OK I tested with pdftocairo and indeed the result is much better. I'm
kinda surprised though, I would have thought that pdftops would be more
specialized and generally better, but it doesn't seem so.
Are there some pitfalls to using pdftocairo for postscript rendering?
Are there major differences compared to pdftops that I should be aware of?
Thanks for the info, you guys are very helpful.
Pierre-Luc
On 07/06/2016 08:10 AM, Adrian Johnson wrote:
pdftops does preserve the original resolution. However if the page
contains anything that can't be converted directly to PostScript
(usually transparency), the entire page is rasterized to a single
fallback image with the resolution specified by -r.
pdftocairo also preserves the original resolution. In the case where a
fallback image is required, pdftocairo will only rasterize the parts of
the page that can't be converted to PostScript. So if PDF images do not
contain transparency they will always be output at their native resolution.
There is a bug open to add support to pdftops for finer-grained
fallbacks. But I'm not sure how to handle different color spaces.
https://bugs.freedesktop.org/show_bug.cgi?id=66056
On 05/07/16 23:19, suzuki toshiya wrote:
Dear Pierre,
Sorry for that I'm posting this before testing by myself,
have you take a look on pdftocairo? I remember, when I
convert a pdf to a svg, pdftocairo does not re-rasterize
the embedded raster image, so I expect PS output from
pdftocairo could be better for your purpose.
Regards,
mpsuzuki
Pierre-Luc Samuel wrote:
Hi,
I've been wondering if there is a feature for pdftops that would convert
pdf to postscript by keeping the original resolution of images embedded
in the pdf? The problem I'm encountering is that my pdfs contain low
res images that are stretched to be big; pdftops converts them by
re-rasterizing the low res images so that they look smooth.
The result is that my images now look extremely smooth, but the
resulting postscript file size explodes. The issue I have with "-r
<dpi>" is that the native resolution of my images is about 150 dpi, but
when using "-r 150", I get reasonable file sizes, but the text sections
of the pdf becomes barely readable.... I would like to suggest a
"-nativeimageres" option that would be independent of the existing "-r
<dpi>" option.
Do you guys think a "-nativeimageres" option is technically possible?
Could you give me indications where to start, in case I need to look
deeper into this?
Thanks,
Pierre-Luc
_______________________________________________
poppler mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/poppler
_______________________________________________
poppler mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/poppler
_______________________________________________
poppler mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/poppler
_______________________________________________
poppler mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/poppler
