View Full Version : "Mini" Beamsplitter Rigs?
October 23rd, 2011, 04:23 PM
A small Beamsplitter Rig could make the newer and smaller cameras, such as the SONY NEX-FS100u, a valuable and affordable alternative to what is presently available. Easier Portability is also an attractive aspect. Although syncing would be dependent on the IR Remote, I have had good look with that with the sony EX1's.
The specs and features of the SONY NEX-FS100u, combined with a small beam splitter could be quite something.
Has anyone run across "Mini" Beamsplitter Rigs?
October 24th, 2011, 09:35 AM
I'm working on a mini beamsplitter right now. I just dropped the designs off at the fabricator and should have it built this week. I'll send you some pictures once it's done.
October 24th, 2011, 10:43 AM
There are small designs like the Dark Country or Atom but a smaller mirror also means smaller field of view.
I have designed very small rigs for cameras like Cunima but the mirror is the limiting factor for short focal lengths and isn't really worth the effort.
October 24th, 2011, 04:32 PM
A small Beamsplitter Rig could make the newer and smaller cameras, such as the SONY NEX-FS100u, ... Although syncing would be dependent on the IR Remote
I just want to put this out there for anyone tempted to try and fit a square peg into a round 3D hole - You must have a way to physically genlock both cameras to assure usable 3D footage. I know people swear that they can "lock" Canon still cameras and others with lanc shepards and other techniques, however, it's been my experience that nothing is worse than the retinal rivalry that accompanies 3D footage that is out of sync by virtue of no genlock or locked shutter sync.
What Tim says about small mirrors also applies to trying to "kludge" a couple of non synchronous cameras into a 3D project - it's hardly worth the effort.
I'm sure I'll be flamed by those who claim success without genlock or shutter sync, but I venture to guess that their "success" at sync only lasts a minute or two before drift begins. And once drift begins, you have to stop and start all over. Hardly a recipe for professional production, but ok for amateur goofing.
And to address your first question, the Atom is the smallest workable unit I've seen but the Hurricane Rig is very affordable and is small and light enough to work with Steadicam and handheld. 3D Film Factory also have a small 3D rig out now too.
October 28th, 2011, 09:59 AM
This specific 3D project only requires in-studio 3D work, and rather close focus work at that , so a smaller field of view is not a issue. Also, the objects and motorized armatures are from 8 inches to 3 Feet away, so the mirror should not be in the way either.
Interestingly, I searched Youtube and happened upon Jess Blanchard's video: "$100. DIY 3D Beamsplitter Rig"
$100 DIY 3D Beamsplitter Rig - YouTube (http://www.youtube.com/watch?v=ggCLjr10eIs&feature=related)
Which would function well with smaller, DSLR type cameras, or even my Panasonic 750 palmcorders.
I also went to B&H and saw Alisters HURRICANE BEAMSPLITTER rig, with is a beautiful and very versatile instrument. Robert and Mickail (?) at B&H say hello, and they offered to set the HURRICANE up with two SONY FEX-100U's if I call a week ahead of time.
If I get a good paying 3D gig soon, I will certainly purchase HURRICANE.
Syncing the SONY FEX-100U's is still not a definite, but I will be able to import the files on set, and decide if shooting again for sync is necessary.
This work process only requires a minute or so of sync footage, so "Drift" is not an issue either.
That being said, In this digital age, I have never found "drift" to occur, even in hour ,or longer sequences, and if it did, the arrant frame could easily be located and removed in post I would think.
Perhaps it is something that only happens in the larger, more involved and complex 3D systems and setups?
Jesse, Looking forward to seeing how the fabrication turns out, Will you be uploading video demonstration??
October 29th, 2011, 12:29 AM
Sync mismatch has absolutely nothing to do with "digital age", it is an analog phenomena, and doesn't need hours to occur. It is virtually impossible to find two identical crystals and place them into identical electrical tank circuit, so that the internal digital camera clock will not start drifting immediately after power up. This drift, if one is lucky maybe small and slow, but it will occur immediately, and be noticeable in 15-20 minutes in best cases. Obviously, it depends on the angular velocity of the object in front of the cameras, no matter whether the object is moving, or the camera is moving, panning, etc. It could be specially bad if the cameras were zooming. The problem is that the drift doesn't occur on per-frame basis, the frame can just occur half a millisecond sooner on one camera vs the other, and the two frames will be therefore sampled half a millisecond appart, and each eye will see the image at slightly different location. If the drift is between about 10% of the frame and 50%, the mismatch can be easily seen, cannot be corrected, and be very annoying. If the camera or scene are not moving, of course the cameras can be 50 frames off, it doesn't matter. The best solution for this is internal or external genlock, clock phase lock, or some other trick, that will at least start the sensor samplings at nearly the same time, but the drift will occur, unless the cameras are specifically built to correct for it.
October 29th, 2011, 12:38 PM
Thanks for shedding light on the sync mismatch topic, and explaining it in clear detail.
My poor use of the word "digital" was a general term, referring to the new solid state Memory card era of or "tapeless recording"..
but that may be muddling up the terms and topic also.
Anyway, genlock used to only be available on the high-end ($$$) cameras.
I seem to recollect that the SONY EX3, is, (or used to be) one of the least expensive Cameras with Genlock, but then I heard even that was not "true" Genlock (??)
Genlock aside, a good deal of people (including myself) are still at the point of just manually (or with a remote) activating the two cameras and aligning them simply by looking for the two closest looking frames in each track during post production.
and if Drift occurs over time, removing a frame periodically to get things back closer in alignment.
On a lighter side, and not to sound frivolous, but perhaps these early 3D videos will not be unlike the basic audio recording techniques of Music "Garage Band" recordings of the 1960's, and become
"3D Garage Vids" and have a certain panashe unto themselves??