Immersive Calibration Pro NO CAM

[bravolima]

Hi all,

- does anybody in the community achieved good results using Immersive Calibration Pro NO CAM?
- willing to share these experiences?

As usual my space is limited and the sim is already built, that`s why the calibration process with CAM is not possible.
The size of the the projection screen is 184 radius and 170 hight. I use a three Beamer BenQ W1080 setting, the outer ones are overcrossed.
I used Immersive Pro for a while but not the workout wasn´t  precise enough, I had distorsion on the side, of course due small diameter. Now I want a most geometrically precise solution and  started to work again with Immersive Calibration Pro NO CAM. Software both is installed on C:  Main Directory and Network setting is established, the Print services software Bonjour is installed on both systems  and the client software is in connection with the rendering client on the FSX machine.

But unitil now I didn´t achieve any result. I get a more or less good result equal shown in page 28 in the tutorial. But when I start the single frustum mapping the result doesn´t look like the one on page 30! My one is distorted, shrunked and displaced to the left. Contrary to the tutorial I don´t see any red flagged control points during the whole process. Support from FlyElise-NG refers only to so called step-by-step tutorial (I doubt that it is one in its classical meaning).

Help would be appreciated

Bernhard (LSZH)

[RayS]

I just re calibrated my screen 2 days ago after installing a new video card.

I used a cross laser pointer, and a 6' ruler marked off in 8 equal segments. It turns out that the 6' ruler made the alignment process a LOT simpler. I basically divided my screen up into 16x8 points across a 12' diameter curved screen (only 2 projectors, no overlap)

For each column I simply moved the ruler to each column and aligned the 8 vertical points on the laser beam center on the column, horizontally aligned with marks on the ruler. It sounds imprecise but the end result turned out amazing.

Where the 2 projector images meet in the middle there's no overlap, and no gap, so the fluidity as you visually cross that boundary is pretty good.

It's tedious, but I think no more tedious than using the CAM function, or any other automated function. (Which I always found needed some tweaking after the fact) 

[MCrevot]

I do exactly the same way ; I drew the graduations corresponding to the grid on the edges of the screen.
75 euros for the laser and its support, don't hesitates, it is extremely efficient.
Michel

[navymustang]

The problem you are having is with the frustrum mapping.  Sounds like you got the matrix alignment OK. Ensure that you have the column and rows numbered on all three views and overlapped correctly, I made that same mistake last week on an alignment.  Got same odd result you are seeing.
Turn on your frustrum display and ensure that it "looks" right. That is, the frustrum cones are equal on each projector.
Jim

[bravolima]

Thanks so far for your recommendations and hints.
After some inital problems and a reinstallation and update of the software I calculated today the first multifrustum. The lateral distorison never come up.
Not bad! After first experiences I recommend not to stay to close on the pictures of  the step-by-step guide. Perhaps my beamer setting is not so perfect as shown and the red points over the colums  seem to be remnand of an older version. Never had to deal with them and didn´t needed them. Secondy keep an eye on equal distance of the column and move the points initially more in colums and do fine tuning with single point movements later. Don´t forget to save stages. inbetween. Tomorrow I will prepare Immersive Display Pro and import the multifrustum to FSX.

[Ridgenj]

I basically divided my screen up into 16x8 points across a 12' diameter curved screen (only 2 projectors, no overlap)

Dear Ray, I have also a 12' curved screen but was planning to use three Optoma GT 1080. I saw a very nice setup with the BenQ1080 from a Swiss sim. At this point I am looking to do an additional 700 USD investment since I have already the three identical IG's in place.
What would you recommend because I hate the  overlap strips.
I am trying to do the calibration before the MIP and shell are in.
My software is Warpelizer.
Hope I am not highjacking the topic, apologies if so.
Cheers
Luis

[navymustang]

Bravolima - you may still have an issue with your alignment.  The test pattern image you posted shows the test pattern low on your screen.  The center of the test pattern should be on the center of your screen.  If you perform a frustrum calibration and let it show the field of view grid, the centerline of that display should be in the center of your screen also.

[sagrada737]

Bernhard,

Interesting results from your second go around with the warping software.  I notice that there is still some slight misalignment issues with the calibrated warping result.  It seems that you are using 20 columns for your calibration adjustments.   You might want to increase that to a higher number of columns, eg 30 or more, and setup for individual adjustment of each point.  This might help to improve the alignment in the finished display result.  Also, be sure to make each column in the grid vertical, and that the distance between each column is exactly the same - same for horizontal.

The frustrum settings are difficult to understand and sort out.   Essentially, you want to set the frustrum coverage to just cover the projected screen size.   Also, for better blending, you want to have around 12% to 15% overlap for good results.   However, blending is highly dependent on the settings of each projector in terms of their brightness, contrast, hue, color temp, and their respective values.  You can check this by checking one projected image against another, and make necessary adjustments to each projector to get them as close to each other as possible.  Then tackle the blending adjustments - your results should be greatly improved.

Also, not all the low-cost projectors are identical in terms of their optics.   You may see some variations at the extreme edge of the projected image.   This is why having a greater percentage of overlap is necessary - to help eliminate these variations during the blending process.

In the end, it is almost impossible to achieve a "perfect display" using multiple projectors.  The good news is that once you are concentrating on flying your Sim, you really won't notice some of the imperfections of the Sim display - especially when you are flying in scattered clouds.  Night flying with a multi-projection display is the worst, with the blended display tending to show bright bands in the overlap areas.  In the end, it is all a compromise to realize what works for your particular Sim setup using a multi-window, multi-projection approaches. 

As a side note, you might want to try just a simple "warp" of your three windows, and forget about the blending.   This however will yield a slight dark vertical line separating each window.   In addition, you will not have the overall correction for perspective that you can achieve with a blended multi-window display, which keeps the horizon horizontal.   However, you may not mind the very slight curved effect of the horizon on each display.   This simpler approach also will typically yield a clearer display image vs. the blended approach.  Here is a video posted by Scott a couple of years back that shows what this type of display would look like:

https://www.youtube.com/watch?v=MiThNYfiwCA

As you can see...   The black vertical lines that separate each projected image are hardly noticeable.

Best of luck on your project.
Mike



Mike

[bravolima]

Once again thanks for the contribution. Here are now my first shots, after modification of the .flt file
The horizon seems to be a bit bend to much.
@navymustang: I made a another test and brought the of the visual centerpoint and the grid centerpoint together, there was no change, but you are right somehow the center of the frustrum seems to be lower, I thought this represents the horizon.



modfyied: It seems that the height of the projection has to be adjusted.

[sagrada737]

Bernhard,

That's an interesting display result.   The curved horizon certainly seems abnormal.  One thing you can check in your warping software, is the Vertical field of view value  (FOV) for each projector (try a value of 68 degrees or so), and the Aspect Ratio value (whatever you set your screen to).  These must be the same value for each projector BEFORE you export the calibration for your FLT File, or you may see such distortion in the final result.   Also, the flight file Window information should be checked to ensure that the Zoom, Translation, and Rotation values are what you expect for Window 2.1   Window 3.1   and  Window 4.1

Mike

[bravolima]

The project with FS Elise NG stopped now. The progam itself is good, I have nothing to say against it. The reason why I stopped are its multiview capabilities. For my needs and my beamer setting with narrow diameter (3.90 m) it needs proprietary configurations and this is beyond my knowledge and usual support

[navymustang]

Bernard and Mike - I believe the recommendation is not to specify the vertical FOV, just the Horizontal and allow the aspect ratio to figure out the rest.  It removes one less place for errors.
Jim