Hi,
I did the comparison that I mentioned and the renderings are identical.
I do see a slight difference in your dropbox images but this goes away
when moving my head, i.e. the difference is due to screen angle.
Colorspaces can appear at many places, in the content stream, in images,
shadings, resources... I had a quick look at your file with PDFDebugger
and it uses ICC colorspaces.
Tilman
Am 05.06.2017 um 13:49 schrieb Derwent Ready:
Hi,
Thanks for the quick response. I just tried your code and put the outputs
from Indesign, Photoshop RGB/CYMK, PDFBox default and PDFBox using the SWOP
profile next to each other and yes there is definitely a difference so
thanks for that. :)
The SWOP is definitely closer to the Photoshop versions than the PDFBox
default. This is more noticeable on one set of images than another.
I've linked a couple of images in a dropbox link that show the differences.
On the first image (purple) the difference is minimal but on the second
(orange) the gradients are really quite different. I'm not sure if this is
a PDFBox issue or a profile issue or what.
https://www.dropbox.com/sh/e80qn8qa2s2ad56/AAA0qKuhAjXeP0e2Qe2AXmFea?dl=0
How would I check that the file has DeviceCYMK colourspace? I've included a
blank version of the PDF in the dropbox folder.
Thanks again,
D
On 5 June 2017 at 10:40, Tilman Hausherr <[email protected]> wrote:
Here's what worked for me, I used it in my rendering tests which detect
differences:
PDDeviceCMYK.INSTANCE = new PDDeviceCMYK()
{
@Override
protected ICC_Profile getICCProfile() throws IOException
{
return ICC_Profile.getInstance("USWebCoatedSWOP.icc");
}
};
The results are definitively different.
If you have a test file, I could render it with and without the change and
check whether it is different or not. Maybe your file doesn't have the
/DeviceCMYK colorspace but something else?
Tilman
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
