Analysis of Window Position and Depth in a Stereo Card

Analysis of Window Position and Depth in a Stereo Card

By David Lee

It is a simple matter to analyze whether the window placement on a stereo card is ideal or whether it is "reversed." (If you read this article without following the instructions, it will seem more complicated than it is.) In addition, the analysis of depth takes only a little more effort.

Start by taking a sheet of letter size paper (a 5 x 7 file card also works well).

1) Place the top edge of the paper at the bottom edge of the stereo image (the stereo card should preferably have a plastic sleeve to protect it).

2) Using a pencil, make a light mark on the paper at the left edge of the left chip. Label this A.

3) Make another mark at the left edge of the right chip. Label this B.

4) View the image in a viewer and make a mental note of where the near and far points are located.

5) Place the edge of the paper so that it runs through the near points with the A under the near point on the left chip.

6) Make a mark on the paper at the near point on the right chip. Label the mark C. (It should be at, or very near, B.)

7) Place the edge of the paper so that it runs through the far (also known as infinity) points with the A under the far point on the left chip.

8) Make a mark on the paper at the far point on the right chip. Label the mark D.

Analysis of the window position

Let us first assume that you do not want any part of the scene coming through the window. (I will consider the case of something coming through the window later.) Point C (the near point in the right chip) should be right on, or slightly to the right of, B. If C is to the left of B, then you have a window reversal. If C is more than slightly to the right of B, then the window is farther in front of the scene than necessary and you will end up having more disparity on the edges (also known as "floating edges") than necessary.

If the card has already been mounted this will be the end of the analysis, but if you haven’t adhered the chips yet, this analysis will tell you what to do next.

If B and C are together, you have a perfect window and don't need to make any changes.

If you have a window reversal (C is to the left of B), then measure the distance from C to B, and trim that amount from each of the outside edges (erring on the side of trimming too much). If you are using a Q-Vue mask then you do not need to trim, but you must get a narrower mask.

If C is more than slightly to the right of B and you want to place the window closer to the scene, then measure the distance from B to C, and trim that amount from the inside edges (erring on the side of trimming too little).

If you have something that you want to appear through the window, simply decide where you want the window to intersect the scene and make that point C (instead of making the near point C). Do everything else the same.

The distance between the two chips will affect the viewability of the image, but will have no effect on the window position.

Also, be sure that you have located the actual near point. In some complex scenes there are weeds or other objects that are closer than is at first obvious. This throws off the window position.

Analysis of depth

The analysis of depth involves the amount of deviation from the near point to the far point. Simply measure the distance from C to D. Ideally this distance should be about 1/8 inch. If it is less than this the scene will probably appear to be flatter than it optimally could. If it is a little greater than this, the scene will start to appear less smooth, and as it continues to increase it will cause eyestrain in the person viewing it.

The other useful information one can obtain from these marks is the distance between the far points. As this dimension exceeds about 33/8 inches (depending on the viewer) the image will become progressively harder to view without eyestrain. I try to make the infinity separation about 3¼ inches.

Back to blog

Leave a comment

Please note, comments need to be approved before they are published.