Page MenuHome

Fix T101211: Image jitters when scaling by large values
ClosedPublic

Authored by Richard Antalik (ISS) on Nov 15 2022, 7:27 PM.

Details

Summary

Issue was caused by imprecise math due to using float numbers.
Use double instead.

No negative performance impact was observed.

Diff Detail

Repository
rB Blender

Event Timeline

Richard Antalik (ISS) requested review of this revision.Nov 15 2022, 7:27 PM
Richard Antalik (ISS) created this revision.
Jesse Yurkovich (deadpin) added inline comments.
source/blender/imbuf/intern/transform.cc
473–474

Seems these lines were left in by mistake?

Richard Antalik (ISS) marked an inline comment as done.

Remove unintentional changes

Modern processors use double math internally. So this patch has minimum overhead. Fine by me. Would just make the return values consistent.

source/blender/imbuf/intern/transform.cc
211

would also make the return type double so it matches the change in the arguments.
Same or modify_v

Jeroen Bakker (jbakker) requested changes to this revision.Nov 18 2022, 9:17 AM
This revision now requires changes to proceed.Nov 18 2022, 9:17 AM

Use double type for UV wrappers

Richard Antalik (ISS) marked an inline comment as done.Nov 18 2022, 5:40 PM
This revision is now accepted and ready to land.Nov 23 2022, 2:41 PM