I like automatic perspective correction
Monday, June 2nd, 2014 at 11:11pm
As a rule I try to avoid spending too much time post-processing my photos which is why I have been generally satisfied with the simple (compared to full Photoshop) adjustment tools within Lightroom. There are of course the occasional exception to any rule.
I have mentioned before about how I went from this (as-shot) image of the Albury train station:
to this brightened, sign and chimney removed, and perspective corrected image:
What I said in the previous post wasn’t quite correct, I actually cloned out the signs and added a gradient before I decided to try to correct the perspective using GIMP. I recall that this took about an hour to figure out how to do, then the image went back into Lightroom to clone out the third chimney and some final tweaks.
That was two years ago. Today in version 5 of Lightroom is a perspective correction tool. You don’t have full control to change the perspective, instead it is a couple of correction settings that work in an automatic way. Since it was introduced it has been a very handy tool.
I don’t know why, but a few days ago I thought of this image and wondered how the built in perspective correction would do. So I went to the original image in the Library (ie all the edits up to the point it was exported to TIFF for GIMP), made a virtual copy, clicked the ‘Vertical’ perspective button, recropped and then cloned out the third chimney to arrive at:
If you open both modified images and swap between them you can see that the only difference is a slight difference to the proportions of the building. But when you consider one was a few seconds to click a button while the other was at least an hour and then the change history is split across two files, I am glad that Lightroom now has built in perspective correction.
I am also hoping that over time the perspective correction tool is expanded into something like the Adaptive Wide Angle filter from Photoshop. While I am wishing for things, I would also like Lightroom to get content aware fill…
