No announcement yet.

OT white leds for camera light

  • Filter
  • Time
  • Show
Clear All
new posts

  • OT white leds for camera light

    I have read some about using leds for camera lighting. I don't mean to light up the camera, but the scene to be shot . I'm wondering about the color temperature, etc, and whether this is a good source of light to enable a typical camera to take natural colored pictures. My camera can take a decent indoor pic, but with a greenish yellow cast to it that is not completely correctable through the adjustments. I want to use some of those white leds to build an illuminator.

    I see on Ebay that some are of extremely high brightness and can be had for mere pennies per, in packs of 50 or more. One could almost literally blast out your retinas at 30 paces with some of these and a couple button cells.

    That's not my intention of course, but 140,000 mcd per led- that got my attention. One package of these and one could strobe light a small town .

    Anyway, the question has to do with helping a digital camera to deliver accurate color, etc. Any comments?
    I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-

  • #2

    This may very well work. I have seen medical cameras that used LEDs for their illumination source.

    As for maintaining true colors, that may be the problem. White LEDs use a combination of several color LEDs, usually red, blue, and green. However, LEDs are more or less single wavelength sources so you have a "white" light that actually consists of only three wavelengths. Real white light consists of a continous range of wavelengths from infared to ultraviolet and actually has components beyond the visible range on both ends of the spectrum. Different objects will reflect different wavelengths of each area of the spectrum differently so a bright red object may not reflect the red of the LED as well as it "should" to produce that true color. Likewise for the blue and green areas. Also, the pickup elements in the camera will each have their own spectral sensitivities and may not corrospond to the colors/wavelengths of the LEDs. Colors that contain light in two or all three of these primary colors will be even tricker. When these two effects are combined the results may be acceptable for most situations but may also produce some surprises (off colors) in others. Filters may help. They could be on the LEDs or on the camera. You may have to experiment.

    Sounds like fun. Let us know how it turns out.
    Paul A.
    SE Texas

    Make it fit.
    You can't win and there is a penalty for trying!


    • #3
      White Leds

      The white LEDs I have seen the most are UV with a phosphor, very much like a fluorescent light.

      They can be very bright. When I tested the boards with 15 LEDs on each board I use a sheet of paper to cut the brightness.



      • #4
        I've used the white led's with a digital camera and because they have a lot of blue color in them it dramatizes metals except for the yellow ones. Great against a black backdrop or a flat white textureless backdrop (like the "I'm a Mac... I'm a PC" commercials". They're not worth a crap for people pix and most natural objects unless you augment it with some incandescent lamps. Ace Hardware offered some mini tripods with white led lamps in them that are ideal for small item photography and I'd buy 5 or more if they had the damn things in stock! but they don't.

        Anyway - if you use real film you're going to have to play with the colors in the studio as they're going to be like using Kodacolor or Ektachrome film with florescent lamps - everything's going to be seagreen.

        Edit: Found the advert for the LED desktop tripod - it's from Stanley for $19.99.
        Last edited by dp; 01-01-2007, 01:06 AM.


        • #5
          I have read some about using leds for camera lighting. I don't mean to light up the camera, but the scene to be shot
          I use white LEDs for key lighting and highlighting in a lot of my shots.

          In this pic I used a LED flashlight to highlight the inside of the spindle bore.

          This shot is key lighted with a 14 LED small battery powered gooseneck lamp that is really easy to adjust as needed.

          This shot uses a LED key light and fill light.

          And, finally, this scene was illuminated by a single white LED flashlight that I waved about during the 30 second time exposure.

          Last edited by Evan; 01-01-2007, 01:56 AM.
          Free software for calculating bolt circles and similar: Click Here


          • #6
            When I saw this thread, I knew Evan would have something interesting to contribute. Nice photos - I'm going to try using white LEDs to illuminate machinery pics now.

            The curse of having precise measuring tools is being able to actually see how imperfect everything is.


            • #7
              Note that those pics of metalwork are not illuminated solely by LED lighting. I use the normal daylight fluorescent lights in my shop as the main sources. The LEDs serve to bring out details and to fill in shadows.
              Free software for calculating bolt circles and similar: Click Here


              • #8
                Evan, you certainly do some fine picture taking. I like the way you used led light to accent your pictures. The night scene is pretty cool.

                My use for the camera inside would be mostly in the workshop. As such it would be more important to get a clearly detailed and contrasted shot than absolute color correctness. As far as people pics, maybe the daylight screw-in fluorescents would give a decent result- ?

                There's still the machinery lighting using leds to consider. With the increased brightness (and in some cases whiteness) of leds these days, it should be more suitable now than it has ever been.

                Something else I've thought about recently is to provide lighting for my 'slot car' room using leds, then modulate that lighting with control signals for the cars. You wouldn't see any fluctuations in light level, but the cars' sensors would pick it up. Only problem with this idea so far is that I can't seem to find out how fast these bright white leds respond. I think I should be able to get a cleanly readable signal out of the lighting system up to 250 khz or more, enough to carry two channels of R/C signals. Then again, visible light sensors are not as sensitive as IR devices, so those might still be the best choice.

                Ah, too many potential projects- still many old ones to finish up first.
                I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-


                • #9
                  Darryl, do you have any pics of your slot car room? What kind of cars and what scale do you have? I used to race about 15 years ago at a local hobby shop but the owner sold the track and I didn't have anywhere to race. I really liked it. Lots of fun and tunability but not near the cost of rc cars.
                  Jonathan P.


                  • #10
                    Has anybody here ever made a ring light using LED's? I would think this would make a good electrical/Mechanical project.

                    Here's a website detailing on how to build one..

                    EGO partum , proinde EGO sum


                    • #11
                      No, no pics of the room- it's still an attic space needing flooring, walls, and insulation. There's a room now to stand up in, an area for the pits, and the length of the house for the track. That part is about 8 ft x 36 ft. I plan to have a dragstrip along one side for the full length, which is 50 ft, and then a layout with at least one large banked turn, probably at the far end so you can still see the cars. There will be hazards, hill climbs, and shorter straights, with some esses. The drag strip will be part of the loop so that will be one long straight.

                      Something like that anyway. I used to be into 1/24 scale (strombecker?) when I was younger, and then had some of the 1/32 scale setups. I made some of my own cars, sans body, with one being a dragster. I hope to be able to ressurect this as a hobby, which will also give me time in the workshop making parts for it.

                      I was barely getting into making my own track, complete with slots and stick-on copper foil, when I kind of gave it up for partying, etc. Late teens, what else is there to say about that Now what I want is to have a slotless setup, self-powered cars probably using AA nimh batteries, and some kind of assisted steering. By that I mean that the cars will be able to track around the layout by sensing a 'center of lane' buried strip, actually a few strips. Then the user can coax the car to change lanes by servo controlled steering. The car will pick up the nearest buried strip and still steer by servo, but by following the sensor strip. As the driver, you would be able to nudge left or right, or with a bit more steering input you would cause the car to 'catch' an adjacent guide strip and then follow that. You could tuck in tight around corners, or go wide- the front tires would actually carry the car and steer it, unlike the slot car.

                      This gets me away from having to constantly steer by remote, which is very tedious and very inexact. It would be no fun if you either couldn't make it around the track, or had to go so slowly around that it wouldn't be racing. With my proposed system, you could change lanes at any time, anywhere basically, and you would stay on track- subject to your skill as the driver, same as the traditional slot car.

                      I currently have those x-mods from radio shack, guessing that they are somewhere between 1/32 and 1/24th scale. I'm playing with them to determine how tight a radius of corner I can navigate so I can design a realistic (pun) track layout that is driveable.

                      My floor area will probably be done in aspenite (14 sheets or so), then painted with a deck coating. The track itself would be a further few layers of deck coating that is blackened with something. Other areas, like grass or dirt, would be defined by including the appropriate fake grass or dirt in the top layer of coating. There were some pretty elaborate setups years ago, including buildings, overpasses, street lights, etc - I'd like to eventually build that kind of thing up.

                      Anyway, I didn't intend to get into a long winded description of my plans, but there it is. I'd like to control the cars by infrared instead of radio, so as to minimize any external interferences, both to me and from me, thus my wondering about using led lighting to both give light and send control signals for the cars. I have considerable homework to do for many aspects of this project, including cost. That alone could be a shocker if I assess it properly.

                      Anybody want to suggest a method of buried sensor strips and sensors? My best idea so far is buried steel wire, and using a swinging magnet arm on the car- balanced of course- to hold position over any particular wire. That will be the subject of an experiment here, soon.
                      I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-


                      • #12
                        Yeah, the ring light. That looks like a good idea. You could use the brightest currently available leds, and make a dimmer circuit for it.

                        You could add more leds- one or more separate circuits could have colored leds for wild effects. You might even have red, green, and blue leds on three separate circuits, with the white on a fourth circuit. They could all be dimmable separately, thus giving one a way to adjust the color temperature of the final light output-

                        Yikes! No more projects! Too many already!
                        I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-


                        • #13
                          White LEDs will have a very poor response time, probably well under 100 hz. All the high intensity ones I have seen are the UV/phosphor variety and the phosphor cannot respond quickly.

                          Not a real problem anyway as a number of ultrabright infrared LEDs will do nicely.

                          White LEDs still aren't adequate for area lighting and won't be sufficient for the size of area you want to illuminate. Regular fluorescents would probably be best. You can get some pretty nice compact fluorescent flood lights now that should do nicely.

                          For following I would run a pair of wires with a spacing about the same as old TV twinlead. Through each side feed a low impeadance audio signal, each a different frequency. Each wire will act as a magnetic field radiating antenna. Use a pickup coil like the old suction cup telephone pickups for a tape recorder and amplify the output with a fet op amp. Feed that to a dual op amp set up as a first order bandpass filter for each frequency and compare the outputs in a comparator. Use that as a steering signal.

                          You could steer or cause a lane change either way by altering the strength of the audio signal in each wire. Given the length of the wires and the effective resistance an ordinary audio amplifier should be able to drive the system just fine. You could even steer using a remote control with a channel balance function.
                          Free software for calculating bolt circles and similar: Click Here


                          • #14
                            This month's Nuts and Volts (just came, so probably Jan 07) has a cover project making a ring barrel mount white LED flash for a digital camera. Looks like a fairly straight forward project.


                            • #15
                              Yeah, I suppose I got carried away with the led idea. I'll use the lighting that's in there now until I get to changing it. That's three strings of clear christmas lights. 420 watts worth.

                              Per the white leds response time- Ok, far too slow to send data. IR leds it will be, and enough to cover the entire area. No big deal.

                              I like the idea of electrifying wires and having a comparator output a steering signal. I wonder about the potential for interferences, particularly because of the motor current. I can minimize that by careful routing of high current wiring, but the coil is going to pick up everything. It might work fine, though. Thanks for the idea, Evan.

                              I did a test with a 1/4 inch diameter round nib magnet and some steel wire. I can get the magnet to follow the wire with a spacing up to about 3/16 inch, but 1/8 inch or slightly less would be better. That's just the magnet on its own. I could focus the field and get a stronger action. It seems like this would work well, and would be immune to interference. The pivoting arm with magnet on it would work with an optical detector to output a steering signal.

                              I don't yet know if I'll use this method, but I'm going to play with it a bit more.

                              I've just been measuring the cars I have, and they are pretty much 1/32 scale. I think I would prefer my cars to be about 1/20 or so. That's more room to incorporate the mechanics and electronics, yet still within bounds for the width of the track. It also suits my ailing eyesight better. I might have to stick with 1/24 scale for the sake of getting car bodies, but I'm going to look around.

                              Well, I better get to bed. Back to work tomorrow. It's been a nice several days off for me.
                              I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-