Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
Please, support PV!
It allows to keep PV going, with more focus towards AI, but keeping be one of the few truly independent places.
LED light using open source hardware
  • “Floyd” is a high quality LED light source with adjustable color temperature. The goal is to make a high quality, flexible light source optimized for photographers who shoot both stills and video.
    Its smarts are powered by an ARM microprocessor, the same kind of processor that runs your mobile phone. The design is open source, anyone can contribute to the design, hack it, make accessories for it, or otherwise bend it to their will.
    It is a highly flexible platform device able to control attachments like remote light sensing, a light painting brush, and more.


    This is a lightweight portable piece of lighting gear that is easy to work with. It is not a flimsy piece of plastic. It is made from lightweight aircraft-grade aluminium alloy and built to last.

    image

    This thing looks really strange. It seems to be actually not working and not finished yet.
    And I doubt that anyone but few guys will be able to make PCB and solder all chips on it, make enclosure, get all leds.
    Seems like huge time waste. But may be I am wrong here.

    http://riftlabs.com/


  • 22 Replies sorted by
  • Hi Vitaliy.
    Morten here, from RiftLabs. Why do you think it is a waste of time?
    The idea is to develop a better light source, esp. for video. We are doing the development in the open and publishing all schematics and documentation. The reason is so other people can hack it, extend it, tweak the firmware, interface it to other systems, etc.
    Making a complex piece of hardware means you have to commit resources to it. The hardware itself is tricky to build and I agree that very few people will want to make their own hardware from scratch. And unless you have your own CNC machine, your own 3D printer and your own PCB production tools capable of making 6 layer PCBs, it is probably going to be more expensive to build your own than to buy one.

    But I'm thinking that making open systems that anyone can tweak or build upon is better than making closed proprietary systems like most other manufacturers.
  • @mhjerde

    My view is very simple.
    Real goal is to solve problems of other people, save their time.
    I don't give a fuck if it is open or closed.
    Current trend of "open" is same thing as "object oriented" in 90s.
    I really like your concept, but I'll like it even more if you could come and provide people a link to the product that they could buy.
  • LOL, cool

    We are still in development. People are participating, talking about what features are important or not important to them. We're working our asses off trying to get this beast ready. I'll be sure to post a link in a few months when we have a product that people can actually play around with. :-)
  • @mhjerde

    Can you tell me what are you real advantages?

    And why are you using RGB leds (as good balanced light design using RGB leds is very complicated).
  • hi mhjerde
    are you at the point of telling us the lumen output of the continous lighting and the gn number of the strobe output? things like recharge rate for flash, angle of spread for continous, cri etc would also be useful.
    or does this plug into something else that outputs the light.
    cheers
  • Sure.
    Say you are shooting indoors and the existing lights are tungsten. Your camera can adjust its white balance to fit tungsten light and you are OK. But you need to add more light. If you add light with a different color temperature, say daylight, your camera has a problem. Adjust for Daylight or Tungsten? Regardless of what you or the camera chooses, some colors in your image are going to have a blue cast or an orange cast or both. The Riftlabs light can adjust its color temperature to any value, so you can mix and match with ambient light where ever you are.

    Our light also has a built-in colorimeter so you don't have to guess what the color temperature is.

    If you are accustomed to working with compensation filters like CTO, CTB, etc, all of these are built-in as electronic filters. When you use gel filters you lose a stop or two. With electronic filters you don't lose any light.

    Our light can emit any color light, not just white light. If you need a yellow wash, no problem.

    You can record, edit, play back and loop light sequences. The wavery light from a camp fire, hard strobes, flickering neon tube light etc.

    The max temporal resolution of the light is about 10ns (currently) and you can vary the brightness over 6 stops (currently) . If you are a developer, you can play around with the light engine and create any effect you like. You can synchronize it to any framerate via the sync port. There is a 10 pin general purpose I/O port, USB for firmware updates etc. It has a lot of extensibility.

    LEDs are more efficient than most any other kind of light. You get (a lot) more light out of your batteries.
    LEDs does not radiate heat. When you are shooting indoors, heat from incandescent continuous light can be a real problem.

  • @sparedog Yes, this thing outputs light, it is not just a controller. There has been some discussion about remote controlling these lights from an iPhone/iPad/Android App. Sounds intriguing, but we don't have Bluetooth or WiFi currently.

    The lumen output is actually a late stage decision. It is basically a matter of how fast do you allow the batteries to discharge and how good is the heat dissipation of the final enclosure design. I can say it is quite a bit more powerful that the typical portable light panels out there. But if you do stills, a good speedlight will be more powerful that this.

    The strobe recharge rate, the way strobing is currently implemented, is quite short. It is less that 1/2000 second. It also means that strobing is not that powerful. You get about one stop more light from the strobe compared to continuous light (roughly... insert lots of hand waving here...). But you start losing power fast when you shorten the strobe length. The short story is, I'm not sure how useful the strobing is, and you won't reach for this light when you need awesome stopping power.
  • @mhjerde

    Can you tell us how you want to get high quality, high CRI, white light using RGB leds?
  • maybe allow mains power also so high lumen output for continous during mains power
    high lumen continous on battery for short period
    medium lumen continous output for longer period

    it's led right so can you just flick on the light when you want to roll, and flick them off once you cut, it's not like they need a cool down period, do they?

    my view is if it is of use to video it should be at least the bightness of a redhead output (800 watts) so you use the tungsten ambience as fill. anyone else have a view on this?
  • I don't understand how a 2" x 3" LED panel (judging from the photo above) is going to be very effective. The concept is excellent, but won't this be little better than a point light source? How will I modify the light with grids or an umbrella? Can I use one device as a central controller/colorimeter (expensive) and then some remote, large LED panels (less expensive) for fill and accent?

    From a cost perspective...wouldn't it be cheaper for me to have some cheap clamp lights and some tungsten or high CRI CFL bulbs in my gear bag to solve the problem of needing more light consistent with whatever existing lights are there? Pretty much, with my monolights, I just use halogens so they can function as strobes for stills and also as a continous light source for video. I suppose I could carry CFLs as well (just not use them for strobing, as the sudden cutoff would probably fry them).

    The coolness of some of the features you mention--the ability to edit, record, and playback light is perhaps the most intriguing factor for me. I can see having some pre-sets built-in making this a "must-have" (e.g. I take B-roll of a neon sign, later using this light with a red neon sign pre-set and shooting somewhere else entirely, simulating as though the location is by the sign).

    Anyway, these are not criticisms, but simply questions and my first thoughts on your device.
  • @vitality I wrote a brief explanation on how RGB light works here: http://riftlabs.com/2010/09/how-rgb-light-works/
    I think I'll try to make a video during the week explaining it in more detail.

    @sparedog Light output equal to 800 watt halogen will require a different casing - we'll have to deal with more heat. And you are right, there is no cool down time. They are pretty robust.

    The current design is intended for lighter work, run and gun, ENG, that kind of thing. So it is designed to be lightweight and really fast to work with. Adding a version with more muscle is definitely in the plans.

    By the way, I forgot to answer your question about CRI. It is at 94 with the emitters we are currently testing.

    @MrAnthony Thanks for you thoughts! There is no reason you can't use umbrellas, grids, softboxes. But the ones you currently have will probably not fit. I can see that that is not ideal. It will be possible to synchronize several units (master/slave).
    This light won't replace all other lights overnight, and probably never :-) Needs are diverse and there is no one-size-fits-all. But you got to start somewhere, right? Our ambition is to take lights a big step further by giving them brains, making them more versatile, more efficient, and making it faster and easier to do really cool stuff.


  • How do you counteract the fast light falloff from LEDs with such a small array and no collimating?
  • @svart

    You have exactly two comments with content that you could cite here.

    I, personally, do not understand your concerns with light faloff as leds can be made quite focused, it is all depends on use leds or additional lenses.
    I also do not share your complains about led lights quality, normally RGB led quality is worse compared with two different color temperatures and good quality white LEDs.
    Right now it is pure concept with basis in the stage lighting that already has almost all attributes (RGB, adjustable color, programmability to some extent) for cheap.
  • the concern with light falloff is with standard T-type LEDs, which have poor focal qualities. Also, mass produced white T-type LEDs match rather poorly so you would need to use higher quality LEDs which cost considerably more.

    Anyway, I was talking specifically about needing to use white power LEDs with collimators and using yellow, red and IR high power LEDs with collimators to adjust the color temp. The white phosphor LEDs have peaks in the blue and yellow ranges so using reds to balance might work well.

    Like this:

    http://www.rpcphotonics.com/collimators.asp

    And I agree that RGB LEDs have no place in emulating white light. I've not seen one that had the right distribution of spectral peaks to approximate a good quality white light.

  • Those are better that T-type LEDs since the LEDs used in those units don't have focal lensing built in. They have flat dies that emit light relatively straight so no reflectors are typically needed. they could still benefit from minor collimating(refraction focusing).

    They are much better than:

    http://cgi.ebay.com/126-LED-Lighting-PANEL-HD-Video-DV-Camcorder-7D-D90-/220608752481?pt=LH_DefaultDomain_0&hash=item335d4e6b61

    Types.

    While I don't have the papers anymore, when I redesigned a robotic vision lighting system some years back, I researched using panels of LEDs similar what I linked above. I found that over 10ft(arbitrary testing distance) 1x 5w collimated(30deg focus) would illuminate a target better than 200x T-type.

    So looking at:

    http://blog.captive-aquatics.com/.a/6a010535f11c3d970c0133f2b133e0970b-pi

    You'll see that a typical "white"(5500K) led has spectral peaks in blue(450nm) and green-yellow (500-600nm). Adding adjustable amounts of red (620nm) could help fill out the spectrum and adjust total color temp biasing hopefully being able to approximate tungsten spectrum:

    http://www.chem.orst.edu/courses/ch361-464/ch461/oceanopticstungsten.gif

    Here's a good comparison I found:

    http://micro.magnet.fsu.edu/primer/lightandcolor/images/lightsourcesfigure3.jpg




  • @svart
    I think that you do not need to be robot vision designer to understand it.
    It is pure simple math and physics of light.
  • True. I just wanted to throw some insight out there and hopefully they can use it to make their project better.
  • @vitaliy
    have you used the yongnuo 160, and if yes, do you find the DSLRKIT VS-1400 are better?
  • Hi Mhjerde, nice for you to do such an open project. I myself have been doing some DIY fluorescent fixtures because the price in those days and the cost is prohibitive in my small island. I did it maily for my photography with some big arrays (8-10 4 foot tube) to get the necessary light output for photography. I had to do some compromise (no dimming, one plug for each light) not really portable as the build was a bit difficult to move, but in the end with some high CRI tube the result was very good.

    One thing I have been seeing lately is perhaps the perfect light. These are the LEP Light Emitting Plasma. They have cri of 94, about twice the output of HMI, good throw and are relatively small to fit existing PAR. They have been very well received in the latest NAB and cinegear expo. But the price are in the $ 3000 to $ 5000 range. The thing is that a company specialising in DIY projector is selling some kit from Luxim (I think most of the video light company are using the same module from luxim) at $ 600. As always in the video world price tend to be much higher. It could be an interesting system to investigate in the future.

    For your LED project what would be nice, is for you to sell kits that people can assemble. Because as said before it will be of no use to the vast majority if it is only a set of schematics that the person has to build his own casing, solder the parts etc. I hope it materialise very soon and I pave the way for some other project like external monitors etc. with the same open-source concept.

    I forgot one thing about the LEP is that it cannot work facing upward. I don’t know how the manufactures have been able to do it but for me very rarely do you point up and can use some reflectors to some extent.
  • @danyyel How do you qualify Light Emitting Plasma being the perfect video light?

    The main (only?) advantage of LEP seems to be potentially lower cost for a high lumen system. The gas discharge lamp is a broadband light (you need gels to adjust color temperature) it is hot, and it takes 20 minutes to turn fully on.

    The light source we are working on runs cool, can emit any color or color temperature without loss and turns on and off in roughly one millionth of a second. We use the fast turn on/off time to create various "digital" effects, useful for video and stills work.

    These are two very different approaches. The LEP light is a big old light bulb with massive power. It is probably awesome when that is what you need! :-)
    We are going down a different road, exploring what you can do with digital "intelligent" light.