Page 2 of 9 FirstFirst 123456 ... LastLast
Results 11 to 20 of 85

Thread: The perfect white balance - feedback requested

  1. #11
    Join Date
    Jan 2008
    Location
    Stockholm, Sweden
    Posts
    799

    Default

    OK, it seems I can finally put the scattering issue to rest (at least for a while).

    The way I see it we have two cases:

    1. No haze/fog in the air. Rayleigh scattering dominates. This is strongly wavelength dependent; the scattering of violet light is nearly 10 times greater than that of red. The scattering is relatively weak however and will not cause significant extinction unless the distance is large.
    2. Haze/fog present. Mie scattering dominates. Not strongly wavelength dependent, so all visible light will be scattered fairly equally. Clouds are white after all, right?

  2. #12
    Join Date
    Aug 2008
    Location
    UK
    Posts
    5,704

    Default

    Quote Originally Posted by JeroenVDV View Post
    That white is from the Jenlas whitelight-laser. The "new" white is made by combining yellow/blue OPS lasers (even nicer).
    I'd like to see the new white. I recently had a private discussion about Jenlas White and how it was achieved, but can't divulge it further. I'm just hoping it might prove fruitful.

  3. #13
    Join Date
    Jan 2008
    Location
    Stockholm, Sweden
    Posts
    799

    Default

    Quote Originally Posted by RGBLaserFan View Post
    I would have to disagree there. Get a paint swatch card and then tell me your eye can't distinguish between the different shades of white. Paint manufacturers make over 500 different shades and you can see the difference albeit most would need to be side by side. However, some whites are whiter and cleaner than others which is why our eyes can see the difference between eg a halogen floodlight and the icy white of a HID headlight.
    On that swatch card, set the value of each white to 100. How many of these whites can you distinguish between without having them side by side now? Only whites with different chromaticity coordinates are interesting here, since the value in a laser system can be varied simply by multiplying a given radiant power configuration by a scalar. In other words, to get a "greyer" white, just reduce the power of all your lasers by the same percentage.

    This discussion is interesting though, it spawned a new idea. Since none of us have unlimited resources, it might be useful to calculate the luminance of different configurations, so that you can compare how bright they appear in relation to each other.

  4. #14
    Join Date
    Jan 2008
    Location
    Stockholm, Sweden
    Posts
    799

    Default Yet another diagram

    Apparently there isn't that much difference in the luminance (brightness) of the white point with different blue laser wavelengths. This is because green is by far the largest contributor to the luminance.

    Setting the luminance of the 473nm based configuration to 1, this is the result.


    Interpretation of this result:

    Configuration 1:
    1.00W 532nm
    1.49W 473nm
    1.75W 635nm

    Total radiant power: 4.24W
    White point luminous flux: 970 lm

    Configuration 2:
    1.00W 532nm
    0.81W 445nm
    1.30W 635nm

    Total radiant power: 3.11W
    White point luminous flux: 821 lm

    Configuration 1 has an 18% higher luminous flux at the white point (D65) than configuration 2, but its radiant power is also 36% higher. Looking at the luminous efficiency configuration 2 is 15% more effective.

    Conclusion:
    I wish 445nm was cheaper?

  5. #15
    Join Date
    Feb 2007
    Location
    Herts, UK
    Posts
    1,254

    Default

    Another thing to consider with fog/haze and forward scattering is that as the power of the red is increased to compensate for longer wavelengths, the higher powered beams will suffer less losses over longer distances than the lower powered green/blue.
    I think the "required red" ratio curve will flatten off a surprising amount as the red beams wavelength and power increases in these conditions.

    Great discussion BTW Tocket

  6. #16
    Join Date
    Aug 2008
    Location
    UK
    Posts
    5,704

    Default

    So if I read what you're saying correctly Tocket, to get the greatest luminance from the white setting, white should be achieved by balancing the projector with the highest level of green power possible.

  7. #17
    Join Date
    Jan 2008
    Location
    Stockholm, Sweden
    Posts
    799

    Default

    If you say luminous efficiency instead of luminance, then yes. One can pick the other wavelengths in such a fashion that the relative amount of green is very high. This would be beneficial from both a safety standpoint as well as an economical (once the 445nm prices have dropped a bit).

    Basically, you get a brighter white using less total laser power with 445nm lasers.

    Another thing to consider with fog/haze and forward scattering is that as the power of the red is increased to compensate for longer wavelengths, the higher powered beams will suffer less losses over longer distances than the lower powered green/blue.
    I think the "required red" ratio curve will flatten off a surprising amount as the red beams wavelength and power increases in these conditions.
    Red is indeed harder to scatter, but to what degree? Does the sun look red through a cloud? Does a headlight look less blue through fog?

    The calculations for Mie scattering look pretty complicated, so I'd rather not go down that road.

    Also, you need to consider that while red light might have a smaller extinction in fog that would also imply that it isn't scattered as much and thus making the visibility of the beam lower. Once the beam hits a surface the red will be more visible though.

    Like I said, scattering is complicated. That's why I was glad to discover that Mie scattering isn't strongly wavelength dependent. For now I would like to consider this large particle scattering to be equal for all wavelengths. To develop a haze/fog correction would require feedback from real world examples, but that is one reason that I started this thread.

  8. #18
    Join Date
    Feb 2007
    Location
    Herts, UK
    Posts
    1,254

    Default

    Quote Originally Posted by tocket View Post
    To develop a haze/fog correction would require feedback from real world examples
    One real world example:

    An RGY laser made up of 1w of 532nm and 2.2w of ~663nm looks very weak in the red department when projecting onto a matt white surface 5m away, green looks "limey" and it's hard to tell green from yellow on parts of most images.

    The same laser when projecting towards an observer 20m away through medium density fog looks very well balanced. The thicker the fog, the more the balance shifts towards the red.

  9. #19
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,446

    Cool

    Hey Tocket;

    Outstanding post and resulting thread! Thanks for taking the initiative re: the research behind all this...

    One thing I believe has been missed, however: The human eye's color sensitivity changes based on ambient lighting conditions. The color peak for photopic vision is around 555nm, but as light levels drop the peak shifts dramatically towards the blue spectrum. When you're watching laser shows, frequently you're into mesopic or scotopic vision. Thus the color balance numbers must shift as well.

    I've demonstrated this myself by tuning separate color palettes using Pangolin's color balance wizard. If you balance everything in dark conditions and then display it in a brightly lit room, the white looks very yellowish. Conversely, if you balance it in a brightly-lit room and then display in a dark room, the white is a cool, icey blue.

    From your initial post, I *think* you started with the peak for standard photopic vision. If so, then I would argue that we should change that to peak for mesopic vision.

    Unfortunately, I've been unable to find a good number for the mesopic peak. Of course, it's always going to be something of a moving target, because light levels will vary from venue to venue, and your own adaptation to it may also change. But if we can reach a consensus on the wavelength peak for lower light conditions, we'll have a better starting point for your calculations.

    Your thoughts?

    Adam

  10. #20
    Join Date
    Jan 2008
    Location
    Stockholm, Sweden
    Posts
    799

    Default

    Thank you Adam. It is good that you bring up the scotopic vision, because it is another thing closely related to the topic. It's been briefly discussed before, but no conclusion has ever been reached to my knowledge.

    I think I've dismissed scotopic and mesopic vision before though, but perhaps without any good reason. Going back, I see that it was the limits of these vision types that was the grounds. Over 3.4 cd/mē (nit) luminance means completely photopic vision. Compare that to a typical lcd screen, which has a luminance of 300 cd/mē. How bright are your beams/projections compared to an LCD? I'd say, not 100 times weaker.

    Now, is this for the whole field of vision or can you for example have zones of mesopic and photopic vision at the same time? Seeing as the beams might not occupy the whole space, how does this affect perception?

    Perhaps your observed shift of the white in different lighting settings is not a result of change from mesopic to photopic vision, but rather because some other phenomena.

    I do indeed use photopic vision for the calculations. The peak for mesopic vision is not fixed, but rather a function (unknown) of the luminance. However, it is not the peak I need for these calculations, but the full color matching functions. Since these are also a function of the luminance it all gets very complicated far too quickly. Thanks to a recent paper it may be possible though.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •