In this test, we're uploading the scanned pages separately. The reason is that the files are very high quality and thus large (each 50-100 MB). One book of 47 files weighs 3 GB.
For the sake of convenience, we should also assemble them into a djvu file of a reasonable size, for those who don't care about the highest quality.
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | Lokal_Profil | T258486 Collaboration with Musikverket – Free music on Wikipedia | |||
| Resolved | Alicia_Fagerving_WMSE | T254466 Work with sample material about old manuscripts from Musikverket | |||
| Resolved | Alicia_Fagerving_WMSE | T254553 Upload sample scans to Commons | |||
| Resolved | Alicia_Fagerving_WMSE | T259244 Upload example manuscript as single djvu file |
Since we want to save color, we can go with
#!/bin/bash for file in *.tif; do outfile=`basename $file .tif`.jpg convert "$file" "$outfile"; done; for x in *.jpg; do c44 -dpi 400 $x; done djvm -c ./result.djvu *.djvu
This gives a reasonably nice quality file. From 3GB of scans we get a 39 MB single file: https://commons.wikimedia.org/wiki/File:Johan_Helmich_Roman_14d_Sinfonia_F_dur_BeRI_10_-_SMV_-_full.djvu
I sorted it in the category with |* so that it's first when you open it.