When small interaxials just don't cut it! - Page 28

I see it has been sold.
email rclark@intervideo24.com
phone 818-843-3624
2211 N. Hollywood Way - Burbank, California - 91505 - United States
Why don't you call them and find out. They bought it from somewhere. Looks like this may be what you need.

At least we know this technology exists.

Congratulations, I hope! Looking forward to hearing your impressions on it. Especially how portable it is.

Then I am keen to hear something about your experiences with this tool. Especially how it performs in synchronisation.

But how will you use that in the field? Looks fine for a studio...but for the field?

The info online was pretty sparse. When you get that synchronizer in, please post any reference to the manufacturer from any documentation on it. I'm guessing this box may be made in Europe.

Wolfgang- on your stefra LANC controller, where do you read the mis-sync in ms? The only digital readout on that controller seems to be just a single digit display to let you know what mode you are in. The web site shows a single digit 3 in the pictures of the device. The Lanc Shepherd has a larger LCD screen that gives the mis-sync in a fraction as well as a decimal number in miliseconds to 3 decimal places.

Well I don't have anything ready to upload yet, either, but I obtained an interesting result from shooting an anniversary clock (or "torsion pendulum clock") with my Sony TD10.

In Vegas, I introduced a one frame offset between Left and Right and rendered a half-width side-by side file. In other software (Virtualdub) I harvested only every 4th frame. This made the effective offset only 1/4 of a frame. I defined the final framerate as 23.976fps.

I wasn't expecting the result I obtained. When moving in one direction, the balls of the anniversary clock appeared to be orbiting the central spindle so closely they were almost touching it. When the balls changed direction, they appeared to orbit at an exaggerated distance, even extending outside the transparent dome!

A marked effect for only a quarter-frame time offset.

I may reshoot this and upload it.

You push the on/off buttom, and get a display information:

There is the range 0ms up to 1ms in 0.1 ms steps:
"." shows a missync below 1 ms
"3." means that the missync is between 0.2 ms and 0.3 ms
Here you have the steps ".", "1.", "2.", "3." and so on.

For the range 1ms up to 4ms you see the number without the ".", so "1", "2", "3" or "4". Also here, "4" for example means that the missync is between 3ms and 4ms

For a missync larger then 4ms the unit displays "E" (error). The controler works up to 10 ms(PAL) and 8.34 ms (NTSC) missync.

The manual states that a missync > 1/250 s = 4ms is too bad for streesocopic recording. But Werner Bloos told me that he would not accept a missync > 1ms.

Thanks, for that explanation. Seems awfully cryptic. The Lanc shepherd is so simple by comparison.

Hey Frank, Just got back from grocery shopping and you just cost me $6.99 for a bag of peanuts. Wife just had to buy a big bag after seeing your video. Hope she doesn't send me off to find a bucket to match yours. Oh well, she didn't ask for a Chevy Volt at least.

No, it is not cryptical. After short trying it becomes quite clear how to read that.

I have tried today to finalize the optical adjustment of my camera pair. Seems quite good, what I know now for sure is that the ballheads keep the adjustment very stable, even if I take the slider down from the igus base, and even if I take down the cameras from the ballheads. Had to decide if I adjust it in full zoom or in full wide angel, and came to the conclusion that it is better to do that in wide angel - since here I shoot and there is still a limited variance at the y-axis in full wide angel if I do the adjustment in full zoom. Similar to what Frank has shown in his video zooming full in.

When I did the alignment of the slide tables with the shims between the 323 and the plate, almost 8 months ago now, I finalized my system at full zoom. It holds accurate in full wide too. Still, my Vegas auto correction for alignment averages .0035 give or take for vertical correction. I was thinking I should recheck the shims again.

If you accidentally unlock the ballheads you lost all your calibrations don't you? That's why I don't like them for this application. They are designed to be readjusted and what I want is a semi permanent alignment. Less I need to do in the field.

Anyway, I'm working on my hyper stereo rig for the really big stage. Got the second tripod in on Saturday and have it all set now with a laser transit attachment. Forgot to order another 323 for the head so I can swap out the cams easily. Ordered some 50 ft extension cables for the sync controller, but I'll still be using sound sync to add the markers. since I'll be too far away to use a clap board for both cameras, I'll be using some small two way radios to send sound sync to the other camera. Haven't tested the latency on that yet. My biggest concern is the UV haze since the distances will be so great. Bryce canyon was pretty clean but the distances there were about 15 miles maximum. I suspect this new location will have subject distances of 25 to 30 miles. I also plan to shoot wider angle 16mm lens and move in closer if the haze is problematic.

Sure, it is similar to the 323 shims. If you screw them down you loose your calibration - and that is true for both the ballheads and the screws.

The optical behaviour is similar to what Frank has shown in his video (where he showed a drive from full wide-angel to full zoom). There is a small vertical variance, for all the systems that I have tested by now. If I allign the system in the full zoom, I find a vertical variance in wide angel - something of 1.5 to 2%. If I aling the system in full wideangel, I see that vertical variance in full zoom. So I had to decide where to calibrate the system - and I do that for the full wide angle since that is the range where we tend to shoot our 3D videos.

The largest correction requirement that I see from the automatics in Vegas is the rotation. That is something where I expect the shims to be superior. But I have not to repeat the calibration every day, since it stays constant. And if necessary I am able to do a new calibration very fast, since that is foreseen by the system.

Maybe important for hyperstereo: I am able both to converge the system, but also to shoot with parallel axis - since there is a separte screw that allows you to adjust the individually.

Parallel vs convergence is the only adjustment I make in the field and I designed the system that way. You are right that rotation and keystone corrections in my clips are near perfect. At full zoom, I get .0015 max on rotation and similar on keystone. It's true that most of my shooting I found to be at full wide to 3:1 zoom but I do some shots with 10:1 too. There is no variance between full zoom and full wide if I do the final parallel at full zoom. I can't recall trying convergence on a target at full wide. I think convergence on a distant target is more difficult, but then I'm not sure cameras are really parallel when I think they are either, after all the distances are so great. Full wide is very forgiving on accuracy.

The intersting point is: you can apply the automatic correction of Vegas also to a zoom drive - I apply the automatic 3 times and have a full correction. Fair enough for the cases where I go into more zoom.

For the convergence/parallel discussion: I am interesting to hear your experience with a hyper-hyper stereo base. When you have 50 or 100 m. If you would not converge here, you will loose a lot from the video left and right. Given the fact that we can correct it in the postpro in a perfect way, it is maybe less important how we shoot - even if it has an impact to the depth bracket and to the amount of horizontal correction.

I wish I would have the time to shoot more - since now I am ready with all the long preparation. One of the camera has been returned today to the dealer, I still have 3 cameras - and at the moment I think it is not an ideal solution, but a solution that will work fine.

Have you been able to make some tests Frank? Does it work for you to multiplex two streams to 3D using hdmi?

And they seem to sell to US only?
Edited by Wolfgang S. - 6/7/12 at 3:05am

Thank you for the offer. I had some short email conversation - and they would send the box also to Europe. The question is what the box is able to do really. Because here in Europe that will be even more expensive - you pay the transport and for sure the additional tax. So I would like to learn what the box is able to do first.

Help is on the way--

mini hdmi right angle

http://www.bhphotovideo.com/bnh/controller/home?O=&Ntt=Mini+Hdmi+Right+Angle&A=endecaSearch&N=0&Q=&InitialSearch=yes

http://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=mini+hdmi+right+angle

Frank, what are the connections that are offered by the stereo combiner? Can you feed in two times 2D hdmi, and is that convertetd to a 3D preview with an hdmi out?

And I am really keen to understand how the synchronisation work - is that a box that is able to force synchronisation with lanc or A/V R?

It should use frame store technology, genlock and Lanc unnecessary to sync the signals. I'm curious as to how Frank will record the combined result. Lanc may still be important as a zoom controller.

Do you have a Blackmagic design capture card for recording the output?

But for 3D you will need specific cards from Blackmagic. There is only the HD Extreme 3D or the Ultrastudio 3D, I think.