Colour and Resolution Targets
In order to accurately calibrate and benchmark imaging devices, it is standard practice to capture colour targets and resolution targets. Capturing a target and assessing the results can give an accurate indication as to how your system is performing. This document lists a variety of targets that are used to measure a capture device's ability to resolve detail or record colour information. Some of these targets are widely used to calibrate cameras and scanners as part of an effective open loop colour system.
One of the most important steps in any digitisation project is ensuring the hardware used for data capture is performing to an established standard or benchmark. This process will involve first calibrating equipment and then employing colour management software to 'characterise' the device, creating an ICC (International Colour Consortium) colour profile that should provide consistent and predictable results from the various devices in an imaging workflow.
In order to accurately calibrate and benchmark imaging devices, it is standard practice to capture colour targets and resolution targets. Capturing a target and assessing the results can give an accurate indication as to how your system is performing. For instance it may reveal weaknesses in a scanner that does not capture the full vibrancy of some colours, or the sharpness of detail that you require.
It is important to remember that all monitors within the workflow should be calibrated for colour, white point and gamma prior to making any judgements on the performance of a scanner (or camera). Once the monitor has been calibrated, a scanned target can be assessed (either subjectively by eye or objectively by colour-photospectrometer) and calibration of the scanner can be performed or an ICC profile created.
It is not always easy to understand the difference between 'calibration' and 'characterisation' of a digital device.
- Calibration is the establishment of the imaging device to a known state. This ensures that the device is providing consistent results each time it is used and with other machines with which it exchanges work. Calibration is always the first and imperative task in any colour management system.
- Characterisation is the establishment of the difference of any colour between what is created and what should have been made. These differences are taken for a large range of colours and can provide a large 'Colour look-up table' (CLUT), which provides the central information that makes up an ICC colour profile.
A colour target, or colour chart, is essentially a hard-copy image containing a range of colour patches of known values that can be compared either objectively or subjectively with their original and then used to calibrate the device or create colour profiles for use within an ICC Colour Management System.
Additional objects such as highly reflective spheres and light traps may also be used alongside a colour target to provide extreme highlights and deep shadows. These white and black points are particularly useful in a workflow using RAW files, during the RAW conversion process the RGB colour values can be measured and adjusted at a range of points from highlights through to shadows
Colour targets are available either as transparencies for film scanners (transmissive) or printed hard-copy for cameras and flatbed scanners (reflective). The target is scanned or photographed and the resulting output is compared to the original hard-copy and/or the target's data files in order to determine the accuracy of the imaging device or create an ICC profile.
As colour can be corrected in imaging applications such as Adobe Photoshop, the point of calibrating a scanner as exactly as possible may not be obvious. However, when corrections are made after an image has been scanned, unwanted effects may be introduced such as contouring (where distinct bands appear across the image) and inaccurate tonal relationships (where colour tones are distorted due to rounding errors in the calculations made within the software). This demonstrates that best practice is to make sure that the original scan is as accurate as possible.
The various types of target available can be divided between those which offer an objective means of comparison and those that work on a more subjective level.
To be useful, objective colour targets have to be manufactured to the highest standards. They are not copies of a master image but are individually produced items. Their colour patches are measured with extremely accurate apparatus and the measurements, or values, are supplied as data files with the target.
The accuracy of the measurements provided can be divided into three levels:
- Generic average values - targets are printed to established values and these are the values supplied. These are the cheapest type of targets
- Batch accuracy - a few targets within each batch produced are measured and an average set of values is published for the whole batch. These are more expensive
- Specific bespoke values - each individual target is read and values are as accurate as possible. These tend to be very expensive
A comparison is made between the known target values supplied by the manufacturer and those recorded by the imaging device. A 'profile' for the device can then be created by way of a 'colour look-up table' which adjusts the device's colour values in order to produce an accurate rendering of the captured image.
For creating an ICC profile the standard target used is the IT8, available as transmissive (IT8.7/1) and reflective (IT8.7/2). An example of an IT8 is Kodak's Q-60 which includes 240 colour patches and a grey scale ramp. It comes in both 35 mm and 4x5 inch formats as an IT8.7/1, and on 5x7 inch paper as an IT8.7/2. The target data files can be downloaded from the Kodak Web site.
Diagram 1. The Kodak Q-60R1 Colour Target
Fuji, AGFA and other manufacturers also produce IT8.7/1 and IT8.7/2 targets, and many new scanners now include them. Many of these are also available from professional photography and digital equipment dealers, e.g. Colour Confidence.
Another industry standard target is the X-Rite/GretagMacbeth ColorChecker. This target only provides 24 patches rather than the IT8's 240 but does come with published values making it comparable to the IT8. It does not provide enough patches or accuracy to be used to create an ICC colour profile but can still provide some objective colour measurement. The smaller number of carefully chosen patches can also be used subjectively by making a visual comparison between the actual ColorChecker and the captured image. Further details are available from X-Rite. See also X-Rite's list of UK suppliers.
Diagram 2. The X-Rite/GretagMacbeth ColorChecker
Subjective targets provide the user with a means of visually comparing the differences between the target and the output image (e.g. by holding up the target to the monitor). They are not usually supplied with values, although the GretagMacbeth ColorChecker mentioned above, which is more normally used as a subjective target, does provide known values for its 24 patches.
Another common example that is the traditional favourite of the photographic trade is Kodak's Q-13 target, a reflective target with only 17 colour patches (two tints of the subjective and additive primaries, RGB & CMY, and a grey scale ramp). In the past these colour targets were perceived as providing enough accuracy for normal photographic use, however they are arguably not really up to the demands of providing any accuracy within a digital imaging workflow. Its limitations are reflected in its cheaper price.
Diagram 3. The Kodak Q-13 Target
Adobe's Ole No Moiré Target
Other subjective targets are provided with imaging software, such as the Ole No Moiré image that is supplied with Adobe Photoshop. However, it should be noted that this image is not an input test image - it does not exist in print form, but is a standard digital image that can be printed to allow visual comparison back to the computer's monitor and to other prints made on comparative devices.
Diagram 4. The Adobe Ole No Moiré Target
Grey scale targets
Grey scale targets, or ramps, are often included in colour targets. These have a number of patches graduating in steps from black to white and the objective is to be able to discern all of the patches on each output device in the imaging system. They can also be used to provide colour information, as each step should provide a balanced level of the three colour channels within it. Like colour targets, grey scale targets can either be employed objectively with reference to data files (e.g. the GretagMacbeth ColorChecker), or they can be used subjectively (e.g. the Kodak Q-13).
A grey scale target can be used to make sure that each step is both correctly grey (equal amounts of RGB), as well as ensuring that each step progresses through the range of values in equal measure, i.e. each step provides a jump of 12 values within the RGB value.
In general grey scale ramps provide the easiest and most revealing test data of any test targets here, allowing colour, contrast, and gamma evaluation. This is particularly true when they are being used subjectively, as the human eye is very sensitive to slight differences and tonal shifts within grey.
The amount of spatial detail that an image capture device can resolve can also be tested for by capturing an image of a resolution target.
Resolution targets contain a wide range of differently spaced black lines, and other geometric patterns, on a white ground. An image of a target is captured and the resulting output (e.g. monitor, printed page) is examined to ascertain the maximum number of linepairs per inch (lppi) or linepairs per millimetre (lp/mm) that are resolved in each direction (horizontally and vertically), as it is common for scanners to have a different resolution in one axis to the other. These targets can also be used to test for any inconsistency in resolution across the capture area.
The USAF 1951 resolution target
One of the most common examples of resolution target is the USAF 1951, which consists of bars organised in groups of three. The resolution of an imaging device can be measured by assessing at which point a group appears blurred and taking an lp/mm reading from a resolution table. Although the test yields precise resolution values, this is a qualitative test which depends on the user's definition of acceptable blur.
Diagram 5. The USAF 1951 Resolution Target
The ISO Camera Resolution Chart
The ISO Camera Resolution Chart can be used to perform various types of resolution measurement, including a visual test with horizontal, vertical, and diagonally oriented 'wedges', and an automated measurement test using slightly slanted black bars. Further details are available from the International Imaging Industry Association (I3A). The I3A lists Armstrong Optical as a UK supplier for resolution charts.
Diagram 6. The PIMA / ISO Camera Resolution Chart
While it is possible for extremely experienced operators to subjectively calibrate imaging devices within a 'closed loop' system 'by eye', it is better practice to utilise the objective abilities of a Colour Management System's profile building software to create an ICC profile using an IT8 target.
Even after you have calibrated your system and established an ICC Colour Management System you may still need to make adjustments to individual images. If an 'incorrectly-coloured' original is scanned you may need to make colour and tonal adjustments at the post scanning stage, but the use of colour and resolution targets will ensure that your equipment is performing (and continues to perform) to an established benchmark.
Note: Whichever targets you use within your workflow, remember that they must be looked after and replaced if they have in any way become damaged. If a target fades, the accuracy of the data files will no longer be reliable, and either a new target will be required or the target data files will need to be recreated. So keep targets clean and out of bright light, and replace them frequently. Keep a record of all targets used and when they were purchased/replaced.
See the JISC Digital Media advice paper on Colour Management for a more in depth look at ensuring colour consistency in an imaging system where a Colour Management System is employed.