“Or else a scan-to-PDF converter which attempts to use a higher-res bitonal image in places where it detects that the content seems to be black text, and a lower-res color image in places where it detects color content. “
That’s bogus... The resolution that the scanner scans is between the scanner and the scanner driver... It has nothing to do with the application. The driver absolutely can’t change resolution in the middle of a scan.Were talking about a TWAIN driver. Where did you hear that crap?
PDF format is the de facto standard for companies to release electronic representations of paper documents. PDF files can contain an arbitrary mixture of bitonal, color, and grayscale bitmaps, along with an arbitrary mixture of text and vector graphics. I don't know to what extent automated tools will take a high-resolution bitmap and decompose it into different parts, but I do know that if utilities don't already exist that would turn a scan of a document that looks like the Obama COLB into a PDF similar to the one that was released, they should. In many cases, storing and distributing 1200dpi uncompressed scans of mostly-text documents would be overkill; in cases where file size is important, analyzing a source image and decomposing the parts that seem to be bitonal, grayscale, and color would greatly improve the efficiency by which a reasonable facsimile of a document could be stored and transmitted.