Printable Version of Topic

Posted by: JimL Feb 27 2009, 12:39 PM

How can I prevent a B&W text capture (prep for OCR) from being instantly converted to deep color?

Thanks

Posted by: Peter Mar 3 2009, 02:19 PM

The screen capture function always uses the bit depth of the display. If the captured data does not contain more than 2 colors, you can simply do a Color->Convert to->Indexed 256 (or Black & White) and you should end up with a monocrome image.

Thanks,
Peter

Posted by: JimL Mar 3 2009, 04:25 PM

The output needs to be black and white, but I get this gaining or losing of edges when I convert out of deep color. Should that vary with the format? So that BPM might work better than JPG?

Posted by: Peter Mar 5 2009, 11:29 AM

JPEG cannot store B&W images. JPEGs are either 8-bit (grayscale) or 24 bits and is a poor choice for storing B&W material.

GIF, BMP, and PNG can store B&W. One of the most effective way to store B&W is to use TIFF with CCITT FAX4 encoding. A page that is a megabyte saved as BMP will shrink to a few kilobytes when saved as TIFF CCITT FAX4. (PMView can do this!)

Thanks,
Peter