Content Aware Fill - Demystification

Willis said:
I think thats what most people have been saying and I just wanted to show people that it's not as easy at the video shows. ;)
Click to expand...

Consider CAF demystified. Now hopefully every topic about needing something photoshopped won't mention it, as if that's all PS can do now :D
 
Willis said:
I think thats what most people have been saying and I just wanted to show people that it's not as easy at the video shows. ;)
Click to expand...

If people actually bothered listening to the videos rather than gawping, they'd have heard that the photos demonstrated provided a good platform for touching up, rather than just being a 1-click-wonder. Adobe never advertised it as such - internet hype did - Adobe themselves said it was a tool to be used in conjunction with the other tools. It just happens to provide a fantastic starting point.
 
simulatorman said:
Nothing sexy, just a very quick test using low res jpegs, it took a couple of minutes:


film80.jpg
film80none.jpg


It’s not a miracle tool but with some work it will do for a quick fix or two.
Click to expand...

I was going to call bull**** on this image as I have been playing with CS5 over the past few days and was massively underwhelmed by this much touted feature.. When I saw the image you posted it was just so unlike any of the results I have had so far.. but before making a fool of myself and posting I thought I would test it out myself..

I got exactly the same result as you.. Incredible when you think that it has extrapolated the data from somewhere and actually created a coherent image. Maybe because it is a `noisy` image you cannot see the flaws it creates..
 
MarkECP said:
I was going to call bull**** on this image as I have been playing with CS5 over the past few days and was massively underwhelmed by this much touted feature.. When I saw the image you posted it was just so unlike any of the results I have had so far.. but before making a fool of myself and posting I thought I would test it out myself..

I got exactly the same result as you.. Incredible when you think that it has extrapolated the data from somewhere and actually created a coherent image. Maybe because it is a `noisy` image you cannot see the flaws it creates..
Click to expand...

It just copies and merges the stand to the right of the middle. It isn't making it up from scratch.
 
greedy123 said:
It just copies and merges the stand to the right of the middle. It isn't making it up from scratch.
Click to expand...
Just? I think it does an astonishing job of making a base point out of an almost entirely random scene!

Obviously to the eye there is a full 4 or 5 rows at the front, but it doesn't know that. It's just made an educated guess using 0s and 1s.

People really have let the hype get to them. :\

Personally, I can't see many uses for the tool in my own work (primarily indoor school portraiture). The one use I can see is in editing the prom photos where there is a year group which can't fit onto the backdrop. Previously I've had to attempt to clone stamp and healing tool more backdrop in, but this could save me many hours of work.

Likewise I often find trouble with creases creeping into my canvas backdrops (something I've almost resolved with bulldog clips and string), which while easily cleared with the healing tools, could be fixed with the 1 swoop with this. It looks perfect for it.
 
You must log in or register to reply here.
Back
Top Bottom