In its current version (v0.11) the fauxD data window will remain blank (or unchanged) until
at least two consecutive frames are processed. Seeking (with the sliders) or cutting directly to a frame is interpreted as a discontinuity by fauxD.
(Thanks for pointing that out, as I had never really thought much about it until now. It'll be fixed in the next version.) Presently there is no real work-around for this (even if you create and use a logfile for the video in question). However, if you frame-step with the right arrow (or play the video) then fauxD should perform as expected.
Incidentally, I primarily use AvsP as an editor. Until now I hadn't often used the built-in video player because it keeps the DLL locked - a nuisance when I'm trying to update it. I normally test my scripts with MPC (by double-clicking the AVS, or dragging and dropping the AVS onto the player window, or through a batch CMD on my SendTo menu).
Also, the
YDifferenceFromPrevious scene-change test can completely ignore fauxD's stereo output, resulting in a flat image even if you see data updating in the dialog. (Fusion across scenes is a real eye-sore.) Since AviSynth doesn't like 'looking-ahead' in the video stream my only safe choice is to use the flat image for that frame.
No 3D will be produced until
both conditions are satisfied.
I am using
MVToolsV2 version
2.21. The latest version at the time of this writing (
2.31) works as well, but can cause a hang-on-exit with some apps on my system.
(Fizick, let me know if you want these direct links removed.)
I'll add an option to keep the dialog 'always-on-top' if it works as intended. Although the dialog is more of a debugging tool than a genuine necessity considering only the X values are reasonably close to accurate (because the analysis clip was vertically reduced).
I still have no explanation for your Mvtools~/Mtools~ issue. I don't know how to help you with that.
Right now I'm trying to implement some kind of 'look-ahead' to maintain 3D at scene changes and preserve the effect for that additional 1/24 or 1/30 of a second. If I can get it working then I'll use it to solve the flat-on-start (first frame) issue you also seem to be describing. I've just finished restructuring and consolidating the script files as well. Now everything is in the form of a
function and all of it is contained in a single file for easy import. I'll post the new version once I've finished updating the doc and succeed with (or temporarily give-up on) this look-ahead thing.
After that I'll nuke the attachment in my previous post.
-eslave
P.S.: Please stop quoting my posts in their entirety.