Category Archives: Glossary

Color depth

The color depth indicates how many different color levels are available for each individual pixel of a graphic. Since the “fineness” of the gradations depends on how much memory is used per pixel, the color depth is specified in bits.

With 8 bits, for example, 256 color shades can be distinguished for one color channel. A color is created by mixing several color channels of a color space. For computer graphics, the RGB color space is usually used, in which colors are composed by additive mixing of the three primary colors red, green and blue. Even most common computer monitors can only distinguish 8 bits per channel.

One speaks of a true-color representation if the current color depth has at least 24 bits, i.e. 8 bits per color channel (red, green, blue). A color depth of 24 bits corresponds to approx. 16 million colors; this means that practically every conceivable color can be reproduced true to life.

Calibration

In the field of color management, the terms calibration and profiling are often used interchangeably, although they are actually two different processes.

In digital imaging, all input devices (e.g. digital camera, scanner) as well as output devices (e.g. monitor, printer) should be calibrated and profiled to avoid color errors due to faulty devices.

During calibration, devices are trimmed to technical boundary conditions (on the screen, for example, compliance with a specific color temperature); during profiling, on the other hand, the devices are measured and an associated ICC profile is created. Strictly speaking, calibration takes place before profiling, but is often performed in a single pass using special tools.

The difference between calibration and profiling can be clearly understood using the example of a digital bathroom scale: When the scale is turned on, it is usually calibrated automatically so that when it is completely unloaded, the display shows 0 kg. Profiling, on the other hand, would mean that reference measurements are created for a certain number of reference weights (10 kg, 20 kg, 30 kg, etc.) and that a correction value is in turn stored in a profile for each reference weight (e.g. that the scale displays 5% too much when loaded with a weight of 10 kg, only 3% too much when loaded with 50 kg, and only 1% too much when loaded with 100 kg).

Color calibration

A professionally performed color calibration serves to optimize and measurably compare the image quality of a scanner.

The perception of colors by the human eye is extremely subjective and also dependent on the surrounding lighting conditions. In addition, every imaging device, whether digital camera, scanner, monitor or color printer, has device-specific color interpretations that are reflected in different color models, color spaces or color tones.

Therefore, a professional color calibration, individually adapted to the respective camera sensor, when installing a scanner guarantees accurate colors right from the start.

Color calibration is usually performed using a color target that contains standardized color patches. This is scanned and then the calibration software compares the colors determined by the scanner with the target colors (the actual colors of the individual color patches on the calibration slide are standardized). From the comparison of the actual colors with the target colors, a device-specific color profile (ICC profile) is created, which is now used for each subsequent scan. In this way, individual color errors of a scanner can be corrected. Such an ICC profile is always created for a specific scanner only and cannot be transferred to another device of the same construction.

The color calibration should always be done using professional color targets with color reference standards that correspond to the common digitizing standards.

In the professional digitizing sector, the creation of a color profile is part of the standard scope of delivery.

Color management

Unlike the human eye, which can perceive almost any number of different colors, each image-processing input and output device has its own finite color space, called a device-specific color space. For example, a normal RGB screen represents colors from a combination of 256 red, green and blue tones each; this corresponds to a maximum number of 16,777,216 displayable color tones. But even with this enormous number, not every color that the human eye perceives can be represented. Moreover, even devices that work in the same color space reproduce colors differently.

Therefore, in order to ensure uniform color reproduction across different devices – for example within the process chain of scanner, image-processing PC, desktop publishing viewer and printer – the image data must be digitally matched or offset against each other. This is done by special color management modules. They create color profiles for the respective devices, which describe the colors in relation to a reference color space. The conversion is done on the basis of ICC profiles with so-called rendering intents.

Color profile

Jedes bildgebende Gerät, ob Digitalkamera, Scanner, Monitor oder Farbdrucker hat gerätespezifische Farbinterpretationen, die sich in unterschiedlichen Farbmodellen, Farbräumen oder Farbtönen bemerkbar machen. Eingabegeräte wie Digitalkameras und Scanner sowie Ausgabegeräte wie viele Monitore und Displays nutzen das RGB-Farbmodell, währen Drucker dagegen das CMY- bzw. CMYK-Farbmodell verwenden.

Ein Farbprofil dient nun dazu, Farben von einem Farbraum in einen anderen Farbraum zu übersetzen, ohne dass die Farbechtheit der Ausgangsvorlage dadurch verfälscht wird. In einem Farbprofil ist gespeichert, wie ein bestimmtes Gerät Farben gegenüber einem geräteunabhängigen Farbraum darstellt. Die Erstellung der Farbprofile erfolgt gerätespezifisch durch Farbmanagementsysteme, die Umrechnung in der Regel auf Basis von ICC-Profilen mit sogenannten Rendering Intents.

Ziel eines Farbprofils ist somit, eine unveränderte Farbwiedergabe auf allen in einer Imaging-Prozess-Kette (z.B. Scanner, bildbearbeitendem PC, Desktop-Publishing-Viewer und Drucker) miteinander verbundenen Geräten zu erreichen.

Im professionellen Digitalisierungsbereich gehören solche Farbprofile zum Standard-Lieferumfang.

UTT

The Universal Test Target (UTT) is a single test chart developed by the Dutch National Library (KB) in collaboration with Image Engineering Dietmar Wueller (IE) and the Fachverband für Multimediale Informationsverarbeitung e. V. (FMI) as part of the Metamorfoze initiative.

The aim was to design a new, uniform test chart that would incorporate the five standard test charts that had been in use until then and thus simplify handling. This was to provide an insight into the overall image quality of the scan results of all types of high-end cameras and scanners, based on current ISO standards. These are captured and analyzed with special software to provide information on technical aspects such as OECF, MTF, noise and color accuracy.

UTT is available in two versions: measured and unmeasured. The “measured” version comes with individually measured reference data for the particular chart purchased. The “unmeasured” version is produced in respect of the Metamorfoze standards and tolerances.

The UTT is available with a variety of options in sizes from DIN A3 to A0. Because the UTT is applicable to all types of digitization projects and preservation, it is particularly important for libraries, archives and museums.

The aim of the developers of the UTT was to save time and improve quality by using the unified target during the digitization process. However, it should be noted that the UTT is extremely error-prone in practice due to its simple design with individually affixed test charts and must therefore be handled with extreme care.

Test charts

Testcharts werden benötigt, um bei der Erstellung eines Farbmanagements die Gerätecharakterisierung für eine Kamera, einen Scanner oder einen Drucker zu erstellen. Sie dienen der objektiven Messung der Genauigkeit bzw. der Eigenschaften eines Bildverarbeitungssystems, um dessen effektive Funktion sicherzustellen und langfristig zu gewährleisten.

Testcharts können aus physischen Vorlagen bestehen oder als Fenster in das Bildverarbeitungssystem integriert sein.

Zur Qualitätsmessung im Digitalisierungsbereich wird in der Regel mit physischen Testcharts gearbeitet. Dabei ist zu beachten, dass diese jedoch Alterungsprozessen unterliegen und ausbleichen können. Sie sollten daher sorgfältig und vor Licht geschützt gelagert werden.

Testcharts beinhalten Linien-, Punkt- oder andere Muster sowie zur Überprüfung von Farben eine definierte Anzahl von Farbfeldern, deren Zusammensetzung aus den Prozessfarben Cyan, Magenta, Gelb und Schwarz bekannt ist. Zu unterscheiden ist zudem zwischen geordneten (visual) und ungeordneten (random) Testcharts, bei denen die Farbfelder zufällig angeordnet werden.

Je nach Aufbau und Komplexität der Testcharts lassen sich damit verschiedene Kriterien der verwendeten Kamera bzw. des Objektivs testen: u.a. die Schärfe in den kritischen Bereichen wie  Bildmitte und Bildrand, chromatische Aberration, Verzeichnung, Vignettierung, Auflösung, Farbwiedergabe, Kontrastumfang, Weißabgleich, Autofokus-Probleme und Bildrauschen bei unterschiedlichen ISO-Einstellungen.

Zur Abgleichung der speziellen Anforderungen werden zumeist verschiedene im Handel erhältliche standardisierte Testcharts eingesetzt. Als Universalstandard wurde zudem auf der Basis aktueller ISO-Standards das Universal Test Target (UTT)  entwickelt, um einen Einblick in die gesamte Bildqualität der Scanergebnisse aller Typen von High End Kameras und Scannern zu erhalten. Allerdings ist zu beachten, dass das UTT aufgrund seiner einfachen Ausführung mit einzeln aufgeklebten Testcharts in der Praxis extrem Fehler anfällig ist und daher äußerst sorgsam behandelt werden muss.

Bei book2net arbeiten wir daher vorzugsweise aus einer Kombination folgender Testcharts:

Gray chart

A gray chart is used to measure the dynamic range or tonal value differentiation of a digital camera, scanner or monitor. It is a test chart in which there is a smooth or multi-graded (gray scale chart) transition between dark black and bright white.

Dynamic range

In photography, the dynamic range describes the difference between the lightest and darkest point within an image. It is given as the ratio of the darkest to the lightest point.

In an original (slide, negative, photo), the brightest point has a so-called minimum density and the darkest point has a maximum density. The difference between the maximum density and the minimum density is then the so-called density range, or dynamic range.

The dynamic range of the subject or the original is crucial for determining the correct exposure: Only if the exposure range of the sensor or film is greater than or equal to the dynamic range of the subject can all the details of the subject be captured. Otherwise, parts of the subject will be imaged in black and/or white.

White balance

The white balance adjusts the scanner’s camera to the prevailing lighting conditions to ensure a uniform color temperature and thus constant image quality. The color sensitivity of the sensor is thereby adjusted to the respective lighting conditions.

The perception of white is strongly dependent on the environment and the respective light source. In daylight, a pure white appears colder than in fluorescent light, for example, where it appears greener. Even in a controlled environment, this can become a problem.

While the human eye or brain can adjust to the respective light situation and quasi intuitively performs a white balance and compensates for color casts, a digital camera mercilessly reproduces them: If the digital camera is set to daylight, for example, and the picture is taken in artificial light, the image will appear reddish.

A user-defined white balance directly during installation therefore guarantees accurate colors right from the start. Since a change in the lighting environment always necessitates a new white balance, a constant and uniform lighting situation should generally be ensured when using planetary scanners.

White balance should be achieved by means of a full-format scan of a surface that is as white or at least neutral gray as possible. Plain white paper is often used for this purpose, but it often contains optical brighteners and can therefore lead to distorted results with color casts.

At book2net, we therefore use high-precision, spectrally neutral white balance targets to ensure precise user-defined, in-camera white balance under local lighting conditions.