garageCube

[BigDan]

Hi Guys,

So i'm looking to get into this field. I'm a long time A/V video guy and am extremely proficient in most of the CS5 production library.

I have two questions:

1) If I go the route of taking a "still " picture of the object I want to manipulate how do I find out what kind of lens is comperable to the projector? most of the projectors I work with are 1.2-1.8 or 1.5-2.0 type lenses, but for stills the numbers are a little different aren't they (not a dslr guys at all).

2) settling up the projector, and taking the picture is all well and good I guess, but how on earth is it possible to do that for enormus buildings using multiple projectors? My first guess would be they don't do something like that, they'd use a CAD of the items they are going to project on and animate that...

if the CAD part is true, what CAD program do you guys normally do? and how do you draw a cad of a huge building when you don't know the measurements? I'd imagine you'd pull the architectual diagrams?


Thanks for the info! I just want to make sure i'm stepping into this right! I'd imagine the larger-type stuff is further down the road, but would be nice to know. I know vectorworks as a CAD program, but that's it. I supposed I can pull out sketchup or whatever people use.

[deepvisual]

lens focal length isn't crucial, so long as it isn't a fisheye or extreme wide angle.

[BigDan]

Thanks for that info.

Here's the real thing i'm not quite understanding yet.. it seems to be a well guarded secret too..

So if I go my route on a single projector show, and I take a snapshot then mask off the object in photoshop, do the animation video in flash/after effects/whatever, then I finalize it with some music in FCP and use modul8 to project it.. it all makes sense to me..

what i'm not getting is the next step, when I start looking at 2 projectors or 3, or 4, or more!

I understand I set it all up, take the camera snapshots from each perspective (right?) then how to I put all these pictures together so I can animate them into after effects or flash? is this how people do it?

If I can figure out that one cruical step I think with a lot of tinkering I can get the rest.

Thanks!

[deepvisual]

while its possible to run video out from one computer,its not easy to align individual outputs in M8. we will have to wait for mad mapper for now or use other apps..
however, you are right to take your photos from each position.
when joining these together, dont.
instead use an architectural feature as the border between outputs - such as a drain pipe or corner and make your images butt up together on the building.

[BigDan]

while its possible to run video out from one computer,its not easy to align individual outputs in M8. we will have to wait for mad mapper for now or use other apps..
however, you are right to take your photos from each position.
when joining these together, dont.
instead use an architectural feature as the border between outputs - such as a drain pipe or corner and make your images butt up together on the building.

Thanks Deep!

One more question on that if you don't mind. So I don't put them together, I just mask off the areas of the photo and import them individually into flash or after effects or whatever I want to work on?

If modul8 can't do multi projector well, what do you suggest? My thought is something like a hippotizer ( http://www.green-hippo.com/products/hippotizer-hd )

basically what i'd do then are these steps:

1) Set up what my objects i'm projecting on are going to be.
2) Set up the projectors at the distances i'll be using for the "show"
3) Take a photo from each perspective
4) import into photoshop, set the resolution of the projectors, etc
5) mast off the areas of the different shots.
6) import into flash/after effects/whatever and do my work
7) import into premiere or fcp and finalize my video
send into a media server like the hippo on show day and split up the video accordingly
9) once fully alligned its showtime!

sound about right?

[deepvisual]

hippos used to have limits to what they can do regarding outputs.
just check first what version you have.
also beware of the multi- point distortion.
you'll only need to use 4 corners if you have been methodical.
I;ve seen people spend two hours lining up a hippo when it could have been done in 3 minutes if they knew what they were doing.