@pingdesigns

Hi,

Well, the setup for equirectangular images is a way to work with the distortion when projecting onto your physical sphere. It is a way to get your content to conform to the circular/spherical shape. To map projection to a complete 360 physical sphere requires 4 projectors. When you use 1 projector you can't avoid the distortion of the images as the projection image tappers around the physical shape. The equirectangular image format is likely to assist with dealing with the outer edge distortions. If the distortions are not critical because of the type/style of content you are projecting a simple circular mask might be all you need.

This is an equirectangular image that is distorted at the top and bottom so that it will wrap around a 3D model sphere at the polar regions.

Here is the same image used with a 3D model sphere in Isadora projected roughly onto a physical sphere using one video projector (earth globe is all I had at hand). Because the physical surface is curved there is inherent deformation of the projection surface. The viewing angle of the audience is also going to be optimal at the same approximate position as the projector.

You can pre-render content and then format the render/re-render as 360 video using a number of different software: Blender, PremierePro, Final Cut Pro

But... If your content is more organic or you don't mind the distortions, as mentioned you could use a simple circular mask shape as an alpha mask on your regular rectangular output.

Best Wishes

Russell

Hi, 

I made this patch to map projector, calibrate color, brightness, contrast, saturation,
you need this ffgl in your glsl folder

GLSL-FX-brightness contrast saturation.iua3

hope you'll like it !

you can change the circle shape by any other one clicking on composite and replacing shape
bests
ps: I'll add quad distrortion asap;-)

touchosc izzymap 2.zip

@mark i think there s also a pb on the zoom if thé output point goes beyond the workplace limit, zooming out doesn t allow to catch the handle back...

@woland

thanks  for explaining ! and thanks so much for the feature request ! 

@michel Great, thanks just tried it out! Thank you for the fast reply!

@woland thank you!

@emtv I don't know your set, but often lighting the thing that does not move is easier than lighting the thing that does, so carefully lighting the set with IR and not lighting the stage could be an easier solution. As long as you have good contrast between the set and the actors you can always adjust the image and invert it. I could imagine trying to always light the actors everywhere on stage and not spilling on the set could be more difficult.

@bennnid said:

I thought It was because of my projector definition

 What is your projector definition? The higher the resolution the better (obviously) but also understand that the larger the image you're trying to project on a surface, the more visible the individual pixels will be. 

@fred

All excellent points. I can see how this would be a very valuable addition to IzzyMap. Thank you for taking the time to give us feedback :)

I have a case where I want to use a user actor for mapping a lot of small pieces, but I want to use the same mapping for an alpha layer, but the pieces in the two need to have different blend modes. It would be great if the projector blend mode had some other options, like when the mapper was engaged the blend modes still worked, so they could be additive, transparent opaque OR mapper. and when the mapper was used and transparent was selected it would set all the mapper pieces to transparent. If it was set back to mapper then it would use the blend setting of the individual pieces. I think there are other options for doing this, but a way of having the smae mapping with different blend modes for using alpha layers with different projectors into the mapper would be cool.

@dillthekraut thank you very much !!

Best,

On Mac you have these options:

@barneybroomer said:

 Sorry! Today all day behind Windows/Mac makes me confuse, so, control-click publish points and command-click adding points.  Thanks

 No worries, I only knew that because I went and checked to verify. It's a lot to keep straight in one's head, especially cross-platform.

That buttermilk suggestion could saved me a lot of money..... Thanks for the tip! I will try in the future.

@woland perfect, That will do the job nicely. thanks.

Although I still don't know why the "complex shape" tool behaves differently for me in the input window than the output window.

it's one of those gaps in understanding that I feel will come back to bite me at a later date when I need to mask something a bit more detailed. 

Oh well, I'm on track for now at least. 

thanks everyone, You've given me some paths to start investigating.

I'll either come back with a bunch more questions or i'll let you know what I settle on.

Hi,
so trying out the suggestions - it would be great to have a dynamic control of the contour of the tiles. - May be here the 3D Quad Distort object has the highest flexibility.
But I am not sure how to tweak all the parameters of "3D Quad Distort" to shape on-the-fly. So it needs a few more experiments.

static version:

dynamic version:

3D Quad distort version:

@pbleopold

No problem. About four years ago I spent three days going absolutely mad trying to figure this exact issue out. The forum pointed me in the right direction and now I include this step in all the shows I make.

Thanks a lot for the quick response. I guess, I better keep coordinates and "offset" separated... ;o) Best regards, T.

or another way, what if the shape with the most number of points
sets the number of points of the other one, with identifiers, 
and we shape the 2nd form (assuming we have too many points but can align some of them), 
to get our result and transposition?