Too dark print with Costco. How to lighten up?

I just started making some 11x14 and 20x30 prints with Costco. I converted all the images to correct profiles that Costco provided with Photoshop CS5. But all prints are consistently too dark when I got the prints back.

I read online that some people suggest to lighten up the images in PSCS5 10% to 15%. Some people suggest to reduce the brightness of LCD monitor. But I can not make sense out of these. I have spent lots of time on my computer to make sure most of my images fit exactly the histogram with no clips or minimum clips in both ends. So regardless of how bright my monitor is, the images are already correctly tuned and they look good on screen.

So, do I just ignore the hostogram? If not, how do I lighten up the images for print ?

Still puzzled,

Eric

This almost certainly means that your monitor is set too bright. If you are not using a monitor that is calibrated at 100 cd/m2 plus or minus 10, and adjust the files to what you see on the monitor, your prints will be too dark. All LCD monitors out of the box are at least twice that and some 3 times that resulting in dark prints.

So your first step is to get your monitor set up for the right brightness and then prepare your photos for printing by any photo printer, whether your own or a printing service. Anything else will leave you with a print that is too dark.

If you have no calibration device and software software, you can at least get closer to the right brightness by going to the bottom of this web page:
http://w4zt.com/screen/
now darken your brightness control until you can no longer see a difference between Y and Z on the last gray scale on the page and then increase brightness by 2% from that.

Now go back into PSCS5 and adjust your photos and send those to print.

Note the above is very quick and dirty for brightness and does not do anything for color. As I stated at the beginning, to really get good results you have to profile and calibrate the monitor using a colorimeter and profile creation software.

EJ.,

Thanks for the suggestion. I have a EyeOne and have a monitor profile setup last year. Note the prints from my home epson 2200 is usually just a little darker, while the Costco ones are more than 1 stop darker.

I remembered read one of your recommendation for LCD monitor calibration that says for modern LCD monitors, we should only adjust brightness and leave the rest to the monitor profile. I will re-adjust the monitor tonight to see whether that improve.

Eric

The problem may be at Costco's end. Did your print have all nnnn's on the back meaning no corrections were done? Did you check "no corrections"? Also, my local Costco has two printers, but only one of the printer's profiles is distributed by drycreek. And if I check "no corrections" that doesn't mean the print will go to the printer whose profile is distributed - the print job goes to whichever printer is most convenient to print that size paper at that moment. There's a huge difference in how the printers print color despite their being the same model - the profile for one doesn't work for the other. I stopped using Costco for anything other than snap-shot stuff because of this. My sense is Costco's key targeted customer is the casual snap-shot shooter.

I converted all the images to correct profiles that Costco provided with Photoshop CS5

This may be the problem. You don't "convert" the image to the profile provided by Costco, you use the profile to soft proof the image on your monitor.

The only profile you should be converting images to for printing is sRGB.

Jim, I think you're partially right. It's always been my belief that one soft-proof's to see how the colors will render given the gamut of the printer/paper combination that will be used, such as looking to see if any colors might blow-out (and thus need tweaking), and to set the white/black points for that combination. And when it looks as good as you can get it soft-proofing you then convert to the printer's profile before sending the file to the printer. If not, then when and how do color corrections happen to accomodate every color change that happens when going from say Adobe RGB to whatever gamut the the printer/paper combination is? Are you changing colors manually to accomodate this while soft-proofing? Read step 16 Here

Best, Peter

The link gives me "not found," but yes, once all adjustments are done and the photo is in sRGB, then turn on soft proofing and adjust the resulting image to look like it did before soft proofing.

You can duplicate the image and have one original and one in soft proofing to help out more. Once I figured out what the adjustment for my lab is, I created an action that "finishes" the photos by performing the adjustment.

Jim, I had trouble getting the link to work - the dang BB truncated the link - but it should work now.

Okay, I see that we're talking about different things. A profile for soft proofing is not the same as the printer profile itself.

Then again, I could have it all totally messed up in my feeble mind.

The process as I described is what I use for WHCC and used to use for Costco until I got tired of crooked cuts from my local store. The results are predictable and quality.

No, the file for soft-proofing is the same file - it's just that there are lots of ways to get from A to B. I just let the conversion do all the work after I tweak the image a bit, and you do more manual work. In the end I suspect we both get where we want be!

Then again, *I* could have it all totally messed up in my feeble mind!

The profile for soft proofing is, in fact, the same as the printer profile. That is the point of doing soft proofing -- you're simulating on screen how the results of the printer color space conversion will look.

In principle, there is nothing "wrong" with converting an image into the printing service's supplied color space. After all, they are going to do that themselves as part of the output procedure. The trick is that if you do the color space conversion yourself, you need to make sure the printing service doesn't do any further color transforms or color / tonal adjustments to the file... in effect you're supplying them with a "printer ready" file that should be dumped straight to the printer, as far as color work is concerned. (Though they may still need to upres the file or do certain other pre-flighting type work.)

If you can't trust the printer service to be able to deal with non-standard color transform workflow, it may be better to get their specific instructions on what they want, and then not deviate from it. If they say deliver in sRGB, deliver in sRGB. If they say they can take a limited set of other color spaces like Adobe RGB as long as you tag the images, then feel free to do that -- but give them also very specific instructions. Don't rely on them figuring things out. If you want to do something even a little bit different, make sure they understand what you are giving them and that they can & will handle it correctly.

Double-profiling during the output process, or other mismatched processing scenarios like Peter mentions, could be one possible cause of prints that appear too dark, and/or color shifted. Though as E.J. noted, by far the most common reason for "my prints look too dark" is in reality "my monitor is too bright". Hence the other thread on monitor calibration...

Royce is right on. Instructions from Dry Creek Photo (http://www.drycreekphoto.com/icc/) who provdes all Costco profiles says covert images to target printer profiles and specify "No auto correction" when order prints from Costco.

Eric

Peter's other comment still applies -- a Costco profile from Dry Creek may not in reality match any given output device at a specific Costco location. If you pre-convert to the printer color space yourself and then instruct the location to skip the color transformation, you may be sending sub-optimal color info to that printer. If you give them the image in some other device-independent color space (e.g. sRGB or Adobe RGB) that they can handle, and they do the color space conversion, that can potentially correct some mismatches. It really depends on whether they know what they're doing, and whether the profiles they're using are good / better than what's available to you.

If in doubt, ask. If they don't understand the question or you don't trust the answer, consider another option unless you can live with the results vs. the cost...

P.S. Now that the other monitor profiling thread has resulted in realizing the old Eye-One device was broken and reading an extremely low luminance for your monitor, this means the monitor was in fact too bright. Bringing it down to a proper luminance should get you back to a better position for matching display to print, as far as that part of it goes...

Hi all,

Thanks very much for your suggestions and advices. I replaced my monitor calibration tool and recalibrate my monitor and it indeed was too bright as EJ and Royce suggested. After recalibrated my monitor, I sent another image to Costco. The print was much lighter and I can use. The print is still darker than my monitor though.

Thanks for all your helps.

Eric

You have to remember that your image is backlit with transmitted light on a monitor and therefore glows. On a print, you are dealing with reflected light that has attenuated a lot from the source (the lightbulb). You would basically need to take the print outside in noon sunlight for a fair comparison.

EJ.,

Thanks very much for the explanation.

Eric

Just FYI, got my updated photo set back from Costco. It all looks great. Thanks for your helps here and from the Digital forum.

I also found my local Costco uses a big Epson printer for prints of 20x30 and bigger. I made 2 20x30 and they are just gorgeous.

Eric

Would it be correct to take that "too dark" print from Costco, and hold it up beside your monitor screen, then adjust your monitor screen brightness, to match the "too dark" print ?? Would future photo files sent to Costco be correct, now ??

Joe, I wouldn't call that "correct". It's a sort of a visual expedient if one didn't have the ability to properly calibrate & profile the monitor but many (most?) people would have too many variables to deal with to properly dial in the display by eye to get consistent, reliable results. For one thing as E.J. described above, the ambient viewing conditions next to the monitor likely are not the best for judging the print -- you probably wouldn't actually want the two to critically match right there. If the lighting is such that the print looks like it "should", this is probably suboptimal for doing processing work or critical image review on the display. And the opposite -- if the conditions are favorable for the monitor, the print will look dimmer & duller than it should because it's not equivalent to outdoor noon sort of light.

Dealing with the monitor involves several discrete steps, and setting the brightness of the back panel normally is the first one. Everything else a calibration & profiling tool does happens after the brightness level is set -- this is because the brightness pins the contrast range of black & white on the extreme ends, within which the tone curves and colors are then adjusted & baselined. If the brightness is juggled afterwards it can throw off all the other parameters that were determined in the ICC profile.

You could re-run the calibration again after determining monitor brightness by eye-ball comparison to the print, a sort of two step manual dial-in process. But even if all other conditions were such that an eye-ball comparison was valid, in all likelihood it's simply going to confirm what is already known about the optimal range of brightness settings for monitors based on a target cd/m2 value.

Finally, I don't really encourage people to try to go for hyper-exact matching between screen & print. An LCD monitor is a radically different technology for image viewing, compared to ink on paper reflecting light. The purpose of calibration & profiling is not and never has been really about making multiple devices or technologies exactly match each other. Rather it is about baselining them all to some acceptable standard for consistent reproduction. Each device or display technology has its own internal optimal "look" and you don't necessarily want to dumb all of them down to the lowest common denominator, which would be something like a print on low contrast matte paper viewed in dim artificial light.

A monitor is brighter & more contrasty than reflected light from a print, and to a certain extent you want it to be that way because the monitor is your working environment. Working on the image is different than simply viewing the image... for the sake of your eyes and certain aspects of your image quality (like shadow details) you probably don't want to take the monitor all the way down to the level of a print reflecting light, because the conditions of viewing a print are not ideal for working on the source image.

The monitor being a bit brighter or the print a bit darker is not necessarily a "problem" as long as both look good and achieve the purpose for which they are intended, in the viewing conditions that apply to them. Other comparisons (whether visual or numeric from a tool) are IMO guidelines to bring into the mix, looking for a good balance.