Re: 16 bits = 15 bits in Photoshop?
Re: 16 bits = 15 bits in Photoshop?
- Subject: Re: 16 bits = 15 bits in Photoshop?
- From: email@hidden
- Date: Tue, 19 Apr 2005 14:31:06 EDT
Marco writes,
>>Dan, there is no pleasing you! You ask for an example of banding, and I
give it to you. No good, because it's a synthetic image!...I give you another
example, this one from a B&W 16-bit scan. No good, because it's not color! You
keep raising the bar. >>
The bar is right where it's been for several years; you just wish to duck
under it. In my very first post to Bruce, I used the phrase "real-world color
photograph". Shortly thereafter, in a message to Michael Shaffer, I used the
phrase "natural color photograph", and added, "If you are inserting your own
computer-generated gradient then 16-bit can be helpful. If you are speaking of an
existing photographic gradient, this has been tested fairly extensively and
nobody has been able to demonstrate any advantage to manipulating it in 16-bit."
Both these messages were written before you entered the thread.
In the phrases "real-world color photograph" and "natural color photograph",
I intended the word "photograph" to mean something shot with a camera, and not
produced by other means. By adding the adjective "color" I intended to
indicate something with chroma variations and channels that were not based on a
common ancestor, and not a black and white image. I apologize if either the word
"color" or "photograph" is unclear. Also, by the phrase "if you are inserting
your own computer-generated gradient then 16-bit can be helpful" I intended to
convey that I did not need to be convinced that 16-bit was useful in dealing
with computer-generated gradients. Again, I apologize if the phrasing was
unclear.
>>And what about the FACT that there are circumstances (extreme and rare as
they may be, which should not matter for the purpose of this discussion) in
which banding DOES appear in 8 bit files, whereas it does NOT in equivalent
native 16 bit files subjected to the same exact same treatment?>>
What FACT is that, please? Who has demonstrated it? I know that I and others
have asked many, many times for anyone who can prove that to be a FACT to
kindly do so, because the test you describe to prove it to be a FACT is very
simple. And every single person--there are about a dozen so far--who have tried to
come up with such a side-by side demonstration has failed. At least, the ones
who have stated in public that they've done so. You can bet that most of the
people who have made the most extreme statements on how critical 16-bit
corrections are have also done so in private with the same results.
If you can demonstrate that it is a FACT, even in rare and extreme cases,
then it will be of great benefit to people like Roger Breton and others who take
the sensible position that there may be a line to be drawn between where
16-bit is useful and where not. Plus, it will get you some free publicity, if you
want it.
All you have to do is the following.
1) Find a 16-bit real-world color photograph that you think will show the
16-bit superiority AND for which you are willing to give a permission to publish
for the purpose of discussion of 8-bit vs. 16-bit.
"Real-world color photograph" means any color photograph that could
conceivably be used in any professional context. It is OK to use a cropped portion of
the image provided that the cropped portion is itself capable of being used
standalone in a professional context. It is not OK to use B/W pictures, HDR
images, computer-generated graphics, images with grossly insufficient resolution,
or heavily retouched originals.
2) Generate an 8-bit file directly from the 16-bit in Photoshop. Do not use
an 8-bit from the camera or from any other source.
3) Carefully recording all your actions, apply any sequence of real-world
edits. The edits must be applied in the same order to each image.
"Real-world edits" means any action or series of actions that could
conceivably be taken by anyone who was making a good-faith effort to improve the image.
It is OK to use editing methods that a professional would know better than to
use. It is not OK to make obvious efforts to sabotage the image, or to make
drastic moves in one direction followed by drastic moves in another.
4) If you are satisfied that the 16-bit version looks significantly better,
pull some kind of proof that illustrates the superiority. Contact me when I
return from Italy and I will give you directions on how to get me a package. If
it *does* show a significant advantage--and I've never had a disagreement on
that point with anyone who's sent me images--I'll publish it, with proper credit
to you if you wish it, and I'll change the portions of my book that indicate
that no one has ever been able to demonstrate any kind of advantage.
It would certainly take you less time than you've spent so far in this
thread. I hope to hear from you on my return.
Dan Margulis
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden