photo of Amethyst crystalphoto of Strassburg in Elsassphotomicrograph of micrasteria algaephotomicrograph of petrified wood

        Georg N. Nyman PhD.

 

IT8 COLOR TARGET - Example: FUJICHROME PROVIA

 

All machines see color differently as do different humans - red is not red and green not green. This very trivial statement means that as soon as a certain color is "transferred" from one machine to another one, say from a scanner to a printer (or from your digital camera to your printer or monitor), that very same color will probably be at least slightly different. For many amateurs, this does not matter too much as long as these differences are not too large, but if a photographer needs or wants to obtain and maintain a certain color - as example the color of a company's logo or the subtle tone of a model's skin, it is necessary to do "color management".

Color management includes, among other tasks and topics, using a standardized color target, a test chart, for understanding how a certain device reads colors and how they are transferred to another piece of equipment in such a way that the original colors remain correct . This test chart is the IT8 Color Target.

The IT8 Target

An IT8 target is either a color slide (IT8.7/1) - available in two film formats, 35mm and 4x5inch, or a color print (IT8.7/2) in 5x7inch size. The slide target is used to profile a transparency scanning system, whereas the color print target is used for profiling the reflective scanning system of a scanner.

For completeness, I would like to add that there are color negative film IT8 targets existing (to my best knowledge, Fuji is the only provider still offering them as part of the Fujifilm Color Kit) as well, but there is no common color negative film standard existing in the same sense as for color slide (color reversal) film. As you know, all current color negative films have an orange mask as base layer and the color of this layer differs from film type to film type and from manufacturer to manufacturer but a color slide (reversal) film has always a black mask for the unexposed areas and the layer itself is clear, colorless transparent.

I presume that it is clear that one should use a Fuji target in case one works with Fuji film and a Kodak target if one works with Kodak film - the best solution is of course to use exactly the same film as the target was made on, but at least one should use the same brand of target and film - it is not recommended to mix a target of brand A and work with films of brand B.

To explain the features of an IT8 target, I use one of the Fuji color slide films - Fujichrome Provia. Here you see that IT8 target:

It declares itself to be consistent with the ANSI (American National Standards Institute) standard IT8.7/1-1993 (Graphic Technology Color Transmission Target for Input Scanner Calibration) and was made as batch 08/1999. The IT8 stands for Information Technology Committee.

One can see two different areas - one large area with 12x22 squares and a smaller, long one, which contains 22 shades of grey plus D(min)=white and D(max)=black. The 84 color elements on A-L and 1-19 and the 36 skin tone elements on A-L and 20-22 are sufficient to accurately define and measure the color response of any suitable equipment. In total there are 288 patches on this target (Fuji, Agfa, Monaco...) - if the skin tone patches are replaced by a portrait (Q-60=Kodak version), then there are 252 patches on that target. In addition, one can see four small grey squares with two inside white borders - one at each corner - they are alignment marks for the profiling procedure.

The rows A-L are for hues (0-360degr - the hue circle divided into 12 sections) more or less equally spaced, the columns 1-12 are for Chroma C* - subdivided into three sets, corresponding to three levels of lightness L*, 1-4, 5-8 and 9-12. The columns 4, 8, and 12 are for that L* the maximum C* which can be achieved with the film used (boundary gamut). For one particular row, as example A - all patches from 1 to 12 are at the same hue angle, each row therefore represents a different hue angle.

Columns 13-19 are for the colors C, M, Y, Grey, R, G and B in 12 different lightness steps. The grey scale in the middle part of the target is evenly divided according to L* - therefore the grey steps appear equal to the human eye.

Columns 20-22 / rows A-L are representing various skin tones. This area can either be filled by squares with skin tones or with an image of a human, as one can see it on the comparable Kodak target. Fuji prefers the color squares, which are also my preference, but there is no standard procedure prescribed what has to be in those fields.

The second grey scale at the lower part of the target is divided in 22 steps and can be used for adjustment of the linearity of a scanner, tone reproduction and similar tasks. On the left end, there is D(min), outside the step 1 and on the right end, outside step 22, there is D(max), thus the minimum and maximum density which this particular film can reproduce are included in the target.

When you get such a color target, you do not only get that slide film or color print, you also get a set of data, the reference file, together with the film and the print - usually stored on a floppy disk.

Below you see a screenshot of a part of the reference file for that Fuji color target:

You can see when this file was done, what lighting conditions, what film material, manufacturer etc. After manufacturing the target, it was most likely measured with a high precision spectrophotometer and the LAB and/or XYZ values for each patch have been determined. The file is a simple text file.

What are X,Y,Z ? They are called the "tristimulus values" and are the result of the multiplication of the curve Reflectance Intensity vs Wavelength for the illuminant (in the case of this target it was light D50 which is standard white 5000K), the sample itself (in our case here each patch) and the "standard human observer".

This standard human observer - you can see the spectral curves on the left - represents the spectral response for red (x), green (y) and blue (z) of a "normal" human eye. The x axis is the wavelength and the y axis the tristimulus values for a particular wavelength.

Therefore, a red patch will have a high X value, a green one a high Y and a blue one a high Z value.

 

 

 

 

As the XYZ systems is not very practical, one usually converts these data into the LAB system. This LAB color system specifies any color by its position inside a color sphere - in 3D coordinates.

LAB separates color information into lightness L* and color information a* and b* on two axes - red/green = a* axis and yellow/blue=b* axis. L*=0 means black and L*=100 means white. Darker colors are located below the center axis, brighter colors above and the more pure a color is, the more it is located towards to outside boundaries. In the middle axis one finds all shades of grey from white to black.

 

 

 

 

A variation of the LAB system is the LCH system. LCH stands for Lightness, Chroma and Hue. LAB and LCH share the same color space, the difference is the positioning of a color. LAB uses a rectangular system of coordinates whereas LCH a cylindrical system (polar coordinates) - but the color remains in both systems at the same position.

 

............................

 

 

 

Now back to the IT8 target...these reference file data are the absolute measurement data of that very same target, you are using to profile your scanner with - by comparing this set of data with the output from the scanner, a program (often a scanner supplied program) calculates the differences between target data and scanned data and this difference is then used as basis for all color calculations on other images and slides.

The results are put together in another file which contains all information necessary to characterize the spectral behavior of a scanner - the ICC profile. Again the ICC profile is nothing more than a set of data and some additional information which is used by programs like Photoshop to understand how a scanner reproduces colors. As it would be beyond the scope of this short introduction into IT8 targets to explain the structure of ICC profiles, I suggest for those, who want to know the details to browse the many available pages here: Color Consortium

To conclude, I want to show you two photos of an ALPA12SWA medium format camera, the left one without color management and the right picture with applying the ICC profile to the images - the difference is striking!

 

 

 

Back to: HOMEPAGE NYMAN

Back to: Color Management

 

Hit Counter

 

 

Send mail to Webmaster with questions or comments about this web site.
Last modified: 19-May-2007