ReConverge

If you’re working on a stereoscopic or multi-view project, the ReConverge node lets you shift convergence (the inward rotation of the eyes or cameras) so that any selected point in the image appears at screen depth when viewed with 3D glasses. This point is called the convergence point. It is the point where the lines of sight from the two cameras meet.

At the convergence point, the different views in the image are aligned and appear at screen depth when viewed with 3D glasses. Anything behind the convergence point appears behind the screen, while anything in front of it seems to pop out of the screen.

To calculate the convergence shift, the ReConverge node needs a disparity field that maps the location of a pixel in one view to the location of its corresponding pixel in the other view. To create the disparity field, you can use O_DisparityGenerator plug-in, which is part of the Ocula plug-in set. Alternatively, you can create the disparity field in a 3D application. Once you have the disparity field, you can store it in the channels of an .exr file, or use the Shuffle node to add the disparity channels in the data stream where you need them.

It is also possible to use the same element as the convergence point throughout the image sequence. To do so, link the ReConverge node with a Tracker node.

To better view the effect of the ReConverge node, you can insert an Anaglyph node between the ReConverge node and the Viewer.

Note: The ReConverge node only shifts views horizontally, not vertically.

Inputs and Controls

Connection Type	Connection Name	Function
Input	unnamed	The image whose convergence you want to adjust. This input should contain at least two views (one for the left eye and another for the right) and a disparity field: • If your input is an .exr file, the file may already contain both views. This is because .exr files support multiple views in a single file. With any other file types, you need to have separate files for the left and right views. You can combine these files into a single input using a JoinViews node. • Make sure there is a disparity field upstream from the image sequence. If the image sequence is an .exr file, the disparity field can be included in its channels. Otherwise, you can use a Shuffle node or Ocula’s O_DisparityGenerator plug-in to add it in the data stream.

Connection Type

Connection Name

Function

Input

unnamed

The image whose convergence you want to adjust. This input should contain at least two views (one for the left eye and another for the right) and a disparity field:

• If your input is an .exr file, the file may already contain both views. This is because .exr files support multiple views in a single file. With any other file types, you need to have separate files for the left and right views. You can combine these files into a single input using a JoinViews node.

• Make sure there is a disparity field upstream from the image sequence. If the image sequence is an .exr file, the disparity field can be included in its channels. Otherwise, you can use a Shuffle node or Ocula’s O_DisparityGenerator plug-in to add it in the data stream.

Control (UI)	Knob (Scripting)	Default Value	Function
ReConverge Tab
Converge upon xy	convergencepoint	0, 0	The x and y coordinates of the convergence point (the point you want to appear at screen level when viewed with 3D glasses). You can also move the point by dragging the convergence point overlay in the Viewer.
Convergence offset	offset	0	Adjusts the offset for convergence (in pixels). To bring all elements of your image forward from the screen level, enter a positive value. To move all elements further away, enter a negative value.
Mode	convergemode	shift right	The view to move to achieve the convergence shift: • shift right - move the right view. • shift left - move the left view. • shift both - move both views.

Step-by-Step Guides

Changing Convergence