WARNING: Some people got freaked out by the awesomeness of my dumb questions and incorrect language. So please ignore the little mistakes, missing knwoledge, be sure that I´ve googled my question before with no results and, please, answer in a gentle way. Thanks.
I want to make some own textures (again, I know that´s not easy) using my camera (Fujifilm SL1000) and Gimp (latest Version). I want to take pictures in 4k and use the middle 2048 x 2048 square of it. Every camera has an amount of lens distortion, so I want to correct it in Gimp. And because of I´m using just the middle I´m somewhat sure that the correction of the distortion is not that visible in the end (correct me if it´s either).
The Fujifilm SL1000 has a sensor size of 1 / 2.3 inch (6.2mm x 4.6mm) and the lens a 24-1,200mm eg.
But… Gimp allows you to adjust the correction of the distortion in percent. Great.
I wonder now how much percent is physically accurate to correct the distortion of my camera?
Thanks all of you beforehand, and a happy new year!
NGCHunter2
I don’t have an immediate answer, numerically, but I have done some test photos using a “target” card which has a simple grid. Take a photo of it in the center of your shot, then correct it in the digital photo program of your choice, with on screen guides to align the verticals and horizontals. Take note of the correction needed for different setups and then go about your texture grabs. I know, not quite the simple numbers you may want, but I don’t have them, and tend to problem solve to get there…
prokoudine, Hello and thank you. I did not know about this particular Filter and look forward to a more accurate method than my ‘hun-n-peck’ approach.
Late addition - Does not load correctly on Windows 10. I noted others had a similar experience while doing a Google search, but I am not focused on programming and compiling so I can’t help at the moment. It loads the database just fine, but I could not find any excutable to go along with the XML files. Thank you for the efforts, maybe someone else can make the last step work…
