Dan Margulis Applied Color Theory
Camera Raw and Color Shift
Date: Thu, 01 Apr 2004 08:32:03 -0000
From: Blaine Franger
Subject: Am I the only one noticing this problem in
Photoshop CS?
I am new to this group, so I hope I'm doing this right.
A couple months ago I upgraded to Photoshop CS from 7.0.1 and I am
having some problems with accurate color with Raw files (Canon and Nikon)
in PS CS. The colors are perfect in PS 7. The colors in CS are washed
out and have a very noticable magenta shift. It kills my blues and makes
skintones very pink. I can't figure out a way to fix it. The shift is there
in the file browser even before I process the image in Adobe Camera Raw
(Version 2.1), which leads me to believe it is possibly bad profiles in
photoshop and not the fault of Adobe Camera Raw. Has anyone else
experienced this? FYI, I have tested it on many different computers with
different copy's of PS CS, and different cameras also and it was there
every time. Please look at the example I posted here- http:
//www.photo.net/photodb/photo?photo_id=2251209&size=lg
Thanks for any help on this issue!
Blaine Franger
___________________________________________________________________________
Date: Thu, 1 Apr 2004 11:32:29 +0100
From: "Andrew Haley"
Subject: Am I the only one noticing this problem in
Photoshop CS?
It's really hard to tell what's going on because I
don't know what the original scene looked like, or what colour temperature
you set when doing the conversion.
Can you post a shot of a colour test card, taken under
a known light
source? Or maybe a grey card?
Andrew.
___________________________________________________________________________
Date: Thu, 01 Apr 2004 10:54:17 -0000
From: "Stephen Marsh" \
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Blaine Franger writes:
The colors are perfect in PS 7. The colors in CS are
washed out and have a very
noticable magenta shift. It kills my blues and
makes skintones very pink.
Initially, there would seem to be two things to look
into:
i) Are the same settings in use between the ACR plug
and ACR2 built into CS?
ii) The settings may be the same, but the colour
description hard wired into CS has changed from the v1 plug.
The second point seems fair, but it is unknown about
the first point.
I can't figure out a way to fix it.
ACR2 offers a calibrate tab...but perhaps there is a
simpler fix.
CR is not something I have a lot of experience with, I
usually only handle 'processed' files.
Good luck,
Stephen Marsh.
___________________________________________________________________________
Date: Thu, 1 Apr 2004 09:39:09 -0500
From: Dragonfly Imaging & Printing
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Hi Blaine,
I've been using CS since last fall, and I am quite
happy with Camera Raw
.
Canon and Fuji imports are better than the plugin, and
I like the new controls.
I'm not sure what's going on with your workflow.
I assume you've reset your camera to defaults etc. etc.
Could it be white point related?
Make sure that you aren't using the prior settings as
your default.
John Toles
Dragonfly Imaging & Printing
http://www.dragonflyprinting.com/
http://www.dragonflygallery.ca/
___________________________________________________________________________
Date: Thu, 01 Apr 2004 17:11:45 -0000
From: Blaine Franger
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Andrew: The original scene's looked like the images
converted in PS 7. PS 7 with NEF plugin did a great job of rendering colors
when processing the raw file. PS CS is doing a really bad job. The color
temperatures were kept the same on both 7 and CS during processing (color
temp. was "as shot")
Stephen: In PS 7, I used the nikon NEF plugin to make
the conversion. I understand this adds a variable, but why are the colors
changed when I preview the thumbnails in the file browser even before I
process them? I cannot get the original colors back using any of the
adjustments in Camera Raw like "calibrate".
I'm noticing this problem is very image-dependent.
Because some conversions the color change between PS 7 and PS CS is hardly
noticable. But a lot of my images are greatly affected, especially ones
with deep blues and bright yellows. Look at the difference in the yellow on
the road sign in that example.
If you have any further insight please let me know, I'm
dying to fix this!
Thanks, Blaine Franger
___________________________________________________________________________
Date: Thu, 1 Apr 2004 10:15:59 -0800
From:Paul D. DeRocco
Subject: RE: Am I the only one noticing this problem in
Photoshop CS?
I've generally found the color better with ACR2, for
Canon 10D raw files.
Ciao,
Paul D. DeRocco
___________________________________________________________________________
Date: Thu, 1 Apr 2004 10:27:34 -0800
From: Lee Varis
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
My guess is that you are processing using the default
"as shot" settings in ACR. It is quite possible that Adobe's
assumptions about the camera settings have changed in ACR v2. It looks like
they are favoring a cooler "auto-white balance" setting. The
beauty of ACR is that you can determine your own custom defaults for your
particular camera. The raw file is actually unaffected by the white-balance
default you set in the camera. The camera setting only tags the raw data
with some basic instructions for the white balance setting as shot. This
info can be somewhat cryptic so Adobe has to guess at a setting based on
just one camera that they purchased to do the tests.
Instead of using the "as shot" default you
can pull down the white balance menu and try out other "guesses"
to see if any of those get you closer. When you do this you'll see the
temperature slider change to various different numbers based on Adobe's
idea of what a good corresponding color temperature should be (5500 for
"flash" for instance) The "tint" slider will also
change. I've noticed that CS ACR puts a +10 tint setting for
"daylight" this is a magenta shift. The "Flash"
setting has "0" tint - negative settings would move you towards
green.
You can think of the "Temperature" and
"Tint" sliders like the B and A channels in LAB respectively.
"Temperature", or the "B" channel, affects the
yellow-blueness of the color and "Tint", or "A"
channel, affects the magenta-greeness. It usually takes me about 5 seconds
to dial in the color this way. Once you have the color dialed for the first
image in a series you can option/alt click (the "OK" button) to
update the settings for that raw file in the browser without opening it.
then select the rest of the images in the browser and apply the
"previous conversion" to them through the automate menu in the
File Browser. Now you can batch those files with the new setting.
If you find yourself constantly increasing the contrast
slider past the default of 25% you can save a new default contrast slider
setting either as "Set Camera Default" or "Save Settings
Subset" using the small triangle menu next to the "settings"
pull down menu.
A small investment in time learning to use ACR will pay
off with huge benefits. The default settings were never meant to be the
last word on the color for any camera only you can decide what that
is. ACR allows you to save out new defaults or apply custom tweaks to
numerous files all at once. CS gives you additional controls through the
advanced tabs with Lens and Calibrate that provide additional benefits over
the original ACR.
regards,
Lee Varis
http://www.varis.com
888-964-0024
___________________________________________________________________________
Date: Thu, 1 Apr 2004 12:19:29 -0800
From: "fotoman"
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Paul:
Adobe CS has a long history of problems in just about
all areas for some computer and almost none for others...The camera Raw is
one of the few that works for me and I would try and reinstall or contact
Adobe if you can't find an answer...PS please be able to hold on the phone
for a Very long time and have no more than One problem for them....They
will not support more than One thing at a time
Doug Jirsa
___________________________________________________________________________
Date: Thu, 01 Apr 2004 20:53:24 -0000
From: Blaine Franger
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Lee,
Thanks for all help on adjusting the image in the ACR
window, but I have tried that many times and I can never get the same
colors that I get in PS 7 ( using ACR 1.0 or NEF plugin). I really think
this is an internal problem with CS, maybe in the profiles or something.
Blaine Franger
___________________________________________________________________________
Date: Thu, 01 Apr 2004 20:06:18 -0000
From: Blaine Franger
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
I just tested it with some 10D CRW files and they had
the color shift as well..
Blaine Franger
___________________________________________________________________________
Date: Fri, 02 Apr 2004 01:01:57 -0000
From: "jafent2002"
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
I haven't seen anyone mention ACR 2.1 Have you tried
upgrading ACR?
Personally I don't use ACR with my Minolta A1.
MRWFormat does a much better job, and extracts all the data that Adobe does
not. But, the above was just a thought.
John Fagerberg
___________________________________________________________________________
Date: Thu, 1 Apr 2004 20:03:32 -0500
From: Scott Southerland
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
Hi Blaine - perhaps you could post one of the problem
RAW files on the web so that we could take a look at it and give more
meaningful suggestions?
Scott
___________________________________________________________________________
Date: Fri, 02 Apr 2004 09:53:06 +0100
From: Shangara Singh
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
It was 1/4/04 9:53 pm, when bfranger82 wrote:
Thanks for all help on adjusting the image in the ACR
window, but I have tried
that many times and I can never get the same colors
that I get in PS 7 ( using ACR 1.0 or NEF plugin). I really think this is
an internal problem with CS, maybe in the profiles or something.
Blaine
I don't think it's a problem. It's just how the two
versions work. You will definitely get different results if you use the
Default & As Shot settings. I would put that down to the hard wired
profiles in the two versions. If you like the look that ACR 1 produces, you
will just have to use Photoshop 7.
Here are for versions of the same file from a Fuji S2
that confirm your results. The settings used on the top two versions were
Camera Default and As Shot, and on the bottom two versions Camera Default
and Auto. None of the sliders were moved from their default positions.
http://www.photoshopace.com/acr_cmp.html
HTH.
Shangara Singh.
___________________________________________________________________________
Date: Fri, 02 Apr 2004 09:48:05 -0000
From: Blaine Franger
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Scott- I'd be up for that....how would I go about
posting a NEF on the web? Sorry, I don't know a lot about web and FTP
stuff. I don't have any online storage space either...
Blaine Franger
--- In colortheory@yahoogroups.com, Scott Southerland
wrote:
Hi Blaine - perhaps you could post one of the problem
RAW files on
the web so that we could take a look at it and give
more meaningful
suggestions?
___________________________________________________________________________
Date: Fri, 2 Apr 2004 16:54:26 -0500
From: Alex Lippisch
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Blaine,
I'm pretty sure your problem is in not using a
correctly defined "Camera default", and using "as shot"
white balance settings in Photoshop CS ACR. My images from my 10D looked
similar to yours at first until I redefined the "Camera default"
by playing with the various sliders and then clicking on the button next to
Settings, going to the drop down menu and selecting Set Camera Default.
As Lee has said, the camera setting only tags the raw
data with some basic instructions for ACR as to the colorspace, white
balance, and all the other settings that it thinks represent the
"Camera default" conditions. My findings are that WB "as
shot" is rarely "very good" and so you need to create custom
settings for batches of images. The ACR camera default sets the WB to
"as shot" which picks up the WB from the camera's EXIF tag,
however the WB is much better if using the ACR definitions of pre-set WB ie:
Tungsten, Flash, etc. or best yet is to use custom and measure the photo
with the eyedropper.
I would start with increasing the saturation level and
that should bring your images a lot closer to looking right.
The previews in file browser will also look better when
you have redefined the "Camera default", because that is where it
gets its idea of what the image should look like.
Let us know if this helps.
Alex Lippisch
___________________________________________________________________________
Date: Fri, 02 Apr 2004 22:03:04 -0000
From: Blaine Franger
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
Well, for those of you who care, I finally got an
answer. I went directly to the man in charge, Thomas Knoll (supposively the
inventor of photoshop). This was his response to my email-
"Adobe and Nikon use different algorithms, and
different color calibrations. When the file browser first displays
the thumbnails, it extracts the small preview image inside the file.
This preview is computed the camera, so it uses Nikon's calibration.
The second pass computes the preview from actual raw data using
Adobe's calibration. This results in a different colors. If you
consider Nikon's calibration to be perfect, you are going to consider
anything else to be wrong. Adobe Camera Raw is NEVER going to generate
EXACTLY the same results as Nikon's, since the underlying algorithms are
different. You can come very close by using the "Calibrate"
tab in Camera Raw's advanced mode and creating a new default for your
camera model. The only change from Camera Raw 1.0 (optional plug-in for
Photoshop 7--do you actually have this or are you using Nikon's plug-in?)
and Camera Raw 2.x (included with Photoshop CS) is Photoshop 7 did not do
the second pass in the file browser."
Bottom line, if you can use ACR and get the colors
right (or however you like it) then ACR is great for you. If you can't get
the right colors, like me (this problem is very image- dependant) then were
forced to use something other than ACR, like NEF plugin or C1...ect.
I have another email out to Thomas asking if Adobe
plans on improving the algorithms to more closely match the Nikon NEF
plugin. We'll see what happens with that.
Thanks for all the help.
Blaine Franger
___________________________________________________________________________
Date: Sat, 03 Apr 2004 13:25:33 +0100
From: Shangara Singh
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
It was 2/4/04 11:03 pm, when bfranger82 wrote:
The only change from Camera Raw 1.0 (optional plug-in
for Photoshop 7--do you actually have this or are you using Nikon's
plug-in?) and Camera Raw 2.x (included with Photoshop CS) is Photoshop 7
did not do the second pass in the file browser."
Well, maybe you missed my post and included link (I
spent some 20-30 minutes creating the page, btw) but, as I said, if you use
the very same "default" and "as shot" settings in ACR1
and ACR 2.1, you get different results from the same file. So, something
else _has_ changed. Nikon's plug-in does not affect my files as they are
Fuji Raw files.
BTW, although the look of the thumbnail changes,
unfortunately, Photoshop CS does not mark the thumbnail to say that it is
getting the preview from the processed Raw file. IMO, a serious oversight
because, if your exposures and colors are pretty good, it's hard to tell
which files have been processed and which haven't. If it displayed a tiny
triangle in one corner, or someSuch, it would alert the viewer.
Shangara Singh.
___________________________________________________________________________
Date: Sat, 3 Apr 2004 11:55:22 -0800
From: Dennis Dunbar
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
One of the big problems with processing camera raw
files is that there is no standard format for camera raw. In many ways this
has some very large implications. The complications range from: archiving
(what if the manufacturers change their format or somehow support for the
old ones goes away? How will you be able to use archived raw images in the
future?) to the difficulty Thomas Knoll has in reverse engineering the
necessary algorithms so PhotoShop can support these files.
I am very certain Thomas and Adobe would love to have
the manufacturers share the info so it's easier to properly support the
various versions of camera raw. But since the manufacturers insist on
having their own secret sauce built into their raw files and since they
largely refuse to share the info with Adobe there will inevitably be some
differences in how Adobe's ACR processes the files and how the individual
camera companies' software does. Blame the manufacturers, not Adobe on this
one.
Dennis Dunbar
___________________________________________________________________________
Date: Sat, 3 Apr 2004 18:35:04 -0800
From: David Cardinal
Subject: RE: Re: Am I the only one noticing this
problem in Photoshop CS?
From: Dennis Dunbar
One of the big problems with processing camera raw
files is
that there is no standard format for camera raw. In
many ways
this has some very large implications. The
complications
range from: archiving (what if the manufacturers change
their
format or somehow support for the old ones goes away?
How
will you be able to use archived raw images in the
future?)
Amen. Kodak has already orphaned the raw formats from 3
cameras. There are upwards of 100 different variations of Raw formats out
there now. They won't all be supported by software in a year, let along 10
or 100 years.
manufacturers insist on having their own secret sauce
built
into their raw files and since they largely refuse to
share
the info with Adobe there will inevitably be some
differences
in how Adobe's ACR processes the files and how the
individual
camera companies' software does. Blame the
manufacturers, not
Adobe on this one.
Frankly this is mostly a different issue. The raw file
itself only contains very limited data on how to interpret the sensor
information. Most raw files do contain information on the camera settings,
but that is about it. And it would be _VERY_ helpful if those were
standardized so those of us who write raw file decoders could get at them
without spending nights hacking.
But, the "secret sauce" in Nikon Capture is
not in the file itself. It is in Nikon's knowledge of how the camera works,
plus their bench testing of the sensor and camera system, plus their own
vendor specific work. Same for any other Raw converter. They rely on their
own algorithms to cleverly interpret the data as the file does not contain
any information that says how exactly to interpret (for example) a sensor
reading of (say) 1234 Red or 1000 Blue or the best way to demosaic or
resample, etc.
So a standard format would at least level the playing
field and hopefully make all of us feel more secure about archiving Raw
files as if they were digital negatives, but there would still be plenty of
room for innovation in the calibration, profiling and interpretation of
images from various cameras.
--David Cardinal
Pro Shooters LLC
http://www.proshooters.com
___________________________________________________________________________
Date: Sun, 4 Apr 2004 15:32:20 +0100
From: "Andrew Haley"
Subject: RE: Re: Am I the only one noticing this
problem in Photoshop CS?
David Cardinal writes:
Amen. Kodak has already orphaned the raw formats
from 3 cameras. There are
upwards of 100 different variations of Raw
formats out there now. They won't
all be supported by software in a year, let along
10 or 100 years.
This is perhaps where free software can come to the
rescue. dcraw supports a bunch of formats, and is not about to
disappear any time soon, even if the manufacturer's own software does.
Frankly this is mostly a different issue. The raw
file itself only
contains very limited data on how to interpret
the sensor
information. Most raw files do contain
information on the camera
settings, but that is about it. And it would be
_VERY_ helpful if
those were standardized so those of us who write
raw file decoders
could get at them without spending nights
hacking.
True enough, but at least the ability to decode the
files isn't going away.
But, the "secret sauce" in Nikon
Capture is not in the file
itself. It is in Nikon's knowledge of how the
camera works, plus
their bench testing of the sensor and camera
system, plus their own
vendor specific work. Same for any other Raw
converter. They rely
on their own algorithms to cleverly interpret the
data as the file
does not contain any information that says how
exactly to interpret
(for example) a sensor reading of (say) 1234 Red
or 1000 Blue
You can get some of that by profiling, can't you?
or the best way to demosaic or resample, etc.
But why would Nikon Capture necessarily have the best
demosaicking? This is quite an active research area, and it's quite
possible that in the future we'll be able to do better than we can now.
Andrew.
___________________________________________________________________________
Date: Sun, 04 Apr 2004 11:40:27 -0700
From: Andrew Rodney
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
The biggest issue hitting digital photography today
(and conceivably into the future) is this huge mess we have with RAW file
formats. We need a standard! We need something a bit like the ICC spec for
profiles that is open but has some private tags that allow those silly
camera manufacturers some idea that they can use their secrete sauce while
allowing converters to be able to process the files into something we can
use in the future. IF a camera today produces this non standard RAW file,
we have no guarantee that the data is going to be readable by future
software. I have far less a concern that a format like Tiff can be read but
RAW as it stands today is a huge potential issue! We need the ability to
process RAW data but we need a standard that allows this process to be
conducted now and in the future.
Andrew Rodney
http://www.digitaldog.net
___________________________________________________________________________
Date: Sun, 4 Apr 2004 09:19:45 -0600
From: Ron Kelly
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
Folks
I have been following this thread with interest. It
seems as though quite a few folks out there feel that raw shooting and post
processing is the way to go.
I have tried this myself but found it insufficiently
rewarding for the cost of admission. I have a few comments and questions if
anyone would care to respond.
What is the big advantage of a raw workflow over jpeg?
Presently, I shoot several jpegs of a given scene very
easily. I have the ability to combine the images if necessary to deal with
excessive contrast. This works very well in situations where I can't get
the exposure right in one shot.
I hear people mentioning that they can change the white
balance if they shoot raw. Well, what's wrong with your ability to get it
right in the first place? Do you have to make such a gross post-exposure
adjustment that it's really significant? Are you expecting not to have to
color-tweak before printing? Why don't you just white-balance bracket when
you find a tricky situation, such as with mixed light sources?
Jpegging does introduce artifacts, admittedly. This is
a very small problem my tests show. Shoot jpegs on the highest setting and
you'll lose very, very little. Perhaps on the very highest magnification
use it could show, but only then. It's very subject dependent, and for my
work (scenery, rocks trees, water) it's practically a non-issue.
Here's what I gain if I shoot jpegs:
-speed
-convenience
-file size
-retain the ability to post-exposure combine images for
contrast control and "creative" purposes
Here's what I gain if I shoot raw:
-can adjust the white balance (minor advantage, IMHO)
-no artifacts on extremely high magnifications
-?
The big problem with shooting raw is the post
processing. I've tried a few different ones, and as this thread shows there
is considerable challenges in post processing the raw files for a variety
of reasons. I must admit that I haven't used raw converters enough to get
very good with them. I don't like my results so far, so I'm actually
getting a worse result than if I make the same photograph from a jpeg and
processed in Photoshop.
The raw converter that came with my camera, a Canon
10D, is insufferably slow. I've also tried Photoshop CS, which by
comparison is much faster, but still a very time consuming task.
However, they ALL take too much time! Get a super fast
computer but you still have to sit there and make choices. Then, your
equipment grinds away on the files, slowing down your computer all the
while. But what then? Do you archive the processed file - no. This would
remove potential advantages for raw file conversion in the future. So,
you've got a job to do now, AND in the future as well. This seems to me to
be a heavy tax to pay.
Furthermore there are many programs that don't work
with raw. You have to process a file before you can show it to anybody, or
in most cases, use it with a filing/browsing/indexing program.
It occurs to me that perhaps a jpeg/raw combination (I
believe my camera allows for this) might offer the best of both worlds.
However, it's going to slow things down in the camera, as well as later, in
the card reading. What is a worthwhile price to pay for bloating your files
system by several times? It's got to be more than just a slight occasional
advantage, doesn't it?
So what's the big deal with raw workflow? Am I missing
the boat here? I just don't get it.
Ron Kelly
___________________________________________________________________________
Date: Mon, 5 Apr 2004 12:13:48 +0100
From: Andrew Haley
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
Ron Kelly writes:
What is the big advantage of a raw workflow over jpeg?
Apart from the white balance issue:
Better exposure latitude, by maybe a stop or more.
Double the chroma resolution.
The colour gamut issue. Some colour prints exceed
the gamut of both Adobe RGB and sRGB in bright yellow and darker
blue/green. This is less important for prepress work, granted.
The big problem with shooting raw is the post
processing. I've tried a
few different ones, and as this thread shows
there is considerable
challenges in post processing the raw files for a
variety of reasons. I
must admit that I haven't used raw converters
enough to get very good
with them. I don't like my results so far, so I'm
actually getting a
worse result than if I make the same photograph
from a jpeg and
processed in Photoshop.
Well, I just open the raw file in Photoshop CS. It
doesn't seem to be very time consuming.
So what's the big deal with raw workflow? Am I
missing the boat here? I
just don't get it.
Does anyone really make the decision in a dogmatic way?
Surely the only sane way to proceed is to use JPEG for fast shooting
and non-critical work but save raw for the special circumstances where you
need the last 10%.
Andrew.
___________________________________________________________________________
Date: Mon, 05 Apr 2004 07:24:16 -0600
From:Andrew Rodney
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
on 4/4/04 9:19 AM, Ron Kelly wrote:
What is the big advantage of a raw workflow over jpeg?
There are a lot! First, when you shoot JPEG, you1re
rendering the color from the Grayscale file and that is now fixed and
locked in stone. Having a RAW file and a good converter allows all kinds of
tuning based on your preferences in how the de mosaic of the chip is
controlled and color is produced. You also get with many cameras, as much
as half a stop more dynamic range to work with and you can use two
different RAW conversions of the same data to pull (and combine) even more
tonal data. As RAW converters get better and better, you can go back and
recreate the color any time in the future. I've seen amazing quality
increases from the SAME data converted with different converters or updates
to the same converter. You1ve got digital clay. The image you shot is not a
color image. It1s RAW. So having the ability to render the color now and in
the future, plus the quality increases (you want 16 bit color, ain1t going
to get it from that JPEG), is a big factor in why so many pro1s shoot and
deal with RAW files.
Andrew Rodney
http://www.digitaldog.net
___________________________________________________________________________
Date: Mon, 5 Apr 2004 09:36:59 -0600
From: Chris Murphy
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
On Apr 4, 2004, at 9:19 AM, Ron Kelly wrote:
What is the big advantage of a raw workflow over jpeg?
Raw is the negative, JPEG is the print. If you are
happy with prints and throw out all of your color negatives, then you'll
probably be happy with just JPEG.
The camera itself always shoots raw. Raw is a dump of
sensor data, plus a bunch of metadata some of which is consumable and some
of it is proprietary. When you ask the camera to just give you JPEG, it
still shoots raw but it processes the result before you ever see anything.
It's like taking an instant polaroid. If you're happy with the results,
great, but if you ever wanted more refined control such as highlight detail
recapture that's something unique to the algorithms that process the raw
file.
I hear people mentioning that they can change the white
balance if they
shoot raw. Well, what's wrong with your ability to get
it right in the
first place? Do you have to make such a gross
post-exposure adjustment
that it's really significant? Are you expecting not to
have to
color-tweak before printing? Why don't you just
white-balance bracket
when you find a tricky situation, such as with mixed
light sources?
There are all kinds of photography in practice. A
captured moment only occurs once and it's possible to just not have the
ability to perfectly expose in the first place. Over time our ability to
understand how sensors react to light, the imperfections of lenses, and the
mosaic pattern of the CCD/CMOS element itself is always improving. A raw
image processed with Camera Raw 2.1 can be noticeably better than even
Camera Raw 1.0, depending on the camera. Moderate to major edits on a JPEG
will expose artifacts as well.
You may never need the negative, you may never need to
reprocess with more advanced technology 2 years from now, or 10 years from
now, or 50 years from now. But there are people who want a digital
equivalent of a color negative and that's essentially what a raw file is.
However, they ALL take too much time! Get a super fast
computer but you
still have to sit there and make choices. Then, your
equipment grinds
away on the files, slowing down your computer all the
while. But what
then? Do you archive the processed file - no. This
would remove
potential advantages for raw file conversion in the
future. So, you've
got a job to do now, AND in the future as well. This
seems to me to be
a heavy tax to pay.
Photography is changing rapidly. It will get easier and
faster. But do you really want to risk important images now by throwing
away the negatives, confident you will never need to reprocess them for
some reason? If you are that confident, don't worry about raw.
Furthermore there are many programs that don't work
with raw. You have
to process a file before you can show it to anybody, or
in most cases,
use it with a filing/browsing/indexing program.
The workflow I imagine we're in is just like that with
film vs digital files. How often do you go back and rescan a piece of film?
Hopefully not very often, but it's there if you ever needed it—unless
you threw it away in which case your SOL, you'll just have to live with the
scan you've got.
The raw file you will process immediately to something
that satisfies you today. You'll work with the resulting TIFF most of the
time. The raw file gets archived. It's not a daily workflow file format, or
an exchange file format. You wouldn't share your negatives with people, and
in the same way you aren't going to share your raw files with people.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
---------------------------------------------------------
___________________________________________________________________________
Date: Mon, 5 Apr 2004 08:04:56 -0700
From: David Cardinal
Subject: RE: Am I the only one noticing this problem in
Photoshop CS?
From: Andrew Haley [mailto:andrew-yahoo@l...]
This is perhaps where free software can come to the
rescue.
dcraw supports a bunch of formats, and is not about to
disappear any
time soon, even if the manufacturer's own software
does.
I sure hope it doesn't, but I'm not betting my photo
archives on it.
But, the "secret sauce" in Nikon
Capture is not in the file
itself. It is in Nikon's knowledge of how the camera
works, plus
their bench testing of the sensor and camera system,
plus their own
vendor specific work. Same for any other Raw converter.
They rely
on their own algorithms to cleverly interpret the data
as the file
does not contain any information that says how exactly
to interpret (for
example) a sensor reading of (say) 1234 Red or 1000
Blue
You can get some of that by profiling, can't you?
Some, yes, but not all of it. Nikon, for example,
characterizes their cameras by measuring their response to direct light at
a large number of frequencies, allowing them to build a
"physical" model of the response of the sensor. You can't get
this just by profiling in a traditional sense.
But why would Nikon Capture necessarily have the best
demosaicking?
This is quite an active research area, and it's quite
possible that
in the future we'll be able to do better than we can
now.
Sure, it'll improve. But the manufacturer knows the
filter material they use to generate RGB, its impurities, the size and
shape of the electron wells and therefore something about the color noise
between adjacent sensor sites, etc., so they have a big headstart (if they
choose to use it). Anyone else would have to spend quite a bit of time and
probably destroy a few cameras to get that data.
Personally, along with standardizing on a format for
Raw sensor data, I'd like to see some of this additional information
standardized and encoded in the Raw file as well. Just like when you load
film into a mini-lab it has data about the specific type of film to help
make the prints look good.
Imagine if Kodak had said they would not provide film
curves for their films so that if you wanted them processed
"correctly" you'd have to use Kodak processing equipment.
--David Cardinal
Pro Shooters LLC
http://www.proshooters.com
___________________________________________________________________________
Date: Mon, 5 Apr 2004 10:52:38 -0700
From: David Cardinal
Subject: RE: Re: Am I the only one noticing this
problem in Photoshop CS?
From: Chris Murphy
Raw is the negative, JPEG is the print. If you are
happy with
prints and throw out all of your color negatives, then
you'll
probably be happy with just JPEG.
For negative shooters that analogy works well, but the
analogy I prefer and I think more accurately represents the characteristics
of a high quality JPEG is that Raw is similar to shooting negatives and
JPEGs are similar to shooting slides.
With both slides & JPEGs we still have quite a bit
of flexibility, but the color and visual image is "pre-developed"
in the slide itself. There is less room to make changes later and they are
both high-gamma formats. Negatives and Raw are both much closer to a
linear/low gamma capture.
OT: FWIW, I'm thrilled that with the D2H I can now
shoot Raw+JPEG and to a great extent get the best of both. For high-volume
action shooting I often shoot JPEG only to save card space and transfer
time, but for almost everything else I can now shoot both formats as
quickly as I used to be able to shoot JPEGs.
--David Cardinal
Pro Shooters LLC
http://www.proshooters.com
___________________________________________________________________________
Date: Mon, 5 Apr 2004 11:33:46 EDT
From: Paul Harding
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
I have not responded to the thread in quite a while.
After my initial exposure to the authors and the content of their
submissions, it became clear that "that world" was way beyond the
capabilities of both myself and my computer/printer set up. So .... I
continued down my path, doing as I had done before, abiet with more
knowledge but not changing my work flow. I still shoot with hi rez JPEG and
have yet to be disappointed .. not only me ( perhaps I don't have the
discerning eye that the rest of the group has) but my clients always ask
"How do you get such good pics?" WHY CHANGE?.
I'll proclaim it again, It's the message .. not the
messanger.
Paul Harding
JIT Graphics
___________________________________________________________________________
Date: Mon, 05 Apr 2004 10:10:24 -0700
From: John Denniston
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
Hi Ron,
The advantages of RAW are many but any work flow must
be tailored to the project you are working on.
In my retirement I shoot a lot of sports, always jpg,
most of the time with auto white balance, because this is the work flow
where the time quality equation is best for me. I also do scenics and
editorial features. These are shot RAW because the number of prints I'm
making for these projects are few and I have more time to fuss over them.
Regards, John
John Denniston
www.dirtbikephoto.com
www.dennistonphoto.com
___________________________________________________________________________
Date: Mon, 5 Apr 2004 20:05:20 +0100
From: Andrew Haley
Subject: RE: Am I the only one noticing this problem in
Photoshop CS?
David Cardinal writes:
Sure, it'll improve. But the manufacturer knows
the filter material
they use to generate RGB, its impurities, the
size and shape of the
electron wells and therefore something about the
color noise
between adjacent sensor sites, etc., so they have
a big headstart
(if they choose to use it). Anyone else would
have to spend quite a
bit of time and probably destroy a few cameras to
get that data.
Imagine fifty years from now. Let's say you had
1. A copy of dcraw source code.
2. A raw file that is a shot of a colour target.
3. Okay, so you need the values of the target's
colour patches.
and a little time. How much of the image quality
do you suppose you could retrieve? 99%?
Andrew.
___________________________________________________________________________
Date: Tue, 6 Apr 2004 02:57:13 -0400
From: John Castronovo
Subject: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
From: David Cardinal
Imagine if Kodak had said they would not provide film
curves for
their films so that if you wanted them processed
"correctly" you'd have to
use Kodak processing equipment.
Wait a minute. I've been printing color negs in a
custom/commercial lab for nearly 30 years, and I haven't seen useful data
from Kodak (or anyone else) yet. We're pretty much on our own to do our own
calibrations. Even our new Agfa digital minilab is totally unaware of film
specs from manufacturers, and for $180,000, there's no place in the
software to even put such information.
Excuse the tangent, this is now O.T.
john c.
___________________________________________________________________________
Date: Tue, 06 Apr 2004 23:47:22 -0400
From: Lee Clawson
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
At 08:19 AM 04/04/2004 , Ron wrote:
So what's the big deal with raw workflow? Am I missing
the boat here? I
just don't get it.
Ron,
I look at image quality needed for a particular end
use. Last year we did signage that was screen printed. Though I didn't ask
for the format the JPEG's had more than enough image quality.
In regard to workflow with JPEG. After receiving images
I recommend working in an uncompressed format. That is, not processing and
(re)saving in JPEG.
Lee
___________________________________________________________________________
Date: Mon, 5 Apr 2004 17:23:06 -0600
From: Chris Murphy
Subject: Re: Re: Am I the only one noticing this
problem in Photoshop CS?
On Apr 5, 2004, at 11:52 AM, David Cardinal wrote:
For negative shooters that analogy works well, but the
analogy I prefer and
I think more accurately represents the characteristics
of a high quality
JPEG is that Raw is similar to shooting negatives and
JPEGs are similar to
shooting slides.
The analogy works in the respect that color negative is
more forgiving and so is raw, whereas chromes are less forgiving and so are
JPEGs. BUT where it doesn't work is if you have any serious editing to do,
high-bit scans of positives are easier to make heavy edits on than JPEG
which will expose JPEG artifacts. IF you're able to get a TIFF from the
camera, you're better off than JPEG if the images were to need lifting.
With both slides & JPEGs we still have quite a bit
of flexibility, but the
color and visual image is "pre-developed" in
the slide itself. There is less
room to make changes later and they are both high-gamma
formats. Negatives
and Raw are both much closer to a linear/low gamma
capture.
Yeah. Raw is full sensor data dump. There is no tonal
mapping yet. You could say that it's exactly linear, depending on the
response of the CCD/CMOS array.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Wed, 14 Apr 2004 06:15:41 -0500
From: "Michael Streibel"
Subject: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
Dan:
I have been following the discussion about RAW storage
and conversion with increasing dismay since my digital work-flow and that
of my students now includes burning CDs and DVDs of all RAW images
immediately after capture and only converting them as needed. If
companies are going to be secretive about their in-house knowledge of RAW
formats and then drop RAW formats as they develop new cameras, they really
should release this in-house information into the public domain once they
drop a format. Since you are a well-known public figure who has
advocated for the right thing time and time again, could you begin a public
campaign for an industry policy of full-disclosure by companies about their
RAW formats once they drop a format from their products? That way,
third party developers could create utilities for the digital imaging
community and we could retrieve digital images with old RAW formats without
having to keep old versions of software, operating systems, and computers
on hand.
Thanks.
Michael*
Michael J. Streibel
The University of Wisconsin-Madison
___________________________________________________________________________
Date: Wed, 14 Apr 2004 07:03:37 -0600
From:Andrew Rodney
Subject: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
on 4/14/04 5:15 AM, Michael Streibel wrote:
Since you are a well-known public figure who has
advocated for the right thing
time and time again, could you begin a public campaign
for an industry policy
of full-disclosure by companies about their RAW formats
once they drop a
format from their products?
Many of us are doing as much as we can to push a
Standard RAW format. Not going to be easy but it really has to be done.
This is a critical factor for all future use of digital cameras and hugely
affects all users!
The very first such discussion of industry power users
was two weeks ago at the i3 forum in Laguna Beach (see http:
//www.i3forum.com/ and check out the guest by invite only list). I'm sure
more about the meetings with some of the industry big-wigs will be forth
coming from the i3 organizers.
Andrew Rodney
http://www.digitaldog.net
___________________________________________________________________________
Date: Wed, 14 Apr 2004 14:03:43 EDT
From: Dan Margulis
Subject: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
Michael Streibel writes,
Dan:
\I have been following the discussion about RAW storage
and conversion with
increasing dismay since my digital work-flow and that
of my students now
includes burning CDs and DVDs of all RAW images
immediately after capture and only
converting them as needed. If companies are going
to be secretive about their
in-house knowledge of RAW formats and then drop RAW
formats as they develop
new cameras, they really should release this in-house
information into the
public domain once they drop a format. Since you
are a well-known public figure
who has advocated for the right thing time and time
again, could you begin a
public campaign for an industry policy of
full-disclosure by companies about
their RAW formats once they drop a format from their
products?
Obviously the companies *should* do this but equally
obviously there is going to be limited interest in supporting a format that
is no longer being used. The fact that they have disclosed the format won't
do you any good if, as likely, nobody develops software to read it.
The idea of *working* with somebody's raw format makes
a lot of sense, for the reasons stated by many of the people in this
thread. The idea of *archiving* in that format and not having any backup is
a bad one IMHO. Save in Photoshop format or standard TIFF and it's most
unlikely that the format will become a total orphan during your lifetime.
Save only in a proprietary raw format and in five years it may become
inaccessible.
As for the lack of a uniform format, this story has
played itself out again and again in many areas. One or more leading
vendors think they will give away a competitive advantage by allowing
everybody else to access their stuff, and thereby hose the remaining 99.9%
of the world who want open standards. Getting a standard that is acceptable
to all the vendors takes a long time, a lot of negotiation, and tends to be
designed by a committee. And, in order to be fair to all, it often tends to
be incompatible with anybody's prior formats. So, while it would be great
if there were a uniform format for raw images, it's quite likely that it
would make anything saved in previous raw formats obsolete.
Dan Margulis
___________________________________________________________________________
Date: Sun, 18 Apr 2004 07:17:18 -0500
From: "Mike Davis"
Subject: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
I'm way out of my field of expertise here. It is
my understanding that a RAW format, being a simple chip "dump"
with some EXIF information added, would be dependent upon the specific
capture device (chip) being used (i.e. it's relative sensitivity to various
wavelengths of light being captured). Just as different film captures
light values differently, the resulting data from various cameras and their
respective chips will be different. A RAW converter must take the
terribly unbalanced captured data (inherent in sensitivity to varying
wavelengths of light due to their energy variations) and
"reinterpret" them to create a "balanced" image,
similar to what we would perceive as human interpreters of those light
values without the camera (or scanner) intermediary.
Conversely, if there were a single algorithm used to
"read" all capture information between the storage memory and a
graphics editor (Photoshop), the output would certainly be different for
each chip, and indeed for each camera due to lens coatings, etc. The
only alternative would be to create additional "in camera"
processing to output a "standard" result from a
"standard" image so that a single RAW format converter would give
everyone "exactly" the same image (whatever THAT is). But
by doing that, you simply move the proprietary conversion from between the
storage memory/Photoshop to between CCD/storage memory, which would slow
operation of the camera. What am I missing here in the discussion to
"standardize" the RAW format? How can you possible
"standardize" a RAW format?
Mike Davis
mldavis2 AT sbcglobal DOT net
___________________________________________________________________________
Date: Sun, 18 Apr 2004 09:46:17 -0600
From:Andrew Rodney
Subject: Re: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
on 4/18/04 6:17 AM, Mike Davis wrote:
What am I missing here in the
discussion to "standardize" the RAW format?
How can you possible
"standardize" a RAW format?
Good question. Let me direct you to a site you should
address this to:
http://www.rawformat.com/
This is a new site that was just put up after the i3
forum last month to encourage this discussion among users and
manufacturers.
My understanding is that a RAW standard format would
not limit the rendering of the color data moot and identical to all
converters. My understanding is that the open standard would allow each
converter access to the data in such a way that there would be no need to
3reverse engineer" a way to get to the data for de-mosaicing and image
production.
I1ve run the same RAW data through a number of RAW
converters what have done this hard work and found vast differences in the
final image rendering and quality. The issue was that each converter had to
come up with a method of doing this since there may or may not be any kind
of SDK from the manufacturer so it was difficult to handle the file in the
first place. My understanding of coming up with a RAW standard is simply a
way in which the basic recipe for getting to this data could be produced
making it easy to support RAW processing now and into the future for every
camera. The alternative is the possibility that a company might decide not
to support some product in the future leaving all the RAW files useless to
future users who may or may not have access to the original converter.
That1s a huge concern.
Assume a RAW format could be developed by someone like
Adobe and we actually could convince the major manufacturers to stick to
some predefined methods of communicating ways to process the data. That
would not guarantee that we'd have a way of handling this data forever but
it would certainly be better than what we have today which is 100+
different, unique file formats we lump together as RAW whereby anyone that
wants to process that data (besides the originator) could more easily
handle.
Andrew Rodney
http://www.digitaldog.net
___________________________________________________________________________
Date: Sun, 18 Apr 2004 12:05:10 -0700
From: David Cardinal
Subject: RE: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
Like many standards, this one will need to have
"levels" of compliance. If you break apart the problem of
standardizing a raw data format for D-SLR images there are many pieces.
Some vendors will want to open up some of them, others all of them, and
some may stick with their current position of opening up none of them. A
simple list of pieces for a standard includes: *
Description of basic sensor data encoding (bits per
sample, padding, ordering of colors)
(some of this is allowed
for in EXIF and used by some vendors already)
* Description of compression or other advanced
techniques used by the vendor to reduce image size
* Description of any alternate representations also
included in the file for\previewing purposes (some of these are in a
standard place already, but many are not)
Those three might be considered "level 1" and
would at least give everyone access to the raw sensor data.
Then there is the potential to standardize the next
level of info, either by embedding it or perhaps using a style sheet like
mechanism that could point to the info for a given camera at a well known
URL/URI.
That information could include:
* Spectral response of the sensor
* Settings on the camera beyond EXIF standard (WB,
etc.)
* Any other "hints" available from the
camera, like color temp recorded by color sensor, etc.
* Other information on the linearity of the sensor, its
native noise characteristics, etc.
The "Level 1" info is pretty well decoded for
all major cameras, but only through the hard work of various folks around
the world and it changes with every new model.
The "Level 2" info is spotty, and mostly kept
proprietary both by the camera vendors and by those who have decoded it for
use in products.
If someone wanted to be _really_ aggressive there is
room for a "Level 3" which would be the tone curves, color
matrices and other corrections to apply to process the raw data the way the
camera would by default. That would allow the most open system for raw file
viewing.
Some vendors have SDKs which allow Raw file processing
(Canon & Nikon, for example)--which is a laudable first step--but they
have several problems:
1) They are usually late--lating 2-6 months behind the
camera bodies
2) They have mediocre and inconsistent interfaces
3) To be multi-vendor and use them you'd need to drag
around a set of changing libraries for each camera vendor and update them
every time a camera is shipped given the way things work now
4) They don't provide open access to the format. They
let you "develop" the image according to certain parameters,
which is great, but if you want to tweak the algorithms they aren't really
designed to do that.
Then there are ancillary issues like standardizing the
way Raw+JPEG is handled, the way voice annotations are handled, etc., which
are already drifting between vendors.
--David Cardinal
Pro Shooters LLC
http://www.proshooters.com
___________________________________________________________________________
Date: Mon, 19 Apr 2004 10:56:24 -0400
From: Lee Clawson
Subject: Re: Re: RE: Am I the only one noticing this
problem in Photoshop CS?
How can you possibly "standardize" a RAW
format?
Andrew,
Adobe, as the leader doesn't impress me, but the model
they used publishing
and placing postscript in the public domain does.
Lee
___________________________________________________________________________
Date: Sat, 24 Apr 2004 00:03:34 -0000
From: Blaine Franger
Subject: Re: Am I the only one noticing this problem in
Photoshop CS?
For all of you that were following this thread, just
wanted to update you on the latest. If you didn't know already,
Thomas Knoll finished the Camera Raw 2.2 update for PS CS. It is
available on the Adobe site.
It turned out that there actually was color problems in
certain situations with the plugin, and I'm not crazy! I just did a
few quick tests on some of the files that were being affected the
worst, and ba-da-da-da! The problem is fixed! With just a few quick
adjustments in the calibration features, the colors I got were
perfect! They actually even looked better than the colors I was
getting with Nikon's NEF plugin for PS 7.
Thanks to everyone for the help, those of you who are
interested in keeping up to date on
ACR features, issues and stuff go to
www.adobeforums.com
Blaine Franger
Adobe Photoshop training classes are taught in the US by Sterling Ledet & Associates, Inc.