UltraPan8 Stereo

[carllooper]

I came across this from 1952. A Bolex Stereo Kit.

http://www.kylegilman.net/2009/02/18/th ... ex-stereo/
http://www.bolexcollector.com/accessories/filter50.html

What would occur when used with UP8? Well, you would get a stereo image film in 4:3 format.

What sort of lens system might be required to shoot stereo and keep the UP8 wide screen format?

One way would be to shoot at twice the normal speed (eg. 48 fps) with some sort of mirror swapping the exposure back and forth between the two lenses.

Each image in the pair would be at a slightly different time. But with the right software you could correct for the difference in time, ie. using motion interpolation to interpolate one frame back in time by half a frame, and the other forward in time by half a frame. MPEG algorithms do this sort of thing as a matter of course. And slow motion algorithms specialise in exactly this sort of thing.

The lens system? That's the hard one. I guess what you need is not a flipping mirror as such, but a shutter that let through one lens and then the other, slaved to the frame speed. Interesting mechanical design challenge.

Another approach, but I have no idea how you would design it, (or even if it was physically possible), would be to use a 16mm camera rather than UP8, but with a lens system that put one image above the other (rather than side by side).

Carl

[bakanosaru]

wouldn't it be much easier as you say to split the frame vertically and use the full 16mm height?

I.e. just use a 16mm camera and have the stereo frames in an over/under configuration. (No need to run the camera at 48fps and use motion interpolation.)

I'm pretty sure there have been over/under 3d setups in the past if not for capture then for projection.

[carllooper]

I'm pretty sure there have been over/under 3d setups in the past if not for capture then for projection.

That'd be the go then. Some sort of mirror/prism arrangement to get an under/over framing. Must have a think about that.

Found a wiki brief on the under/over format: http://en.wikipedia.org/wiki/3D_film#Re ... rip_format

Have just been looking at the Loreo stereo "lens cap" system, which is an interesting idea. I was thinking you'd need two lenses for a stero system, but in fact you can just use an existing single lens, where you put a mirrors/prism system in front of the lens. Means there's bit of a blurry line between the images.

SINGLE LENS LEFT/RIGHT SYSTEM


SINGLE LENS UNDER/OVER SYSTEM

[bolextech]

I am not a fan of 3D in general, but if you're going that route, the best system in my opinion is the 2-camera 2-projector system.
Such a set-up was constructed using Bolex cameras by J-E Nystrom and is illustrated on his website: http://www.sci.fi/~animato/

In the early 80s, I had the privilege of witnessing a live demonstration right here in Montreal by Lenny Lipton of his 3-D super8 system and I have to admit that it was quite impressive. We all know what later came out of that experimentation. . .

Cheers,
Jean-Louis

[carllooper]

My 8 year old daughter has been very much into the latest offerings of 3D films, but lately she's been asking to see 2D versions of 3D films instead. When I asked her why, she said she didn't really know why.

Which was actually a relief because I was getting tired of the 3D myself.

I suspect 3D really is just a cyclic thing. Some have argued that 3D was cyclic because the technology hasn't been good enough. But recent 3D is about as technically perfect as it can possibly be. And yet one can still get tired of it.

[carllooper]

What began in another thread, belongs in this one:

UP8.3D

One can do side by side 3D on UP8 - it just involves including an anamorphic in the lens setup. And then adding in a readily available 3D adapter (or building one's own - mirror box etc). It would simply look like this on the film.



And then write some software to convert the digital copy of the film to BlueRay3D for viewing on either a 3D TV or a computer screen with readily available 3D glasses for such.

I'm thinking of giving something like this a go. But not for doing 3D as such: in my experimental digital work, a 3D signal can be used to construct a better looking 2D signal. An optical flow map is established between the two images, from which a single 2D image can then be synthesised, but also a corresponding depth map extracted. But if this was all that was happening there would be no point. One could do it in 2D in the first place. However by starting the process in 3D it provides important information about how the resulting 2D signal can then be enhanced further. The depth map (from the 3D signal) can be used to disambiguate tricky optical flows across time - particularly at the edges of objects - where one object passes in front of another. The 3D signal allows one to tunnel deeper into the 2D signal (over time) in order to pull out more information than could otherwise be pulled from a signal that began in 2D.


Carl

[Nicholas Kovats]

Carl,

Check out this fascinating blog entry from Lenny Lipton (March 24, 2010) discussing his new 'film' based 3D projection system whereby one frame = 2 adjacent images. They managed to eclipse the low light output of the standard DCP 3D projectors, i.e.
http://lennylipton.wordpress.com/2010/0 ... rojection/

And his technical assessment (October 24, 2009) of why he thinks 35mm film based over-under 3D is not up to snuff, i.e. http://lennylipton.wordpress.com/2009/1 ... r-and-out/


Jean-Louis,

I am very jealous. I have been fascinated by Lenny Lipton's twin S8 camera/projection since the late seventies.