Stereo 3D Rig: Compositing with S3D-Footage & C4D-Renders

Question

Hello @ all!
I do a research for stereoscopic workflows with the new Stereo 3D Rig in Adobe After Effects CS5.5.
I really searched the web and think there is no resource about stereoscopic workflows anywhere that I've not seen.
I even bought the Video2Brain-Tutorial hosted by Angie Taylor.
But though there are few questions remaining!
1. How can I achieve a horizontal image translation (HIT) with a parallel camera setup (without converging) to adjust my parallax?
background information: I try to compose some text-layers with my real-footage shot with a parallel setup. I know how to set the footage-parallax indepently with expressions - this is not the problem!
I don't want the Stereo 3D Rig to converge when adjusting the parallax for my compositing elements (the text-layers).
I tried it with a mix of using the stereo scene depth parameter and the scene convergence parameter, also tried the expression which avoids moving the convergence plane, but i can't manage to get the result that i want. This is because i have to decrease the stereo scene depth to about 0,1% which is too less, but otherwise the deviation of left and right view is too big. The coordinates of the text-layers are ok.
Of course i could just use the Converge to option (as everybody does this in all the stereoscopic workflow resources that are available in the web), but this creates strong keystoning effects and in a way works against the principles of stereoscopy.
Also this is not recommended by the Adobe Help Sites:
You can also use parallel virtual cameras. This technique is useful if you need to match live footage and add digital elements to that scene. Keeping the virtual camera orientations consistent with the cameras used in the footage helps to keep the perspectives of the digital elements and that of the stereo footage aligned.
[…]Real scenes are almost always shot with parallel cameras. Keep this in mind if you are trying to mix and match live footage with digital elements. If your scene consists of only 3D elements in After Effects, then it is probably safe and preferable to use converged cameras.
[Source]
Understanding stereoscopic 3D in After Effects
http://kb2.adobe.com/cps/898/cpsid_89868.html
2. What about the Stereo 3D Rig for compositing purposes with footage rendered out of Cinema 4D?
I try to figure out a stereoscopic workflow in After Effects allowing me to make use of the composition-project-file that i exported out of Cinema 4D.
After importing the file i have my left and right camera and some Nullobjects as placeholders which are supposed to be replaced by some other 3D-elements in After Effects (e.g. text-layer).
Actually, the cameras have already the right data and could be used, but the of course creates left and right Eye View automatically based on the Master Cam.
Is there any idea for using the Stereo 3D Rig in a stereoscopic workflow to make a compositing with data exported out of Cinema 4D?
Or is it better to create, in this case, a structure of compositions which only uses the cameras exported from Cinema 4D?
Any help is appreciated!
Best,
Paul

Answer 1

Hi Paul,
1.
Sounds like you're hitting a limitation of 3D glasses effect. The HIT (or scene convergence in 3D glasses) works on the pixel level and does not calculate sub pixel.
Here are some things you can try.
Use this expressionin in 3D glasses scene convergence to cancel out your stereo scene depth more robustly and allow you to either change your scene convergence in pixels, or in % of comp width.
if ( effect("ADBE Stereo 3D Controls")("ADBE Stereo 3D Controls-0006") == 0 ) {
    offset = effect("ADBE Stereo 3D Controls")("ADBE Stereo 3D Controls-0003");
    if ( effect("ADBE 3D Glasses2")(5) == 1 ){
        offset *= thisComp.width / 200;
    value - offset;
} else {
    value;
Now you can switch your 3D glasses units to pixels and change your convergence by 1 pixel at a time, instead of 0.1% of comp width.
If that is not enough control another option if you need sub pixel accuracy; you will need to recreate what 3D glasses scene convergence is doing manually using precomps. Try Precomping your left and right eye comps and then then hook up an expression to those comps x position, left comp moves right the right moves left. Put that slider right next to 3D glasses, and now you should be able to adjust the convergence by sub pixels. Another benefit of this method is it doesn't crop, you will get only one image on the edges though.
(This is a bug in 3D glasses, it should really operate with sub pixel values but we ran out of time to do this feature).
2.
I'm not as familiar with the C4D workflow. Can you import their stereo cameras directly in AE? If so, you can set up your Master cam to follow one of their cameras and set the mode to be Left or Right hero. Then match their convergence settings. You could try ploppling their cameras into the left and right eye comps above the AE camera so it uses theirs as the active camera if you don't want to deal with matching settings, but then you can't make quick changes to the camera without changing both cameras exactly.
If you're exporting left and right footage, you can follow the guidance on the document you mentioned.
Let me know if this helps out or you have further questions.
Amir
After Effects Engineering (the guy that made the stereo rig and who is responsible for your pain)