Hi! Yes, tried that and it kind of works. I used a checkerboard, let Photoscan apply the same distortion than on the real frame, and used that in Nuke with the grid analyze mode of the lensdistortion node. Problem is: the algorithms are different, so both results don't match exactly. Would be better than applying no distortion, but using a node with identical algorithms should give more precise results.
Abraham Am 13.05.2014 um 14:05 schrieb Elias Ericsson Rydberg: > You could export your undistorted plate from Photoscan, match that distortion > in nuke with the LensDistortion node. Then change the input on the node to > use your CG and have it distorted. > > If you could use any distortion values from Photoscan that would be nicer, > but this could also work? > > Cheers, > Elias Ericsson Rydberg > > > 2014-05-13 12:02 GMT+02:00 Schneider, Abraham <[email protected]>: > Yes, I know about that. But I don't want to undistort my plates. I'd like to > distort the CG renderings and merge them with my original plates. So I have > to 'invert' the undistortion that Photoscan does and apply this to CG > renderings. I'll get the parameters of the undistortion inside of Photoscan, > but would need a node in Nuke where I can use these values. > > Abraham > > > Am 13.05.2014 um 11:30 schrieb Misho Ristov: > > > You can export undistorted photos from photoscan and use that directly in > > Nuke > > > > > > from memory: > > > > tools->undistort photos > > > > > > Misho > > > > > > On 12.05.2014 23:07, Nathan Rusch wrote: > >> I think one of the 3DEqualizer distortion models is based on the Brown > >> model. If that's the case, there's a good chance you can use their > >> distortion plugin to get the same result. > >> > >> Details here: > >> > >> > >> http://www.3dequalizer.com/index.php#?site=tech_docs&id=110216_01 > >> > >> > >> > >> -Nathan > >> > >> -----Original Message----- From: Schneider, Abraham > >> Sent: Monday, May 12, 2014 5:25 AM > >> To: Nuke user discussion > >> Subject: [Nuke-users] Photoscan lens distortion in Nuke? > >> > >> Hi there! > >> > >> I'm just playing around with Agisoft Photoscan ( > >> http://www.agisoft.ru/products/photoscan/standard/ > >> ) to deal with a large set where I have to do some CG additions in several > >> shots. Most of the shots are with locked camera. So having a rough temp > >> model of this outdoor set (big square with some buildings surrounding it) > >> would be really helpful to correctly position the CG buildings. > >> > >> Reconstructing the set with Photoscan from my DSLR images worked quite > >> nice. The resulting model should be more than enough to help creating > >> these shots. My idea now was to add single frames from each of my locked > >> camera shots (ARRI Alexa) into Photoscan, align these images as well and > >> export the found camera positions with the model and use these virtual > >> cameras for my renderings. > >> > >> Problem now is: to align all the images, Photoscan uses an complex (?) > >> lens distortion algorithm and undistorts all images before doing there > >> alignment. So the undistorted images match fine to the model from > >> Photoscan. But to do my comps, I'd have to it the other way around: > >> distort the CG renderings and merge them with my original Alexa footage. > >> So I'd need a node in Nuke to apply the same lens distortion as Photoscan > >> does. I get the values from Photoscan, but which node can I use to do the > >> distortion? > >> > >> I think, Photoscan uses the same model as they use in their free lens > >> distortion software: > >> http://www.agisoft.ru/products/lens/ > >> , which mentions a 'Brown's distortion model'. Anyone of you knows if > >> there is a node in Nuke (or Nukepedia or so) which uses the same > >> algorithms? Or any other idea? > >> > >> Thanks, Abraham > >> > >> > >> Abraham Schneider > >> Head of VFX pipeline / VFX Supervisor > >> > >> > >> Türkenstr. 89, 80799 München / > >> Phone +49 89 3809-1096 > >> > >> EMail > >> [email protected] > >> > >> > >> Visit us on Facebook!________________________________ > >> > >> > >> ARRI Film & TV Services GmbH > >> Sitz: München Registergericht: Amtsgericht München > >> Handelsregisternummer: HRB 69396 > >> Geschäftsführer: Helge Jürgens, Josef Reidinger > >> _______________________________________________ > >> Nuke-users mailing list > >> > >> [email protected], http://forums.thefoundry.co.uk/ > >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > >> > >> > >> _______________________________________________ > >> Nuke-users mailing list > >> > >> [email protected], http://forums.thefoundry.co.uk/ > >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > _______________________________________________ > > Nuke-users mailing list > > [email protected], http://forums.thefoundry.co.uk/ > > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users _______________________________________________ Nuke-users mailing list [email protected], http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
