DIY DVD-R Spectroscope – Looking at lights with a camera
Over the last couple of years, I have had the fortune and curse of shooting with a lot of different lighting gear, and in a range of situations for both video and still photography. This range of experiences has improved my instinct for the quality of light in front of me. I have used everything from broadcast, tungsten lighting and ultra high-end strobes, to highly touted LED light panels and off camera flash kits. I’ve shot video and stills in churches with stain-glass filtered light, in dark back rooms, and in all manner of mixed light. It’s been a great education, and has presented an enjoyable set of challenges for me to solve.
During all this shooting, I started getting curious about why all these lights are so challenging to master. They all seem to have their own quirks and problems, but in general, light behaves the same way no mater where it is, or where it is coming from. Strangely, many photographers become jaded toward one type of light or another, though. Why, I wonder? Young artist often claim to only shoot with “natural” light. One local TV cameraman I talked to swore off any sort of LED lighting as “too green.” Another expert wrote off one brand of strobes, because their trigger system is “crap.” I started thinking and paying attention.
I’ve had my share of fouled up footage and lost moments, but I can’t say I’ve become all that passionate one way or the other about any particular type of lighting. Each lighting system/source is meant for a particular type of job and provides its own quality. Since I don’t focus on one type of shooting, I guess I just try to adapt to what I have in front of me. Often, I manage to get great light even without that one perfectly designed light for the job. That is the part of photography I love.
Thinking about how much jaded information is out there, and all the bottled up expertise no one shares, I started to play around with my own ideas on lighting to see if I could develop a better, simpler approach toward lighting technology. I decided to see if I could adapt some non-photographic LEDs and lighting setups into something predictable for the camera. (Here is a video from one of my first experiments.) It raised more questions than solved problems.
There are a few variables I consider in lighting sources after the choice between continuous or strobe (for obvious reasons):
Is it affordable?
Since I don’t own a studio, nor shoot studio style work more than a few times a year, I can’t justify buying ProFoto’s or Arri’s for many thousands of dollars. I need tools that will pay off in low numbers of shoots.
Is is portable?
Again, I don’t own a studio, so any equipment I own needs to be tucked away, and portable enough for me to setup shop in a living or conference room. In addition, if it won’t be plugged in, I need it to run on small batteries.
Is is controllable?
This point is a little more broad, but lights need to be flexible in terms of modification, intensity, and placement. A flash kit does me wonders in small rooms using a couple of umbrellas and maybe a gel. But, for example, if I shoot formals at a wedding with thirty people, I’m going to need some mono-lights or a pack of some sort to light up a big area. That’s when I lug in my big cases, and/or rent a professional light pack and kit.
Likewise if I am doing an interview, I can get away with plugging in some LED or other bulb type “hotlights,” then bouncing the light around with modifiers. If, on the other hand, I needed to light a whole set in detail, I would need something bigger like a big set of HMI lights. That situation hasn’t come up for me, (since I don’t shoot music videos, yet) but I have had a chance to use that sort of setup. With digital control boards, massive power needs, and truck loads of gear, that sort of setup enters a whole other ball game. I am glad to have been able to shoot a bit in a broadcast studio, but those situations remain out of my league for the present, nor really capture my interest.
My experimental follow up
After utilizing some LED spot-lights as modifiable, continuous lighting for interviews, and quick video productions, I tried to answer the question of how a camera sees the color of these lights. I looked up how to build a simple spectroscope out of a DVD disc, and started looking around. By no means is what I present here scientific, or even remotely accurate, but it does demonstrate the variations in light colors.
Here are the spectrographs as my camera saw them:
Each image shows the actual image at bottom, and the channel histogram on top. The right side of the histogram shows the more intense light, the left is less intense. The peaks demonstrate color spread (e.g. a sharp blue peak on the right side demonstrates a lot of pure blue light in the image.). I shot these with my Canon SLR pointed at the DVD disc tilted at 60 degrees, and light passing through a sixteenth of an inch slit. To process them, I fixed all images at a temperature shift of 7000K which is minus a lot of blue, and a green/magenta shift of 70/150 which is 25% toward magenta. I left the camera raw calibration at Adobe Standard, which is wrong for shooting, but I figured would work as a benchmark as long as it was consistent. Finally, I set all color and contrast adjustments to their zero/defaults and only adjusted the “exposure” enough to fill the channel histogram to the range offered by the software.
I think these images are informative about two things:
1. An SLR sensor seems to favor green over what the eye sees. I saw much more red and purple in the spectrum when I looked through the spectroscope. The camera always captured the most green.
2. The quality of light coming from various light sources isn’t as shifted or complicated as I might have guessed. The only light source that shows a clear split spectrum was the CFL.
Looking at the histograms, it is hard to know how balanced each light sources is, but it is clear that the flash is very close to the sunlight spectrum, and most of the others aren’t too bad, either. I think I would only be concerned with a light like the CFL, which is missing all sorts of colors, and a “warm” type light which is missing a lot of blue.
My goal with this wasn’t to capture a raw spectrum from each light source, but more-or-less get a sense of how the light might get to the camera. I did things like point the spectroscope through the glass in light fixtures, and leave windows open, and other lights on. In the case of the sun, I bounced the light off a white card, since direct sunlight entering a camera is rare. Also, the spectroscope-camera combination seems to favor greens heavily (which matches the technical descriptions from manufactures). To my eye, as opposed to the image form the camera, each spectrum was subtle and contained a lot of differences in the red end and the blue end. This shows how heavily digital cameras process the light that hits the sensor. I, as a photographer, imagined that I was getting mostly what I saw through the view finder, but the reality is digital images are not very close to what I see with my eyes (a quick discussion of the topic). Post-processing, then becomes that much more important to get all those color perceptions back. Having an adequate light source is only the first step.
I think what I have learned is that most lighting setups work OK considering how the camera is getting color out of a scene. When I notice that something is off in a light’s quality, it is probably so far shifted, the camera won’t be able to compensate for it. Those situations cause problems for photographers. Daylight bulbs seem to give my cameras, at least, enough color information to create good balanced images and videos. If I can see a color shift, particularly in the greens and blues, it is probably time to find a different light. That’s good info to have.
I’ve gained a little more understanding of human-made light with this experiment. That is important in photography. It is a technical craft, that when controlled and mastered opens up doors to artistry and expression. NOt getting all the tools working to their fullest will hinder that goal. It also gives me some direction in my cheap DIY lighting explorations. I like that.