Thursday, January 3, 2013
Something Completely Different: A New Look at Robert E. Lee
My childhood best friend happens to be a very accomplished writer and historian who is just about to release his fifth book, on the Battle of Gettysburg. I'd already done artwork for the audio book publishing company, where he also works, and he asked me if I might be interested in painting portraits of Generals Lee and Meade as illustrations for his book. We discussed the idea of colorizing original photographs, which had been done for both Meade and Lee - but not very convincingly - by other artists.
I'd attacked methods of colorizing historical black and white images over the years, trying to bend my brain around how color and light work and why most professional colorization projects had, in my opinion, failed to look 'right'. The answers that I came up with did help: how shadows were not merely darker than adjacent areas but also less saturated, and how the tones of various colors were never consistent due to myriad factors. Black and white photography captured images differently than color photography. Failure to incorporate these realities in any attempt to colorize an image were often responsible for some of the lackluster results I'd experienced.
I played around with traditional ways to colorize images while trying to implement these lessons, and produced some work that I was happy with - insofar as I was willing to offer them as commercial products, and they sold well. But I still wanted better results.
The human face is arguably the most difficult thing to colorize. There are two main reasons for this: the human face is made up of almost every color on the wheel - even those that seem illogical and those that we don't consciously see - and the human face is so familiar to us that any error in its reproduction by an artist is glaringly obvious even to the untrained eye.
I considered how to overcome these obstacles.
Photoshop can do many things, thank God, and among these things is the ability to analyze an area of color in a way that is not circumvented by our own brain's annoying habit of optical interpretation. Our brains don't see that many human ears contain the color green, for example, but Photoshop does. By sampling color images in this way we can 'see' what our eyes do not - at least not directly - and can then match our paints to mirror what is actually there.
In the case of 'colorizing' Lee (using quotes because we're now going beyond simply colorizing), I decided to try to match areas of his face to color samples of white men of his approximate age and completion. To make a longer story short: I painted over, and over, over various translucent layers (77 of them in total), until my Lee painting technically matched my color samples, and I refused to let my brain interpret anything. When I pulled back and observed the completed image: I was very pleased!
I sent my completed Lee to my friend, who in turn sent it to a colleague of his who'd previously had published a best-selling book of colorized Civil War photographs, and the latter remarked that my results were "far superior to anything [he'd] ever seen." I followed up my Lee painting with one of Meade, and both are being published within my friend's Gettysburg book.
I hope to do more of these! As an artist who has focused almost exclusively on aviation, it's both nice and commercially wise to venture out of the comfort zone and try new things.