My last few weeks have been insanely busy – next week is the end of classes, so of course this week is the end-game for all projects & assignments. Sarah wanted her pictures edited so she could give some to her parents for Christmas (which made me happy :) ) so I squeezed in some editing time. I’ll show the full finished set in another post – right now I wanted to share an interesting observation about how color channels work when editing portraits.
For those of you who don’t spend too much of your lives in the Adobe suite, every color in the subset of colors allowed in the RGB color space can be represented as percentages of red, green, and blue. Those of you who have ever worked in web design are familiar with this concept (though you may not know it) through hexidecimal color values. These are hexidecimal (base 16) numbers with 6 hex digits. The first two represent the level of red, the next two green, the last two blue. The highest, #FFFFFF, is white – every color is saturated to the max. The lowest is, logically, black represented by #000000. Every combination in between results in a color, such as #9A32DF, #CC6688, or #123456.
(There’s also the CMYK color space, which represents all colors in mixes of cyan, magenta, yellow, and black. RGB and CMYK are both widely used, but RGB is typically how colors on the web are displayed while CMYK is the scheme usually used for printed images.)
This is the idea behind color channels in photoshop – you can view an image which displays, in greyscale, only the red, green, or blue values in the image. The best way that I know of to turn a colored image into a black and white one is to use Photoshop’s monochrome channel mixer. You can use the different RGB values as brightness levels and mesh them together by percentages to get a properly balanced B&W image. Here’s an example to give you a general idea.
This is the original:

These are the channels:

Here’s what happens when you go to grayscale (though I have no idea why you’d want to with this image) through different methods:

Not only is the black/white balance nicer, but the channel version preserves a lot of the detail in the seeds and veins of the berries that are missing in the mode conversion. I also balanced it to lose some of the less important, distracting details in the meringue base.
So, after a lengthy explanation, to my original point. My previous post about photo editing involved a lot of nit-picky steps: fix the skin, fix the eyes, fix the lips, etc. When I went to turn one of the portraits into the traditional B&W actor headshot, I found most of it was done in a single step. Behold:
The original:

The channels: (from the top down: red, green, blue)

Notice anything about the red channel? Yeah, the skin is flawless. Contrast isn’t bad either. A little washed out of course, but it’s not going to take much to fix that. Here’s the final, mixed version:
This is 80% red, 20% green. It could definitely use a little bit of touching up in spots, but the mixing itself did most of my work for me.
Which, as you know, is always a plus.