Analysis of a Simple Glamour Photo: Annja Enhorn

(click on any images for a full-up version in a new window)

In this walkthrough I'll describe the shooting and post-processing of a simple black-and-white glamour nude. Producing the image was a mixture of following creative intuition and straightforward production techniques. Wherever possible I will try to separate the technical from the non-technical, because I think that perhaps the most important part of any art-making process is understanding the dividing lines between where you release and maintain control over what's happening. There's some kind of continuum between "control everything" and "control nothing" and you, as an artist, will come to understand your own comfort range and where you usually fall within it.

Equipment, Concept and Set-Up

Equipment used: 2 White Lightning 800w/s mono-lights, with large soft boxes, normal ("tall") light stand and ground-level ("short") light stand, Fuji FinePIX S3 Pro, incident flash meter, photoshop. The lovely Annja Enhorn from Allentown, PA, posed in the center on a simple futon draped with some wrinkled white sheets for texture.

The analysis chart shows the lighting that was used - a horizontal variation of the "2 softbox" technique. In this set-up, a pair of softboxes are used to create the effect of a giant soft light by separating them slightly and then angling them so that their output creates a large "soft pool" of light. Since the light in the soft pool is coming from slightly different directions it's kind of like light from a softbox - it's diffused but directional, and seems to "cling" to the edges of the subject.

Annja's a dancer and is very comfortable posing and moving on her own. So, for this shot, I set the lighting up and explained to her that she was going to be stretching and rummaging around in the covers, like you might if you were waking up on a lazy sunday morning. In other words, I split the creative control over the photo between myself and Annja - by establishing the lighting and scenario, I had built the basics of the shot, but Annja's interpretation of the scene was going to have just as large an effect on it as my initial conditions. I suppose my final "creative moment" was deciding when to hit the shutter on my camera. But, as you can imagine, I had the camera on autofocus and was cheerfully firing away every time Annja's position looked good, or even semi-good. The creative process, thus, gets divided into several stages:

  1. Setting the scene (lighting, concept, composition)
  2. The model's interpretation of the scene
  3. Capturing the photographs
  4. (Later) figuring out which photo(s) to work on
  5. Post-production

In this walkthrough, I will try to touch, at least lightly, on all of the different stages of this creative process. What I enjoy so much about photography is that there are really interesting details at each stage of the process, to which you can devote a great deal of attention - or ignore nearly entirely. I try to pay attention to the details that I like, because I like them, but I'm acquainted with photographers who've chosen completely different sets of details to consider, and their choices are just as valid as mine.


Here you can see the setup of the scene and lights. The exposure is all wrong, of course, because I shot this hand-held in programmed mode and my camera's built-in computer underexposed because of the bright white object in the center of the scene. You'll see the main elements of the scene, though: the paper sweep, the futon, the 2 softboxes stacked and angled, and the photographer's shooting position. I put my tripod with its 3D adapter head in the shooting position, for reference. You'll notice that the studio wall to the right of the softboxes is painted white. If you were shooting these scene and wanted nearly the same look but didn't have a pair of lights and softboxes, you could use a single stobe and a reflector, and bounce the light off the white wall, using it as a diffuser.

When I am working in the studio with strobes I virtually always use an incident flash-meter to compute my exposures. Once I've metered the scene, I "fudge around" to get the base exposure that I think will work correctly. Let's look at the scene above - if I recall correctly, the light meter, when placed in the center of the futon, read F-11 @1/125 second with the camera set at ASA 100. Because the lights are so close to the scene, however, the right (from our perspective) edge of the futon is half as far from the lights as the center. Let's not go into the inverse-square law and all that but it's important to remember that:
the rate at which light falls off is more significant with lights that are closer, than if they are far away.
So having 2 big lights like this right up against the side of the scene is going to not only affect the luminance of the scene, it'll affect its contrast. Sure enough, when you meter by the right-most edge of the futon, the light was a full stop brighter. So, we have our first technical decision/challenge! If I used the reading for the center of the futon, I might blow out the highlights on the white sheet at the right side of the futon. If I expose to get the right side of the futon safely within my contrast range, the left side is going to be pretty dark. I decided to follow the old black-and-white group f/64 philosophy:
Expose for the shadows, develop for the highlights
Of course we're shooting digital, but the concept is the same. If I were shooting this on black and white film, I'd make sure that the shadowy area on the left side of the futon would hold some detail and then I'd process my film in such a way that I "stretched" the brightness values to their maximum at the right side of the futon, giving me a correct contrast range. Back in my zone system days I would have spot-metered the left, center, right sides of the futon and computed the contrast range, adjusting development to make sure it fit if necessary. In digital photography, if my image was too contrasty, I might shoot in RAW mode and adjust the contrast by choosing a slice out of the total image contrast-range - however, in this case, I was lucky and the whole image fit within the contrast range I expected my camera would handle effectively. So, my image, when I shoot it, is going to look just a touch under-exposed - and I'm going to "bump" it into the right range in photoshop. There's another option which I didn't consider - namely, letting the right hand side of the futon get overexposed into a white blob, and then add texture into it using the clone brush or healing brush or other photoshop tools. I didn't choose that approach because it takes a lot more time fiddling in photoshop for me to achieve a good result than to just expose the image where I want it and not have to spend a lot of time fixing it.

Since digital film is "cheap" I usually do a test shot and further check the exposure with my camera's built-in histogram display and my "mark I human eyeball light meter."


Doing the photos is the easy/fun part. Basically, I sat on a comfortable milk crate for 15 minutes while Annja stretched, yawned, rolled around, and so forth. At one point she announced "I'm bored" and pretended to be asleep. After a second she laughed and started to haul herself upright and I thought that her positioning looked good and mashed the button.

That's the raw image! As you can see, there is a pretty bright highlight on the right side of the futon, but it's not blown out. The left side is fairly dark; this is right in the range I expected when I computed my exposure.

There's an important stage in the process that I just glossed over! During the 15 minutes or so that I was shooting Annja in this lighting, I shot about 100 frames ("3 rolls of film") in various poses. Why did I pick this one? That's a hard question - I don't think I can answer it! "It looked right" is about as far as I can explain. How did I pick this one? I've gone into that topic a bit elsewhere, but basically, I looked in the file browser and looked first at the composition and shapes, and then selected a few images to look at more closely. I sometimes take days or months to pick an image to work on.

I still think of photography from the perspective of a film/darkroom worker. You shoot a few rolls of film, develop them, then let them dry. Later, you make contact sheets from the negatives, and select a few images from the contact sheet, see if they look OK, and make work-prints from those. I used to carry my work-prints around in a big folder and let them creep into my subconscious for days or weeks, and, when I had the time to really attack darkroom work, I'd spend a day printing fine prints of the best of the work-prints in my folder. That's pretty much how I approach digital, too - I simply cannot afford (in terms of time) to look at every pixel of every digital image that I've shot; a lot of my digital negatives go into the electronic equivalent of a book of negatives on a shelf. Perhaps I will look at them someday and go "wow! here's one I missed!" Several of my favorite photographers have written articles about the stuff they sometimes find in their piles of "old stuff" - how increasing skill and personal change/growth might encourage a photographer to go back and re-assess an old photo that lay fallow for years. They're completely right! Every so often I go back through my electronic archives and browse. Whenever I find something old that I see in a new way, it gets copied into my "working folder" for post-processing.


From here on I'll be discussion digital operations in photoshop. If photoshop workflows make your eyes glaze over with boredom, you need read no further.

Usually, my first processing step is to check my exposure using "Levels" to make sure that the image is not blown out or too dark.

This is the histogram from "Levels" on the original image. As you can see, the levels at the upper right (bright/white) end of the spectrum does not reach all the way to the edge of the histogram - this means that there was no "clipping" or blowing out of the highlights on the white sheet. The (dark/black) end of the spectrum also looks good; this image appears to have fallen within the exposure range, as intended. I consider it a "show stopper" if I pull the image's Levels up and it's obviously going to want a lot of adjustment; using Levels right away is a very important sanity check for me, so I don't spend any time getting excited about an image I plan to throw away eventually, anyhow.

If this were an image I intended to leave in color, I would use Levels to stretch the image's luminance values through the entire range, which would give it a bit more contrast and dimensionality. I'm planning to do a monochrome conversion, which will adjust those values; it's best to wait and work on the image with its original settings intact.

Next, I do a dust/scratch/hair cleanup. I try to work methodically, scrolling the image across and down so that I look at every screen-full of the picture for a little while. I usually use a mix of healing brush and clone brush for this stage of operation. This is also the stage where I remove moles or skin blemishes, etc., from my subject. Depending on the image and how closely I want to reflect reality, I also remove tattoos and piercings, etc. Cleaning up CCD dust bunnies is a real pain in the neck, since the effect of dust bunnies depends on the zoom factor at which you examine the image. In the example above, you can't see the dust bunnies because they're too small (unless you bring up the full-size version) but other times you'll get large splotches that don't show up if you're zoomed into the image, but are readily apparent when the image is viewed from a distance. I start zoomed in at 100% image size and do a clean-up pass, then go back and look at it at 25% where I'll see the larger stuff clearly.

When I'm done cleaning up an image, I always save a copy in case I want to second-guess myself about cropping or monochome conversion. Dust bunny hunting is really boring and I've found that I never do as good a job the second time I clean up an image as the first. So I take my time and do a good job on cleanup and then save a copy of the file as whatever####-cleanedup.jpg This is another reason why I avoid messing with levels or performing non-reversible alterations until later.

Now, it's time to start thinking of how to crop the picture. Cropping is another one of those mysterious things that I don't think about very closely. Cropping a photo can be done for lots of reasons, including:

  1. There is garbage in the image you don't want to clean up
  2. You want to "strengthen" the composition
  3. You want the image to be a predetermined size
  4. Some other reason

In this case, I cropped the image down to enhance its lengthiness and up to hide the foreground rucks in the paper sweep. That saved me having to fix them in photoshop and I like the way the image looks; the narrower framing seems to make it closer and more intimate, somehow. Of course that is all purely subjective; I've spent a lot of time talking with photographers who really sweat and worry about composition. I'm darned if I can tell the difference between carefully composed images and ones that are done haphazardly - the composition either works or it doesn't, as far as I am concerned.

Since I wanted this image to be black-and-white, it's time to bring up channel mixer and convert to grey-scale. There are lots of tools in photoshop, and a zillion plug-ins, that do black-and-white conversion for you. I use channel mixer because it's easy, free, and it gives me exactly the amount of creative control I want - and no more. In the example above, I've got the image up in channel mixer and checked the "Monochrome" option box. Channel mixer shows me a black-and-white version of the image based on rendering the selected mix of color channels as a grey-scale. You'll notice that the first color (Red) is 100% - which means that we're buidling our grey-scales totally out of the red channel. Again, it's time to be intuitive - I don't like the way that looks! So it's time to experiment. Usually I experiment with a few settings and then refine. If the 100% red doesn't immediately grab my attention, I try 33% Red, 33% Green, 33% Blue. Or 50% Green, 50% Blue. Go with your instincts!

When you're playing with channel mixer, you're also implicitly altering the image's levels, as if you'd gone into the different color channels in the Levels command and adjusted them. In theory, if you have 33%/33%/33% in a channel mix monochrome conversion, then you're not changing the image's levels at all. But if you were to do a 100% Blue in channel mixer, on a scene that didn't have much blue in it, you're effectively clipping off the top of the image's tonal range. Some photographers tell you "never go over a total of 100% in channel mixer" but, hey, it's your image - do what you want with it. If you take an image that's thin in the blue spectrum, and channel mix 200% blue, you're going to get the same effect (pretty much) as if you did the monochrome conversion and then brought the image up in levels.

Here's what the image looks like with the blue channel as the dominant color. I like this version a lot better!

One thing to remember: shadows are full of blue - the dominant color in shadows is more blue than anything else; so if you're doing a monochrome conversion with delicate shadows that you want to preserve, take a good look in the blue spectrum. Humans (who tend to be pink-ish in color) sometimes look really cool if you emphasize the red spectrum. Usually I force myself to fiddle with a couple different options in the channel mixer before I click OK. Then I usually like to take a look at its histogram and see how it looks. You can see that we lost a little bit of contrast during the conversion but nothing too major; that'll come right back either by adjusting levels or with shadow/highlight adjustment.

Here's our image after monochrome conversion.

I remain unsatisfied with it; it looks a bit too dark.

The shadow/highlight adjustment tool is one of the coolest things that was added as part of Photoshop CS. Basically, it allows you to separately adjust the tonal range of highlights separately from shadows. You can get the same kind of effects using curves, if you understand curves, but shadow/highlight makes it really easy. The "Amount" controls the degree to which the filter takes effect; you'll notice that I've left the highlights alone but boosted the contrast in the shadows a little bit. I'm happy with the image, now, since it's easier to discern the curve of Annja's hip from the grey of the sweep.

As a final bit of tweaking, I decided to try adding a softening/blur layer to the image. In a blur layer, what you're doing is overlaying the image with an out-of-focus version that you use to affect the underlying image. It's the same technique that I use in my "virtual enlarger diffusion" technique, except that here, instead of giving the image a black flare, we're going to give it a flare that pushes the image's contrast a bit more, and which smooths the skin textures.

First, we duplicate the background layer by bringing up the layers browser and right-clicking on the background layer in the browser, then choosing "Duplicate layer". Once the layer is duplicated, we then blur the duplicate using gaussian blur.

You can make this effect extremely subtle by using only a small amount of blurring. Here you can see the effects of the gaussian blur on a sampled part of the image. Since gaussian blur is based on pixel radii you need to adjust the level of blurring depending on the underlying size of the image.

Once you've blurred the layer, right click on it in the layers browser and choose "Blend Mode" - it will bring up a whole dialog box that lets you control how one layer affects the layers underneath it. There are all manner of tricks you can play with blend modes and layers, but let's stay focused on our mission here - we set the blending mode to "Soft Light" and then fiddle with the transparency of the blur layer. If you make the blur layer completely transparent, the effect is turned off. Care to guess what happens if you set the blend mode to "Soft Light" and then adjust the contrast or clipping of the layer with "Levels"? This is a very cool technique for performing detailed but subtle alterations on your images. Going in with an eraser and erasing the blur layer over the model's eyes, for example, will turn the effect off in just the areas that you erased.

As a final step, I flatten the layers and adjust the overall levels of the image. I guess it's the remains of my zone system indoctrination, but I enjoy seeing a nearly complete contrast range. You'll notice that, in spite of all the pounding I've done on this photo, the overall tonal range hasn't changed very much. Because I like my photos to look pretty realistic, I take it as a good sign that I haven't completely shifted the image's tones all over the place. That's a matter of personal preference, of course! I like my photos to be an "interpretation of reality" rather than a "departure from it."

If you look at the final photograph and compare it with where we started, you'll see that my "interpretation of reality" is substantially different from what my camera captured!

Dulles Airport, gate d5 and Denver Airport Gate B61 Jul 19, 2005