The resolution describes the ability of a recording system to reproduce fine details. This can be determined for the image center and the four outer regions with the Siemens stars located on the lensTESTER. The calculated result is an indication of the resolution that is determined with respect to the contrast. This in turn describes how well light and dark lines can be displayed. Ideally, this is 100%. However, with increasingly finer details the contrast performance decreases. Below 10%, light and dark lines can no longer be distinguished.* Therefore, this value is set as a threshold for determining the maximum achievable resolution (limiting resolution in %).
LP / PH:
The resolution for lenses is usually stated in LP/mm. The specification describes how many black&white changes a lens can display per millimeter. However because the test method refers to the performance of a complete camera system, the size of the camera sensor is integrated into the assessment of image resolution. If the measurement of resolution here is set as “LP/mm”, a comparability is no longer given. To illustrate this, look at the following example.
A lens that can display 40 line pairs per mm is used in two cameras with different sized sensors.
(Lens resolution: 40 LP/mm | Sensor A: 36 x 24mm | Sensor B: 24 x 16mm)
In combination with Sensor A, 960 line pairs are shown at a sensor height of 24. In contrast, Sensor B is at 640 line pairs per height.
This makes it clear that the resolution is dependent on the sensor height and cannot be considered on an equal level at a setting of 40 LP/mm. Therefore, a statement in line pairs/picture height (LP/PH) is used for a comparison of the resolution, regardless of the height of the camera sensor.
Limiting resolution in %
The specified limiting resolution in percentage gives a result as one indication of the achieved performance of the recording system used. It refers to the theoretical maximum resolution of the recording system, which is described by the Nyquist frequency. If a camera achieves a resolution of 1600 LP/PH, it covers 80 percent of the theoretically possible performance at a maximum vertical sensor pixel count of 4000 pixels.
Vertical pixel count sensor = 4000 → max. 2000 LP/PH possible (2 pixels per light-dark transition)
measured resolution: 1600 LP/PH
proportionally: 2000 LP/PH = 100% | 1600 LP/PH = x%
(1600/2000) *100 = 80%
The graphical representation in the spider diagram illustrates the resolution achieved for the image center and corners. The results are indicated numerically for the central Siemens star as well as individually and as an average for the corners.
The edge sharpness is expressed by the so-called acutance. A statement about the visually perceived sharpness impression can be made.
The acutance is determined with respect to the Contrast Sensitivity Function (CSF). Here the CSF describes how the eye perceives spatial frequencies, i.e. the number of line pairs per visual angle, under a given viewing condition in relation to the contrast.
The acutance is determined as a percentage for a defined viewing condition. The higher the score, the better the perceived sharpness impression, based on viewing conditions. The result refers to a 100% monitor display (96ppi) at a distance of 0.5 meters. The ideal value is 100%, for values greater than 100% images are often oversharpened.
Because of how lenses are constructed, the quality of the imaging performance can vary due to various factors. Firstly, the optical elements have characteristics that lead to aberrations or field curvatures. On the other hand, there is the construction of the lens system itself. Well-bonded lens groups and precisely placed lens elements are decisive quality aspects. If the lenses are minimally displaced from their optimal calculated position, known as decentering, the optimum beam path is no longer given. This appears in the image result, for example, when the resolution significantly drops to one side and the image appears more blurred than on the opposite side. Reasons for decentering may be shocks or in rare cases a faulty lens (misalignment).
The index “Centering” provides information about the deviation of the resolution based on the four outer regions. The result obtained is to be regarded as a reference value and is not intended as a precise description of a decentering. Ideally, the Centering value is "0".
The reasons are that in addition to the lens system, a minimal-angled camera sensor may be responsible for the deviation to the edges. In addition, the orientation of the camera, the focus and the light setting could affect the measurement results.
A higher drop in resolution to an outer region is displayed in the test image. There is a high centering value in the evaluation. With a good camera alignment, a uniform illumination and precise focus, it stands to reason that one part of the recording system is responsible for the deviation. Now it should be determined whether the camera or the lens is responsible for this decrease. To determine this, take test images using other lenses. If the results are not affected by the same uneven performance, it seems likely that the lens is the error source.
Digital SLRs generally have a phase autofocus for the autofocus. The accuracy of the autofocus system depends on the interaction between camera and lens.
The reliability of the recording system’s autofocus can be determined by an auto-focus shooting series. The determination of the autofocus accuracy can be analyzed for both the phase autofocus and the contrast autofocus.
In the results overview the results are presented in form of a graphic representation. Each number of the x-axis represents a recording. The y-axis displays the results of each image for the center and the corner (mean value). The higher the value, the higher the calculated limiting resolution. Based on the camera system used, the information is represented as percentage of Nyquist.
Now, multiple statements can come together with the help of this visualization. On the one hand, it can show how harmoniously the autofocus is working. If the horizontal line is nearly straight, it shows a fairly consistent performance, and vice versa. On the other hand, the presented resolutions indicate how accurately the system is operating. It may be worthwhile to compare the results with those of a manual-focus test recording. If the results are similar, the auto focus is precise. Otherwise, it should be considered whether it is preferable in this case, or when there are large fluctuations within an AF series, to use the manual focus mode instead.
DSLR: Comparison of phase and contrast autofocus
Photographers who own a digital SLR are likely to be interested in a comparison of the phase autofocus with the contrast autofocus active in Live View mode. The tests carried out here are done once in the deactivated and again in the activated Live View mode. Therefore, you can check how much the two versions differ from each other and which one is best for the majority of your shots.
Aperture, focal length and ISO series
The graphic representation in the result overview provides the possibility of displaying the calculated limiting resolution with respect to aperture and focal length settings. A conclusion can be drawn concerning the performance of the lens. The respective graphs show the central and the average limiting resolution with respect to the aperture and focal length settings you have set.
The graphic representation is complemented by the numerical results display. The calculated results are displayed for the center and the average of the corners. The numerical values describe the best resolution results for the whole measurements series - overall, for the center and the corner. In addition the associated f-stop number of the image is given.
The imaging performance of lenses varies with the set aperture. Edges often show a drop in the case of resolution with a large wide aperture. If the aperture is further closed, the result in the image field becomes more homogeneous. In addition, the performance is limited by the so-called diffraction blur at a certain aperture setting.
The aperture series will provide information about how the lens reacts at the different aperture stages. Thus, you can better target your results in practice.
Focal length series
You can determine how the resolution behaves in the zoom range of your lens through a focal length series. The graph contains the resolution in percentage, plotted against the focal lengths of the test images.
In some cases weaker lighting conditions require setting a higher ISO level. However, with increasing sensitivity the risk of noise also increases. Depending on the camera sensor’s size and pixel density as well as the in-camera processing, disturbing effects often occur with increasing sensitivity. The often integrated noise reduction as well as the noise itself can lead to weaker results with respect to the resolution. Record the test image using different ISO settings and test the behavior of the camera. Which sensitivities are the best for your shooting situations can be seen using the calculated results. The presentation of the results overview is shown in a graphic representation. The x-axis contains the image number and the y-axis the associated results for the center and corner (mean value).
Tip: Narrow the ISO sensitivity of your camera. For example, in order to avoid high ISO settings in the auto ISO, several cameras allow to narrow the sensitivity series.
Choose "Sandbox" and create your individual test series. Find out, how the resolution behaves at different camera settings. For example, adjust the sharpness level and evaluate the numerical and graphical results. Since the measurement type "Sandbox" has no restrictions in aperture, focal length etc., also images, which have no metadata, can be evaluated.
MTF curve (Premium & Business)
The Modulation Transfer Function (MTF) represents the modulation function in relation to the Spatial Frequency (SF). It is used to determine the limiting resolution of the camera-lens combination.
The Modulation Transfer Function is represented as a line graph. The Spatial Frequency (SF) is represented on the x-axis. It describes a progressively finer structure based on the sinusoidal star. A black and white line next to each other yield a line pair. The Spatial Frequency is given in line pairs per picture height. The y-axis indicates the modulation, that is, the contrast in respect to the maximum to minimum intensity. Ideally, the value should be at 1 (100%). The maximum resolution is given by the highest frequency, which results in a modulation of 0.1 (10%).
The MTF presentation contains additional lines for visual support and a better evaluation. These characterize the Nyquist Frequency and the 10% threshold contrast (MTF10).
The image results converted by the camera and the resulting MTF curves can be displayed in different views. It is possible to simultaneously view the curves of all of the Siemens stars for center and corner. Moreover you can also choose only the central or the corner result display. In addition to assessing the calculated measured values, the test recording can be consulted to support the visual comparison.
Reading the MTF curve(s)
In an ideal case, the generated recording would reflect the original 100%. The corresponding MTF would be represented as a line at a contrast ratio of 1. Since various factors such as the lens and camera image processing have an impact on the generated image, the contrast decreases with increasing Spatial Frequency. The true result of the calculation is characterized as a curve.
Thus, conclusions can be drawn for the respective camera system about the loss of contrast as a function of Spatial Frequency.
- The flatter the curve is over the Spatial Frequency, the better the result is.
- Increases in the curve (over 1) indicate a sharpening from the camera's internal processing.
- Small dips in the curve can occur, for example, by artifacts appearing in the image.
As a control measure, look at the recording in order to check whether the result is also reflected visually.
Moreover, the maximum resolution can be seen on the MTF curve, meaning the limiting resolution reached by the camera-lens combination used (for the selected camera settings). It is located at the point where the curve intersects the 10% contrast line*.
A conclusion about the power of the reproduction of detail can be drawn in connection with the specification for the system-related Nyquist Frequency. The closer the limiting resolution to the Nyquist Frequency, the better the ability of the camera-lens combination to reproduce fine details.
* The value is determined based on the Rayleigh Criterion. According to this criterion, it is defined that the two points are still just then identifiable if the intensity between the two maxima of the Airy disk is below 81%. In correlation with the contrast, we thus set a limit of 0.1.
R = KL = Imax-Imin/Imax+ Imin = (1-0.81) / (1 + 0.81) ≈ 0.1