Doom9's Forum - View Single Post

ai4spam · 29th June 2005, 05:39

By user request, here is my Char Matrix.
http://sr1.mytempdir.com/132702
It has some 8000 characters in it, and works for some 90% of the DVDs. Beware, although I did my best to type in things correctly, there may be occasional times when a character is detected incorrectly, so make sure to set the OCR sensitivity very high, even at 1000. There are DVDs with which it doesn't work, in which case you can either add your own characters, or make a new matrix file. If some subs are really "esoteric" (as in, a font I've never seen), I prefer to make a new matrix for them. If only a few characters are not recognized (such as diacritics and the like), I just add them to this big file as they show up.
This file will help you with the "best guess" feature too: most of the time a character is detected correctly, but SubRip is not 100% sure. Just press the "Use" button if the character is the one you want, or Ctrl-Enter to copy it in the edit box if the style needs changing (such as, if it's italic in the matrix, but not italic in the sub), then Enter to confirm.
There are only a few "compound characters" (like "%") because I've cleaned this matrix up recently, after I implemented extended selections. You can use the "<<" and ">>" buttons to shrink/expand the selection, or Ctrl-Left arrow and Ctrl-Right arrow.

29th June 2005, 05:39	#149 \| Link
ai4spam Programmer Join Date: Sep 2003 Posts: 382	By user request, here is my Char Matrix. http://sr1.mytempdir.com/132702 It has some 8000 characters in it, and works for some 90% of the DVDs. Beware, although I did my best to type in things correctly, there may be occasional times when a character is detected incorrectly, so make sure to set the OCR sensitivity very high, even at 1000. There are DVDs with which it doesn't work, in which case you can either add your own characters, or make a new matrix file. If some subs are really "esoteric" (as in, a font I've never seen), I prefer to make a new matrix for them. If only a few characters are not recognized (such as diacritics and the like), I just add them to this big file as they show up. This file will help you with the "best guess" feature too: most of the time a character is detected correctly, but SubRip is not 100% sure. Just press the "Use" button if the character is the one you want, or Ctrl-Enter to copy it in the edit box if the style needs changing (such as, if it's italic in the matrix, but not italic in the sub), then Enter to confirm. There are only a few "compound characters" (like "%") because I've cleaned this matrix up recently, after I implemented extended selections. You can use the "<<" and ">>" buttons to shrink/expand the selection, or Ctrl-Left arrow and Ctrl-Right arrow. Last edited by ai4spam; 26th August 2005 at 12:22.