Modelling Prints in Post Processing:Part 2 White Balance

In Part 1 I showed some success in correcting the darkness in my initial prints. The result while seemingly matching the luminance of the image on the screen also was much warmer than the screen version. I have to admit I am at a loss for how this can happen. Increasing contrast increases saturation but that doesn’t seem to be what is happening. I experimented with a few adjustments including saturation, specific color levels and other tools. Nothing was as satisfactory as changing the white balance. At this point I am accepting this warmer tone without explanation and now seeking to attenuate it while trying to stay true to the original photo’s look. 

Desired Image

Print after gamma (.625) correction.

I began this next correction in much the same way as the first part. I added a new layer for white balance correction. I then set the gamma modelling layer on and adjusted the white balance until I got the screen image to match the initial test print. This for me seemed to be +50% since the print is much too warm! Now I have a model for how the print looks. 

I then add another white balance layer which I will use to correct the white balance problem. Unfortunately I do not understand the mathematics for white balance correction, so I have to experiment until I have cancelled out the other white balance effect. In this case -65% seemed close enough. 

Again I grouped these two layers together so that I now have a gamma group and a white balance group. These two groups I included in another over-arching group. Now by descending into the groups I can turn on all the models and corrections and by toggling the super group on and off I can see how closely it matches the original image.

I then apply my gamma and white balance compensations/corrections and export a new image to print. 

Exported image with .625 gamma and -65 white balance.

Print with gamma and white balance compensations

This print is very close so the experiment was pretty satisfactory. 

The next test of the validity of this is to apply these compensations blindly (without a test print) to an image to see if it hits the mark in terms of the print outcome. 

Original image of ‘blind’ trial

Next I applied the same compensations for white balance and gamma. 

Compensated image. 

Resulting print

Again apologies as my photos of the prints don’t do them justice. The print is not a perfect match but it is much closer than an uncompensated print. 

Incidentally I had a C-print made at a color lab in Spokane and their print ended up a little dark but much lighter than the original Walmart inkjet. The dark clouds lost their cool blue tone in favor of a more neutral gray tone. So the C-print was not any better uncompensated even though it is touted as their premium print service where they make some adjustments. It is much better than Walmarts unadjusted version of the same image. The prices are right around $3 for an 8x10 print in either case though I have to pay shipping from Spokane. 

All in all my recent journey into sourcing prints from different vendors is very frustrating if one expects to get consistent results. In the end I am going to have to settle on the best option and either live with the differences or develop compensations.