On 2013-11-30 02:30:23 +0000, (Floyd L. Davidson) said:

bd wrote:

[...] DCRAW works fine, but... the results
are not as good looking as those that the camera delivers in jpg format,
since, I suppose, the latter go through one or more enhancement processes
such as sharpening, improving contours and contrasts through the use of
various filters or in altering gamma factor etc..

There are many things that are simply impossible (sharpening) or
difficult (any adjustment that isn't automatic) with DCRAW. It
is not meant to be used as a general tool.

With my new EOS 600D, I use "ufraw", which works as an interface for
dcraw. The results that I obtain are neither better nor worse than those
with the G2.

Note that UFRAW is a lot more than and "interface" using DCRAW.
It uses DCRAW as the basic converter engine, but has it's own
processing for many other of the enhancement tools you need.

However, it is a RAW converter, not an image editor. Hence
there is more to post processing that it can do alone! The next
step is to process each image with GIMP.

This is certain to get a bit lengthy, but let me walk you
through some suggestions for your workflow.

First, you want UFRAW to set a "base" configuration that you
start with. That's a little complex because when working you
want it to save all setting each time you save an image. When
you start work on a group of files, the first thing is to check
the configuration of UFRAW. Here are my defaults, from top to
bottom:

---- Pre-Configuration ----

At the top:
Enable Raw histogram

2nd line:
Set Exposure to 0.0 (see Note)
Set Restore Details for Negative EV to "HSV space for sharp detail"
Set Clip Highlights for Positive EV to "digital linear"
Set Auto Adjust Exposure to disabled (see Note)
Note: Clicking the last icon will set exposure to 0 and disable
Auto Adjust.

3rd line:
White Balance -- set to either "auto" or "camera"
Grayscale Mode -- None (which means color)
Lens correction (optional, may not exist) -- None
Base Curve -- Straight Line (use reset button on right)
Color Management -- A. Input ICC Profile: No profile
B. Gamma: 0.45
C. Linearity: 0.10
D. Output ICC Profile: sRGB
E. Output intent: Perceptual
F. Output Depth: 8 (16 sometimes)
G. Display ICC Profile: sRGB
H. Display intent: Perceptual
Correct Luminosity, Saturation --
A. Contrast (Optional, may not exist): 1.00
B. Saturation: 1.30
C. Manual curve
D. Click on both left and right bottom reset buttons
Lightness Adjustments -- Does nothing on mine (if it exists, set
everything to 0)
Crop and Rotate -- Click on two reset buttons, and "lock" icon.
Note that there
are bugs associated with the functions in this
menu, and on
occasion odd crops may show up that require
these resets to be
reset to get a normal view.

Save -- THIS IS IMPORTANT!
A. The "path" should be your current working
directory.
B. Select the output file format (I would
recommend only TIFF).
C. Set JPEG compression level as desired
(less than 92)
C. Enable JPEG progressive encoding
D. Enable TIFF lossless compression
E. Enable Embed EXIF
F. ** Set Create ID file to ONLY **
G. Set save image defaults to Always
H. Disable remembering output path
I. Enable overwriting files without asking

Bottom of page:
Enable Live Histogram
Check the indicate box for Overexposure
Uncheck the indicate box for Underexposure

---- End of Pre-Configuration ----

Obviously there are some items you might want to set
differently. I would suggest waiting until you work with those
for awhile first, and then slowly start changing things to match
your specific needs *after* you get an idea what the purpose is
for each. However, it is also true that sometimes you'll want
something different for a specific set of images, so if it
actually makes a difference, change the default on a case by
case basis.

The effect of setting UFRAW to only write an "ID" file is
important. I start UFRAW like this:

ufraw *.nef

You will want to change *.nef to whatever suffix is appropriate
for your raw files. But it will do each raw file in sequence.
If you don't want to do a given file, click on "cancel" instead
of "save" and it will skip to the next file. Clicking on "save"
will *only* write a *.ufraw file, which happens almost
instantly. It then goes to the next RAW file. You don't need
to wait for it to convert each file.

When you've gone through all of the files you want to process,
convert them to TIFF format files with a batch process like
this:

ufraw-batch *.ufraw

You can then take a coffee break, read Usenet, or whatever while
all of the time consuming number crunching is done.

But lets go back to the first RAW file you process, and follow a
typical sequence for making the adjustments. I usually just put
the exposure slider somewhere near correct, and then adjust
White Balance. You can switch between different presets, or go
to manual and adjust it yourself. Then I click the "Color
management" icon and adjust gamma and linearity plus exposure to
get the look I want. If the image is going to be processed by
GIMP, leave the output bit depth at 8. If you will process with
software that can deal with 16 bits, change it to that.

The next step will be processing with an editor. If you use
GIMP or any other 8 bit editor you'll want to get the gamma and
brightness very close to perfect with UFRAW, because changing
either in GIMP can cause posterization. But GIMP is fine for
cropping and most other edits. You can do minor adjustments to
contrast and brightness, in particular to local area selections.
Local sharpening and blurring is also done with GIMP, and then
before writing the final output file to disk it should be 1)
saved as an XCF formatted intermediate file, 2) scaled to the
appropriate size, 3) apply Sharpen and Unsharp Mask, and finally
4) saved to disk (perhaps as a JPEG format).

Your first efforts may not match the perceived quality of images
produced by the cameras JPEG engine, but... rest assured that
with practice you can develop the skills needed to always
produce a better result than the camera does. The reason is
fairly simple too. Your computer can at least equal what the
camera's computer can do, but it has the advantage of 20-20 hind
sight too. The camera has to be configured before you make an
exposure. You only get one guess at what is right for each
exposure, and that's it. With post processing you take the
exact same raw sensor data the camera used, but you get to try
any and every possible variation on configuration until you get
not just something close, but rather the precise configuration
you like the best. Many times every single shot is slightly
different, and you do get better results for each one!

Just don't expect to do that immediately, and do expect to save
your raw files because I guarantee that in 2 years, not to
mention 5 years, you'll be better at editing!

Also be aware that with Linux if you become proficient at writing
shell scripts there is just no end of ways to improve productivity.
The ImageMagick tools are fabulous for editing. And there are
many ways a shell script can speed up your workflow. For example,
I preview my images, as JPEGs, with a very customized version of XV which can
sort them into various directories. The JPEG images I don't want to convert
with UFRAW go into one special directory, and then a shell script moves
the RAW files to the same directories where the JPEG is now at. Then
I run UFRAW and it never loads a file I don't want to process. Plus
when I want to run the batch on all of them, I use a script that does
odd things like automatically setting wavelet noise reduction depending
on the ISO it was shot at, and it determines how many CPU cores are available
and proceeds to keep each CPU busy with a different process (which with
as many as 12 cores can make a huge difference in how fast a few hundred
RAW files can be converted to TIFF files).

I stand amazed that you actually believe that the procedure you just
expounded, is in anyway efficient and productive, just to justify the
use of one OS over another.
....and in the end you still only have 8-bit editing & adjustment.


--
Regards,

Savageduck