Page MenuHome

Documentation: Add Tracking Video Clip Requirement
Needs RevisionPublic

Authored by Justin Jones (jjones780) on Oct 21 2019, 3:11 AM.

Details

Summary

Blender Manual paragraph to help users avoid issues with movie clips and shifting frames.

See closed ( but not solved ) issues:
https://developer.blender.org/T47720
https://developer.blender.org/T31215,
https://developer.blender.org/T35133,
https://developer.blender.org/T44675.

Diff Detail

Branch
/trunk/blender_docs
Build Status
Buildable 5432
Build 5432: arc lint + arc unit

Event Timeline

Justin Jones (jjones780) added a comment.EditedOct 21 2019, 3:40 AM

My best effort at writing some documentation.

I considered putting this in the "clip" section but feel that it needed to be up front.

It _may_ eventually be fixed in code as T47720 mentions ("TODO to build timecodes automatically") but I can't find it on any todo lists that I've been able to locate thus far.

Justin Jones (jjones780) set the repository for this revision to rBM Blender Manual.
Justin Jones (jjones780) retitled this revision from Tracking Video Clip Requirement to Documentation: Add Tracking Video Clip Requirement.
Sergey Sharybin (sergey) requested changes to this revision.Oct 21 2019, 10:23 AM

Having the mention of time codes is fine, but i am not happy with the current placement of the information and its wording:

  • Such information does not belong to an Introduction section, but more like Best Practices or so.
  • I wouldn't recommend converting video file to an image sequence this has implications down the road. Rather, build timecode.
  • I don't think use of such strong imperative statement is something what belongs to a manual.
This revision now requires changes to proceed.Oct 21 2019, 10:23 AM

"I don't think use of such strong imperative statement is something what belongs to a manual."
Thanks for being so diplomatic. In my defense I was recovering from a visit to the dentist.

I've dug into the source code and am testing a few ideas. I've gained an appreciation for the issues ( many edge cases, fallbacks required etc ). The issue in my particular case was the dts was slightly before the calculated "guess" and so it would end up _two_ frames early when seeking to a new area.
I'll let you know if I come up with any concrete techniques worthy of further investigation and will leave any documentation changes until then.