Process for making scanned text pages for the Web:

MdBR 20 May 2001 onwards. [Back to the index page...]

Of course, best is to have pdf files with the fonts embedded.

 

Otherwise:

 

Scan image as "colour image"

 

Then in Photoshop:

Crop, leaving a bit of extra space

Change Mode to Greyscale (Mode - Greyscale)

Rotate to straighten, if necessary, then crop again. (Image - Rotate - Arbitary)

Scale image to 31% [in this case] (click constrain proportions, but NOT constrain File size) (Image - Image Size - Percentage)

NOW sharpen (and MUST be BEFORE postrizing) (Filter - Sharpen - Sharpen)

Posterize to Level 4 (Image - Map - Posterize)

Edit out anything that needs it (stray lines, etc.)

To be safe, go to Image Size, and click constrain proportions AND files size and change to 72 dpi, not 300dpi.

Save as a GIF file. (Make sure no icons, previews, etc.)

The file can be made slightly smaller (just 2K or so): go to GraphicConverter application, open the file, and Save As a gif file in another name. It will be slightly smaller. (Incidentally it also automatically makes it 72 dpi)

 

The following makes larger files (double the K size) but only slightly clearer:

greyscale

crop, rotate, etc.

Posterize level 4

_Then reduce_ to 31%

Sharpen.

back to the index page...