You need to print to Verify your edits

I'm just home from my first printing workshop. We had a great time and as much as workshops are there to teach my participants, I always learn a lot too. No one's learning is ever complete.

At the start of the week I explained to my group that although having a tightly calibrated / profiled monitor that matches what we see in print is something we strive for: it is not ideal. The truth is, that the only way to verify what we have in our files is to print. I am not alone in knowing that even with a highly profiled monitor I can be mislead. It is only the print that is honest: it shows me what is right and also what is wrong. Errors that I did not see on the monitor become evident in the print, and once I return to the monitor to check if they were there also, I see they were there all along.

The great American photographer Charlie Cramer has often said that 'computer monitors have their own reality distortion field. The more you look at them, the more your eye adapts. The only way to see what is really in your file is to print'.  I'd sure love to attend one of Charlie's workshops sometime. He is someone I have heard consistently great things about: he sounds like a great teacher.

The thing about profiling computer monitors is that you can't trust the software. It is always 'aiming for the target you set, but often finding out that it can't reach it'. So when your calibration software says 'your monitor is calibrated', what it is often saying is this; 'I did my best'. There are a few reasons for this: firstly, depending on the monitor hardware, it may have a difficult time trying to reach the white-point and luminance levels you are asking for. I know my old Eizo lost shadow detail when I tried to calibrate it down to 100 candles. It was too low for the monitor.

I use BasICColour's Display 5 software. Below it shows you how 'close' it got to what I was aiming for (known as the delta). You can see that my colorimeter and software got very very close indeed. But this still only means that the software got close to what I aimed for. But you may be aiming for the wrong result......

 BasICColour Display Calibration & Profiling software shows you just how much of a delta there was between what you aimed for, and what you got when you calibrated / profiled. You can see I chose a luminance of 100cdm, and the black point of the monitor can't reach absolute zero, so I've set it to what it's physically capable of reaching (0.26cdm). Even with this report showing me the delta, I still need a verification test proof to compare with my monitor: the only way to confirm your profiling is visually.

BasICColour Display Calibration & Profiling software shows you just how much of a delta there was between what you aimed for, and what you got when you calibrated / profiled. You can see I chose a luminance of 100cdm, and the black point of the monitor can't reach absolute zero, so I've set it to what it's physically capable of reaching (0.26cdm). Even with this report showing me the delta, I still need a verification test proof to compare with my monitor: the only way to confirm your profiling is visually.

You need to have something to verify against. Just because your software says 'I did it!', means nothing. If you are finding that your prints look warmer than your monitor, then you are probably using the wrong white-point setting. To find out what that should be, requires you compare your calibration with a day-light viewing booth. On the image below I have a daylight viewing booth (colour temperature is D50 - 5000K) and to match that, my computer monitor is around 5,800K. Each monitor will vary. Some may be higher in colour temperature while others may be lower. Just because I asked my calibration software to reach D65 (6,500K) means it is only a target it is aiming for. In truth, D65 on a monitor is far to cool.

You can't trust the numbers, only the visual inspection. That means iterating around the profiling / calibration software looking for a white-point that matches a viewing target. Once you find that colour temperature for your monitor, you now have a place to evaluate your prints.

 Even though my monitor is tightly profiled and calibrated to match my GTI viewing booth, I still see errors in the final print that were actually present in the monitor representation. I now feel I still have to learn to 'interpret' what my monitor is telling me, and not to trust it too much.

Even though my monitor is tightly profiled and calibrated to match my GTI viewing booth, I still see errors in the final print that were actually present in the monitor representation. I now feel I still have to learn to 'interpret' what my monitor is telling me, and not to trust it too much.

Once I have my monitor showing a close representation of what is under my viewing booth may I evaluate my prints. And this is where the fun begins: this is when you will find tonal distractions, colour casts and other distractions in the final print that you 'thought' weren't on your monitor. Looking back at your monitor to review, you will find they were there all along. 

The human eye is highly adaptable. One thing I have learned is that my visual system is constantly trying to lie to me. Monitors only get me so far. I have to print to verify what I think is in the file. Even if the calibration and profiling of my monitor closely represents what is on print.

I'd go one step further by adding in what Charlie Cramer has to say about the printing process:

"Poor images can look great on a monitor but will always look bad in print. Whereas great prints always look great on a computer monitor"

Your images aren't complete until you've printed them, and then further optimised them.

You have to print.