Pages: [1]   Go Down

Author Topic: r-g-b flash?  (Read 19289 times)

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
r-g-b flash?
« on: March 15, 2013, 10:07:30 am »

I would like to gel my flash for the occasional shots where you want to limit the amount of gear/intrusion and keep the overall mood of the existing light. But I never get around to it. I guess there are enough parameters to worry about...

Did anyone ever think about (or see) a flash that could vary its color temperature? Having 3 smaller bulbs, each filtered by corresponding filters (somewhat like the Philips Hue home lighting system), where the contribution of each is under software control sounds like a neat thing.

At the bare minimum, fiddling with gels could be replaced by a great big dial on the back of the flash called "Temperature [Kelvin]" (perhaps a smaller one called "tint"). Even more relaxing would be an ETTL-extention that allowed the camera to "see" the general ambience temperature of the scene, dial up corresponding settings in the flash, and ensure that heavily lighted objects/shadows had the same color temperature as the scene average (or highlights or some other sensible statistic).

-h

http://www.meethue.com/en-US
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: r-g-b flash?
« Reply #1 on: March 15, 2013, 06:59:41 pm »

Did anyone ever think about (or see) a flash that could vary its color temperature? Having 3 smaller bulbs, each filtered by corresponding filters...

Interesting idea.

If using Xenon tubes, efficiency and size would suffer. The smaller the chromaticity range in the output, the more efficient you could make the system. If a straight-line approximation to the black-body curve were adequate, you could get away with only two light sources.

If using LEDs, you could have a small, efficient flash with three nearly spectral emitters. However, there would be a likelihood that some metamerism problems would surface with the spectrally peaky primaries.

Jim

David Sutton

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1345
    • David Sutton Photography
Re: r-g-b flash?
« Reply #2 on: March 15, 2013, 08:32:03 pm »

Most folks I know have gone to variable colour temperature LED studio lights.
Logged

wolfnowl

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5824
    • M&M's Musings
Re: r-g-b flash?
« Reply #3 on: March 17, 2013, 03:29:53 pm »

Most folks I know have gone to variable colour temperature LED studio lights.

Exactly what I was going to suggest.  Here's one: http://fiilex.com/

Mike.
Logged
If your mind is attuned t

RFPhotography

  • Guest
Re: r-g-b flash?
« Reply #4 on: March 17, 2013, 04:34:54 pm »

Those may work in some static environments, but not in a lot of situations where speedlights are being used. 

I'd expect that colour temperature tuning will be the next development in some of the smaller LED grid lights.  Problem with continuous light is that it doesn't have the same action-stopping capability of flash.
Logged

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18090
  • When everyone thinks the same, nobody thinks
    • My website
Re: r-g-b flash?
« Reply #5 on: March 17, 2013, 06:26:01 pm »

... Did anyone ever think about (or see) a flash that could vary its color temperature? Having 3 smaller bulbs, each filtered by corresponding filters (somewhat like the Philips Hue home lighting system), where the contribution of each is under software control sounds like a neat thing....

Pssst... patent it before someone else does :)

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: r-g-b flash?
« Reply #6 on: March 18, 2013, 04:01:54 am »

If using Xenon tubes, efficiency and size would suffer. The smaller the chromaticity range in the output, the more efficient you could make the system. If a straight-line approximation to the black-body curve were adequate, you could get away with only two light sources.
Yes, I am guessing that two fairly spectrally smooth tubes would do. One "warm", one "cold" as long as they could be lineary mixed to produce fairly accurate intermediate color temperatures. If the flash is limited by total heat or total delivered power anyways this might not even be so inefficient as one might initially believe (two tubes and two power delivery circuits though).

Is it at all realistic to automatically control color temperature? Did not Nikon use to have a separate ambient WB sensor on their cameras? Put one on this flash and estimate ambient WB right before firing the flash. Might not be perfect for mixed-lighting, but might be good enough for 80% of the shots...

-h
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: r-g-b flash?
« Reply #7 on: March 18, 2013, 04:02:46 am »

Pssst... patent it before someone else does :)
I guess this post counts as "prior art". As long as I can buy the product, I am happy :-)

-h
Logged

theguywitha645d

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 970
Re: r-g-b flash?
« Reply #8 on: March 18, 2013, 10:58:06 am »

I wonder why flash tubes are daylight? Perhaps it is difficult to vary the spectral output? Then to build in either a color meter or to get the camera WB information would seem very complex.

To be honest, how difficult is it to carry a few color compensating filters to stick onto a flash unit?
Logged

David McCaughan

  • Newbie
  • *
  • Offline Offline
  • Posts: 29
Re: r-g-b flash?
« Reply #9 on: March 19, 2013, 01:50:04 pm »

I can buy a lot of gels for what this is gonna cost.  ;D
Logged

jvb

  • Newbie
  • *
  • Offline Offline
  • Posts: 20
    • http://www.vanbecelaere.com
Re: r-g-b flash?
« Reply #10 on: March 28, 2013, 12:37:19 pm »

Nice idea. The old color enlarger technique jumps to my mind. A Dichromatic filter that can be dialed in gradually?
Must be possible to couple it with the WB of the camera.
Anyhow, not really a DIY thinkpath, but something for the "Toyotas" .

johan
www.vanbecelaere.com

Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: r-g-b flash?
« Reply #11 on: June 07, 2013, 04:54:12 am »

Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: r-g-b flash?
« Reply #12 on: September 11, 2013, 04:58:56 am »

Yes, I am guessing that two fairly spectrally smooth tubes would do. One "warm", one "cold" as long as they could be lineary mixed to produce fairly accurate intermediate color temperatures. ...
Apple obviously reads ll :-)
http://en.wikipedia.org/wiki/IPhone_5s
Quote
(the iPhone 5S camera) also has dual "True Tone" LED flashes, optimized for warm and cool color balancing providing the first variable color temperature flash on any camera.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: r-g-b flash? Or white-amber with auto-color from Apple
« Reply #13 on: September 12, 2013, 07:02:34 pm »

Apple obviously reads ll :-)
http://en.wikipedia.org/wiki/IPhone_5s
I was going to post about the quirky fact that Apple had just launched an innovation in flash photography. It is more rudimentary, offering only one degree of freedom (color temperature?) but on the other hand, it does it with automatic color metering.

But maybe it is not so quirky: maybe we should get used to the fact that companies like Apple are selling more cameras, and generating far more profits on them, than any traditional camera maker, so at times, they will have the resources and profit potential to invest heavily, and drive innovations from the bottom up. Another thing to watch out for is that Apple has integrated the design of optical hardware, custom subprocessors within its ARM-based ”system on a chip", and software for that hardware, to support features like this flash system and the new version of "sweep panorama", which merges 60fps information, rather than just a few frames. Samsung, Sony, and Panasonic might have similar capabilities; Canon, Nikon, and Olympus not so much, at least so long as they each muddle along developing in-house proprietary microprocessor hardware (Digic, Expeed, BIONZ, TruePic, Venus Engine, etc., etc.) requiring an entirely in-house software effort to go with it.

High-end gear might not be directly affected by these mere conveniences and gimmics, but the makers of mainsteam cameras, up to the "non-pro" system cameras, might want to get more serious about the features that their signal processing and camera control software and hardware can support, and maybe drop their in-house "IT", which I doubt can match the pace of development of ARM SOC designs and software.
« Last Edit: September 12, 2013, 07:30:15 pm by BJL »
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: r-g-b flash? Or white-amber with auto-color from Apple
« Reply #14 on: September 13, 2013, 03:47:17 am »

I was going to post about the quirky fact that Apple had just launched an innovation in flash photography. It is more rudimentary, offering only one degree of freedom (color temperature?) but on the other hand, it does it with automatic color metering.

But maybe it is not so quirky: maybe we should get used to the fact that companies like Apple are selling more cameras, and generating far more profits on them, than any traditional camera maker, so at times, they will have the resources and profit potential to invest heavily, and drive innovations from the bottom up. Another thing to watch out for is that Apple has integrated the design of optical hardware, custom subprocessors within its ARM-based ”system on a chip", and software for that hardware, to support features like this flash system and the new version of "sweep panorama", which merges 60fps information, rather than just a few frames. Samsung, Sony, and Panasonic might have similar capabilities; Canon, Nikon, and Olympus not so much, at least so long as they each muddle along developing in-house proprietary microprocessor hardware (Digic, Expeed, BIONZ, TruePic, Venus Engine, etc., etc.) requiring an entirely in-house software effort to go with it.

High-end gear might not be directly affected by these mere conveniences and gimmics, but the makers of mainsteam cameras, up to the "non-pro" system cameras, might want to get more serious about the features that their signal processing and camera control software and hardware can support, and maybe drop their in-house "IT", which I doubt can match the pace of development of ARM SOC designs and software.
Apple have full control over their content/application/OS/hw stack, and a lot of resources/expected sales. This gives them great incentive and possibility to develop new tech that can increase customer willingness to pay, or reduce component costs.

I believe that cellphone cameras come from a starting-point of "not good enough", while hordes of customers (regular people) are lining up to buy the next camera if it has a better camera. And they will do it again in 18 months. Compare this to the enthusiast camera market where people hold on to their camera for 3-5 years or more, and complain that new cameras are only slight improvements over older ones.

I assume that the general public are more easily tempted by what camera nerds might see as gimmicks or marketing by numbers. And the general public might see certain camera nerds as old-school relics.

I assume that "Digic" etc are often just branding of camera-optimized ARM SOCs. They throw in 2-4 ARM cores for regular work, then a custom on-chip package for image pipeline, compression etc (possibly only low-level acceleration of said functions). That is similar to what cell-phones do (custom imaging hw for improving battery-life/latency), only that the pixel rates are probably higher in something like the Canon 5Dmk3. For big players (Canon, Sony), it might pay off to do semi-custom design (purchase the general components, design camera-centric components or hire someone to do it for you, then get TSMC/Global Foundries to manufacture the silicon).

Compare this to the "general computing" market. Who is it that drives innovation in general computation these days, Dell or Apple? Computers seems to have become "good enough" for most people, price and nice looks seems to be primary differentiators. For cell phones, there is a need for more computing power, and Apple (&competitors) are happy to deliver...

-h
« Last Edit: September 13, 2013, 03:49:41 am by hjulenissen »
Logged

pshambroom

  • Newbie
  • *
  • Offline Offline
  • Posts: 13
Re: r-g-b flash?
« Reply #15 on: October 10, 2013, 10:18:01 am »

I came up with the same idea many years ago, even pre-digital. I bought a special Minolta enlarger head with individual RGB flash tubes, hoping to modify it for studio use. Never really got any traction with putting it together. Now in the digital age I've been hoping that some manufacturer would would make a 3 tube RGB handheld flash with auto color temp sensing from the camera. I was surprised and somewhat delighted that Apple was the first to do so (in a limited way.) My latest idea is that flash/ambient color balance matching could be accomplished with any single tube flash through in-camera processing of two exposures in quick succession. The first would be the flash at a high enough sync speed to minimize ambient exposure, The second would be ambient only. THese could be matched with in-camera processing and baked into a jpeg. Ideally, both exposures could also be stored in a RAW format with post-processing options available for custom blending. THis is can only work when fully electronic shutters are available that can shoot 2 photos, store data, and process it in close to real time. I don't know what the current state of that technology is. I wrote about this on some blog a few years back and there was some replies stating that this shutter technology is a few years off. ANy thoughts or current info about this?
Logged
Pages: [1]   Go Up