What's new in MTF mapper 0.5.19 Beta
Apr 20, 2017
- New MTF/SFR curve plotting functionality. Just left-click on the MTF50 numbers (usually in cyan) on the "annotated" output to pop up the SFR window. If you <shift>-left-click, up to 3 curves can be displayed concurrently.
New in MTF mapper 0.5.15 Beta (Mar 15, 2017)
- More user-friendly output if you ask for "chart orientation" output, but the input image does not contain valid fiducials (such as using the incorrect chart type).
- Much improved thresholding algorithm, which should not require any tweaking.
New in MTF mapper 0.5.14 Beta (Feb 16, 2017)
- A new output type to be used with the "lensgrid"-style charts that will help you to align your chart so that it is parallel with the camera sensor.
- MTF Mapper also now recognizes many more raw file types, including Fujifilm, Panasonic, Olympus, and Ricoh/Pentax.
New in MTF mapper 0.5.7 Beta (Jul 5, 2016)
- The GUI has been updated to QT5, and now works correctly under both Windows 10 and Linux/Wine. The included dcraw.exe also compiled with better optimizations, so overall operation using the GUI is significantly faster.
New in MTF mapper 0.5.2 Beta (Jun 27, 2016)
- New chart type (and corresponding processing mode) to produce a curve representing the depth of field of a camera. This mode can be applied separately to each colour channel to produce a visualization of the longitudinal chromatic aberration of a lens. The core edge position / orientation and slanted edge ROI has been overhauled. This leads to improved accuracy on wider PSFs (smaller apertures). This code is still relatively new, so please let me know if you are experiencing any problems with this version.
New in MTF mapper 0.4.24 (Nov 17, 2015)
- Fixed some scaling bugs in "lens profile" plots --- these plots are still a work in progress.
New in MTF mapper 0.4.23 (Nov 7, 2015)
- Version 0.4.23 is fairly experimental. It differs from 0.4.21 in that it does away with one of the libraries that MTF Mapper relied upon (FFTW). This makes MTF Mapper more portable. The accuracy of the new built-in FFT routine has not been proven yet under all conditions, but testing so far indicates that it produces identical results to FFTW.
New in MTF mapper 0.4.21 (Nov 7, 2015)
- Version 0.4.21 has a new experimental "lens profile" output mode. From the command line
- This produces a lens profile chart similar to those published by lens manufacturers, i.e., a plot of contrast (at a fixed resolution) as a function of distance from the centre of the lens (sensor). For best results, use the new chart type produced by: "mtf_generate_test_chart.exe -t lensgrid -o lensgrid.svg"
New in MTF mapper 0.4.19 (Aug 10, 2015)
- Version 0.4.19 has a new experimental raw output format, in addition to the existing raw output format. The new "-q" switch produces a file called "edge_mtf_values.txt", with the following format:
New in MTF mapper 0.4.18 (Aug 10, 2015)
- Version 0.4.18 has improved apodization and smoothing options. This should lead to more robust MTF50 estimates (smaller standard deviation). NB: because of the SFR smoothing (enabled by default) the output of 0.4.18 will differ somewhat from previous versions. Use the --nosmoothing option to disable SFR (and MTF50 estimate) smoothing. NB2: ESF and PSF curves will also appear smoother than before owing to a smoother ESF binning process. As a result, ESF and PSF curves are *NOT* true representations, but rather the convolution of a Gaussian and the true ESF/PSF. A future update will add a toggle to change this behaviour.