System：
- Mac mini M4 24G RAM macOS 26
- MacBook Air M2 16G RAM macOS 26
- DEVONthink 4.1.1，OCR Setup
- PDF：103M，550 Pages
Problems with OCR in Chinese.
-
Problems with a single large file (500 pages +).
- There is almost no way to OCR. The layout is chaotic.
- The display is abnormal. The default is double-page display. After OCR, it becomes a single page, and every time you jump back to the first page.
-
The problem of splitting it into medium files (270 pages).
- The layout has improved, but it still can’t be used.
-
Accidental discovery: it may be caused by the addition of Link (DEVONthink’s item link) added to the source file. In addition, new problems: deleting the link will be stuck for a long time - actually directly OCR - the effect is quite good.
- The OCR of source files without DEVONthink’s item link is fine
PS：Adding links to large scanned files also takes a long time (possibly more than 30 minutes). After testing, it was found that when adding DEVONthink’s item links to scanned files, OCR (which has a different effect from self-conducted OCR) is actually performed first before adding the links?