Power LED's for Photographic Lighting (redux)

Other Hardware General Electronics
1

I decided that the new board needed a new discussion on the topic.. another bit of commercial/goof around shooting done last night, this time in my kitchen. Once again, a small "recessed lighting" fixture with a Warm White (2700k) 10w CREE LED was instrumental in getting some of the best shots.

Up until I got this particular LED flood (which was purchased for $20 from Home Depot), I had been disappointed quite severely with the actual usefulness of LED's for photography- the monochromatic, very cold "white" LED's that have been the market until recently are only useful under very limited conditions- Macro photography, etc.. and even then, the light was inferior to halogen, only more usable because LED's don't heat up like a Halogen.

That being said, LED's have a lot of advantages if they can be made to work well- the first being power consumption. Though the fixture I'm using is 110vAC, of course it's just a 3.4v LED inside, pulling 3A. That's a power level EASILY provided by portable battery sources for hours.. and throwing 600ish Lumens of light that shoots very much like a 75-100w halogen.. except (bringing us to benefit #2) that the whole thing is cool to the touch, even after an hour of continuous use. I can place this flood INCHES from a face, and it's not uncomfortable to the model.. a very nice diffuser is part of this bulb setup. Benefit #3 and #4 are combined.. simply put, the cord and socket weigh as much as the bulb, making a total of about three pounds. Each one of the halogen heads weighs about 12 pounds, and is fragile. The LED was dropped numerous times, because it's light enough to be handheld (number three).. No filament to break... benefit number four.

The shots below were lit 45 degrees off shooting angle (more or less) on both sides with 70w of studio halogen through white nylon bounce diffusers from about three feet on the right and five on the left. All the direction and accent lighting is being done with that 10w CREE. As soon as a little spare cash can be rustled up, I want to pick up four (minimum) of the power LED's, to see how well they can be used to replace the heavy and fragile halogens completely. The CREE was being held by the hair/makeup artist (John Manning.. a great friend and one of the best on the east coast of the USA.. trust me, it's not the camera work as much as it is his work.. I'm a passable photographer, but he'll be doing Elton John's makeup next month pre-show here in Boston. His portfolio includes Cheech and Chong's recent tour, GaGa's Girls (Lady GaGa's backup dancers), and a disturbing cross section of pretty much anyone good looking from 1990 on. I've already pleaded and begged.. not a chance that I'll even be able to get a backstage pass, much less meet or actually take shots of Sir Elton, though..)

I'm of the belief that LED lighting is about to become very important to professional photography, and we may be able to come up with some interesting ways to play with these new power LEDS that outdo any other light source, and quite soon, technology wise.. as a hobbyist in both electronics and photography, I see great things on the horizon.

All shots taken with a Canon EOS Rebel XT, a Canon f1.8 50mm Prime series II ("Nifty Fifty", the $99 plastic one!) - these shots have not been color balanced or edited, only resized to post... the color rendition is VERY workable... this is right off the memory card.

The second two are of Heather, but the first is Mrs. Focalist.. never forget which side your butter is breaded on..

Since we have a decent population of photography enthusiasts around here, let's see what you've done with photographic LED lighting.. or ideas for their use.. Under current consideration is building a high-speed, short-duration strobe for stop motion use.. LED's are capable of emission times in the nanosecond range..

k1LED.JPG
k1LED.JPG640×960 155 KB

h1led.JPG
h1led.JPG640×960 140 KB

h2LED.JPG
h2LED.JPG640×960 139 KB

2

What is the Arduino related question? Do you want to build a Arduino controlled Flash Light? Arduino has a clock frequency of 16Mhz (~60 nsec) so it is more in the micro-second range than in the nanosecond range but micro or even milli may be fast enough to get some nice effects.

Best to get started with this one - http://arduino.cc/en/Tutorial/BlinkWithoutDelay - as similar logic makes it possible to switch on/off multiple leds with different timings, e.g. 10 msec warm white followed by a 1 msec Yellow etc. LIght intensity can be controlled by PWM - http://arduino.cc/en/Tutorial/Fading

Hopes this helps,

3

Not really a question per se.. more of a conversation starter. From a photo perspective, LED lighting presents a lot of challenges but some potential huge rewards.. and the key lies in their control. Control, LED... hmm, this sounds like a job for Arduino!

One possible direction for the discussion is use of LED's for high-speed MCU-triggered stroboscopic effects- whether it's a single pulse (in a more traditional photographic flash type setup) or rapid timed succession of pulses.. or, as you point out, PWM-dimmed. The shots are really the first step, in that the light must be usable in spectrum and intensity before any other design time is considered.. and it appears technology now has brought us to the "reasonable color rendition" point where we can start considering the technology for actual use.

Stop motion is one possible use- Xenon tubes actually fall down in terms of flash duration (too long) for stopping a bullet mid-flight, for example. In theory, an LED is not. LED's may allow for better stop motion shots that any xenon strobe, and without having to fiddle with potentially lethal capacitor charges, and more cheaply and durably.

Photography begs for Power LED technology.. and Power LED technology begs to be Arduino-ized at every turn... be it color mixing, strobe effects, or precision timing..

4

Photography begs for Power LED technology.. and Power LED technology begs to be Arduino-ized at every turn... be it color mixing, strobe effects, or precision timing..

Well unless there is some kind of feedback coming back to the Arduino in the form of color and/or intensity and/or duration sensors, the Arduino's function in this kind of application is kind of trivial and boring.

Lefty

5

In some respects, you are very correct- Arduino can do some amazing things- but what's not appreciated if you are not a photographer is that simple functions cost thousands of dollars when it comes to photo equipment. My first Arduino (adafruit BoArduino) is now part of a frankencontroller and intervalometer for photographic use, also incorporating audio and light triggering as well as strobe and shutter control systems.. similar in some ways to the Camera Axe project elsewhere on the site. If I were to set out to purchase hardware to accomplish what I've done with Arduino (simple as it may be), it would cost several thousand dollars- yet it's all been built from the ground up for under a hundred.. and best of all, I learned a lot doing it.

Though I am still early in the design stages, one goal is to use Arduino as a full lighting-systems controller. Even the simple ability to store and recall dimming levels on a series of lights, to be able to rapidly load and switch lighting balance groupings- that type of functionality costs thousands where we as Arduino folk can design and build such a device for pennies on the dollar- and have fun and learn while doing it, to boot. For portrait work for example, even having fading control with seven channels with memory (two each side, top, bottom, and backlight) would be a godsend. For an amatuer on a budget, like me, Arduino opens up functions that although "trivial" are hardly boring by any stretch of the imagination. Actually, Arduino is a perfect platform for this type of system- plenty of horsepower for the buck, simple, and open source.

Deeper and more complex is the feedback section, as you point out. A thought I've been mulling over is using an infrared laser (or even just a narrow emitter infrared LED) to specify the lighting balance points and have the lighting system be able to rejigger the balance simply by pointing the laser at the desired light balance point(s).. and we're still within Arduino-being-overkill design level. Obtaining relative location on three axises can be done pretty painlessly, two phototransistors per axis, your distance is the ratio between the axis sensors and not actually raw intensity (at least in my mental design of the beast.. we'll see if life imitates art trying to imitate life)..Having a smart controller in the middle of it all, that Arduino- makes intelligent lighting possible.

LED light sources only make this all the more viable and useful.. the projects may not be as complex as OpenCV or something that's never been done before, but they may put creative tools into the hands of people who may not otherwise ever be able to afford the functions, if you have to buy the pro equipment to do it...

6

A 10W Cree, eh? I shall have to be on the lookout for such an item. I didn’t realise they went up that high. I’ve also been looking at Philips units, particularly interested in colour rendition, but they cost too much.

One interesting application of such LEDs as you’re using is that they could be used to build a ‘bank’ of lights, so that the effect is of a large light area, but as you say, quite close to the subject, only just out of shot. That way it’d appear like a large diffuse light source, like a softbox.

Another interesting aspect is, as you say, they’re so small and light – low mass. Could you imagine a sort of arduino-controlled robot arm type of arrangement on top of the lighting stand, and possibly about three or four of them, which could be used to position the lights from the camera position? Even during a sequence of shots?

7

Exactly. The halogen lighting is actually not that different at 70w at a yard away, vs 10w LED inches out of frame, as in these shots. Halogen has so many drawbacks, only made worse by need for additional strobe lighting- I just see these LED's beginning to make a real inroad. Now DMX protocol already exists, but to be honest I'm not sure it's what I'd use.. basically need to research it a bit more. DMX is a protocol designed for lighting systems control, and sticking to a standard might be a useful idea- but no need to, if it would be more cumbersome than useful. I get the idea that DMX would be better implemented on a single smart fixture as part of a controller system (commercial) rather than a purpose-built system like I envision. Like the idea of an "active" RGB wash/fade DURING a shot? Can't see a reason it couldn't be done pretty cheaply.. best of all, done use a level of power that makes a portable setup a realistic idea. Ideas abound.


10 Watt CREE Warm White LED recessed can fixture bulb.. $20 from Home Depot. I just use it with a recycled desk lamp cord and socket most of the time. The white "shell" is opaque plastic, a plastic diffuser sits an inch or so in front of the LED, it provides amazingly smooth light, nearly softbox smooth. There's not a very discernable hot spot as with Halogen, even with no other diffusion, held just out of frame. The bulb base and aluminum heatsink are most of the weight.. for most purposes, I'd be inclined to leave the shell and diffuser in place. The "step" in the shell is the location of the diffuser, if anything, taking a Dremel to the rest might be a decent idea if you didn't want extra plastic in the way. They are sold as "EcoBulb" brand, right next to the insanely expensive Phillips jobs (which do have major hotspot problems and cost three times as much). I've not used it with any type of snoot or lightguides- but given the low heat and design, at a minimum, hair and background lighting with these ought to be a dream compared to portable halogens..at a fraction of the price.

I would expect that if using short-throw lights, a whole "arm" might not be needed, a lot could be done with tilt/pan brackets, which can be made from standard hobby servos and controlled via Arduino. I am imagining and designing a frame, to keep support out of shot as much as possible- probably make it from PVC pipe to make it easy, light, and cheap. A multiple DOF arm would be awesome.. but I'm sure also immensely more difficult... but who knows? At the moment, I'm using a commercially-packaged 10 watter, but a quick check shows CREE sells emitters into the HUNDREDS of watts now- energy levels that start to reach commercial strobe bank output levels- a tight angle emitter tossing a few tens of thousands of lumens might just make Xenon obsolete. CREE and others are now making emitters with amazing CRI, and being able to actively tweak spectrum with some lower power RBG elements would make them darn near perfect for close-range work like portraits. OR.. how about your "modeling lamp" being nothing more than PWM-tamed LED? Your "strobe" is merely letting it go full power.. and maintains EXACTLY the same color balance as the modeling lamp. Actually seeing what you're going to get... might be useful, in fact, skip the "might". In fact, since you are really only fooling the eye by using PWM, a secondary strobe is unnecessary if the shutter timing can be synced... We don't see the way the camera does, and this is one of those places where difference between an eye and a camera may be pivotal.

I can even envision some very interesting effects that could be done with a swept-lit ring light.......and nothing at all needs to move...