Extracting images from pdf files, was Re: [BL] Re: Interrupts and isapnp
keesan at sdf.lonestar.org
Fri Jun 10 15:04:22 EDT 2005
I tried to print to the dot-matrix printer with ghostscript (device epson)
in both linux and DOS, gs 8.50 and 5.10. It prints (both pdf and ps
versions) at four times proper size (90 instead of 36 dpi) even specifying
Another older pdf file printed properly so I figured they had done
something odd to the pdf format. The site insists that you need Adobe 5
to read its files.
I converted the file to two .pbm files (pbmimages) and converted one to a
.gif (with pictview).
In DOS, I used printgf for DOS to print with the switch for 1:1 pixel (one
printed dot per dot in the image file). It printed four times the usual
size. So there is something about the image format used when scanning
that is confusing 'epson' printers. (I have not yet tried with stcolor
escp/2 Epson inkjet printer, which does not have fan-fold paper).
I then set the switch to print 8x11" and adjusted the paper over to the
left and down to the bottom to avoid margin errors. It printed (in over
15 minutes for one page at LQ) perfectly. But since there was a 300 dpi
to 360 dpi conversion involved, the lines are smudgier and the double
lines are fat singles again. This Panasonic is eminently unsuited for
printing these pdf files.
In the process I discovered that printgf lets the user move the image
around the page, with 0,0 being upper left corner as default, so I could
just print to an HP with it from DOS after measuring each badly printed
page to determine the needed adjustment (rather than doing a mechanical
I will see if gs also lets me adjust the margins this way so I can print
from linux without converting to .pbm and .gif first. It would have to be
done CLI since each page needs a different adjustment. Or perhaps xv can
be used to print (via gs), or netpbm's pbmtoepson allows for image
placement on the page. Extracting the pbm is quick and easy.
On Fri, 10 Jun 2005, sindi keesan wrote:
> I am now able to use pdfimages to extract images. I must have been doing
> something wrong. It produces 300 dpi pbm files which I hope to be able to
> edit, but it would still be simpler to just adjust a dot-matrix printer to
> get the page printed properly, rather than cropping and padding top and
> bottom and left and right of each page separately, which would probably take
> more time than scanning and printing a copy from the library.
> pdftopnm produces ppm files from the same pdf files, much more slowly, and at
> 150 dpi resolution.
> BasLinux mailing list
> BasLinux at lists.ibiblio.org
keesan at sdf.lonestar.org
SDF Public Access UNIX System - http://sdf.lonestar.org
More information about the BasLinux