-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
when camera_type is equirectangular, add spherical metadata to rendered video #1151
when camera_type is equirectangular, add spherical metadata to rendered video #1151
Conversation
eed5d40
to
8848729
Compare
8848729
to
c4dade4
Compare
4715e03
to
38325b8
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for digging into the details for this functionality. It's a great addition!
Note: video render width and height were unspecified for spiral trajectory; this uses values for Nerfacto
It seemed inefficient for all rendered images to be held in memory just to encode video, so that is no longer done. As a curious side effect, I noticed that non-filename trajectories such as spiral don't have a defined render width or height. That didn't matter before when passing all the images at once to be encoded,, but it does now that we create the video writer before we have the first image (and its size). The values for Nerfacto are currently the default for spiral trajectory, but if there is a better way to know what resolution should be expected, that would be preferable. For later perhaps: it should be possible to reuse camera path files, but override the width and height for a given render command. (That might also help eliminate the ambiguity for non-filename trajectories.) |
1eb9723
to
c6487f9
Compare
c6487f9
to
963758c
Compare
(okay, finally got isort / black / pylint to agree...) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, this restructure looks good to me. Thanks for working on this.
Looks good, feel free to merge. You should now have access (check your email) |
Done thx |
…ed video (nerfstudio-project#1151) * when camera_type is equirectangular, add spherical metadata to rendered video * remove unnecessary CONSOLE.print * refactor to make pylint happier * get os.stat size properly; document camera_type argument * write video on the fly to avoid holding all images in memory Note: video render width and height were unspecified for spiral trajectory; this uses values for Nerfacto * fix images output * doc fix
…ed video (nerfstudio-project#1151) * when camera_type is equirectangular, add spherical metadata to rendered video * remove unnecessary CONSOLE.print * refactor to make pylint happier * get os.stat size properly; document camera_type argument * write video on the fly to avoid holding all images in memory Note: video render width and height were unspecified for spiral trajectory; this uses values for Nerfacto * fix images output * doc fix
Implements #1149.