Red 3D (Crosseyed)

[Jim Cummings]

Hey Y'all,
This is Jim Cummings and I'd like to post some RED 3D footage for Crosseye viewing.

http://vimeo.com/12036538

This is not from the unreleased Red 3D rig, this is from a single red with a beam splitter.

I used an adapted Russian SKR-1 3D rig, and fit it to my cheap Nikon lenses.

Does anyone know of a rig with larger mirrors that will get a perfect split of the 4.5K sensor?

[Pedro Guimaraes]

cool,

You should anaglyph it and post it on youtube 3d player.

Yes I do know of one, we were testing it last week. A prototype. Unfortunately I cannot release pics.

Also slight technical correction. The rig you describing if I'm not misunderstanding you is technically not a "beam splitter". Rather more of what I would call a "image splitter". A "beamsplitter" type rig would have a mirror with a reflective and a transmissive side. A image splitter is a device that usually uses high quality front surface mirrors .

It's a dual mirror device right? 1 camera, 1 lens, 2 or more mirrors producing 2 images on a single sensor?

have any pics?

[Jonathan Stevenson]

Holy headache batman!! But looks cool :)

[George Tsai]

had no clue youtube had a 4d channel till now haha

[Thor Melsted]

no need to anaglyph for the youtube player
just post the same crosseyed video and tag the video with the proper yt3d tags and you'll get your choice of 3d viewing methods upon playback, including anaglyph.

cool video, btw

[Robert McGee]

What if the lens could be turned sideways? Instead of being side to side, the image would be top to bottom?

[Dan Hudgins]

What if the lens could be turned sideways? Instead of being side to side, the image would be top to bottom?

It is possable to do over-under with like ANA mode (1.2:1 in REDONE), but you need maybe 6+ mirrors not 4 in a side-by-side fixture.

If you have the optical centers above-below you get no stereo unless you watch the film sideways!

If you want to do that with a 4 mirror rig, you turn the CAMERA sideways, and keep the two optical centers side-by-side like the viewers eyes. You can then rotate the frames 90 degrees later in software (like mine) as part of the grading. To swap the images around, you can cut them out of the camera frames, rotate them, and then put them back into a 4K over-under for DCI projection. I plan on working on some better cut and past options to do that sort of re-fitting of stereo images. For now I do have image cut out to R/L frames, and stereo formatting working L/R to various formats, but not quite as flexable as it might end up later.

With a 4 mirror rig, you can use TWO scope adapters and get side by side squeezed format, then unsqueeze and get 2.39:1 stereo 3D. With the camera sideways, you can use two DOVE prism to rotate the images so that the width goes up and down, which will be side to side when the image is rotated in post.

http://en.wikipedia.org/wiki/Dove_prism

Quote: [Dove prisms have an interesting property that when they are rotated along their longitudinal axis, the transmitted image rotates at twice the rate of the prism.]

[Dennis Guskov]

adapted Russian SKR-1 3D rig

where did you find it? link plz?

[Omar Saad]

where did you find it? link plz?

Just wondering if anyone has found the link for this rig yet?

Thanks,

Omar

[Paul Hubert]

Just wondering if anyone has found the link for this rig yet?

Thanks,

Omar

I think it's a SKF 1 Stereo Kit not SKR 1. You can find them on ebay for cheap. Just search SKF 1 or Stereo lens and you should find some.

Cheers,

--Paul

[Kobe Van Steenberghe]

If it is the SKF 1 then you can find some info (including link to ebay) here:

http://www.viewmaster.co.uk/htm/skf1.asp

[Pietro Impagliazzo]

Am I the only one who can't view crosseyed 3D?

[jimmycthatsme]

I think it's a SKF 1 Stereo Kit not SKR 1. You can find them on ebay for cheap. Just search SKF 1 or Stereo lens and you should find some.

Cheers,

--Paul

That's the one Paul.
The russian packaging makes it look like SKR, but the english is SKF.
Thanks for clearing that up.

[Jason Goodman]

That test looks nice and sharp and the 3D looks good, but I advise the list to take it with a grain of salt. I have been shooting 3D for 17 years. Prior to December 18, 2009 (when Avatar made 3D mainstream) average people learning of my involvement in 3D would want to give me advice offering some kind of technological improvement, either eliminating the glasses or somehow making 3D better. I would tell these people, 'anything you can think of that is optical or mechanical has already been tried'.

Those attempts that worked well (1/2 mirror beam splitters etc) remain in use. Those that didn't have fallen by the wayside for the most part. The only things that are advancing 3D technology at this point are advances in digital technology. Better image sensors, higher resolution, smaller cameras etc. I advise you to be wary of any "new" lens attachments or other opto-mechanical devices for 3D shooting. While they aren't necessarily bad, these prism or mirror based lens attachments have been around forever. Applying them to large high resolution digital sensors is not bringing anything new to the table.

If you can deal with the extremely vertical nature, or if you can crop out your 16:9 frame (or choose a normal aspect that suits you) and still have enough resolution and 3D effect to satisfy you, maybe this is an interesting device. Otherwise, it is an entertaining distraction with little practical use. Using additional optics to make an over/under format has also been done. This is more sensible as it preserves a useful aspect, but each of these approaches have challenging issues.

Zoran Perisic has done a lot of work in this area with some good results. John Rupkalvis has as well. These devices like all 3D cinematography equipment have their pros and cons. The most severe limitations include fixed IO (generally), limitations on focal length, light loss, optical uniformity and overall optical quality, resolution loss etc.

It would be interesting to see this type of device on a Canon 5D sized sensor (vista vision format) presuming that sensor could output greater than 1920x1080 at 24fps. This would at least provide a more normal and usable aspect ratio without the need for a tremendous amount of glass in front of your lens. You could also try using anamorphic lens attachments to optically squeeze more info onto those doorway shaped frames, but that has been tried as well and the experts of these devices seems to run for the hills at that suggestion.

Obviously the small size of the 3D camera this lens attachment allows is appealing, but be aware this may not be the magic bullet many hope for.

Jason Goodman
CEO
21st Century 3D New York
505 8th Avenue #1006
New York, NY 10018
212-244-8585

21st Century 3D Hollywood
3450 Cahuenga Blvd W #509
Los Angeles, CA 90068
323-799-1046

http://www.21c3d.com
jason@21stcentury3d.com

[Rodolphe Pellerin]

Does anyone know of a rig with larger mirrors that will get a perfect split of the 4.5K sensor?

Hi Jim,

I am also interested in a splitter like this...

If you have a NIKON mount you could also try the LOREO 3D Lens in a Cap (Model : 9005) designed for APS-C DSLR with 1.5 or 1.6 crop factor.
The problem is probably the quality of the 2 lenses used in this product (plastic polymer composite). But maybe we could improve it...

http://www.loreo.com/pages/products/loreo_3dlenscap9005-apsc_spec.html

http://www.4kparis.com
rodolphe@4kparis.com

[jimmycthatsme]

Yes,
Sorry to all, the packaging makes it look like SKR-1 but it is in fact the SKF-1 Russian Rig. Thanks for the Correction.

I haven't been able to get a proper Loreo Lens in a Cap 9005, but I am working on it and it seems that it will cover the full 4.5K spectrum.
If this is the case, then we will be able to pull (2) 16X9 side by side images at 1080p and still have room to converge in post.