(click on any images for
a full-sized version in a new window)
In this walkthrough I'll explain the steps I followed to produce the photograph of model RC that you can see above. I'll try to describe both technical details, artistic choices, and some of my motivations at each step. I believe it's very important to separate one's creative choices from practical details, wherever possible - it keeps one honest and neither belittles nor aggrandizes the role of technology in one's art. Put differently, I will try, wherever possible, to distinguish Art from Craft.
The photo we'll be working on in this walkthrough is one of my "Women on Pedestals" series. It's a series that happened by accident after I did one shoot with Samantha Stine that resulted in some photos that became instant favorites. When I was shooting her, I had her posing against a paper sweep lit with a couple of softboxes; a very simple set-up. It was a really cold day and Samantha was huddled under a blanket while I got the lights set up and metered the scene. In order to give myself something to focus on while I checked the lighting, I put this plaster column into the scene, because its white unpainted surface gave a really nice clear highlight to base my exposure on. When I was ready, Samantha came out and immediately hit a series of beautiful and dynamic poses, balancing on the column.
(Sam, Morrisdale Pa, 2004)
Since then, a number of the models I've worked with have asked me to "shoot one of those column pictures of me" and, after a while, I decided that I'd just make it a series. Now, it's almost a given that I'll pose my models with "that damned plaster pedestal" (as Olaf Starorypinski would say). So, what has become a series started as a lighting test and some inspired posing by a model. As the artist, how much credit do I get to take? I don't let it bother me much; I'm just happy to be taking photographs that please my eye. Anyhow, as the series evolved I established a few rules for myself, namely that each image needed the original shot's soft glowing light, the pedestal, the simple background, and some kind of slight color-toning. To maintain the original emotional concept of Samantha's pose, I have decided to let each model interpret the pedestal in her own way with a minimum of direction. Usually, the models laugh, think a few minutes, and do some pose that is somehow unique to them. I've been fascinated with the way the series has evolved as various models reflect their own personalities through it.
The lighting scheme used for the pedestal series is a variation of a lighting setup I came up with when I was working on a still-life back when I first started experimenting with studio lighting. I had 2 small strobes (a Novatron "Fun Kit") with little soft-boxes and wanted to shoot a "product-style" image with a "big soft box effect." As it happened, I put the 2 softboxes together and angled slightly inward and discovered that I really liked the effect; it gives a nice soft luminance with an extended "edge effect" that gives a lot of dimensionality to the resulting image. Note - I am not claiming to have "invented" this lighting setup, or anything like that. I'm sure it's been used before by zillions of other studio photographers. One of the things that I personally avoid in art is worrying about being "derivative" or "copying" - I figure that if I just charge ahead and do what pleases me, I can just smile and nod if someone looks at one of my photos and says, "that's the same lighting that Joe Whoever used to use!" I'm just happy to be having fun taking photographs that please me.
The setup for the lighting is shown above. You'll notice the pair of softboxes on the right of the sweep, angled slightly inward toward the column in the center of the scene. There is another strobe on the left hand side of the sweep, but it's turned off and won't contribute anything to the scene. The paper sweep is "thunder grey" from Savage. The color of the sweep is mostly irrelevant because I am going to do a monochrome conversion and colorize the image; I've used colors ranging from white to warm brown as sweep backgrounds for other images in the series.
It doesn't get much more simple than that, does it?
Computing exposure for a scene like this is ridiculously easy: you just hold an incident flash-meter right on the top of the colum and take a reading. Because the lights are fairly close to the scene, there's a chance that the bright edge of the column will be slightly over-exposed - I deal with that by taking a test-shot and looking at the histogram in my camera. If I were shooting film I'd either expect to do a little burning on the pedestal at print-time, or I'd put the light-meter on the bright side of the pedestal and adjust my development time a little bit to bring up the shadows.
RC hit about 4 different poses for the scene and I shot about 2 dozen frames all told. Usually, I go through the results of a shoot in file browser and grab 3 or 4 that catch my eye immediately. At this point, I like to be instinctive about what I'm doing. Usually I'll carry around the images in a work folder for a couple of weeks and let myself choose one subconsciously; I'm not particularly fond of thinking about composition because there are a tremendous number of intangibles - and no vocabulary that works (for me) to describe them. I look for harmony and flowing lines or interesting curves and triangles, balance, tension. I know there are Art Intellectuals that have spent tremendous effort to define "rules" for composition; personally I can't think of a bigger waste of time. I had a teacher, once, who was a big fan of the "Rule of Thirds" which, as far as I could tell, he used to dismiss images he didn't like and to make excuses for images that he did like. Look at the images in the file browser above; the second one over doesn't have the dominant interest in the center of the photo; a "Rule of Thirds" photographer would feel morally and intellectually obligated to choose that one - but personally I like the flow of her curves in the first image: the way her hips and torso play off the verticals in the pedestal. So - "Screw the Rule of Thirds" I say. I chose the first image based on the "Rule of Curvy Hips" I guess.
This is the un-modified image before I started to work on it in photoshop. For the sake of the walkthrough, all the images I'm putting up are 1000pixels on the long side, to keep them small (and to keep from having to worry about someone ripping off the full 12-megapixel image). So, you can see all of the images flaws clearly - the column isn't straight, there are smudges on the floor, and there are a few digital dust bunnies to deal with.
First let's deal with rotating the column to vertical. The actual plaster column I use has little rubber feet on the base, and they're not even, so it's tipped a bit to the side - I've learned to just not worry about it, and fix it in photoshop. If I were shooting film I'd invest a few hours levelling the plaster base.
Here I've created a guide (View -> New Guide, vertical 1000pixels) to give me something to align on. I then select the entire image (Select -> All, or Control-A) and rotate the selection area (Edit -> Transform -> Rotate) until it looks right. You just grab the little doodlies at the corner of the image with your mouse and fiddle with it until it looks right, then click the check-mark in the top menu bar.
Now that the image is rotated to satisfaction, I can consider cropping it. Again, there are no rules for cropping images except "crop to strengthen your composition" - whatever that means! I did feel there was a bit too much blank paper at the base of the image and that it'd look a little better if the viewer's attention was brought even more squarely on the model, so I moved the edges and top in a bit and the bottom quite a lot. Does it look better? I'm happy with it and I guess that's good enough.
Next comes the hunt for digital dust bunnies! I methodically scroll back and forth across the image and use healing brush or clone brush to erase all the little flecks and flaws from CCD dust. Usually I do this as a 3-pass process. First, I clean up dust bunnies from the background and make a fast pass over the entire image. Then I go back and clean up flaws on the subject - this can include extreme measures like tattoo and piercing removal (if I feel like it) as well as moles, skin blemishes, etc. Since my objective is completely ruled by my aesthetics at this point, I don't have any remorse about making whatever alterations suit me. I've been known to use Liquify filter to smooth out a curve here, or narrow a waistline there. In this case, my work is minimal because her skin is flawless, and so are her curves.
Now it's time for black-and-white conversion! I bring up Channel Mixer (Image -> Adjustments -> Channel Mixer) and check the "monochrome" box then fiddle with the sliders until the photo has the tonal range I want. This is governed by the skin tone of the model, the overall mood I want in the picture, and a bunch of other intangibles. When I'm doing monochrome conversion I seldom adjust the tones to over 100% (which would effectively be adjusting the levels for a particular color channel) I try to get the overall tonal balance where I want it, and then fix the level setting as a separate step. In this case, I pretty much hit the exposure perfectly so I didn't bother with adjusting levels at all.
"Toning" the image is done using the Hue/Saturation adjustment (Image -> Adjust -> Hue / Saturation) with the "Colorize" check box selected. I find this stage to be really enjoyable because you can quickly get a whole lot of different perspectives on your photo. I usually leave the saturation level where it is, then slide the Hue selector around until I see something I like.
I've noticed one rather funny thing in the column series - I seem to tone the image to sort of match what the model's skin tone looked like. Of course what I am really dealing with is my own memory of what the model looked like, idealized through the process of black-and-white conversion.
The final step in prepping the image for posting is to downsample it. I save the full-resolution version (in case I want to make prints someday) under a different name from the original file (in case I want to revisit the original file someday) and then downsample using the Image size dialog (Image -> Image size). Downsampling Jpegs can sometimes have the effect of softening them slightly. If you feel that the image would not benefit from being a little bit softer, you can choose different resampling modes in Photoshop. Bicubic Sharper works well for this image. I use a personal rule of thumb for upsampling and downsampling images:
Upsampling your image basically means that your image isn't big enough. My preferred way to handle that is to get a bigger native image, not to try to "stretch" a bunch of data until it turns into a blocky mass of interpolation artifacts. If I need a 12000 pixel-per-side image I'll shoot a color negative in my 8x10 and have it scanned by a service bureau using a high-end scanner. Then it'll print with adequate sharpness for most purposes.
I decided on a 1000pixel-on-longest-side dimension for posting because screen sizes and resolutions are improving, and the resulting image is still not quite good enough quality for someone to print. I am actually less bothered by the idea of someone stealing my work than I am by the idea of someone doing a bad job printing my work. If you want a print of one of my photos, ask me, you'll find it's pretty easy to get one.
I'm happy with this result. I hope you've enjoyed this walkthrough, and maybe learned something from it, too!
Aspen Airport, and United Airways flight from Denver to Dulles International Jul 20, 2005