Mini Old School Class on White Background.

Ad shot from the 1990's.  A Quick and clean revisiting of White Background Lighting Techniques from the film days. (ie: before you could just cut stuff out in PhotoShop).

If there's one subject that comes up again and again on lighting forums it is: "What is the correct way to light and prep for a white drop out background?"  Time and time again the bold rush in to suggest everything from blasting the background with one huge flash to shooting against black and just cutting it out with Gimp Tools XP10.  But there is actually a method that used to be taught in all the photo schools or learned at the feet of the guys who did it before you.  And it made sense back then.  

I thought I'd put this up for three reasons:  I like the image of the skater.  I was playing with the pen tool I've had for ten years and remembered how much I like to scrawl things across photos.  And finally, a younger photographer, who will have to do many, many of these kinds of shots asked me to.

In the shot above I'm starting by rolling out a nine foot wide seamless backdrop of Super White paper.  I roll it so that the front end (right behind the skater) is at least 12 feet from the plane of the hanging paper in the far background.  This allows me put an even light on all of the background but leaves me enough room to scrim the background light off the skater with black panels.

I'm using five lights on the background.  There are two flash heads on either side of the set and all four of them have white umbrellas with black backings on them.  They are aimed at the opposite sides of the seamless so the light feathers across the surface.  I've also added a center light, high up to both clean up the middle of the background and also provide additional light (via careful feathering) to the paper that's spread across the floor behind our model.  I always attempt to light the backgrounds so that every surface that shows to the camera is within a 1/3 stop of everything else.  Having too much light in a spot is just as bad as not having enough when you don't have the luxury of hands-on post processing.

In front of the background paper I've placed four sturdy milk carton holders to elevate my model's platform.  I've placed a stout piece of plywood under the shiny, white Pleixglas for support.  Having the platform raised means that, at the angle I want to shoot my dancer and the focal length I'll use, I'm actually seeing the far end of the white paper at her foot level instead of the part of the paper nearest the camera.  This is pure white and gives me a great reflection back on the shiny plexiglas surface which works to obliterate detail and go as white as possible.  It also gives me a nice reflection of the skater's blades and shoes right in front of her.

The one boom arm you see coming into the frame on the left hand side is tightly secured to a solid tether and it there so the skater can reach out and steady herself if she starts to lose her balance.

I'm using a large (54 by 72 inch) Chimera softbox from about 45 degrees to the left, in front of the skater and just far enough away so that it won't show in camera.  The only other source of illumination is a white fill reflector from the opposite side.

I meter (with an incident meter) the background and get a reading.  In this case it was probably f11 and 1/3 stops.  I then meter while adjusting the distance of the softbox to the model until I get a reading that's one third of a stop darker on the model.  In this way I am assured that the background will go pure white but I'm equally assured that the least amount of light will spill forward from the background lighting to contaminate (and lower the contrast of) the model at the front of the set.  In this situation all we left for the color separator was to clean up the area around the model's feet.

Even in the zenith of our digital days I can think of several reasons (all lighting and lens oriented) to maintain the same lighting practice.  Less spill means less veiling flare.  And, as I've written, less unwanted contamination on the subject.  The even-ness of the background means that, even though you will be using the selection tool in Photoshop you'll have less issues to deal with and will spend less time with "refine edge."  Finally, if you get used to doing it correctly you'll see that it also works just as well for video.  And it's easier than "green screening" everything and fixing it all up in post...

I won't get into the argument about "incident meter versus reflected" or exactly how to hold your light meter.  I think that's too personal to talk about on a public blog.  And I do think you'll figure out your own technique.  After all, you can see the results right away now.  But can you see 1/3 stop above white, on your camera's histogram?  Meters are still relevant.

We took the day off today.  Back tomorrow to discuss portfolios for the new age of screen dominance.  Maybe.


Simon Lupton said...

Thanks Kirk.

I've never been much into taking portraits but I think you're starting to get to me.

I'm all for getting it right in camera so any time you feel like showing off 'old school' stuff like this just go right ahead.

Patrick Green said...


Interesting article.

Could you explain why when you take a picture of a scene containing something white, such as a building, the picture show the something as white. Yet, when you take a studio picture of a person or an object with a white background, the white background comes out grey unless you light it with several lights and overexpose compared to the object.

Many thanks

kirk tuck said...

Patrick, when shooting outside the same quantity of light hits everything equally (unless some object is shaded or sits in different light). The regular range of tones is easily reproduced. In most available light photography the goal is to have some texture or tone left in the white areas. Otherwise you'll have "blown" highlights.

In the studio you are starting with a blank slate. If you light an object in the foreground with a light that's in front of the object the strength of the light falls off as it travel (inverse square law). So a big umbrella light illuminating a person five feet away will have one quarter the amount of light on a background ten feet away. To get the light to be equal between the foreground and background you'll have to bring up the level on the background to match the foreground. The common way to do that is to light the background separately.

If you want some detail in the white areas (like you would in a typical scene) you are trying to match light levels between the foreground and the background.

But the goal of a "white out" background is to effectively take out all of the tone and detail in the white background in order to cut out the main subject from the background. To do this you need to make the background at least 1/3 brighter than pure white, hence the need to overexpose (slightly) in comparison with the foreground subject.

Patrick Green said...

Kirk, thanks for the explanation.

Gregg Mack said...

Kirk, this is an excellent "how to" blog!
That white reflector on the right seems to be doing a fabulous job of bouncing back a lot of fill light. Do you remember how big that white reflector was? Did you aim the big 54" x 72" Chimera softbox directly at it?

Debbi_in_California said...

I love to see the 'old school' methods. They work today too! Thanks for sharing this!

Anonymous said...

Could you clarify how you are measuring the +1/3 overexposure? Incident facing camera vs incident facing background vs reflective?

Kirk Tuck said...

Incident dome always faces camera!