Using eSFR ISO Part 2

Display options (below Plot) contains settings that affect the display (units, appearance, etc.).

MTF plot units selects the spatial frequency units for MTF plots for for the summary plot. Cycles/pixel (C/P), Cycles/mm (lp/mm), Cycles/inch (lp/in), Line Widths per Picture Height (LW/PH), and Line Pairs per Picture Height (LP/PH) are the choices. (Note that one cycle is the same as one line pair or two line widths.) If you select Cycles per inch or Cycles/mm, you must enter a number for the pixel size— either in pixels per inch, pixels per mm, or microns per pixel. For more detail on pixel size, see the box below. note: for using picture height (LW/PH) or (LP/PH) units with cropped images enter the original picture height into the more settings dimensions input

Maximum MTF plot frequency selects the maximum display frequency for MTF plots. The default is 2x Nyquist (1 cycle/pixel). This works well for high quality digital cameras, not for imaging systems where the edge is spread over several pixels. In such cases, a lower maximum frequency produces a more readable plot. 1x Nyquist (0.5 cycle/pixel), 0.5x Nyquist (0.25 cycle/pixel), and 0.2x Nyquist (0.1 cycle/pixel) are available.

Secondary readout controls the secondary readout display in MTF plots. The primary readout is MTF50 (the half-contrast spatial frequency). Two secondary readouts are available with several options. The first defaults to MTF30 (the spatial frequency where MTF is 30%). The third is used only for Lens-style MTF plots.

Clicking Change opens the window shown on the right. Secondary readout settings are saved between runs. Choices:

• The first radio button (MTF) for each readout selects MTFnn, the spatial frequency where MTF is nn% of its low frequency value.
• The second radio button (MTFnnP) selects MTFnnP, the spatial frequency where MTF is nn% of its peak value: useful with strongly oversharpened edges.
• The third radio button (MTF @ ) selects MTF @ nn units, where nn
  is a spatial frequency in units of Cycles/pixel, LP/mm, or LP/in. If
  you select this button, the pixel spacing should be specified in the Cycles per… line in the Plot section of the input dialog box, shown above. A reminder message is displayed if the pixel spacing has been omitted.
• The fourth radio button (MTF Area) selects MTF Area, the integral of the MTF from 0 to the Nyquist limit. MTF Area tracks MTF 50 (it increases as sharpening increases), up to the point where an overshoot appears in the sharpened MTF curve then it remains relatively constant. It therefore does not reward excessing sharpening with better numbers. 
• The fifth radio button (Mean ROI level) selects Mean ROI level. This is the average digital number across each ROI. 
• The sixth radio button (Peak MTF) selects Peak MTF, which indicates the level of sharpening (peak MTF > 1) in an image. 
• The seventh radio button (Angle from Tangential SFR) selects the angle from tangential SFR. Edges oriented radially from the image center will have a tangential SFR and will have angles close to 0. Edges oriented tangentially from the image center will have a sagittal SFR and will have orthogonal angles, close to 90. 

Edge plot selects the contents of the upper (edge) plot. The edge can be cropped
(default) or the entire edge can be displayed. Three displays are available.

1. Edge profile (linear) is the edge profile with gamma-encoding removed. The values in this plot are proportional to light intensity. This is the default display.
2. Line spread function (LSF) is the derivative of the linear edge profile. MTF is the fast fourier transform (FFT) of the LSF. 
3. Edge pixel profile is proportional to the edge profile in pixels, which includes the effects of gamma encoding.
4. Edge linear, unnormalized is similar to Edge profile (1.), but not normalized. Useful in diagnosing situations where one channel may be saturating (and affecting MTF measurements).

Wavelength (um) for diffraction-limited MTF plot. If a number is entered (0.555 um is the default for visible light systems) and pixel spacing is also entered, the diffraction-limited MTF is displayed on the MTF plot and the diffraction-limited edge or Line Spread Function (cross-section of Airy disk) is shown on the edge plot as a faint dashed line.

Settings affect the results as well as the display.

Speedup Speed up calculatons by eliminating some calculations that many users don’t require, including noise statistics and histograms. Checking Speedup can significantly speed up calculations.

Edge roughness Calculate edge roughness (for including in CSV output). Slows calculations slightly. Required if Edge roughness plot is to be displayed.

MTF noise reduction (mod apod) Reduce noise using the modified apodization technique. Improves MTF accuracy, especially with noisy images, but not an ISO standard calculation. Generally recommended.

Channel specifies the primary channel to display (all channels are analyzed). It’s normally left at it’s default value of Y for the luminance channel,
where Y = 0.3*R + 0.59*G + 0.11*B. R, G, and B color channels can also be selected. Selecting one channel-only can significantly speed up calculations and may help in rare instances where bad data in one channel causes a crash.

Color reference and Color space are used when a color analysis is to be performed (if the image contains the color pattern and the Color chart box is checked in the eSFR ISO setup window). The Color reference dropdown menu allows you to browse for the
reference file, typically a LAB data file in CSV format, supplied with color eSFR ISO charts. It should be stored in a convenient location. The last selected file is saved.

Linearization method specifies the method to use for linearizing the slanted edge ROI image data for edge SFR analysis.

Options for eSFR ISO include:

1. 1. Input gamma value (default) – The inverse of the value entered into the “Gamma (input)” field is used to linearize the data.

      • Note: The input gamma value represents the “forward” (encoding) gamma value (typically around 0.46 = 1/2.2 for sRGB or Adobe RGB color spaces); the inverse (display gamma) is used for linearization. More information on gamma can be found below.

   2. Gamma calculated from chart contrast – The value selected for the “Chart contrast” setting is used, in combination with the contrast measured from the edge ROI region, to calculate a value for gamma on a per edge ROI basis. The inverse of this gamma is used to linearize the data.

      • The gamma value is estimated from the ratio of the light and dark pixel levels in the slanted-edge region (away from the edge) if the chart contrast ratio (light/dark reflectivity) has been entered (highly recommended to be 10:1 or less for this calculation). Equations for this calculation can be found below.

      • Note: The calculated gamma value(s) are included in the JSON output as the field “gamma_from_chart”, on a per edge ROI basis

   3. No linearization – The data is not linearized.

      • Note: “Input gamma value” of 1 is equivalent to “No linearization”.

   4. Linearly interpolated LUT from step chart (requires step chart analysis to be checked) – A linearly interpolated look-up-table (LUT), representing the inverse of the measured OECF, is calculated from the mean of the grayscale patch (step chart) ROIs and the nominal patch densities (or those from a grayscale reference file). The LUT is used to linearize the data.

      • Note: A grayscale reference file contains measured patch densities and can be loaded from the Rescharts more settings window

   5. Gamma calculated from step chart (requires step chart analysis to be checked) – The mean of the grayscale patch (step chart) ROIs and the nominal patch densities (or those from a grayscale reference file) are used to calculate a value for gamma. The calculated gamma is the slope of a linear fit of the patch densities and log(grayscale patch means). The inverse of this gamma is used to linearize the data.

      • Note: A grayscale reference file contains measured patch densities and can be loaded from the Rescharts more settings window

      • Note: The calculated gamma value is the same as that which appears on the Rescharts Tonal Response plot, also corresponding to the gray dotted line on the plot (the gamma value is the slope of this line). Note that there is also a gray dashed line that represents a 2nd order polynomial fit.

      • Note: The calculated gamma value is included in the JSON output as the field “gamma_from_stepchart”

      • Note: The range of patches is limited for the calculation of gamma, specifically using the patches that have a mean within the central 80% of the total range of patch means. Therefore, some darker or lighter patches may not be used for calculation of this gamma.

Gamma is the average slope of log pixel levels as a function of log exposure for light through dark gray tones). It is used to linearize the input data, i.e., to remove the gamma encoding applied by image processing so that MTF can be correctly calculated (using a Fourier transformation, which requires a linear signal). Gamma defaults to 0.5 = 1/2, which is typical of digital camera color spaces, but may be affected by camera or RAW converter settings. Small errors in gamma have little effect on MTF measurements (a 10% error in gamma results in a 2.5% error in MTF50 for a normal contrast target). Gamma should be set to 0.45 or 0.5 when dcraw is used to convert RAW images into sRGB or a gamma=2.2 (Adobe RGB) color space. It is typically around 1 for converted raw images that haven’t had a gamma curve applied. If gamma  is set to less than 0.3 or greater than 0.8, the background will be changed to pink to indicate an unusual (possibly erroneous) selection.

A value for gamma can be estimated and used for linearization via the Gamma calculated from chart contrast option (uses the entered chart contrast and the pixel levels from the edge ROI) as well as the Gamma calculated from step chart option (uses the tonal response curve, or OECF, measured from the step chart pattern). 

A grayscale step chart pattern is included in SFRplus, eSFR ISO and SFRreg Center ([ct]) charts. Gamma is calculated and displayed in the Tonal Response, Gamma/White Bal plot for these modules. Gamma can also be calculated from any grayscale stepchart by running Color/Tone Interactive, Color/Tone Auto, Colorcheck, or Stepchart. [A nominal value of gamma should be entered, even if the value of gamma derived from the chart (described above) is used to calculate MTF.]

Gamma, Tonal Response, and related concepts contains a detailed explanation of gamma: why it’s used, how to measure it, and much more.

Incident lux (for ISO sensitivity calculations) When a positive value of incident light level (not blank or zero) in lux is entered in this box, ISO sensitivity is calculated and displayed in the Stepchart noise detail figure. More details are on the ISO Sensitivity and Exposure Index page.

Zone weights Weights of the center, part-way, and corner zones. Used for calculating weighted means of key results, displayed in the Multi-ROI and 3D plots. The defaults of 1 (center), 0.75 (part-way), and 0.5 (corners) are for typical pictorial photography; corners should probably be given more weight for technical photography. (Note: The boundary between center and part-way is 30% of the center-corner distance. The boundary between part-way and corner regions is 75%, except for eSFR ISO, where it is 68%.)

Reg. mark vert. spacing in cm (Registration mark vertical spacing) is used to calculate the Field of View (FoV) when Pixel spacing (pitch) has been entered. It is optional (not normally entered). FoV is displayed in Image & Geometry and reported in the CSV output file if calculated.

Lens-to-chart distance in cm is used to calculate the actual lens focal length when Pixel spacing (pitch) and Registration mark spacing have been entered. It is optional. The distance from the lens to target is most conveniently measured with a laser measuring device. Focal length is displayed in Image & Geometry and reported in the CSV output file if calculated.

eSFR ISO noise settings  opens the Noise and Visual Noise settings window described in Multicharts/Multinoise and eSFR ISO Noise measurements. This sets which types of noise are calculated (affects speed), the display x-axis (Density units, EV, dB, etc.), and Visual Noise viewing conditions.

Standardized sharpening Three settings are available:

See SFR instructions for more detail.

Reset restores the settings in eSFR ISO More Settings to their default values.

Additional parameters (all optional) for Excel .CSV output contains a detailed description of the camera, lens, and test conditions. EXIF data is entered, if available, but can be overridden by manual settings. Description & settings is particularly useful for annotating the test system (it is displayed in MTF Compare).These settings are optional but can be useful when several tests are run for different lenses, focal lengths, apertures, or other settings. The settings are displayed next to the MTF plots. They are saved and reused in subsequent runs for files with the same pixel dimensions. If EXIF data is available (currently, only in JPEG files) it overrides the saved settings. The Reset button clears all entries.

The percentage of clipped pixels is not a reliable index of the severity of clipping or of MTF measurement error. For example, it is possible to just barely clip a large portion of the image with little loss of accuracy. The plot on the right illustrates strong clipping, indicated by the sharp corner near the “shoulder” on the black line. The MTF measurement is better than reality. The absence of a sharp corner indicates that there is little MTF error.

Clipping can usually be avoided with a correct exposure— neither too dark nor light— and by avoiding high contrast targets (like the old ISO-12233 chart). The maximum recommended edge contrast is 10:1; 4:1 contrast (recommended in the upcoming revision to the ISO-12233 standard) is even better. Low-contrast targets are more reliable overall: in addition to better exposure latitude (reduced risk of clipping), they tend to have less sharpening in cameras with variable signal processing, and MTF results are less sensitive to errors in estimating gamma.

Clipping warnings

Pixel size

Pixel size is closely related to image quality. For very small pixels, noise, dynamic range and low light performance suffer. Pixel size is rarely given in camera spec sheets: it usually takes some math to find it. If the sensor type and the number of horizontal and vertical pixels (H and V) are available, you can find pixel size from the table on the right and the following equations.

pixel size in mm = (diagonal in mm) / sqrt( H² + V² )
pixel size in microns = 1000 (diagonal in mm) / sqrt( H² + V² )

Pixel size in microns (microns per pixel) can be entered directly into the SFR settings box. Example, the cute little 5 megapixel Panasonic Lumix DMC-TZ1 has a 1/2.5 inch sensor and a maximum resolution of 2560x1980 pixels. Guessing that the diagonal is 7 mm, pixel size is 2.1875 (rounded, 2.2) microns.You can find detailed sensor specifications in pages from Sony, Panasonic, and Kodak.