I like images that use natural light and added light. They seem harmonious.

This image of a radiologist was lit with the light of the screen in front of her and also by a small flash in the back of the room that bounced off the ceiling and boosted the overall illumination and added an accent light that separated her hair from the dark screen of the monitor behind her.  I used a Fuji S5 camera to make the photograph and I still marvel at the sharpness and dynamic range when I look at the full sized image.

I shot at f2.8 and, of course, the camera was on a tripod for the exposure.  The tripod often being the most important tool in creating good shots.

Lately I've become very interested in using small LED panels to take the place of flashes.  Part of the reason is the general compulsion to keep learning and to keep commercial photography interesting for me.  But another part of the equation is the belief that these light sources will be become the ubiquitous light sources of the future.  At some point flash might become the specialty tool and LED's the day-to-day lighting instrument of choice.  Maybe not, but there's no real cost for experimenting.

I did a portrait today and I lit it the same way I would have with flash or with tungsten but this morning I decided to use LED panels as my primary light source.  I set up a nine foot wide gray canvas background and lit it with two conjoined, small, battery powered LED panels.  Like these:  Little LED. They made a nice soft glow that surrounded my subject.

I used the six foot by six foot PhotoFlex light panel with diffusion that you've probably seen me use for Zach Scott portraits and other favorite work.  Over to my left and positioned at about a 45 degree angle to my subject.   It's no secret that I love huge, soft light sources. It's a beautiful way to light faces. Behind the large diffusion panel I used two of the ePhoto 500 LED panels.  The photo shoot was very successful but I learned some of the limitations that come with using inexpensive (read: not very color accurate) LED panels.  And I learned that the shortcomings are in no way insurmountable.

Seems that no matter what the distributors say there is a nice big spike in the green spectrum of the lights.  If you do a custom white balance you'll be pretty much okay but you might find some anomalies in the color balance that lead to a few splotches.  I shot my Canon DSLR in RAW so I was able to lower the saturation and increase the luminance of the green channels (and, to a certain extent, the blue channels) in order to compensate.  But here's what I learned through subsequent trial and error:  adding a 1/4 strength minus green gel filter does a reasonably good job compensating for the aberrant color spike.  The name of the game is get the light as close to daylight as possible.

If you don't have a color meter handy you can always set your camera white balance to daylight, shoot a white target, use the color eyedropper to correct to white and note the numbers you get in the Lightroom develop panel.  You're looking for two separate but related parameters.  You want to see how close to 5500 degrees kelvin the color temperature is and you'll want to note how many points of green or magenta have been dialed in to get a neutral white target.  You'll likely see a swing over to the magenta side of the scale which means you'll need to add some magenta to compensate.  If Lightroom indicates that it requires 30  points of magenta to render neutral white you'll probably have a filtration starting point of between 1/4 and 1/2 minus green filter.  That's actually a magenta filter that takes out green.

When you filter you're going to loose some power and that's important with LED panels.  They don't put out a tremendous amount of light and the light they do put out isn't collimated into efficient columns of focused light like you might find in a well designed tungsten fixture.  You may need to move the panels closer in to the diffusion material.  Don't worry.  Nothing will catch on fire.

So, why go through this exercise if you already have tons of great flash equipment that works well?  For one thing, the quality of continuous light is different than flash.  There's also.....no flash.  And that means fewer anticipatory flinches and blinks.  You get into a motordrive rhythm that's heavenly.  And with modern DSLR's you never need to stop.  There's ample light for focusing and the ambient light (after you've figured out the filter factor) makes nice fill light.  It's also new and different.  And for me that's enough.


jwolf said...

Hey, Kirk....enjoyed the article.....I was just wondering if there is any chance you will try out the new Oly E-5?

kirk tuck said...

jwolf, I still have some fun Oly lenses so why not. I'm really into my LEDs right now but........

Wolfgang Lonien said...


when I first saw that photo of the radiologist on your site, together with your explanations, it knocked me off my feet. She's such a beauty, and you photographed her in a perfect way, showing her at her workplace.

As for the LEDs, well exactly my experience. And I'm not sure if you can exactly meter them, because these spikes don't seem to be constant (except maybe like in "constantly changing"). Too difficult for me at the moment, but maybe yours have longer (slower) times for the spectral changes.

I've tried but found LED lighting look quite unnatural. Haven't tried any gels on them, because the only ones I have just fit my small flash.


Alan Fairley said...

Hey, Kirk, about that green spike. Could you not shoot a ColorChecker and use the Adobe Camera Raw Profile editor to create a profile specicially for that kind of light? That would automatically accomplish what you're doing with the color channel sliders.... Not sure how that would work with mixed source lighting, but you'd have the same problem using channel ajustments in that anyway (though you could mask off the areas needing adustment, I guess. But great to see an old dog learning new tricks!

Brian S said...

These guys must have realized the problem you mentioned because they send a minus green filter with the light.


kirk tuck said...

Alan, I suppose you could mask the area and make an adjustment. I do know how many mireds the adjustment needs to be but that only works if the LED's are the only source. Since most stuff takes place in mixed light it's more efficient to globally correct the lights themselves so they are a close match to daylight and go from there. Especially when shooting projects where a large number of images are needed. At some point correcting each image would become odious.

kirk tuck said...

Brian, I would also mix in an 1/8 CTO for a little more warmth.

e_dawg said...

Nothing more annoying than correcting for multiple light sources when the light sources are not continuous spectrum (fluorescent and LED).

But at least you can gel your light sources. Worst is when you're at an event and have to make do with mixed lighting sources indoors and (can neither gel nor ask someone to shut off the fluorescent / mercury / sodium vapour lights)...