Re Re: A response to Dan's challenge


Lee Varis
 

Hi all,

The thing is that in this workflow, Jann has applied a gradient. That
is an artificially smooth artifact. Try this same workflow but
changing these steps:
From what I read below, he applied a curve to a selection, is that the same
as applying a gradient?
Selections and alpha channels are 8 bit even in 16 bit mode.

2.I made a round selection of the corners of the image . Exactly the same
selection ( I duplicate it from 8bit to 16 bit ) on both files .
If the selection is feathered there will probably be a significant
amount of banding in it to begin with. I'm actually not sure why his 16
bit file shows no banding unless it has something to do with an
interaction with the "dither gradients" option. In my own experience
scanning naturally produced "spotlight" gradients on Ektachrome
transparencies I see about the same amount of banding with 16 bit as
with 8 bit manipulated files - the resulting bands are just in slightly
different positions within the gradient areas - I STILL need to apply
significant amounts of noise to mitigate the effect! (believe me I wish
16 bit mode worked better, at least with my equipment - it doesn't)

--
regards,

Lee Varis
varis@varis.com
www.varis.com
888-964-0024

Join colortheory@groups.io to automatically receive all group messages.