<
  News     DIFP   Benefits Profiles Account  


forums

Recent Posts



(New threads or threads with new replies - ↑ Links to top of thread)

 Sony Artisan and Windows
Apr 21, 2004 7:52 PM
Posted by John Eakin

 Preferred printing direction (Enhanced Matte) & IP with roll paper
Apr 21, 2004 7:31 PM
Posted by John Vitollo

 Kodak 14n or Fuji S2 for stock?
Apr 21, 2004 7:20 PM
Posted by Andrew Rodney

 Confused about layers in PK Sharpener
Apr 21, 2004 7:11 PM
Posted by Thomas Nash

 Sony Artisan with Monaca Optic XR
Apr 21, 2004 5:05 PM
Posted by Jeff Schewe

 Epson R 800
Apr 21, 2004 4:43 PM
Posted by Sharon Brunson

 Fuji S3 and verticals
Apr 21, 2004 3:18 PM
Posted by Peter Burian

 Applying in camera parameters and RAW files
Apr 21, 2004 3:11 PM
Posted by Peter Burian

 Is the D70 a good deal?
Apr 21, 2004 3:08 PM
Posted by Peter Burian

 Camera for raw
Apr 21, 2004 3:02 PM
Posted by Peter Burian

...see more posts

articles

Weekly Chatter:Need for Correct Exposure in RAW
Wednesday November 26, 2003

Section 1: Need for Correct Exposure in RAW






Digital Capture > Image Processing & Workflow > Need for Correct Exposure in RAW




































AuthorSubject: Need for Correct Exposure in RAW  
Robert Daniels
Posted on Sat Nov 22, 2003 11:41 AM

I gather that correct initial exposure is essential and that underexposure or overexposure can't be corrected later -- without a price being paid in image quality. In fact, I've heard that you lose half the information the image could have had, if you have an initial underexposure of even a single stop -- and that you can't get it back later through a manipulation in RAW conversion or Photoshop.

However, a respected local photo supply store, with some knowledge of digital, is telling it's customers that exposure errors in RAW files can by corrected (by as much as 2 stops) after the fact -- either in raw conversion or later in Photoshop -- without any significant loss.

What's the real story on this? Having heard Jeff Schewe speak on the subject recently in New York, my impression is that correct initial exposure is quite critical. And that the information we get in on camera histograms and flashing overexposure indicators can't always be accepted at face value. So a good deal of care -- and even bracketing -- may be called for in getting the initial exposure right.

Thanks!


Jeff Schewe
Posted on Sat Nov 22, 2003 1:30 PM

Robert Daniels wrote:

> However, a respected local photo supply store, with some knowledge of digital,
> is telling it's customers that exposure errors in RAW files can by corrected
> (by as much as 2 stops) after the fact -- either in raw conversion or later in
> Photoshop -- without any loss.

They are full of horseshit. They are parroting some of the marketing hype
some of the camera companies have put out and it's disinformation-pure and
simple.

Yes, you can get a "usable image" by using exposure correction on raw
processing. But, how do you bring down highlight detail that is 2 stops
overexposed? If you underexpose, you can correct for the midtone but you
will get severe noise and the real risk of banding-even from the raw file.

> What's the real story on this? Having hear Jeff speak in New York recently,
> my impression is that correct initial exposure is quite critical.

What, you didn't believe me? Test it yourself and see who is full of crap.
Do an exposure ring around and see how far you can go to correct off
exposures. I have. You do NOT have +- 2 stops of latitude, not if you want
any resemblance to quality capture. Yes, blowing out highlights is surely
something to stay away from-but the knee jerk reaction to under expose to
"save the highlights" leads to noise and banding.

With the wider dynamic range of digital captures, it's easy to fall into the
bad habit of post exposure corrections. . .and yes, up or down a 1/4-1/2
stop is a no-brainer. But, digital sensors are _NOT_ like chrome or negs,
it's a whole new recording device with it's own tendencies and weaknesses
and strengths.

But the worst place to get accurate information about just about anything is
a camera store. Jeeesh, give me a break :-)


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Robert Daniels
Posted on Sat Nov 22, 2003 6:23 PM

Jeff, I didn't doubt you for a second -- just needed confirmation to take back to a person who was a direct reciepient of this misinformation.


Jed Wormhoudt
Posted on Sat Nov 22, 2003 6:35 PM

Ah! What a delight, what a strengthening pleasure it is to have Mr. Schewe taking us by the hand and gently guiding us, the unitiated, the dim, the unknowing, the ever questioning, towards the light.


Sergio Bartelsman
Posted on Sat Nov 22, 2003 7:30 PM

I have noticed that big moves in levels in C1 Pro do deter image quality, but what do you think about WB? Is it lossy or lossless to move it around?


Ian Lyons
Posted on Sun Nov 23, 2003 5:34 AM

Sergio,

> I have noticed that big moves in levels in C1 Pro do deter image quality, but what do you think about WB? Is it lossy or lossless to move it around?

There is NO free lunch - look at the histogram AND the image. If you find yourself needing to make large adjustments of ANY kind to get the desired result then you you can take it as read that data got tossed as part of that process. WB may not result in the same degree of degradation as say over or under exposure but you can't expect to make an adjustment for free. This is why you REALLY do want to make your edits on the 16-bit raw file.


Mitchell Reuben
Posted on Sun Nov 23, 2003 9:15 AM

I've been following this thread and now need to ask the following question. If you can't really correct exposure and white balance without image degradation, why shoot in raw at all? Couldn't you just PS levels and curves adjustments with similar results?


Robert Daniels
Posted on Sun Nov 23, 2003 10:34 AM

Regarding WB, my understanding is that when you're shooting in RAW there is no loss from adjusting WB later. That, in fact, the WB setting you use in shooting is effectively only information appended to the RAW file (as is the color space your shooting in... ie Adobe RGB etc.). In other words, if you had WB perfectly accurate, the main advantage would be your images would pop up in File Manager imediatly looking great. But, if your WB was off -- even way off --you could arrive at exactly the same place, with no loss in image quality, by making color correction adjustments in the RAW conversion or later in Photoshop. This, however, is not true of exposure. This is because (if I've gotten this right) the WB and color space information that -- comes along -- with the RAW file are effectively notes on how to interpret it, and not part of the RAW file itself. Because of that, you can effectively choose to interpret the RAW file another way -- without any loss in image quality.

Jeff -- did I get this right? This was something I was very excited about hearing recently at one of Jeff's presentations in New York. So I hope he'll jump in and clarify how this works.


Jeff Schewe
Posted on Sun Nov 23, 2003 12:00 PM

Robert Daniels wrote:

> Jeff -- did I get this right? This was something I was very excited about
> hearing recently at one of Jeff's presentations in New York. So I hope he'll
> jump in and clarify how this works.

Correct. . .I think Ian must have overlooked the fact that WB adjustment is
a "normal" required adjustment for any raw capture.

While using the exposure slider will indeed alter the maximum quality you
can get out of a raw capture, normal white balance adjustments do no real
hard to the image. Yes, it is an algorithm and thus results in a certain
degree of imperfection, but it's required so you really have no alternative.
And yes, the _ONLY_ camera setting that actually effects the raw capture at
the time of taking is the ISO setting. The WB setting is merely a tag that
software may or may not use. Camera Raw does see most camera's WB setting,
but not all.

As far as what Ian wrote regarding exposure, yes, any adjustments will
result in less than optimum file quality. And yes, that's why you want to
work in 16 bits/channel.

I would also point out that the working space also plays a major role. I've
pretty much decided to bring images in from Camera Raw as ProPhoto RGB to
make sure there is no significant channel clipping and to make sure I've got
the full color gamut of the capture in file.


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Stephen Kennedy
Posted on Sun Nov 23, 2003 1:00 PM

Jeff Schewe wrote:
> I would also point out that the working space also plays a major role. I've
> pretty much decided to bring images in from Camera Raw as ProPhoto RGB to
> make sure there is no significant channel clipping and to make sure I've got
> the full color gamut of the capture in file.
>
Jeff,

Can you expand on this statement with respect to the use of a Canon 1Ds vs Adobe RGB (1998)?

Stephen Kennedy


Jeff Schewe
Posted on Sun Nov 23, 2003 1:10 PM

Stephen Kennedy wrote:

> Can you expand on this statement with respect to the use of a Canon 1Ds vs
> Adobe RGB (1998)?

Not sure what you mean. . .shooting raw ignores any camera settings except
ISO. WB, color space and other camera settings are ignored (or may be used
by conversion software) but do NOT alter the raw captures.

Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Stephen Kennedy
Posted on Sun Nov 23, 2003 1:19 PM

Jeff,

Your choice of ProPhoto RGB has everything to do with the final use of the photo, probably converted to CMYK and printed on a press? Is your goal not to excite a client with unprintable colors?

Stephen Kennedy


Robert Daniels
Posted on Sun Nov 23, 2003 1:59 PM

Would the advantage of bringing a RAW file into Photoshop in ProPhoto RGB be:
1. That some manipulations can be better accomplished in this larger color space.
2. That when you need to convert to a smaller color space -- Adobe RGB or CMYK -- you can clearly see what colors are being clipped and possibly even have more control over how they are expressed in the smaller color space.

I guess that if you bring RAW files into Photoshop with a smaller color space (Adobe RGB) you never see what the chip initially captured and what was clipped. And, you'd have no control over how the larger color space of the chip was compressed into the smaller color space.

I'm also curious as to whether you can go to an RGB printer -- like the epson printers -- with Pro Photo RGB -- or do they only accept (or convert to) Adobe RGB. Of course, this would probably only be interesting if the printer profile was a little bigger -- at least in some dimension -- than Adobe RGB.

Come to think of it , shouldn't the master file be independent of the restrictions of any particular output device. Creating a master file in Pro Photo RGB, where a smaller color space would clip the colors captured by the chip, might make sense -- anticipating future use with output devices that could express the full range of color. Shouldn't a master file -- "be all that it can be."

Jeff's idea of bringing RAW files into Photoshop in Pro Photo RGB is intriguing.



Jeff Schewe
Posted on Sun Nov 23, 2003 4:00 PM

Stephen Kennedy wrote:

> Your choice of ProPhoto RGB has everything to do with the final use of the
> photo, probably converted to CMYK and printed on a press? Is your goal not to
> excite a client with unprintable colors?

Doesn't matter, I turn CMYK softproofing on before they ever see the image
anyway.

The reason that ProPhoto RGB is useful is that absolutely _NO_ gamut
clipping can occur going from raw to ProPhoto RGB. ProPhoto RGB is so large
it contains color you can't see.

The point I'm making relative to the larger gamut is that ProPhoto can
contain EVERYTHING in the capture where even Adobe RGB and certainly
ColorMatch & sRGB will clip.

You still need to be concerned about exposure because bigger color gamut
does NOT equate to larger dynamic range-you're still limited to 12 bits of
precision in 16 bits/channel. That doesn't change with ProPhoto RGB.


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Jeff Schewe
Posted on Sun Nov 23, 2003 4:05 PM

Robert Daniels wrote:

> 1. That some manipulations can be better accomplished in this larger color
> space.

No, not really. . .

> 2. That when you need to convert to a smaller color space -- Adobe RGB or
> CMYK -- you can clearly see what colors are being clipped and possibly even
> have more control over how they are expressed in the smaller color space.

Yes, this is the reason. . .you control the gamut clipping in the conversion
from ProPhoto RGB > Adobe RGB or whatever space you use down the line.

It's that first raw linear to ProPhoto RGB where the added gamut is useful
to avoid any clipping.


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Brian McKinlay
Posted on Sun Nov 23, 2003 5:09 PM

Sorry, Jeff, I have been following this thread and just want to make sure I have this perfectly clear.

I shoot with a Canon D30. Until now, it has been my understanding that its color gamut delimits the capture to sRGB. From what I understand from this thread that is not true, at least if I am shooting in RAW? But I don't want to replace one possible misunderstanding with yet another.

Thank you,

Brian


Jeff Schewe
Posted on Sun Nov 23, 2003 5:15 PM

Brian McKinlay wrote:

> I shoot with a Canon D30. Until now, it has been my understanding that its
> color gamut delimits the capture to sRGB. From what I understand from this
> thread that is not true, at least if I am shooting in RAW? But I don't want
> to replace one possible misunderstanding with yet another.

The _ONLY_ camera setting that effects the actual raw capture is ISO. Color
space, camera sharpening & tone settings _ONLY_ effect camera made jpg's.
All other settings are merely metadata tags that post processing software
may or may not use.

So if you set your camera to Adobe RGB and you are shooting raw, the Adobe
RGB setting is ignored.

However, changing the ISO from 100 to 400 DOES effect the raw capture.


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Sergio Bartelsman
Posted on Sun Nov 23, 2003 6:06 PM

Jeff, for the sake of curiosity,in the case of the Canon 1Ds what color space are the raw captures in?


Jeff Schewe
Posted on Sun Nov 23, 2003 6:50 PM

Sergio Bartelsman wrote:

> Jeff, for the sake of curiosity,in the case of the Canon 1Ds what color space
> are the raw captures in?

It ain't in a "color space". The raw captures are linear and the sensor has
a color mixing function based upon the exposure and the white balance of the
shot. One could finger print the spectral response of the sensor at various
lighting conditions and ISO settings-Thomas did that with Camera Raw. He
checked the spectral response at 3200K and D65 and formulated a tweener
blending between the two responses-remarkably accurate between the two
points and useful even beyond the two points.

But, get over the thought that a sensor has a "color space". It doesn't. It
has a spectral response-that varies based upon the lighting and exposure.


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com




Robert Daniels
Posted on Mon Nov 24, 2003 4:09 PM

Jeff, if the gamut of epson ultrachrome printers is larger than Adobe RGB, wouldn't your suggestion of bringing RAW files into Photoshop in ProPhoto RGB save some unnecessary clipping (assuming the chip captured an image with a color space larger than Adobe RGB -- shooting vibrant colors etc.).

Also, is there any disadvantage to doing corrections and editing in ProPhoto RGB? In other words, could one stay in ProPhoto RGB from RAW conversion to sending the file to an Epson 2200, 7600, 9600, 4000.

Thanks... The idea of using ProPhoto is interesting.


Jeff Schewe
Posted on Mon Nov 24, 2003 4:35 PM

Robert Daniels wrote:

> Jeff, if the gamut of epson ultrachrome printers is larger than Adobe RGB,
> wouldn't your suggestion of bringing RAW files into Photoshop in ProPhoto RGB
> save some unnecessary clipping (assuming the chip captured an image with a
> color space larger than Adobe RGB -- shooting vibrant colors etc.).

I don't think UltraChrome is bigger than Adobe RGB. . .it's bigger than
ColorMatch RGB in some colors. But yes, bringing in in ProPhoto RGB does
preclude any gamut clipping.

> Also, is there an any disadvantage to doing corrections and editing in
> ProPhoto RGB? In other words, could one stay in ProPhoto RGB from RAW
> conversion to sending the file to an Epson 2200, 7600, 9600, 4000.

As long as you stay in 16 bit, there's no more risk of editing ProPhoto RGB
than any other working space ecept for the fact that no monitor can display
al the colors of the file. But that's a problem with any color space larger
than a monitor.

However, ProPhoto RGB is not for 8 bit files. . .edit an 8 bit ProPhoto RGB
file and you'll get banding real quick.


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Stephen Clark
Posted on Mon Nov 24, 2003 5:56 PM

> Jeff Schewe wrote:
>
> As long as you stay in 16 bit, there's no more risk of editing ProPhoto RGB
> than any other working space... edit an 8 bit ProPhoto RGB
> file and you'll get banding real quick.

This is the same ProPhotoRGB Kodak came up with back in Y2K?

Bruce Lindbloom's table (http://www.brucelindbloom.com/index.html?WorkingSpaceInfo.html) ranks it at 91.2% Lab Gamut Efficiency (highest listed) where AdobeRGB(1998) is only 50.6%.

So it's still an appropriate "working space" choice in this CS age?


Jeff Schewe
Posted on Mon Nov 24, 2003 7:40 PM

Stephen Clark wrote:

> So it's still an appropriate "working space" choice in this CS age?

Yes, in 16 bit. . .


Jeff Schewe
www.schewephoto.com - www.pixelgenius.com - www.imagingrevue.com





Bruce Fraser
Posted on Tue Nov 25, 2003 1:50 AM

The gamut of Ultrachrome certainly contains colors outside of Adobe
RGB, especially the darker cyans and greens. I've been using ProPhoto
for the majority of my work for about 5 years now, with no problems
(or at least none that are attributable to my choice of working
space)...

At 4:10 PM -0600 11/24/03, IR - Robert Daniels wrote:
>Jeff, if the gamut of epson ultrachrome printers is larger than
>Adobe RGB, wouldn't your suggestion of bringing RAW files into
>Photoshop in ProPhoto RGB save some unnecessary clipping (assuming
>the chip captured an image with a color space larger than Adobe RGB
>-- shooting vibrant colors etc.).
>
>Also, is there an any disadvantage to doing corrections and editing
>in ProPhoto RGB? In other words, could one stay in ProPhoto RGB
>from RAW conversion to sending the file to an Epson 2200, 7600,
>9600, 4000.
>
>Thanks... The idea of using ProPhoto is interesting.
>--------------------------------------------------


Robert Daniels
Posted on Tue Nov 25, 2003 8:17 AM

Bruce, That's great information... I assume, from what you wrote, that ultrachrome printers can take ProPhoto RGB input --and make use of its larger gamut (over Adobe RGB). While not all images would use the extra color space, this would insure that all the capibilities of the ultrachrome printers would be used when the imaged called for it, without any unnecessary clipping.

Also, in the name of "output agility," I take it that, independent of the output device, it would make sense to use the larger gamut of Pro Photo in making the "master file."


Bruce Fraser
Posted on Tue Nov 25, 2003 8:15 PM

Ultrachrome printers can take whatever you throw at them. Obviously,
different source gamuts will map differently into the
printer/paper/ink gamuts, but ProPhoto has never given me a problem.
I don't try to create colors more saturated than the ones I've
captured, though!

Ultrachrome on Premium luster or glossy has a considerable chunk of
dark saturated cyans and greens that lie outside Adobe RGB, and a
bigger chunk of saturated oranges and yellows that also lie outside
Adobe RGB.



At 8:20 AM -0600 11/25/03, IR - Robert Daniels wrote:
>Bruce, That's great information... I assume, from what you wrote,
>that ultrachrome printers can take ProPhoto RGB input --and make use
>of its larger gamut (over Adobe RGB). While not all images would
>use the extra color space, this would insure that all the
>capibilities of the ultrachrome printers would be used when the
>imaged called for it, without any unnecessary clipping.
>
>Also, in the name of "output agility," I take it that, independent
>of the output device, it would make sense to use the larger gamut of
>Pro Photo in making the "master file."
>--------------------------------------------------





Other Sections Available if you login:
calendar

April 2004
S M T W T F S
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18
19
20
21
22
23
24
25
26
27
28
29
30
 
May 2004
S M T W T F S
            1
2 3 4
5
6
7
8
9 10 11 12
13
14 15
16
17
18
19
20
21
22
23
24 25 26 27 28 29
30 31          
 

Upcoming Events


MAKE YOUR WORKflow (A D-65 event - led by IR expert Seth Resnick)
    4/18/04 at 9:00 AM to 4/24/04
The Digital Portfolio, (at John Paul Caponigro's studio, Maine)
    4/19/04 at 9:00 AM to 4/23/04
Article - Book Review
    4/24/04 at 5:00 PM
Santa Fe Workshops - Master Photoshop
    4/26/04 at 9:00 AM to 4/30/04
Article - Tutorial
    4/30/04 at 5:00 PM

...see more upcoming events

imagingrevue Members receive special discounts on most events listed in calendar

  


customer.support@ imagingrevue.com
Privacy Policy
Terms of Service