Originally Posted by undermined
I did a quick search of the thread but didn't see anyone mention that a "high-end" digital camera deosn't even capture RGB color anyway.
Cmos and CCd sensors used in most digital cameras do not record RGB information for each pixel either but instead have to demoisaic the pattern called a Bayer pattern, to interpit the color for each pixel
From the wikipedic entry:
"The raw output of Bayer-filter cameras is referred to as a Bayer pattern image. Since each pixel is filtered to record only one of three colors, the data from each pixel cannot fully determine color on its own. To obtain a full-color image, various demosaicing algorithms can be used to interpolate a set of complete red, green, and blue values for each point."
So the way a digital camera captures light isn't even setup up to measure light but instead it interpits the light in a scene during the the sensor was exposed.
digital camera's sensors don't even measure white balance or meter the light for exposure, that is why you can set any white balance you want based off what you tell the camera you would like to be neutral white in a scene and it will offset the rest of the color information accordingly.
and digital camera's use a light metering sensor to tell them how bright or how dark a scene is, the imaging sensor has no idea of this, it only captures the light acording to how you set the exposure. it does'nt know if a scene is to bright or too dark, too red or too blue it just records the light that hits it's sensor and the camera attaches vaules to that data.
A meter reads the data and tell you that information based of a known standard were as with a camera you decide the standard and it captures the scene based on what you told it it saw. Then there is the problem of shutter sync for the refreash rate of the display and the issue of needing a calibrated display to look at the photos taken by the camera to get a idea what colors the were being captured.
Just like I can look at a fire engine and say it is red, but I don't know if that red is the same red your eyes see or how red is that red a camera is just not the right tool to measure a display for color reproduction, maybe to measure speed or distance traveled but otherwise not.
This is a good thread for color measure measurement I just wish it didn't take so much info for people to understand why there is a reason one uses a meter to measure.
Good points, but because I am not measuring color on a per-pixel accuracy, but instead averaging out the RGB values of 50% of the captured image, the bayer-filter (if my camera uses that) doesn't affect me, as it's still capturing RGB, just not on a per-pixel level. The overall average of an area of the image will still be the same regardless of the bayer filter or not.
Also, as I've posted before, it doesn't matter what the specific luminance of the colors are, but the ratio of luminance between each of the R, G , and B values, which determines chromaticity. As long as the exposure is not too high (which results in RGB values > 255) then the ratio can be used to determine the chromaticity of the color. Overexposed values distort that ratio and therefore can not be used. As for shutter sync, I make sure I use a slow enough shutter, along with other appropriate ISO, F-Stop, and EV values to avoid any darkening of the image due to bad vsync.
I set the white balance based on a white card in overcast daylight (which Wikipedia refers to the best natural source of D65). You can accurately determine the color of something using a camera as long as the white balance is properly set. If you have a 100% red card next to a 100% white card in the same lighting, and you white balance the camera to the white card, then the red card will read as 100% red in the RGB levels. The problem comes (as referenced above) when the SPDs of the objects you are shooting differ. I believe the range of error between SPDs will be small enough that using a camera to calibrate will result in a better greyscale than stock settings, and that meters would get you even closer.
Obviously it depends on the quality of the digital camera, and all cameras differ, but so far, for me, the calibration resulted in a better image. If you perform the same experiment and the calibration looks worse to you, just reset your settings. While most people can't accurately calibrate their TVs by just using their eyes (especially without a proper D65 reference point), they usually can tell whether the change in settings made the image look better or worse. For me, this made the image better, and the image will look even better once I do a proper calibration (once I get the tools necessary), but for now I'm satisfied enough.
I understand why people use a meter, and I would as well if I had one, but right now I don't, so this is the best I can do.