Re: 16 bits = 15 bits in Photoshop?
Re: 16 bits = 15 bits in Photoshop?
- Subject: Re: 16 bits = 15 bits in Photoshop?
- From: bruce fraser <email@hidden>
- Date: Mon, 18 Apr 2005 05:36:45 -0700
At 1:20 AM -0700 4/18/05, Marco Ugolini wrote:
In a message dated Sun, 17 Apr 2005 17:23:52, Ulf Skogsbergh wrote:
Images that have a lot of detail or noise and grain may not show the
difference between 16bits and 8bits, but will not stand up to close
inspection. 4/c reproduction for the most part sucks and 8bits is
probably enough, but when I work on images I see a distinct difference,
especially on images that have a very high tonal range. 4color
reproduction is not the only destination for images.
Very nicely put. I agree completely, Ulf.
Yes, this is a key point that brings up a related issue.
Most of the people Dan works with are concerned with a single-use
reproduction of an image on a single output process, which is usually
but not always 4c press.
Most of the people I work with are the owners of the images, and they
generally like to sell them many times for many different uses. For
those people, a sensible workfllow strategy is to create a master
image that's edited for idealized output, and produce versions from
the master that are optimized for different outputs. It could be
press, it could be fine-art inkjet printing, it could be digital
chromes, duratrans, or a host of other outputs.
The benefits of high-bit images are probably considerably more
obvious in the latter scenario than in the former. (The image
originator is, after all, the one person who knows what the image
should look like.)
That said...
Just about any image we use today has had some manipulation done in
high-bit space. I don't know of a single scanner or digital camera
that takes the linear high-bit capture data and converts it to linear
8-bit. I doubt that anyone would be happy with a device that behaved
that way. At a bare minimum, the linear-to-gamma-encoded conversion
is done on the high-bit data. If you use the controls in the scanner
driver, or let the scanner do a negative-to-positive conversion, or
shoot something other than raw with a digital camera, you're making
further edits on the high-bit data. One of the hidden variables in
Dan's "challenge" is how much of that work has already been done in
high-bit space. If you start with a linear, uninverted scan of a
color neg, for example, there's simply no contest between the 8-bit
and 16-bit versions. But nobody in their right mind would do such a
thing except to prove a point.
So the question has never been whether or not to use high-bit data,
but merely at what point you choose to downsample to 8 bits per
channel. For me, it's final output (unless I'm printing through a RIP
that uses the high-bit data, in which case, it's never), but I've
said many times that everyone needs to find their own pain point.
Where that pain point lies depends on many things besides image
quality.
--
email@hidden
_______________________________________________
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