It's an interesting idea, and I'm glad someone is working on that... but the corrected images look terrible, they remind me of cellphone cameras from a few years ago. Extremely soft focus and lots of color glow / blur.
Personally I think what they've done is show the reason that multi element lenses are the way to go, but every work has to start somewhere. So long as they can massively improve the results, I hope they continue to work on it :)
LostBoyNZ: It's an interesting idea, and I'm glad someone is working on that... but the corrected images look terrible, they remind me of cellphone cameras from a few years ago. Extremely soft focus and lots of color glow / blur.
Personally I think what they've done is show the reason that multi element lenses are the way to go, but every work has to start somewhere. So long as they can massively improve the results, I hope they continue to work on it :)
I think you make a fundamental mistake comparing the deconvolved image with your perception of what an ideally performing lens should deliver. You should instead compare the deconvolved image with the original. Then consider that if a corrected lens was used with known point spread function, deconvolution could be used with that as well.
Sorry, I didn't understand it as well as I thought :/ It's a vast improvement over the original yes. I'd be very curious to see them use it on a medium quality DSLR lens.
LostBoyNZ: Sorry, I didn't understand it as well as I thought :/ It's a vast improvement over the original yes. I'd be very curious to see them use it on a medium quality DSLR lens.
There are some sample images http://www.cs.ubc.ca/labs/imager/tr/2013/SimpleLensImaging/standard_lens_and_multispectral_results.html using a Canon EF 28-105 macro. I'm not sure if that's what I'd call a "medium quality" lens - I think probably pretty poor quality by modern standards - it looks very soft compared to even relatively inexpensive modern consumer quality zooms. But if the lens is sharp, then there's nothing to be gained by deconvolution. I understand that one maker (Olympus?) claim that they have a method for deconvolution which works to "correct" diffraction. I'll believe it when I see it.
Very interesting! It'll be good to see where this goes.
In the meantime, I'll continue to invest in high quality optics. Although if mirrorless cameras keep improving I think I'll end up having to buy new lenses all over again.
Twitter and LinkedIn »
Follow us to receive Twitter updates when new discussions are posted in our forums:
Are you subscribed to our RSS feed? You can download the latest headlines and summaries from our stories directly
to your computer or smartphone by using a feed reader.
