| Summary: | Why are some metadata updated "automatically" ? | ||
|---|---|---|---|
| Product: | [Applications] digikam | Reporter: | Caro <caro.portal> |
| Component: | Maintenance-Metadata | Assignee: | Digikam Developers <digikam-bugs-null> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | normal | CC: | caulier.gilles, metzpinguin |
| Priority: | NOR | ||
| Version First Reported In: | 8.7.0 | ||
| Target Milestone: | --- | ||
| Platform: | Microsoft Windows | ||
| OS: | Microsoft Windows | ||
| Latest Commit: | Version Fixed/Implemented In: | ||
| Sentry Crash Report: | |||
| Attachments: | Configuration | ||
|
Description
Caro
2025-09-08 16:25:34 UTC
I can't reproduce it with a quick test here, using a JPG and HEIF file. What are your exact metadata settings regarding reading and writing with ExifTool? Please note that we're still working with Exiv2 and ExifTool uses a so-called EXV container for writing, so more metadata may be written than if you just add a keyword. Reading with Exiv2 may result in a different interpretation/rounding of the values, which may then lead to changed Exif values when writing back. Maik Created attachment 184822 [details]
Configuration
Hi Maik, I have added a screenshot of my configuration and I have Exiftool 13.04 installed. Here is the link to the picture I'm using for tests: https://we.tl/t-JJLqzkWryv Thanks for your help ! Nice day, C Hi Maik, Did you get the shared testing file from comment #3 ? File is not online anymore... Gilles Hi Gilles, The problem can sometimes be reproduced with any file, depending on certain metadata in the Rational format. Exiv2 rounds fractions differently than Exiftool. Since we write all the Exif metadata back using the EXV container, slight changes occur at the rounding level. I don't see it as a critical issue. The only solution would be to write the metadata completely natively using Exiftool, writing only the changes that were actually made. I already have some ideas on how we could add this step by step for specific metadata. Maik *** This bug has been marked as a duplicate of bug 481755 *** |