96” X 36” a one beamer I just got the impression of being detached from the action.

The Youtube video is impressive but does this not limit you to one type of body style? What happens when you want to drive a saloon or any other game using a different style body?

The graphics in today’s games although not perfect can offer you a good reproduction of the cockpit (I find driving from the nose is the most immersive) and without doubt out of my previous attempts this seems the way forward.

I have just ordered a 60” P95 type acrylic dome in the hopes of a closer 1:1 image reproduction I will post the results as soon as I get it set up

Hi Brightonuk,

I've been following your simulator evolution for quite a while, great work! Its only after looking at your results with the dome that I actually started to think that a dome would be worth while. I do have a few issues with the dome though; amount of distortion, scale and the fact that I'd have to sit the dome ontop of the tub rather than try to merge the virtual image to tub. Also, how far into the dome do you position your head to get the right immersion effect? With a motion platform, would you run the risk of disorientating the driver seeing as his head would be moving relatively close to the dome projection?

Would going for the larger dome mean you would have to switch to a 3 PJ config or would one PJ still cope? If the results are positive (which I think is highly likely), I might be willing to take the other half off your hands if you're interested in selling it. The only thing is that the open wheeler I intend on basing the sim on has a front width (tyre to tyre) of 1.835m (72). Your dome being 1.5m in diamter has a base circumference is 2.356m, would settings the FOV in rfactor to get a 1:1 be possible?

I'm still thinking if I'm going to go down the 2DOF route too, as I don't have a great deal of knowledge in this area and am waiting for a few opinions on whether it would be 'do-able' with a tub arrangement rather than just a racing seat.

I understand what you mean about the body style limitations, but I'm just out to make the best 'formula' simulator that I can (its been something that I've wanted to do for while!)