Wah Wah, need help with stereo

[bengvc]

Jesse, I'm vying for your attention, am hungry for your input, see below, and thanks!

So, my suggested stereo template of 3.8 for each value — screen distance, image width, and nearest distance — yielded comfortable 3D throughout my first rendered animation of Bulbox. My original testing, pointing the protruding horn of Tricom at screen center, placed the tip at screen plane, hence my nearest distance parameter matching that of screen distance. But, when I applied this template to Bulbox (where the camera is never pointed at protruding objects, rather flying about inside them), I'm seeing that matching nearest distance to screen distance pushes everything back by an amount. Doesn't look bad, but ideally I'd like to have the nearest point of walls, ceiling, and floor be at screen plane, so I'll soon test by how much I need to lower the nearest distance value relative to screen distance to achieve that.

What works in Bulbox, however, doesn't exactly port to other formulas, the only example I have at present being a new scene I'm playing with based on Hal Tenny's TUT of the Tuberland formula (thank you Lenord and Hal!). One difference in my stereo settings here, in an effort to compensate for the slight rear projection of the nearest walls, ceiling, and floor, was using 3.8 for both screen distance and image width, but 3.2 for nearest distance. I rendered left and right images in 1.7.8, worked up an overlay in After Effects and spotted a zoom disparity (see Tuberland02(1.7..jpg, second pic below). I then launched 1.7.3, opened the same scene and rendered the same frames left and right, saw the same problem, but with yet greater zoom disparity (see Tuberland01(1.7.3).jpg below). So, Jesse, you indicated an adjustment to stereo in 1.7.8, which appears to have helped (at least, so it appears in this particular scene, but I'd be so appreciative of your taking another look. FYI in this scene the camera is travelling left to right, no zooming at this point.

Again, I'm still learning to crawl with M3D, VERY much loving it, my focus on the stereo side is nothing more than wanting that output from the git go, so I"m simply unsure if it's unrealistic to expect the stereo to work reliably in every 3D fractal. Thanks in advance for your help.