Ernie Misner Loc: Lakewood, WA

Pretty hilarious. First you say the histogram does not offer any legitimate information and can not be depended on at all. Then you say you certainly can depend on a histogram and it's bell curve to show whether you have under or over exposed. I would be sad if someone who is starting out took you seriously about the first statement and did not bother to learn using his histogram. I have been depending on mine for years to give me some good information to go on, mostly about over exposure. Yes, just a tool with limitations, but an excellent tool. We need to be cognizant whether there are some bright highlights in the scene that are actually okay to over expose and blow out so that our main subject will not be underexposed. In that case the bell curve will actually, and correctly, climb the right hand edge of the histogram. And the bottom line, as always, even in the day of film, is to bracket your exposures have have options available back at home.

floral43

You seem to be intelligent to me, how ever you shouldn't make such general statements. like above none of that is true,
I looked at the schematics and read the reports you sent me and i appreciate you haven to take the time to collect this information.
3D color matrix metering system.
1) An improvement over kodacs method for establishing their 18% gray card.Kodac took 10,000 images approx. Nikon achieved a better way of averaging (brightness intensity [gray scale]} based on their matrix metering system, brightness data is converted from 1005 pix RGB sensor. try to grasp this COLOR is EXTRACTED from BRIGHTNESS. That's what sensors do,cc, cmos, etc. Brightness value is measured on a scale,that scale{no matter what the pix depth is goes from 0 black to 255 white with 254 tones of gray.reflected by a histogram. BV data is grouped into over lap areas.
Sigma cameras tried a similar approach with their foveon sensor.
please read my article on how a computer processes a raw gray file produced by a camera. any camera.
The statement( light meters don't measure the temperature of light because their multifunctional Absurd. If you look through your menu or mode what ever! you will find a k (kelvin scale) a method that kelvin discovered to measure light. light temp controls color. Where have you been.
I appreciate you having the intelligence to continue this string and to hang in there with me, after all how else can we learn.

Apaflo Loc: Anchorage, Alaska

That was not a general statement, it was very specific. Not one of the specific statements quoted was true. None.


Again, lets be very specific: the above is not true.

The 0-255 scale is precisely one thing, and that is the effect of using 8 bit depth data. If we use 16 bit data, 12, bit data or 4 bit data it is not a range of 0 to 255.


The light meter does not measure the temperature of the light, only the intensity.

The light meter uses data from a very different sensor than the image data is read from. The camera evaluates color temperature using data from the image sensor, not from the light meter sensor.

floral43

The intensity is the BV that is determined by temperature of light. has nothing to do with color. Color has not entered the picture yet.
No matter what pixel depth is used, the binary representations interpret that 0 is black and max value (255 8 bpp, 65,535 at 16bpp etc. is white.
i have come the conclusion that your not reading previous explanations or just don't understand them.
I will try to explain the function of a sensor to you and it doesn't matter whether it's a bayer pattern or some improved version of it.
A sensor contains columns and rows of pixels! the pix are made up of wells that have depth. on top of each pixel sit color filters.RGB.
For the purpose of this example we are going to remove the color filters, we are left with the pixels (wells). In other words we now have a sensor that will collect varying degrees of light intensity only with out the perception of color,black with varying tones of gray to white.255
Take a picture of a red rose, what is red, the rose absorbs approx most of the other color frequency's and reflects red frequency's varying in intensity back to the camera sensor, that strike the wells. What is this reflection made of. {varying in intensity, frequencies,and wave lengths, The intensity determines how deep the light penetrates the wells that store it. although we have the frequency that determines red stored in the sensor, it will appear as a dark shade of gray.like a film camera. because we removed the RGB color filters.
Now lets place the RGB filters back on top of the PIX wells the light frequency that represents red stored in that pix will penetrate the red filters only and you will see red.
So as i said from day one the sensor doesn't see color, color is the perception brought about by the color filters, placed on top of the sensor pixels.
I see now that i tried in vain to use a kaleidoscope in an attempt to simplify this concept. but i went way over your head.
judging from your quick responses back to me you either don't understand or care not to.
I have provided proof of what i am saying, you have provided lip service but no proof. In fact that article you provided me with although informative has nothing to due with what I'm saying.

SonnyE Loc: Communist California, USA

You really are a useful idiot....

*
ignore

Apaflo Loc: Anchorage, Alaska

Brightness has nothing at all to do with "temperature of light'.

http://en.wikipedia.org/wiki/Color_temperature


So what you said previously, that it was always 0 to 255, is clearly not true.


Your previous "explanations" were incorrect. As has been shown multiple times... Going around in circles isn't going to make those errors correct the second or third or forth time you say the same thing.


Good enough until you said "255". There is nothing about sensor locations that is relative to "255". The sensor locations are analog devices. They have a DC voltage output, not a digital value. It is at the output of the Analog-Digital-Converter (ADC) that we have a digital value, and at that point 255 is never white because cameras use greater that 8-bit depth at the ADC.


Intensity does not determine "how deep the light penetrates". Intensity is a measure of how many photons are captured by the sensor well.


You've just said that when the filter is in place the sensor location filtered for red can only see red. Hence it does see color, and it records the intensity of just one specific color in that specific location.

We need to be careful to understand that sensor location does NOT record the color for a pixel at that location. It takes at least 4 sensor locations (though virtually always cameras use 9 or more locations rather than the minimum possible of 4) to determine the color for any given pixel.


The kaleidoscope analogy is inherently invalid. A kaleidoscope is illuminated by white light, and what you see are colors from the kaleidoscope not from the illuminating light. The Bayer filter is not illuminated only by white light, and the purpose is to see colors from the illuminating light, not the filter.

floral43

Your first statement.
brightness has nothing to due with the temp of light,
That is laughable any sunny day in the summer disproves what you just said, the brighter the sun becomes the hotter it gets,and how due we measure that heat? with the temperature.the site you sent me to has not been validated. at least give your statements some thought.
you see it's easy to play the devils advocate.
Put your hand on a light bulb use a dimmer switch turn up the brightness and let me know if you burn your hand.Wow

Apaflo Loc: Anchorage, Alaska

The "temperature" of light is not related to how bright the light is.

The temperature relates to the color of light emitted by a black body at that temperature. Brightness has nothing to do with that color, and for any source of light other than a black body the actual heat temperature that you would feel with your hand is also unrelated.

Once again, read the cited reference;

http://en.wikipedia.org/wiki/Color_temperature

Note the references at the end of the article.

floral43

a black body can only reflect between 7 and 9 % of the light reflected ( reflectance) upon it . the rest of the light is absorb by it. Stick to the kelvin scale. a white body can reflect approx 32%.

Color filter array cameras a two - dimensional area array to collect the photons that are recorded in the image. The array is made up of rows and columns of photosensitive detectors -typically using either ccd (charged couple devices) or Cmos technology- to form the image. In a typical set up each element of the array contributes contributes one pixel ot the final image.
The sensors in the array whether ccd or comos, just count photons, they produce a charged proportional to the amount light they receive.
WITHOUT RECORDING COLOR INFORMATION, the color information is produced BY color filters that are applied over the individual elements the array is a process known as [striping] the term striped array. Most cameras but not all use bayer pattern arrangement for the color filter array, alternating red green,and blue filters, twice as many green as red and blue.( because are eyes are more sensitive to green.
Other color filter array configurations are possible
RAW FILES ARE GRAY SCALE
NO MATTER WHAT THE FILTER ARRANGEMENT, the raw file simply records the luminance value for each pixel. So the raw file is a gray scale image.
it contains color information - the characteristics of the color filter array are recorded. so that raw converters know whether a given pixel in the raw file represents red green or blue luminance, or what ever colors the specific camera array uses. BUT IT DOESN'T CONTAIN ANY THING HUMAN CAN INTERPRET AS COLOR. OBTAINING COLOR FROM THE RAW FILE IS THE JOB OF THE RAW CONVERTER SUCH AS CAMERA RAW.
THE RAW CONVERTER INTERPOLATES THE MISSING COLOR INFORMATION FOR EACH PIXEL. (DEMOSAICING) .
MANY DIFFERENT FILTER SETS ARE IN USE WITH DIGITAL CAMERAS.SO THE RAW CONVERTER HAS TO ASSIGN THE CORRECT COLOR MEANINGS TO THE rgb PIXELS, USUALLY IN A COLORIMETRICALLY DEFINED COLOR SPACE SUCH AS CIE,XYZ WHICH IS BASED DIRECTLY ON HUMAN COLOR PERCEPTION, AND REPRESENTS COLOR UNAMBIGUOUSLY.
A converter does much more than this re other task.
This is what I stated at the beginning of this thread not in so many words. I didn't think i had to get to this example. my original statement was that a camera sensor doesn't see color. than you took off on a tangent misleading and twisting my words, with no regard for articles or proof that i have provided.
Than you started playing the devils advocate, I try to keep things simple you like to complicate issues.
When are you going to learn that COMPLEXITY IS THE REARRANGEMENT OF SIMPLICITY. THIS ISN'T A COURSE IN PHYSICS.
i AM CLOSING THIS THREAD, i DUE APPRECIATE THE TIME you invested and the information you provided, I take issue with the way you criticized the articles i provided in the beginning of this thread.

Apaflo Loc: Anchorage, Alaska

The "color temperature" does not refer to light reflected from a black body radiator. A "black-body" with a temperature of 5000 degrees kelvin is a light source. It is the color of the light, not the intensity, that is significant.


Each sensor location does not contribute one pixel to the image. Instead a matrix of sensor locations, usually 4x4 or greater in size, provides the information for each single pixel of the image.


The sensor does record color information... in the matrix of sensor locations. Color is not "produced BY color filters". The filters form part of the system used to encode the colors in the light that is measured. Luminance values are stored in one way, and color values in another way.


Color information is in fact recorded. If it were not we would never be able to have a color image produced from a RAW file.


Accurate information is important. Your articles are not.