Paul,
Great find for elements users!
The histograms can sometimes _look_ much better. In the examples I tried,
comparing the images, one over the other, the pixel difference seemed like
only one bit, so 9 vs. 8 for a single "move" like a huge curve manipulation.
In real images, it was extremely difficult to see the difference; toggling
the images you could see the pixels affected, but the overall impact to the
image was not much in the ones I compared. Of course, more than one large
move starts to make big differences, where 8-bit would fall apart quickly.
A while back Tyler commented on how it is not so easy to create lots of
tonal levels in a photoshop gradient in 16-bit mode, while film scans have
lots of levels, and of course, he is right.
I recently tried my hand at how many levels I could create from scratch,
without dither, in a gradient across a 1000x1000 pixel image. I got up to
just over 32,000, roughly 15 out of 16 bits used, but it took a few tries.
Best regards,
John Moody
Show quoted textHide quoted text
-----Original Message-----
From: DigitalBlackandWhiteThePrint@yahoogroups.com
[mailto:DigitalBlackandWhiteThePrint@yahoogroups.com]On Behalf Of Paul Roark
Sent: Friday, December 02, 2005 4:34 PM
To: DigitalB&WPrint
Subject: [Digital BW] 16-Bit PS Elements support
PS Elements has limited 16-bit file support. One limit appears to be that
it cannot convert an 8-bit file to 16-bit. However, there is a way to do
this. Simply copy and paste the 8-bit file into a blank 16-bit file. To
facilitate this, I've posted 2 small, blank 16-bit files at
http://home1.gte.net/res09aij/Test-files.htm . One is grayscale, one is
RGB.
All that Elements users need to do is download these and, when the need
arises, resize them to accommodate the target file, and paste it in.
In one test I tried, an typical original 8-bit Jpeg RGB file was pasted into
the RGB 16-bit file. It was then converted to grayscale. Comparing the
histograms of grayscale files of identically and significantly manipulated
images, where one was converted to 16-bit first and one was kept in 8-bit,
the file that had been converted to 16-bit first appeared to contain
substantially more information.
Paul
[Non-text portions of this message have been removed]