Advertisement

Depicting a Famous Painting as a Portrait Shoot

by
Student iconAre you a student? Get a yearly Tuts+ subscription for $45 →

Have you ever looked at a painting or illustration and wondered, "how could I turn it into a photograph?" Well, in this article we're going to do just that. We'll study an illustration by J.C. Leyendecker and turn it into a photograph.

For this shoot we'll use pre-production, production, and post-production techniques you'd find in a major shoot.


Illustration Inspiration

An illustration Leyendecker did for Arrow Collars during the early 20th century.

So here is the illustration we're going to convert into a photograph. Aside from the obvious things - props, models, and wardrobe - the important thing to study is the lighting. The lighting will usually be the most challenging part because unlike painters who can just invent their own light with paint, photographers have to actually create it with a source.


Lighting

Simple sketches like this help make complex scenarios easier to manage.

Since we can't just paint a shadow or highlight into a photograph, we need to figure out where the light is and brainstorm how we'd recreate it.

We'll need to ask the following:

  • Where is the light coming from?
  • How many lights are there?
  • What is the quality of each light (harsh, diffused, etc.)?
  • What is the overall color of the light?

Firstly, we can determine where the light is coming from by the shadows in the illustration because shadows are always on the opposite side of the source of light. This is further helped by the highlights, especially contrasty ones because they give us better hints on directionality.

Secondly, once we've deciphered where the light is coming from, we need to decide just how many lighting sources would generate them. The first light is plainly obvious. The lamp on the table gives a fairly hard light upon the model to the left and an edge to the model on the right. The second light is coming from high and to the right, giving the right-facing surfaces their dues. However, based upon my experience, and the highlights on the lefthand model's hair, hand, and shoe, there is a third light. This final light brings the remaining surfaces into view.

Thirdly, we'll need to look at the quality of light each of these sources is producing. This will give us a clue as to what lighting modifiers we'll be using to recreate the look as well as begin to figure out our lighting ratios. The lamp light is semi-harsh as evidenced by the hard shadows on the left side model's face and the edging on the right model's profile. The upper right light is softer, but with directionality. And finally our left light is of a similar quality to that right-side light.

Lastly, we can be pretty sure that the scene is going to be a warm color because of the tungsten white balance of the lamp on the table. We also have to take into account that this is an old illustration and paper tends to yellow with age over time. So, it will be warm, but not as yellow as seen in the illustration. A few colored gels should help us achieve the look we're going for.


Caveats

With regards to the quality of light and how many lights exist in the illustration or how many would be needed to mimic it, that is partially a matter of taste. I'm taking into account Leyendecker's crisp lines in his shading and compensating for it because fabric is softer than that. And since I believe the scene to be a study or lounge area of some sort, I think the lighting should be softer than the hard sources his drawings suggest.

The goal here is not to robotically replicate a scene, but make into a photograph. It's fine to take some artistic license as well as consider gear, location, and other production limitations.


The Equipment

Here is where photography can be most challenging. For most of us who don't have big budgets to buy or rent all the gear and space we'd need to create this shot with ease, working with what we've got and doing a little DIY is perhaps our only reasonable option. Doing some creative problem-solving in pre-production and maximizing what gear you have has more to do with a shoot than having all the fancy cameras, lights, and grip you heart desires. Nearly every shoot requires you to be an astute problem-solver, especially if equipment fails or isn't available for some reason.

For photographic gear, I used what I already owned along with some household items to help shape my light. I then made a trip out to Home Depot and picked up a few items to substitute whatever I didn't have on-hand, but would be useful in the future:

  • Canon 1D Mark IV, on tripod
  • 24-70mm f/2.8 lens
  • (3) 580EX II flashes
  • (2) Einstein 640ws monobloc flashes
  • 46-inch Photek Softlighter II
  • 43-inch shoot-thru umbrella
  • 43-inch silver reflective unbrella (controls backspill from fill light)
  • Silver reflector
  • Copy paper
  • Aluminum foil
  • Cardboard gobo, 30x40-inches
  • (4) heavy-duty umbrella swivels
  • Super clamp
  • (4) light stands

This was a lot cheaper than renting a backdrop kit and a heck of a lot easier to carry.

Aside from the props (lamp, books, table, etc.) that's all the stuff outside the frame. For my background, I bought a large painter's canvas and to save a few more dollars, I used heavy duty painter's tape rather than gaffer's tape. Using a uniformly colored background allows for easier compositing. The heavy duty painter's tape was strong enough to hold up the canvas but not strip the paint off the walls - important if you're using rented/borrowed space.


Software & Workflow

For this shoot, I went a little fancy with my set up, shooting tethered to my laptop into Capture One Pro 6. Using Capture One or even shooting tethered isn't necessary, but being able to apply basic density and color adjustments as well as check for details as pictures come in is a great time-saver. I also like Capture One because it is a very robust RAW processor and outputs very nice, neutral, and flexible files for further refinement in Photoshop.

A reference print above Capture One Pro 6 helps make sure I'm working towards my goal. I also used it to show my models what I wanted them to do.

Mounting the camera to a tripod is also a good idea if there is a good chance of compositing images, or swapping backgrounds. Keeping your perspective the same makes taking pieces from different shots that you like and compositing much easier and faster.

Since I have a mixed bag of flashes and not enough Pocket Wizards, I had to use a mixed method of triggering them. This presented another challenge as the optical sensor on the Einsteins would be tripped whenever the 580EX II on my camera sent out a pre-flash to trigger the other 580EX II's, making the Einsteins fire just before the shutter opened.

To solve this, I used a sync cable to one Einstein and set the other to optical slave. I set my on-camera master to flash as well. However, I didn't want my on-camera flash to contribute to my exposure, so I set it to 1/128th power, way below what the camera would pick up. Now, all my lights will fire properly.

A meeting room area converted into a temporary studio.

How I triggered my flashes. The on-camera flash tripped the other 580EX II's and sync cable tripped the Einsteins.

I established my backdrop, props, camera settings, and lighting and had everything dialed in before my two models arrived. Being nearly 100% ready for the shot was important as they only had about 20 minutes to spare. So, I didn't have the time to run test shots while they were present. All they had to do was walk on, pose, and I took photos making only minute adjustments to their position and my lighting.

A lot of photographers will say something like, "I only had 5 minutes with this celebrity," but what they don't tell you is that they spent hours and hours setting up and testing to make sure those few moments were of the highest quality. This doesn't include the weeks or months of conceptualizing and planning before that. Good planning and preparation goes an extremely long way especially if you don't have all the gear and help you'd like.


Camera & Light Settings

I'm working well above the ambient because I don't want any shadows or highlights that weren't created by my strobes. I had to use a high f-stop, fairly high shutter speed, and low ISO to black out the bright sunlight spilling through the windows and backlighting my canvas backdrop. I then filled in that darkness with my lights, creating the look I was after without contamination.

These settings forced me to pump more power through my strobes in order to bring everything back into view, especially my 580EX II flashes. Since I was about 3 stops over ambient, I decided to use the powerful Einsteins as the main illuminators and the small flashes as accents. This helped keep the lights at a workable recycling speed.

It's important to notice how distance as well as the light modifier affects the light (quality and output) when you're limited on resources. This helps in the decision-making process about where you will place your lights, what modifiers to use, and which strobes you will use for which job.

For the camera, I chose a lens and focal length that was wide enough to capture the scene, but without distortion. On the 1.3x cropped sensor of the Mark IV, the 24-70mm lens experiences less edge distortion than it would on a full-frame body. A longer lens would have given me nicer compression, but would sacrifice depth of field (DOF). I didn't want to close-down my aperture further and then start taxing my 580EX flashes to a point where it would be 6 seconds for a full charge.

Here are the camera settings, chosen to work above the ambient as well as work within the limitations of my small flashes:

  • 1/160sec, f/8, @ ISO 100
  • 24-70mm lens, @ 52mm (equiv. 67mm)
  • Flash WB

Here are the flash settings I used:

  • Main light (Einstein) 1/4 power
  • Fill light (Einstein) 1/32 power
  • Lamp light (580EX) 1/2 +0.7 power, 1.5 CTO gel
  • Rim light (580EX) 1/8 power, 1.5 CTO gel, gridded

As you can see, the main light and the lamp light are doing most of the heavy lifting. I had to add +0.7 stops of power to the lamp light because the gels sucked away about that much power. The fill was to be invisible, lifting the shadows without having its own signature. The rim light was added mid-shoot as the lamp wasn't providing enough separation.

The 580EX II taped to the inside of the lamp. Paper and foil as reflectors. This is better than a bulb as I can precisely control the brightness.

I also added a reflector on the floor to bounce the main light onto the underside of the shoes. Again, a tiny, invisible kick.


The Shoot Itself

We now have all our lights, props, and equipment in place and dialed into almost exactly where we want it. If possible, use stand-ins wearing the same kind of clothing as your models to better judge how everything will act with people in the scene. I didn't have that luxury, so even with the models in the right pose, I had to shift lights around.

For example, the main light - an Einstein in a Photek Softlighter II - was spilling onto my background. This brightened it too much and would make cutting out the background later more difficult. I added the gobo (cardboard) blocking the light from my background.

The main light gobo-ed to keep it from spilling onto my background, making the illumination uneven.

Another little detail was making the wardrobe behave. Shirts naturally become untucked over time, producing a "muffin top", making even the slenderest models look wide. I did a little ad hoc tailoring with my heavy-duty painter's tape to eliminate the pouf.

The use of clothes pins, tape, or clamps is trick used on fashion shoots to make the clothes look right on the model for that particular shot.

Because I spent so much time planning, setting up, and testing the shoot itself took only 20 minutes and about 50 shots. To keep things going smoothly for that short time, I gave clear instructions and kept an eye on the details on my laptop such as someone blinking. The larger screen of the laptop allowed me to pick up on the little things that may not be easily seen on a 3-inch LCD.

I had spent about 5 hours in preparation, I wanted to get as much right in-camera so I wouldn't spend so much time in Photoshop retouching things I should have gotten done during the capture. It would be a lot of wasted time and effort to have "the shot" ruined because somebody blinked and I didn't notice.


Post-Production

Since I placed a lot of emphasis on getting as much of the image correct in the camera, post-processing the image was pretty simple and minimal. I shot tethered into Capture One Pro 6, applying custom WB, sharpening, levels to every shot I took as I took it.

Once that was done, I exported the file as a 16-bit TIFF in Adobe RGB color space into Photoshop. I wanted to maintain as much information as possible from the RAW capture while I worked on it. An 8-bit image would have fallen apart more quickly.

Although I used 18 layers, the layer that did the most dramatic work was the masking I did to swap out the background. Everything else was a lot of precise dodging & burning to different parts of the image. I used 6 different Curves adjustment layers to non-destructively dodge and burn as well as vignette my image.

Looks like a lot, but each adjustment layer gave me precise, non-destructive control.

The Curves adjustment layer is a really amazing tool because of the great amount of control it provides. Not only can you adjust the curve itself, but also utilize masking, blending modes, and opacity to fine tune the effects.

As I said earlier, because I got a lot of the image right in-camera, all those layers were simply fine-tuning and further shaping the light that was already there. It took about 3 hours, but it was efficient and not wasted on trying to fix earlier mistakes.


Lessons Learned

Speaking of mistakes, post-production didn't go as smoothly as it could have. Hindsight is always 20-20, but it is something I should have paid attention to a little more to save me even more time. The canvas background I chose was too close to the color of my model's hair for Photoshop CS5's Refine Mask tool to feather properly.

After several minutes of adjusting every slider in that tool, and even trying out the older channels method, I ended up using a 2px soft brush to manually mask the hairs properly. It wasn't difficult to do, but after hours of pre-production (starting at about 4:00am) I was a little tired.

So, the next time I need to swap out backgrounds, I will choose a background that will give me more contrast.

I also found a great, cheaper alternative to gaffer's tape: heavy duty painter's tape. Gaffer's tends to be about $18/roll. The tape I got was $8/roll. Hardware stores, particularly the larger ones are great places to find cheaper, but very effective alternatives to their photography-specific counterparts.

Finally, I got experience in doing a fashion shoot while learning to turn an illustration into a photograph. A lot of planning goes into something like a fashion or commercial shoot. These things aren't happenstance. It is also a lot of hard work when trying to do everything solo as I did. Having just one assistant would have saved me a ton of time during setup and light testing.


Conclusion

The result. Took me only 20 minutes ... plus 5 hours of setup and 3 hours of post. No biggie.

I hope you enjoyed this behind-the-scenes how-to and got some good ideas that could help improve your next idea, especially if it pertains to fashion or commercial work. If anyone says something took "5 minutes," it probably isn't entirely true. Sure pasta only takes a few minutes to cook, but waiting for the water to boil is another matter. This kind of photography is a lot like that.

Advertisement