Comments on: Shooting FIFA 2010 in 3D… Lessons from Day 1

By: 3D Focused

3D Focused — Fri, 29 Oct 2010 18:06:09 +0000

[...] This site had a wonderful review of the first day of proceedings, as well as information about the preparations, finishing off with some recommendations. Yes, I realize I’m 4 ½ months late on this. Still great info! [...]

By: Anonymous

Anonymous — Wed, 29 Sep 2010 08:31:00 +0000

Boris, your one of the professionals who’s work I have followed in S3D photography in the past. Thanks for the insight into the lens calibrating on zooms and especially in the time critical live event arena.

3Ality rigs are engineering marvels. What I particularly like about them is the foresight into meta-data preservation.
I think they use modified Preston controllers I was told for I/A and convergence.

Regards
Clyde

By: Boris Starosta

Boris Starosta — Tue, 28 Sep 2010 22:39:00 +0000

Clyde,

That’s a fine video; very generous of Sony to reveal the technological side. I worked on a live TV project with 3ality in 2007, where we (convergence pullers) used a very similar controller, with slider for interaxial and knob for convergence control.

Sony video guy does not mention how or if the zoom lenses are calibrated. The practical fact is that the lenses have a certain amount of “slack” or “lag” – this is the mechanical slack produced by microscopic gaps between teeth of worm gears, etc. THink of the short delay when you switch from zooming in to zooming out – the gears have to move through this slack before the lens reverses motion. It is unavoidable from a mechanical engineering standpoint. Anyway, the lag is different for all lenses, so even if the focal lengths are theoretically matched, if you do some zooming in and some zooming out, back and forth, etc. pretty soon the lenses will be at slightly mismatched settings. The operators ideally should rack the lenses all the way in or out, each time they want to change direction, to keep the focal lengths matched – how practical is that, covering fast moving action?

So that’s where the mismatches come from and the Sony black box can’t correct that, at least not with the procedure as described in the video – that procedure only calibrates for a theoretically mechanically perfect (although mismatched) set of lenses.

Cheers,

Boris

PS
thanks for your wonderful blog!

By: Rego James

Rego James — Wed, 28 Jul 2010 09:58:04 +0000

It was not the world's first in Sport Events. The first in Sport Events was NFL in US shot in 3D and earlier this year April Indian Premier League 4 cricket matches were covered and broadcast in 3D

By: Rego James

Rego James — Wed, 28 Jul 2010 09:58:00 +0000

It was not the world’s first in Sport Events. The first in Sport Events was NFL in US shot in 3D and earlier this year April Indian Premier League 4 cricket matches were covered and broadcast in 3D

By: Paul55

Paul55 — Wed, 07 Jul 2010 21:12:57 +0000

I just watched the World Cup semifinal between Spain and Germany at my local theatre in 3D (using a chromatic system, not polarizing filters). The picture was of poor quality due to the large amount of crosstalk between the lens, which meant that in long shots of the pitch from the side (i.e., most of the game) each player had a very noticeable shadow. Even the lines (especially the halfway line) had noticeable shadows. This problem must be resolved before 3D will be accepted by most viewers.

By: clydeD

clydeD — Fri, 25 Jun 2010 16:11:36 +0000

Sony released a video on their 3D zoom calibration, convergence and other 3D live controls. It may be interesting to view
http://www.youtube.com/watch?v=6ud04UVFTPE&feat...

By: stereofinland

stereofinland — Tue, 22 Jun 2010 07:58:07 +0000

Thanks, very nice article!

By: clydeD

clydeD — Mon, 21 Jun 2010 19:49:55 +0000

Boris,
Thank you for explaining this. It's not something I considered at all. I assumed that these rigs would have matched lens sets (maybe not 100%) but didn't think they would show such variation.

All the more reason I am all for tools that do the job, e.g an Si2k 3D rig. (won't mention other working rigs as I read in another article that this is exactly what was not wanted) as here: http://bit.ly/bFzvbv

Agreed live stereo has some way to go, and cutting some slack to the pioneers is in order. (provided modesty prevails)

By: Boris Starosta

Boris Starosta — Mon, 21 Jun 2010 19:19:04 +0000

Vertical disparity and depth gradient or skew across entire image, if not the result of 2D>3D processing, is most readily explained by mismatched focal lengths in the image aquisition. The resulting mismatched image scaling, and hence disparity scaling, results in the perceived depth gradient.
If these cameras use zoom lenses that are not perfectly synchronized, a mismatch in focal lengths is plausible for some shots.

For this reason and many others, live stereographic image aquisition is most difficult. The situation in broadcast and cinema business may or may not favor the early producers... remains to be seen. Regardless of stereoscopic "talent" available to the production crew, technological and other issues are likely to be the limiting factors at FIFA.

CHeers.