Lately I've been testing a few "jpg artifacts reducers". Up to now, the winner seems to be DFine by nic.
Yet not one is really perfect.
I usually (you know me, so you know what will follow) go to LAB mode where I have a separation of brightness on one hand, and thwo colour channels on the other. With some selectioning and a lot of patience I get acceptable results.
But I wonder how you work.
What do you do to make enlargements of jpg's that suffer as little as possible from the usual artifacts?
LAB makes very clear that the different apps that reduce filesize when jaypegging follow different methods. I once had a very good article on this, but I have to look it up...
Some jpegs don't suffer in the lightness channel. This makes it easy to blur the colour channels a and b to get acceptable results. But some jpg's have badly damaged lightness channels as well as chroma noise. These are quite difficult to mend.
Just some ideas I throw on the fire here.
Let's have a chautauqua on this...
Yet not one is really perfect.
I usually (you know me, so you know what will follow) go to LAB mode where I have a separation of brightness on one hand, and thwo colour channels on the other. With some selectioning and a lot of patience I get acceptable results.
But I wonder how you work.
What do you do to make enlargements of jpg's that suffer as little as possible from the usual artifacts?
LAB makes very clear that the different apps that reduce filesize when jaypegging follow different methods. I once had a very good article on this, but I have to look it up...
Some jpegs don't suffer in the lightness channel. This makes it easy to blur the colour channels a and b to get acceptable results. But some jpg's have badly damaged lightness channels as well as chroma noise. These are quite difficult to mend.
Just some ideas I throw on the fire here.
Let's have a chautauqua on this...