On Thu, Nov 15, 2018, at 16:15, Robin Kramer wrote: > Why don't you just use the calibrate_lens_gui.exe executable from the > hugin suite.
Well, it doesn't actually replace any part of the script, just the work in hugin... which you have on hand if you have calibrate_lens_gui anyway. You still need a controlled method of producing TIFFs from RAWs with no cropping or correction applied... and if you're working with a purposely-shot image with a specific line or two that you intend you use, the normal hugin control point method will usually give the most accurate results... but if the GUI tool is working well for you, then great! :) > I find the interface is very nice. You kind of have to read between the > lines to figure out which parameters are which, but its still easier > than installing and running python and modules. Installing python modules isn't generally a problem, but specifically scipy (and more specifically the optimized math libraries it depends on) had some hitches the last time I did it (including a circular dependency loop, IIRC). > I wish there had been some documentation for this, method. https://wiki.panotools.org/Calibrate_lens_gui http://libregraphicsworld.org/blog/entry/creating-lens-distorsion-models-with-hugin-lens-calibrator > It doesn't do vignetting but softwares don't support that anyway The software I use does... and it is actually possible to get vignetting parameters from hugin, it just requires stitching a (preferably 360-degree) pano for each focal length/aperture/distance set... it's probably the most accurate way for a specific user to get data for a specific use case, but... not very practical on a production scale. ;) > and it > would be better to use gray frame every time, because you can correct > for dust on the sensor. Again not supported by any open source > software. I certainly wouldn't mind seeing flat-field correction added to darktable, it's obviously quite useful. I'm not sure it's a great method for distributing a large database of generalized correction parameters that have been abstracted from any instance-specific details as much as possible, though... just my two cents. :) -- jys _______________________________________________ Lensfun-users mailing list Lensfun-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lensfun-users
