Tools › WebP conversion
Re-encode the pages of comics already in your library as WebP to save space. New downloads are handled by the option in the post-processing settings.
This rewrites files in your library, and the re-encode is lossy.
- ComicInfo.xml and every non-page file in each archive are kept byte for byte, and no page is ever dropped. A page that can't be read, is CMYK, is a low-quality JPEG, or wouldn't get smaller stays exactly as it is.
- Each new file is checked (every page decoded at the same size, page order unchanged, all other entries identical) before it replaces the original. A file that fails a check is left alone.
- Keep the originals in a holding folder until you've looked through some converted issues. Only .cbz files are converted.
- Pause Kavita's folder watching while this runs, and scan the library once at the end.
Each file a run looks at is remembered, and isn't opened again until it changes. The largest files go first.
Looks again at the files that had nothing to gain, for example after lowering the WebP quality.
A little each night
Converts part of the library each night, largest files first, with the date, quality, CPU threads, files at once and holding folder chosen above. New downloads are already converted when they're filed, if that's turned on in the settings. It waits while downloads are running or being filed, and doesn't run at the same time as a conversion started here.
Between
and
Server time. The times can cross midnight, like 23:00 to 05:00.