Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Partially fix score difference for CUDA (bpc >=10) #1217

Merged
merged 2 commits into from
Aug 3, 2023
Merged

Partially fix score difference for CUDA (bpc >=10) #1217

merged 2 commits into from
Aug 3, 2023

Conversation

gedoensmax
Copy link
Contributor

Superseding #1195

Text from 1195:

Greatly lowers score difference on CUDA for high bit images.

VMAF score for 1000 frames of a 4096x2160 10bpc sequence in FFMPEG:
with libvmaf_cuda (GPU):
[Parsed_libvmaf_cuda_0 @ 0x559ffac22b00] VMAF score: 76.101209

with libvmaf (CPU):
[Parsed_libvmaf_2 @ 0x55a7e344cf80] VMAF score: 76.101269

Score difference is now dependet on the sequence used.

JamesHutch and others added 2 commits August 2, 2023 19:19
@JamesHutch @gedoensmax
fix integer_adm_cuda for bit depth > 8
6fb8d5c 
Shift in scale 0 used to assume bit depth of 16, this is now variable.
A typo in the call to adm_dwt2_16_device caused i4_ref_dwt2 not to be
written.
@gedoensmax
code simplification and reducing warnings
9cb796f
@kylophone kylophone merged commit 35aeb1b into Netflix:master Aug 3, 2023
2 of 8 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@gedoensmax @kylophone @JamesHutch