From: blockedofcourse@foo.invalid   
      
   On 12/8/2025 9:10 AM, legg wrote:   
   >> Other thing is OCR: it seem to work quite well with actual   
   >> text (as opposed to mixture containing also figures and formula)   
   >> when scan quality is high enough.  OCR is valuable if only to do   
   >> searches.  Scanning at low/medium quality throws out information   
   >> that is hard or impossible to restore.   
   >   
   > There's OCR, and there's OCR.   
   >   
   > I was having the devil of a time getting pdf scans with searcheable   
   >   text. Blaming scanner and source - till I upgraded the scanner   
   > software. Voila !   
      
   I have a reasonably fast (30 double-sided pages per minute in mono)   
   scanner that I use to scan all my insurance, bank, investment, bills,   
   etc. statements.  Drop a stack of pages in the "feeder", press one   
   of three buttons to determine which presets to apply, then fetch   
   the OCRed PDFs from the server that I've set up to "catch" them.   
      
   The hardest part is then appending these to the "annual" PDFs   
   that I maintain (as I don't want lots of silly little statements   
   to keep track of but, rather, one for each account-year.)   
      
   [Acrobat makes this incredibly easy; I can just select the files and   
   say "combine in Acrobat" from the right-click menu -- even if they   
   aren't PDFs!]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)