Scientific figures PNG→TIFF: reviewers and repositories want reproducible parameters, not vague “lossless” claims
`png-scientific-tiff` ingests collaborator PNG figures into pipelines or repositories that mandate TIFF. Conversion does not recover quantization already discarded; Western blots, microscopy, and volumetric slices need explicit statements on whether densities may be measured. Emit converter versions, parameter summaries, and checksum pairs tied to lab notebook IDs to survive reproducibility audits years later.
Research workflow: document bit depth and color space, bind experiment IDs, then validate scale bars and annotations in both ImageJ and final layout exports
- Log whether each figure is illustrative or quantitative, whether upstream compression occurred, and whether faces or geo-sensitive pixels appear.
- Disable display auto-enhancements; capture LUT names and percentile clamps in sidecar JSON; embed scale bar lengths tied to methods text.
- Upload to institutional buckets with script hashes; segregate restricted TIFFs from public PNG derivatives and wire access logs to IRB paperwork.
PNG to TIFF (publication & data ethics) FAQ: pseudocolor, scale bars, densitometry traps, privacy metadata, and build pinning
Journals demand TIFF but the asset is a PowerPoint PNG—what compliance risks appear?
Disclose upstream processing in supplements; never imply TIFF equals raw acquisition; provide higher-fidelity downloads when ethics allow.
Does integrating band plots on 8-bit PNG after TIFF conversion improve densitometry?
No new grayscale levels appear; perform quantification at the native bit depth; TIFF mainly helps metadata and archival policy compliance.
Map screenshots may leak coordinates via EXIF even after pixel mosaicing—true?
Strip GPS and serial tags before release; audit both pixels and TIFF fields; assume attackers read metadata even when imagery looks redacted.
Reproducibility officers want byte-identical outputs, but TIFF bytes differ by OS—what do we report?
Pin libtiff/libpng builds and document nondeterminism; if residuals exist, show they sit below measurement uncertainty with quantitative reasoning.
A collaborator hid steganographic watermarks inside PNGs—could we be liable after converting to TIFF?
Require written licensing and watermark policies up front; scan for hidden payloads; never bulk-publish unchecked third-party art.