Runcible Blog

People sometimes ask me how I decide to convert a photo to black and white, and I don't have a simple explanation. I'll try to illustrate the thought process on a typical image.

There are some simple reasons for converting to monochrome:

1. The scene is already desaturated or otherwise monochromatic.
2. Color in the scene detracts from the content or message. (i.e. a bright red truck in the background)
3. You're going for an old-fashioned look.
4. You want the image to seem "serious," as in a journalistic style.

Sometimes the reasoning isn't as simple, and an otherwise perfectly fine color photograph might work better as black and white. Let's look an image straight from the camera (or as close as Aperture will lead me to believe):

A candle-lit portrait with auto white balance usually spells ORANGE, so let's try to correct the white balance:

Now the image is merely murky and flat. I prefer more contrast, and my general rule is to increase contrast until I've lost too much information in the image, then bring it down a bit. So, completely black shadows are usually fine (the eye can make up the difference), but completely white highlights are to be avoided because the eye fixates on the brightest part of the image, detracting from the rest of the tones. Although I try not to go overboard with saturation, a little bit helps to bring out the natural color in faces.  And I brought the exposure up about 2/3 of a stop to make the skin tone closer to reality.

 This image is a little too red in the right cheek, but we'll let it slide for now...

The question is, should I keep the image as color or convert it to black and white? In this case, even though the reds are vibrant, I don't think color adds enough to the image to keep it around. The blue window and yellow walls (with red stripe) in the background are a little distracting. The tricky lighting situation makes her face unnecessarily ruddy. On a nitpicky level, I took the photo at ISO 1600, which leaves some chroma noise that can be irritating at large magnifications (or even in print). Let's convert the image to black and white using Aperture's monochrome mixer at 90% green and 10% red. 100% green channel is too unnatural, but I like the look of this 90/10 mixture:

You can play around with different channels to get a more even-toned monochrome image. Since I like relatively high contrast and try to approximate the look of my favorite film images (this one is an extreme example), I go with the green channel and bring the black point up a bit. I may even increase contrast further.

Finally, the light in the top left ruins the illusion that this could be a film image (because the highlights are blown linearly, as only digital can produce).  Let's crop it out:

And there you have it. The photo isn't perfect, but we end up with a rich monochrome image with a reasonable amount of tonal gradation and texture, and no important information lost. If you look closely, the pupil and iris are still distinguishable, which is also something to take into consideration when fiddling with the contrast.

There are a lot of little details to think about when processing digital images, and the decision to convert an image to monochrome shouldn't be made flippantly. Sometimes a dull color photo can be made more interesting by converting to black and white (just as a dull image can be jazzed up by blasting the saturation or going crazy with Photoshop filters), but the challenge is to apply adjustments thoughtfully, understanding how the end result might affect your viewers' perception of what you're trying to show.

Whatever you do, don't click "convert to grayscale"!

I read all of Cliff Stoll's books 8 years ago, and I often channel his contrarianism about social networking, computers in education, and even my own choice of current occupation. He's influenced my life; it's a shame that his passionate questioning 10 years ago has faded into irrelevance, steamrolled by the billion dollar drive to get everyone connected and clicking away. With the impending failure of the OLPC, I wonder if anyone will take the opportunity to think about whether the number of man-hours spent producing green plastic doorstops for impoverished Peruvians could've been better spent thinking about ways to cure malaria or simply training more teachers to work in remote villages. I don't know.

In this hour-long presentation from 1996, Stoll raises important questions about computers in society, in his own frenetic style. Throughout, he maintains that he loves computers and the internet, but he wants people to question the assumption that computers and networks change everything for the better. What happened to the folks who said that the transistor radio would replace teachers and textbooks? Or televisions in classrooms? Or CD-ROMs? (yeesh!)

I particularly liked his comparison of the information superhighway to the interstate highway system built in the 50's. Back then, he said, the interstate highway was sold as something to make the country stronger, bring people together, spur jobs, growth, and cheap goods through easy transportation. All of those things came true, but at what cost? Whether it's the dissolution of communities due to suburban alienation, a looming ecological disaster from a polluting car-culture, or dependence on foreign oil which drags the country into perpetual resource-wars and economic ruin, maybe it would've been worth questioning the costs of that new infrastructure at the time. Similarly, why not question the effects that computers will have on our society? Lots of people have (here's a recent Frontline episode), but Cliff Stoll is a unique contrarian -- a pioneer of the internet and admitted "propellerhead" who thinks that maybe we should take a step back before trashing art class for Microsoft Word and bowling with friends for chatting online.

On a personal note, I think about what kind of value I'm producing in my job as Software Designer Level IV for a Fortune 11 company. There are so many stories to tell. Valuable stories that could make a difference. In comparison, what is the value of another multi-threaded webserver or a faster website for buying souvenirs? Well...

I'm at Logan airport, waiting for a delayed flight to Chicago, heading to PyCon. I'd say more, but blogging on the iPhone is tedious. More later.

 Bring these brave soldiers home, and punish them.

R.I.P Crystal (~1994 - 2008)

When I was young and stupid, I wasn't very nice to Crystal, but she always forgave me.  Ever in search of familial love, she had a habit of sucking on people's earlobes.  That was her way of getting to know someone. 

I think the best cats are other people's cats, which was true for Crystal when I moved away from home.  She'd be eager to see me when I visited, if only to serve as a brief respite from the chaotic dogs in her life.  As she grew older, she seemed to become more mellow and at peace with the world – far above all the riff-raff of people and animals coming and going from her life.  Just taking life one day at a time.