Re: [Lensfun-users] Accuracy and testing of lens profiles

2016-11-22 Thread Oliver Bedford
Am Dienstag, den 22.11.2016, 21:52 +0100 schrieb Torsten Bronger: [...] > It is probably okay. You may check the residual errors. For this, > I open the preview in full-screen. If the error is smaller than > 1.5px, it is okay. What do you mean by "preview in full-screen"? Ctrl-P + F11? I don't

Re: [Lensfun-users] Accuracy and testing of lens profiles

2016-11-22 Thread Torsten Bronger
Hallöchen! Oliver Bedford writes: > [...] > > Example values: > > a; b;c; computed using 4 images: > > 0,032214; -0,105173; 0,069074 > > computed using only one of the images: > > 0,031828; -0,100712; 0,058477 > > Is this normal or an indication of bad images used for > calibration? It is probab

[Lensfun-users] Accuracy and testing of lens profiles

2016-11-22 Thread Oliver Bedford
Hi, I'm doing my first steps in lens calibration. Currently I'm trying to get distortion correction data from Hugin. What I've noticed is that working with different images I get different values for a, b, c. As Hugin seems to average all values for individual focus lengths (???) I did not keep