Repository navigation
Scale PGM and PPM samples to the full range before converting them to the pixel type - #3227
Merged
Merged
Conversation
… the pixel type A non-standard maxval was applied with a brightness pass after the samples had been converted to the destination pixel type. A 16-bit sample with a small maxval was narrowed to 8 bits first, so decoding to an 8-bit pixel type such as the default Rgb24 kept only a few distinct levels (maxval 1000 gave five). The plain and binary decoders now scale every sample from [0, maxval] to the full 8-bit or 16-bit range with rounded integer arithmetic before the pixel conversion, and the extra brightness pass over the image is removed.
4 tasks done
4 tasks done
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Prerequisites
Description
PGM and PPM samples were packed into the destination pixel format before a separate brightness pass applied Maxval. Narrow destinations lost precision before that pass could normalize the values. The initial fix also rounded low-Maxval byte samples to bytes before converting them to higher-precision pixels.
Non-standard ranges now normalize into a reusable float row before the existing bulk pixel-packing API stores the result. Integer rounding occurs at the destination precision. Full-range inputs retain the existing byte and ushort packing paths. The separate brightness pass is removed. Traversal remains scalar; SIMD work is deferred to integration into the ICC branch.
Sample saturation is independent of
SegmentIntegrityHandling. Early EOF returns partial output. Available channels in an incomplete pixel are retained where possible, and missing bytes are zero-filled rather than read from pooled memory. Decimal sample overflow saturates instead of wrapping. Overflowing header fields remain fatal before image allocation.Binary header parsing also stops at the raster boundary. The PGM specification and PPM specification require one separator before binary samples. Whitespace-valued raster bytes and
#must remain sample data.Tests load stored inputs. No headers or rasters are generated at runtime. Cases cover P2/P3/P5/P6 at Maxval 100 and 1000, exact byte and ushort output, saturation in every integrity mode, partial rows and pixels, decimal overflow, and raster bytes that resemble header whitespace. Existing reference images and tolerances are unchanged; the new precision cases use exact assertions.
All 140 PBM tests pass in Release on .NET 10 and .NET 11, with no failures or skips.
Fixes #3239
V4 backport: #3260.