|
|
Thread Tools | Display Modes |
#11
|
|||
|
|||
Great missions STS-122 & Expedition 16
On Feb 22, 8:52 pm, BradGuth wrote:" Another
interesting second or third hand rant, even though you are intent upon playing word games, rather than offering us science that can be peer replicated as to what I had previously specified." laughing your quoting numbers like "spec. candy", as if you really understood what the numbers mean, much less that the ccd is but one component of the digital imaging system. Ok lets try it this way humans are either going to view the image in the print form or on a monitor, both of which have their own dynamic ranges from black (min), to a white (max), that is independent of the ccd's dynamic range. Now venturing into the area of human color perception is necessary if someone wants to reproduce what the human "sees", as the sequence goes from humans capturing the image, image processing and finally viewing of the image on paper or on a monitor. Now even if you take humans out of the image capture part of the equation, you still have to have the image processing for viewing, meaning the ccd's capabilities are just part of the system, just like the silver halide crystals on film are but one part of the sequence that reproduces what the "eye sees". So I think its funny that you seem to be stuck on comparing a devices capabilities against the human visual system without understanding the latter So actually I did answer your question, you just did not understand the answer, nor how it directly relates to your constant ranting about dynamic range, that is why you go from regurgitating numbers of a chips specifications, to questioning the results of the image, when you really don't know what is going on in the optics of the device or the human visual system. The human visual system specifically analyzes and compares the colors of the electromagnetic spectrum with photo pigment responses (short medium and long cones, with contributions from the rods) spanning a range of about 1.3 electron volts of energy covering from reds to blues, and through primary and secondary comparisons our brain perceives what Newton termed extra-spectral hues, completing a connection of low energy reds to the high energy blues. The concept of human color perception is not just describing eye candy, as the red/ green and yellow/blue opposition responses in the human eye are what make up the cie Lab color space axis where ALL the data from the ccd image is mapped. Cie lab space is a 3d Cartesian plane based on the results of experiments that studied how the human visual system responds to stimuli. Luminosity is represented on the z axis going from absolute black to white (0 to 100), with red/green yellow/blue opponency represented on the x, y axis +/-100 in either direction: the x axis is lower case a, going from greenish (-a), toward being reddish (+a) (red vs. green), the y axis is the lower case b and goes from bluish (-b) towards yellowish (+b) yellow vs. blue. Now given all the data captured by any ccd has to be mapped to this coordinate system, and given the fact that same system is based on the human visual system you need to learn a lot more because cielab space itself does not contain all of the colors the human eye can detect, i.e. its missing some greens and extra spectral hues (see lab gamut display brucelindbloom.com) The Most expensive professional digital cameras on the market do not allow the image to be mapped to custom color spaces or color profiles but instead utilize srgb, or adobergb, which does not encompass all the volume of the cielab space, resulting in an even smaller range of colors than what the human "eye sees". (On a side note possibly the reason for colormatch being the profile for the messenger mercury probe images, is simply because even though colormatch color space is small, no data is lost in image transfers after the image has been mapped to that particular color space) . The capacity of a ccd can be reduced to specific wavelength ranges by manner of filter wheels, or template overlaying on the ccd, but you are still counting photons in bins, which as I previously stated is different from how human color perception is achieved through relative comparisons of photo pigment responses (CIE 1931 XYZ color matching functions see nasa ames color research lab), and comparisons of those comparisons. One of the unique aspects of the human visual system is that it attempts to preserve an objects color even in lighting conditions that change from approx 10,000 Kelvin's (midday bluish), to about 2700 Kelvin's (sunset and sunrise reddish). The illuminant whether it is natural or artificial will have a spectral power distribution that can be represented by a tristimulus value, which will be the "white" in our field of view. The tristimulus of the illuminant can be represented in chromacity diagrams, showing the white point of that illuminant, 5000 Kelvin's = d50, and 6500 Kelvin's = d65, indicating the warmness or coolness of the particular white. Now the white point of the color space that the ccd image is mapped to is set by the color profile itself (see brucelindbloom.com for profiles and info), and therefore even though digital imaging devices allow the user to set the color temp, and white balance, is still is using discrete settings, which will then be mapped based on the color spaces white point, which is much different from how the human eye adapts to changing light conditions. An image captured on film or on ccd, is metered so the energy received over a period of time does not overexpose the best detail in the desired subject, which is accomplished by adjusting the exposure time or the lens iris diameter (f/stop), based on the films speed or equivalent digital settings. No imaging system is perfect, as optically there are trade offs, smaller lens iris diameters yields larger depth of field, but less resolution and longer exposure times, while larger lens iris diameter settings yield faster exposure times and high resolutions due to the greater amount of light being received, but at the expense of the images depth of field. All of these components whether they be manually set, or auto metered, determine what objects will be the shadows, highlights, and mid tone ranges of the image, (meaning those variables set the "blackest black" minimum luminosity or "whitest white" maximum luminosity in the image. Therefore an images range from maximum to minimum luminosity is not a function of the ccd or films range alone, but the amount of light received from the objects being viewed based on a number of variables (lens, f/stop, film speed/ccd specs, exposure time, desired zones, developing times for film, and print/image manipulation. Please see "zone system", and you will find that any image is a balance of capturing the subject's details in zone V, while at the same time not blowing out the details of the shadows zones III,II and highlights zones VII,VIII which is much different than the specs of the film or ccd (see luminous-landscape.com below for the simplified zone system description). Ok now, the whitest or most luminous object will be mapped as the highest point on the cielab z axis for that image, (with the slight biases introduced from the tristumulus value), the black or least luminous object will be the lowest point on the z axis with some biases, where the dynamic range is the difference of luminosities, minimum to maximum, and the logarithmic relationship from one to the other is the gamma (the connecting grey values in between). Printer's dynamic ranges are determined by the ink/media relationship or how much ink can be placed on the media usually making a CMY black, without running bleeding or buckling. A specific profile for printing is not just unique to that device, but is unique to that paper and ink set as well and requires setting the ink limits (described above) and is then followed by the careful balancing the colors and grays, that make the full color range or palette the printer can produce, meaning the images produced from a printer have a dynamic range that is a function of the inks and paper, and not the ccd. A monitors dynamic range is determined by the quality of the blackness of the screen, as compared to the best balanced and whitest white that can be achieved from the phosphor emissions, but the problem is a monitors phosphors change over time, meaning that the dynamic range is, dynamic, no pun intended, but that's the problem with monitors and keeping them calibrated. An images dynamic range on film can be determined by measuring the differences in the film densities in the most dense region (least luminous) setting the minimum, compared to the least dense (most luminous) setting the luminosity maximum, on a densitometer, which will show the final dynamic range is much less than the film is capable of producing, and was not just determined by the films range, but by the specifics that adjusted the exposure setting to the lighting conditions when the image was taken. Which once again shows that the human perception and color constancy are pretty unique attributes of human adaptation when were are compared to a device like a ccd or material like film. Now the greater amount of bits, the more information, but that information does not increase the dynamic range, it only parses the grey scale in between the minimum and maximum into finer sections, resulting in slight differences in the shadows, and mid-tones especially, but the same cielab space is utilized with the same limitations, the differences between 8 an 16 bit images are the 16 bit data is just parsed a little better resulting in smoother transitions, meaning 16 bit images aint all that you make them out to be. So therefore analyzing colors (or what you want to imply is missing) from an image strictly based on the ccd's or films specifications alone is not logical, and will yield incorrect results because that analysis does not the completely take into account system involved with producing the image. So brad, yes ccd's have great capabilities, but humans capturing the image can describe the object with words in such a way that compliments what the ccd produces, as humans are part of the viewing end of the equation, so therefore they should be on the image capturing end to better qualify the observed phenomena. Therefore once again it is human nature to creatively/subjectively describe events and sights (an observed events colors) with words that present a feeling to the reader that is far beyond the characters composing the text, and that is why humans must be part of space travel... Color Research Lab NASA Ames Research Center http://colorusage.arc.nasa.gov/lum_and_chrom.php Rochester Institute of Technolgy Munsell Color Institute http://www.cis.rit.edu/mcsl/ Information on color spaces, color conversions, etc. Bruce Lindbloom's website http://www.brucelindbloom.com/ Simplified Zone System http://www.luminous-landscape.com/tu...e_system.shtml Exposure value calculations The Science of Photography http://johnlind.tripod.com/science/scienceexposure.html |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Great missions STS-122 & Expedition 16 | columbiaaccidentinvestigation | Space Shuttle | 191 | March 11th 08 10:39 PM |
Expedition 14/Expedition 13/Ansari Farewells and Hatch Closure | John | Space Station | 0 | September 28th 06 09:58 PM |
Expedition 13/ Pontes/ Expedition 12 Joint Crew News Conference | John | Space Station | 0 | April 4th 06 03:42 PM |
Expedition 13/ Pontes/ Expedition 12 Joint Crew News Conference | John | Space Station | 0 | April 3rd 06 10:05 PM |