I bought a Loreo mirror beam splitter lens a long time ago and tested it quite successfully on a RED One. Been waiting for someone to create an adjustable (and affordable) version ever since.
|
|
I bought a Loreo mirror beam splitter lens a long time ago and tested it quite successfully on a RED One. Been waiting for someone to create an adjustable (and affordable) version ever since.
Hey Petri,
Which model you have tested, 9005A or 9004T?
Can you share some of your test (Footage)?
Had that Idea to use the Loreo 3D Lens in a Cap awhile ago. Its really cool but leaves a blurrrrrryyy edge! I think Stereo toolbox has FCP plugin fix for it!! like 100 bucks on ebay! Lots of fun!! Yeah Gotta keep people like 6 feet back from the lens for good 3D.. Similar to the Gopro 3D system..
[IMG][img=http://s17.postimage.org/qd4x3mqq3/owen_korb_3d.jpg][/IMG]
It's the 9005. It's a bit problematic on Nikon mounts because it wants to lock at a 30 degree angle, and needs to be slightly modified because of that.
Oh, it was more than a year ago, I don't think I have the files anymore, sorry. In addition to the locking angle, it had some other pressing issues, like quite a soft image and a massive thirst for light. It ran out of light on a sunny day in the middle of summer. I wouldn't try it at anything even remotely professional, it's more like a fun toy. But the idea of shooting 1080p 3D on a single RED is a good one, which is why I've been waiting for someone to build a professional version.Can you share some of your test (Footage)?
I've been down this road with the Phantom 65 4 years ago.....
other issues which we had planned to motorize.....IO and focal length adjustment in relation to mirror size needed. Also limiting crosstalk area between each image via a "septum" down the barrel of the lens or right against the back of it.
Here is me machining the "septum" we had in our system..
Here is our completed system on the front of a rally car mounted to a Phantom 65 with a hasselblad lens.
Our system during it's first test for a 3D Rally movie.
Here is me mounting it on the roof of the rally car
during another test shoot a monitor shot.
here is the result on the RED one....pictured is designer and 3D pioneer John Rupkalvis
The other major issue....is post processing of the images created.
We did it with a Quantel Pablo system. We created some "scripts". I guess you can do similar things with AE.
We would shot a chart then from this chart we would remove any distortion that was created from the mirrors. These coordinated would then be applied against all the footage.
Then another "script" would "SLPIT" the image and place it into L & R folders.
The result would be 2 folders with L&R images free from distortion and ready to assemble into a 3D timeline for review.
Biggest issues...... possible usable focal lengths, lens availability, mirror sizes, mirror motorization, mirror mounting(flatness), maintaining mirror alignment during any adjustments, developing an internal septum.......OMG........the list goes on and on......
The system worked. We proved our concept. What this idea needs to turn into a usable "product" is a a couple hundred thousand to couple million in R&D.
Even then it will not address every situation you encounter during a 3D production. It can be a very useful solution if developed further.
I wish you luck and I know how much fun it is to experiment and "invent" (or re-inventing).
But for me the fun stopped when the funding stopped! Good luck in your endeavors.
the phantom 65 has that z3d system now which pretty much does what you accomplished, except it's all in one dual lens system now. the results are pretty damn good.
it's not the same......it's using 2 lenses. that a pretty big difference.
That design limits IO and can introduce other issues like color disparities, possible focus mismatches and other issues.
The main issue is that you have to use these tiny miniature lenses with questionable quality.
In the system we worked on you could use superb quality cinema lenses or in our case amazing hasselblad lenses. Both images go thru the same lens. This makes focus track perfect and also give more equal colorimetry between eyes.
the Zepar is useful but it has a very small IO and once again limits its use. The ultimate idea is to make a system like these (dual or single lens) with a variable IO/IA
Hey I finally put my Loreo 9005 back together and will post up some pics, files and an AE (CS 5.5) workflow comp sometime this weekend for everyone to start playing with!
AE with TrueComp duplicator plugin makes splitting and aligning a snap. Bring the new footage in and drop it into the original comp and then jump over to the 3D HD comp to preview & adjust.
Now if only RED made one of these lenses we'd be good to go! Wink wink nudge nudge...
Interesting! I do believe the future of 3d will not be bulky beamsplitters with two cameras, but rather very sophisticated optics.
Keep us posted on your progress. All the best in your endevor!
Cheers,
Rich
LOREO 9005 w/ wide angle converters
WALL-E?
![]()
| « Previous Thread | Next Thread » |