Fix black lines in dynamic render - #247
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Dynamic rendering often shows one row of black pixels at tile boundaries, as outlined in #228. The main issue is that
ImageProcessor.getPixelInterpolatedhas a hard cutoff at[0, width-1) x [0, height-1). Whenever a pixel is outside that interval, it's set to 0. Due to floating point errors and how pixels are treated in render, often the rightmost column / bottom row would be mapped to the interval[width-1, width)/[height-1, height), respectively, triggering the cutoff.This PR fixes the issue by clamping the
x, yvalues to the valid range. It also updates the tests to the new mapping, and mostly reverts changes introduced in an earlier attempt at fixing this (#223, which fixed the floating point issues, but not the larger gap due to finite pixel size). The performance cost of having the additional clamp was benchmarked to be about 5%.Below are dynamic renderings from the same region before (left) and after (right) applying the fix. @trautmane, let me know whether you think this is sufficient. If so, this PR also closes #228.
FYI @StephanPreibisch