How do I determine the contrast problem I have with my prints?

The problem I have is that all my prints turn out to be too contrasty...I think. Or maybe they are just too dark. The colors looks quite accurate (as compared to the soft proof in CS2), taking into consideration the lower brightness of the print.

I have a color managed workflow, using Adobe RGB as my working space and Spyder 2 as my calibration/profiling tool. The room is not set up for specific lighting though (I'm just an amateur ). I print on an Epson R1800 using Epson provided profiles, and mostly on the Premium Glossy paper.

What sort of tests can I perform to find the source of the problem? How do determine if my calibration is incorrect (Colorvision's software it is after all based on human input), or maybe the profiles from Epson are incorrect? Or maybe something else completely?

My initial thought was to just find or create a simple gray-scale gradient image and print it, but will that tell me which component is the source of the problem? I'd think not.

Any help much appreciated.

I have an r1800, and use my own profiles with differant inks and papers, so I can't give a specific answer re. epson paper/profiles.

It coiuld be that you have not got the driver settings correct for the paper and profile. If this has always happened, have you set the profile correcctly, e.g. allow cs2 to set profile, not the printer? Try taking the print outside, look at in daylight, is it OK there? fwiw, If it looks OK on my screen in the editor, in my lighting, then the next time I see the same colour is when its on the paper, I never use soft proofing, and the print preview is only used to check I have the correct cropping.

assuming the 'nozzle check' is OK, of course.

I do not think a grey scale image will help, but maybe an image with a number of colour gradients would - there are plenty on the web to try, including some sort of standard ones.

Best wishes,

Ray

Thanks Ray. I use Qimage for printing and I'm sure the printer drivers and Qimage are all set up correctly (i.e. no CM in printer driver, correct paper profile, etc). The print is too dark/contrasty in daylight as well.

The problem with printing any test target is that the results would again just tell me if something is wrong or not, not *what* is wrong (display profile, printer profile, conversion, etc.). I guess I would need a test image that would point out, on the print itself, the limits for my printer/paper. I suppose finding such a target is going to be hard

FWIW, the softproof shows the image as much less contrasty than the regular screen rendition, but I understand that this is normal behaviour.

Hi Kris,

I was going to suggest qimage! I also use profile prism from the same supplier for profiling the printer. If you read the tutorial/help files on his site, I think it explains his take on monitor profiling, etc. I set mine manually, as per his instructions and got good results. Later I bought the optix xr, but found I did not need to alter much from my manual monitor adjustment.

At the moment, I don't know if it is your printer out of adjustment, or your monitor, maybe one of your monitor settings has gone awry. A simple test that may show something.

When I load in a jpeg in cs2 (one with no embedded rgb profile), I only get the the colours in cs2 looking the same as if I opened it in Irfan view (or other non colour managed image browser, I guess), if I assign profile to srgb, and then convert document to working rgb. Do you get the same? In my case, if I leave as is, or assign working rgb, I get a colour cast.

If within cs2, I save a jpeg with an embedded profile (other than srgb), then the colours are messed in Irfan view.

I'm wondering if you have the correct profile on your monitor - i.e. have you adjusted brightness/contrast since calibrating (is the calibator any use??) Have you still got adobe colour on startup, or have you assigned the correct profile to the monitor? (I think there is a limit to the number of profiles you are allowed in the folder - something like maximum of 32, iirc.)

btw, I use raw images, normally, and I guess you do, but I thought the jpeg explanation would possibly show up double calibration, whatever. If the print is dark, compared to your monitor (making allowance for the fact that one reflects, t'other transmits light), then either the printer or the monitor or both are in need of colour profiling.

Best wishes,

Ray

just curious...what type of monitor do you have? i find with the apple lcd monitor that i am sometimes lulled into thinking my shots are somewhat brighter than they are and the print (r2400) seems darker although, as you say, the colors are accurate. has this contrast difference always been there, or was everything once fine and now isn't? i am just trying to eliminate some of the peripheral issues.

Marian Howell said:

i find with the apple lcd monitor that i am sometimes lulled into thinking my shots are somewhat brighter than they are and the print (r2400) seems darker although, as you say, the colors are accurate.

Click to expand...

Assuming the profiles are correct, the viewing conditions are an other variable. It makes a lot of difference if you view your display in relative darkness or in semi-bright surroundings.

Some profiling solutions allow to tweak output profiles after creation, sometimes it can help to fool the profiler by adjusting the printfile, but first make sure that the comparison is under similar/typical conditions.

Bart

> Or maybe they are just too dark.

Like Marian suggested, a common cause of dark prints is a monitor that's too bright. LCD or CRT? and what white point luminance are you calibrated to?

- DL

Thanks for your responses guys!

Ray, I will try your tests tomorrow as its getting late over here. I also use IrfanView and I do get consistent colors with sRGB files. I have recently switched to FastStone though as it is CM aware and almost as fast as IV. But I will test tomorrow just to make sure. As far as using the correct profile I am sure I do. I only have one, which comes from the Spyder 2 (I overwrite every time I re-profile). I never touch the controls on the monitor after the calibration/profiling. Adobe Gamma is not used, just double checked.

Marian, I use the LaCie electronBlue IV CRT monitor. I think this is a fine display

As I said originally, my room does not have controlled lighting. I try calibrate in the afternoon, when the lighting is semi-bright (drawn, sun-blocking curtains), but I work during all times of day. In the evenings/nights I use a single diffuse light that seems to create roughly the same brightness (although that is very difficult to judge). I calibrate to 6500, 2.2. Luminance levels are set manually though. This is obviously the weak point in the calibration and profiling process and I'm starting to think this is where I get it wrong, setting the black-point too high maybe.

> setting the black-point too high maybe.

More likely that setting the WP too high is the problem. In a dark room I had to calibrate down to 85 cd/m2 to correct the dark print problem. In an average lit room I use 90 cd/m2 (trinitron CRT).

- DL

I'm actually is very interested in the troubleshooting approach to this problem. Not just "what's wrong" but how do you diagnose the problem.

Kris is using a decent CRT, but has no luminance measurment in the software, so he can't do it by numbers... Chances are it's not too high though.

It's Windows since Adobe Gamma and Qimage are mentioned, so there's always the LUT loader issues to be concerned about. And possible Qimage colormanagement settings problems.

There are printer profiles that are not based on actual measurment of his particular printer....


Where do you start? I don't know a systematic approach. But still..


Try assigning sRGB as a monitor profile. See what you get.

Try calibrating to 5000K / 1.8, see what you get (Don't use monitor buttons for this - indicate you don't have them so the software uses vc LUTs. It's just for testing).