Here is an interesting Article on the Production of this 3D Film:

Wim Wenders' Pina's 3D Finish | Studio Daily (http://www.studiodaily.com/main/technique/tprojects/Wim-Wenders-Pinas-3D-Finish_13236.html)

It's a beautiful film I really enjoyed it, looking to see it again when it next comes close by

I saw most of it at the Toronto Int'l Stereoscopic Conference last month and Wim talked about it for a couple of hours. He said that he had been trying to think of the proper way to photograph Pina's choreography since the early 90's but it wasn't until he saw U23D at Cannes that it occurred to him 3D might be the solution he was looking for.

The chosen camera positions and steadicam work were great, the editing was excellent and it was definitely the best 3D dance film I've ever seen. However, my only criticism is that they used very wide interaxials and had a pronounced dwarfism effect throughout. At least it was consistent.

Since the film's DP has stated that he used Methode Derode throughout the filming, what percentage of BG divergence would you guess he set the rig at?

I haven't been lucky enough to see the film yet, but I would guess he got out into the 3% - 4% range judging from your description. Just trying to work the thing backwards into logic. .

I would positive parallax was typically 2% and often more in the scenes Wim showed us. I was sitting way back in the theatre and it was only a 30 foot screen so my angle of divergence per eye was low but I still felt it. They didn't utilize negative parallax very much at all, but when they did they would tend to brake the stereo window quite often.