Doug Kerr
Well-known member
I have in the past used the original ColorVision Spyder monitor calibration/profiling system, latterly with the OptiCAL software suite. I had not been using it much lately, and when I decided to get my display up-to-date, various things suggested that the old system was inaccurate. (I did not have the proper instruments to really confirm that, but some creative comparisons seemed to confirm my suspicions.)
I decoded to buy a modern version of the system, in the Spyder 3 line. There are two basic packages:
• The Spyder 3 Elite system (~$250.00 street). This is intended for professional workers , and includes an extensive capability to "customize" one's work, and many supporting features, such as detailed reports of the results, with both tables and curves, the ability to use the instrument to make luminance and color temperature measurements on an arbitrary screen, and so forth. (I have essentially all this in my older system, with OptiCAL.)
• The Spyder 3 Pro system (~$150.00 street). This is intended for "serious amateur" workers (go figure), and (as we will see shortly) has limited ability to "customize" one's work, and essentially none of the supporting features I mentioned for the Elite version).
Owing to a general budgetary crunch, and for safety's sake (as I wasn't sure I would really like the Spyder 3 at all), I bought a Pro system. (One can upgrade the software for $99.00.)
My first work with it went smoothly.
First, so as to clarify the scenario, the calibration aspect of the process hopes to bring the display chain (video board plus monitor proper) to operate in a certain color space (chosen by us), which we call the target color space. This is done by manipulation of the color lookup tables in the video board. For various practical reasons, this cannot usually be attained.
Instead, the display chain, after calibration, exhibits a color space "something like" our target color space. Whatever it is, this is the color space than affects the monitor behavior when not operated from a profile-aware application (and for such basic things as displaying the desktop). I call this color space the "adjusted native color space" of the display chain.
Then, an ICC profile is developed for the display based on this adjusted native color space (that is, what is actually exhibited by the display chain after calibration). That profile allows profile-aware applications to transform an image from the working color space to the adjusted native color space of the display chain.
The Spyder 3 Elite software presumably allows the user considerable freedom in defining the target color space (as does my earlier system, OptiCAL). The Spyder 3 Pro software gives more limited freedom, essentially only allowing the user a choice of several different white points (all on the Planckian locus, and defined by color temperature) and several "simple" gamma curves (for gamma of 1.8, 2.2, and 2.4).
It does not, for example, allow choice of the true sRGB color space as the target (its white point is D65, not the same thing as 6500K, and its gamma curve is a more complicated one).
Note that this does not result in a compromise of the working of the display under a profile, since that profile is predicated on the actual adjusted native color space, which almost never is the target color space anyway.
But it might result in the actual adjusted color space not being as close to what we want (probably sRGB) as if our desire could be set as the target. (More on this later.)
Another aspect of target definition is the black and white luminance values. With Spyder 3 Pro, we do not have the ability to set our preferences there.
The instructions, for LCD displays, recommend that in general the monitor's own controls be set to the factory defaults.
Then it seems as if the software gives a white luminance for the adjusted display that is the highest possible in light of the balance between R, G, and B that are needed to attain the target white point.
In any case, once this is all done, we are asked to gave a name for the resulting profile, and we are done.
The LUT entries that provide calibration are stored in a special place (the vcgt - video card gamma table - tag) in the profile. They are not part of the profile itself (are not used by a profile-aware application).
The process is very fast - I think less than 10 minutes from start to finish.
The software package provides a "check" mode, in which the adjusted native color space is compared with the definition of the original target color space. It reports whether the gamma curve and white point are within some unmentioned tolerance of those defined for the target. The report screen has spaces for a report on black and white luminance, but in this situation, since we did not set those to particular values (we couldn't), no report on those (value or conformity) is given.
After the calibration/profiling process, a utility program is set up for automatic start at system startup. this provides the following functions:
• Loading of the calibration LUT information (from the vcgt tag in the active profile) to the video card at system startup.
• (If we want) Periodic monitoring of the ambient luminance (the Spyder 3 sensor has a sensor on its back for that, and we presumably have left it in its little stand on our desk). We can, during calibration, elect to have the calibration optimized for ambient luminance. With this feature of the utility, that can be automatically updated periodically based on measured ambient luminance.
• (If we want) Continual surveillance of the LUT settings in the video card, and, should they be changed by some other program (how dare it!) the ones described in the active profile's vcgt tag are promptly replaced.
I have commented that the Pro system allows limited flexibility in choosing a target color space. The available roster of combination of gamma and white point are carried in a set of target definition files. Presumably, if we added such files for other desriable target color spaced, they could be involved.
I have several made with OptiCAL, and the format seems compatible. It may be that the manufacturer offers target definition files for addition to the repertoire. I'm looking into that.
Will I upgrade to the Elite software? Well, I really enjoy report tables, and curves, and adjustment knobs, and the ability to make measurements, and all that. But will that help my actual result? I doubt it. So I'll probably slow-roll that until the stock market improves.
Of course, the real bottom line here should be, "how are my editing results with my updated display". That's another matter, on which no definitive result can yet be given. (For one thing, I am awaiting a supply of 5000K CFL lamps to put in my office lighting!)
What if we want to have the display "uncalibrated" and only generate a profile, based on the "unadjusted native color space". (This is recommended by various of our colleagues, most prominently Andrew Rodney.) Spyder 3 Pro does not give this possibility. The closest it comes are target definitions that bring the system gamma to one of three values but leaves the white point as "native".
It may be that the Elite version allows choice of this mode (OptiCAL does). I will look into this, as well as of ways to force it (perhaps by hand-coding of a target definition file).
I decoded to buy a modern version of the system, in the Spyder 3 line. There are two basic packages:
• The Spyder 3 Elite system (~$250.00 street). This is intended for professional workers , and includes an extensive capability to "customize" one's work, and many supporting features, such as detailed reports of the results, with both tables and curves, the ability to use the instrument to make luminance and color temperature measurements on an arbitrary screen, and so forth. (I have essentially all this in my older system, with OptiCAL.)
• The Spyder 3 Pro system (~$150.00 street). This is intended for "serious amateur" workers (go figure), and (as we will see shortly) has limited ability to "customize" one's work, and essentially none of the supporting features I mentioned for the Elite version).
Owing to a general budgetary crunch, and for safety's sake (as I wasn't sure I would really like the Spyder 3 at all), I bought a Pro system. (One can upgrade the software for $99.00.)
My first work with it went smoothly.
First, so as to clarify the scenario, the calibration aspect of the process hopes to bring the display chain (video board plus monitor proper) to operate in a certain color space (chosen by us), which we call the target color space. This is done by manipulation of the color lookup tables in the video board. For various practical reasons, this cannot usually be attained.
Instead, the display chain, after calibration, exhibits a color space "something like" our target color space. Whatever it is, this is the color space than affects the monitor behavior when not operated from a profile-aware application (and for such basic things as displaying the desktop). I call this color space the "adjusted native color space" of the display chain.
Then, an ICC profile is developed for the display based on this adjusted native color space (that is, what is actually exhibited by the display chain after calibration). That profile allows profile-aware applications to transform an image from the working color space to the adjusted native color space of the display chain.
The Spyder 3 Elite software presumably allows the user considerable freedom in defining the target color space (as does my earlier system, OptiCAL). The Spyder 3 Pro software gives more limited freedom, essentially only allowing the user a choice of several different white points (all on the Planckian locus, and defined by color temperature) and several "simple" gamma curves (for gamma of 1.8, 2.2, and 2.4).
It does not, for example, allow choice of the true sRGB color space as the target (its white point is D65, not the same thing as 6500K, and its gamma curve is a more complicated one).
Note that this does not result in a compromise of the working of the display under a profile, since that profile is predicated on the actual adjusted native color space, which almost never is the target color space anyway.
But it might result in the actual adjusted color space not being as close to what we want (probably sRGB) as if our desire could be set as the target. (More on this later.)
Another aspect of target definition is the black and white luminance values. With Spyder 3 Pro, we do not have the ability to set our preferences there.
The instructions, for LCD displays, recommend that in general the monitor's own controls be set to the factory defaults.
Then it seems as if the software gives a white luminance for the adjusted display that is the highest possible in light of the balance between R, G, and B that are needed to attain the target white point.
In any case, once this is all done, we are asked to gave a name for the resulting profile, and we are done.
The LUT entries that provide calibration are stored in a special place (the vcgt - video card gamma table - tag) in the profile. They are not part of the profile itself (are not used by a profile-aware application).
The process is very fast - I think less than 10 minutes from start to finish.
The software package provides a "check" mode, in which the adjusted native color space is compared with the definition of the original target color space. It reports whether the gamma curve and white point are within some unmentioned tolerance of those defined for the target. The report screen has spaces for a report on black and white luminance, but in this situation, since we did not set those to particular values (we couldn't), no report on those (value or conformity) is given.
After the calibration/profiling process, a utility program is set up for automatic start at system startup. this provides the following functions:
• Loading of the calibration LUT information (from the vcgt tag in the active profile) to the video card at system startup.
• (If we want) Periodic monitoring of the ambient luminance (the Spyder 3 sensor has a sensor on its back for that, and we presumably have left it in its little stand on our desk). We can, during calibration, elect to have the calibration optimized for ambient luminance. With this feature of the utility, that can be automatically updated periodically based on measured ambient luminance.
• (If we want) Continual surveillance of the LUT settings in the video card, and, should they be changed by some other program (how dare it!) the ones described in the active profile's vcgt tag are promptly replaced.
I have commented that the Pro system allows limited flexibility in choosing a target color space. The available roster of combination of gamma and white point are carried in a set of target definition files. Presumably, if we added such files for other desriable target color spaced, they could be involved.
I have several made with OptiCAL, and the format seems compatible. It may be that the manufacturer offers target definition files for addition to the repertoire. I'm looking into that.
Will I upgrade to the Elite software? Well, I really enjoy report tables, and curves, and adjustment knobs, and the ability to make measurements, and all that. But will that help my actual result? I doubt it. So I'll probably slow-roll that until the stock market improves.
Of course, the real bottom line here should be, "how are my editing results with my updated display". That's another matter, on which no definitive result can yet be given. (For one thing, I am awaiting a supply of 5000K CFL lamps to put in my office lighting!)
What if we want to have the display "uncalibrated" and only generate a profile, based on the "unadjusted native color space". (This is recommended by various of our colleagues, most prominently Andrew Rodney.) Spyder 3 Pro does not give this possibility. The closest it comes are target definitions that bring the system gamma to one of three values but leaves the white point as "native".
It may be that the Elite version allows choice of this mode (OptiCAL does). I will look into this, as well as of ways to force it (perhaps by hand-coding of a target definition file).
