[khin99]

Originally Posted by willus

@Khin99 -- Thank you for the attachment. Because the scan is poor quality and the position of the pages changes a fair amount from page to page, a brute force cropping or gridding is best, I think. If you don't care about chopping up the title page, you can just do this:

k2pdfopt -grid 2x1x0 -mode fitpage source.pdf

-grid 2x1x0 breaks each page into 2 x 1 boxes.
-mode fitpage puts each of the gridded regions onto a single page

In the MS Windows GUI, you can just select "fitpage" for the conversion mode and put -grid 2x1x0 in the "Additional Options" box.

If you want to preserve the title page, it's a little trickier, but not too bad:

k2pdfopt -cbox1 0,0 -cbox2- 0,0,.5s,1s -cbox2- 0.5s,0s -mode fitpage source.pdf

-cbox1 0,0 creates a boxed region on page 1, upper left corner at 0,0 and (by default), extending the full size of the page.

-cbox2- 0,0,.5s,1s creates a boxed region on pages 2 and up, upper left corner at 0,0, with width 0.5 x the source page (0.5s) and height 1.0 x the source page (1s).

-cbox2- 0.5s,0s creates another boxed region on pages 2 and up, upper left corner starting half way (0.5s) across the source page and at the top (0s), and extending to the remainder of the source page (by default).

I tried both and found both output to be acceptable quality. As the original scan was not good quality the output reflects the poor input. As for the command options they are excellent and did what needs to be done. I will be using this to read this ancient book. Thank you so much.