<
  News     DIFP   Benefits Profiles Account  


forums

Recent Posts



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

 Confused about layers in PK Sharpener
Apr 21, 2004 9:01 PM
Posted by Bruce Fraser

 Sony Artisan and Windows
Apr 21, 2004 8:44 PM
Posted by Thomas Fors

 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

 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:inCamera and/or ACR
Friday November 7, 2003

Section 1: inCamera and/or ACR






Color Management/WorkFlow > Hardware and Software > inCamera and/or ACR




































































AuthorSubject: inCamera and/or ACR  
Eric Hiss
Posted on Wed Aug 20, 2003 2:22 AM

Hi,

Now that I've got my printer and display profiles under control, I'm thinking about camera profiles for my Canon 1D. I've read very favorable reviews about a sofware called inCamera. What do you guys think about that? Also I use ACR mostly for the RAW conversion....can that be configured to accept the custom input profile? I know that C1 can but I'm on a Mac and don't think its as fast as ACR (yet).

Thanks,
Eric Hiss


Jeff Schewe
Posted on Wed Aug 20, 2003 3:05 AM

Eric Hiss wrote:

> Also I use ACR mostly for the RAW conversion....can that be configured to
> accept the custom input profile?

Nope. . .sorry, Thomas Knoll doesn't like camera profiles. . .he prefers to
dynamically adjust between two color profiles (3200K and 6500K) to get
optimum white balancing of conversions. . .which is pretty much a
breakthrough (depending on who you talk to and who's trying to sell what).

I have no opinion of "InCamera" specifically, nor any of the current crop of
camera profiling. I do have some opinions about camera profiles in general.
The further away from the exact conditions that existed at the time you MADE
the target capture to make the profile, the less useful that profile will
be.

So, if you have a "standard lighting setup" for certain types of shooting, a
camera profile may work very well. Change the lighting substantially and all
bets are off (so will the profiles be).

If you shoot in changing light (outside) and have no control over things
such as color temp or dynamic range, I seriously think camera profiles are
useless. . .

My preference is to use the small Macbeth digital camera target and shoot a
frame at the beginning of a series of shots. I'll use the second from the
brightest patch to set the white balance tool in Camera Raw. . .then use
droplets to batch convert whole series of shots using the custom settings.


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




Andrew Rodney
Posted on Wed Aug 20, 2003 7:50 AM

on 8/20/03 1:25 AM, IR - Eric Hiss wrote:

> I've read very favorable reviews about a sofware called inCamera. What do you
> guys think about that?

I've had reasonably good results from inCamera but keep in mind I've yet to
find ANY camera profiling solution that's even close to 100%. InCamera is
easy to use and it's damn inexpensive so I'd give it a shot since as I said,
I've seen it work more than not. I'd check and see if you can get say a 30
day money back guarantee (I'd want that with any camera profile package).
You might ask them if they recommend using the GretagMacbeth ColorChecker DC
since that raises the cost quite a bit more. InCamera does work with the
standard 24 patch Macbeth and you might be happy with the results using that
target. But camera profiling seems to be hit or miss...


Keep in mind that if you wish to use Adobe Camera RAW, you don't need (and
can't use) custom profiles. But that's not an InCamera issue.


Ian Lyons
Posted on Wed Aug 20, 2003 8:21 AM

>
> I've had reasonably good results from inCamera but keep in mind I've yet to
> find ANY camera profiling solution that's even close to 100%. InCamera is
> easy to use and it's damn inexpensive so I'd give it a shot since as I said,
> I've seen it work more than not. I'd check and see if you can get say a 30
> day money back guarantee (I'd want that with any camera profile package).
> You might ask them if they recommend using the GretagMacbeth ColorChecker DC
> since that raises the cost quite a bit more. InCamera does work with the
> standard 24 patch Macbeth and you might be happy with the results using that
> target. But camera profiling seems to be hit or miss...
>


inCamera is one of the better solutions and for the Mac I would suggest the Plug-in rather than the standalone version. The plug-in also allows you to build scanner profiles.

I've not found any benefits to using the more expensive DC target and so would suggest trying your luck with the basic ColorChecker.

You can see what I thought of inCamera here:

http://www.computer-darkroom.com/incamera-pi/incamera-pi.htm


John MacLean
Posted on Wed Aug 20, 2003 12:21 PM

> My preference is to use the small Macbeth digital camera target ...
>
> Jeff Schewe


Jeff,

Do you mean the Mini MacB
www.bhphotovideo.com/bnh/controller/home?O=productlist&A;=details&Q;=&sku;=222176&is;=REG

or the DC
www.bhphotovideo.com/bnh/controller/home?O=productlist&A;=details&Q;=&sku;=222175&is;=REG

I'm assuming the former.

I've got the full size regular Macbeth, but it is a pain to carry around.

Andrew, A while back you mentioned on one of the forums about a larger sized card of just the #20 - Neutral 8 patch color, or something similar. Is there such a beast?

Thanks,
John

www.nmdigital.com


Jeff Schewe
Posted on Wed Aug 20, 2003 12:55 PM

John MacLean asked:

> Do you mean the Mini MacB

Yes. . .the Macbeth MINI COLOR CHECKER is what I use. . .I just hold it in
front of the lens and do a quick shot-as long as the lighting hitting it is
the same as on the subject, it's an excellent tool for setting the Camera
Raw white balance tool. I also use the black and white patch as reference of
scene luminance range-although I do NOT try to force the black patch to be
zero RGB and I don't try to force the white patch to be 255. I like to keep
a lot of head room in CR and do final point setting in Photoshop.


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




Andrew Rodney
Posted on Wed Aug 20, 2003 1:00 PM

on 8/20/03 11:25 AM, IR - John MacLean wrote:

> Andrew, A while back you mentioned on one of the forums about a larger sized
> card of just the #20 - Neutral 8 patch color, or something similar. Is there
> such a beast?

Yes but no one outside of GretagMacbeth and a few testes have one but it's
coming...


John MacLean
Posted on Thu Aug 21, 2003 10:23 AM

> Yes but no one outside of GretagMacbeth and a few testes have one but it's
> coming...

Andrew,

Can you alert us when it's available?

Thanks,
John
www.nmdigital.com


Andrew Rodney
Posted on Thu Aug 21, 2003 10:26 AM

-->Can you alert us when it's available?

But of course! I'll also ping the boys and girls in Europe and see if they have a date in mind yet.


Andrew Rodney
Posted on Thu Aug 21, 2003 4:13 PM

The targets will be released by Photo Expo or sooner according to my sources inside of Gretagmacbeth.


John MacLean
Posted on Fri Aug 22, 2003 10:23 AM

Thanks,

When is P.E.? I'm looking forward to your alert!

Best,
John


John Vitollo
Posted on Fri Aug 22, 2003 10:34 AM

http://www.photoplusexpo.com/photoplusexpo/index.jsp

New York, NY
Jacob Javits Convention Center
October 30 - November 1, 2003


Allen Pacheco
Posted on Fri Aug 22, 2003 10:49 AM

I remember reading somewhere that you could use a very wide space like ProPhoto as the output destination for Camera Raw when processing a camera profiling target. Then you'd develop a profile and and in order to use it you'd continue to use ProPhoto and simply apply the custom profile "over" it in PhotoShop. Is this far-fetched or does it carry any weight?


Also, although I have a good deal of respect for those who have given inCamera glowing reviews, they do not demonstrate several very distinct limitations of the program. Limitations which I have never been able to overcome. I have the stand-alone rather than the plug-in so I am assuming that the profile construction routines are very much alike. The ability to reproduce the target that is used to actually build the profile does not show the whole picture, so to speak. I would like to see one demonstration of a decent profile constructed from inCamera that could be used with confidence.





Bruce Fraser
Posted on Fri Aug 22, 2003 10:55 AM

-->Is this far-fetched or does it carry any weight?

I'd say it's far-fetched. To the extent that it worked at all, you're essentially tossing all the advantages of shooting raw. Better to simply use ACR as designed.


Ian Lyons
Posted on Fri Aug 22, 2003 11:18 AM

Allen,


>
> Also, although I have a good deal of respect for those who have given inCamera glowing reviews, they do not demonstrate several very distinct limitations of the program. Limitations which I have never been able to overcome. I have the stand-alone rather than the plug-in so I am assuming that the profile construction routines are very much alike. The ability to reproduce the target that is used to actually build the profile does not show the whole picture, so to speak. I would like to see one demonstration of a decent profile constructed from inCamera that could be used with confidence.
>
>

Maybe you'd be so kind as to describe what you see as limitations. What part of the process are you missing or don't understand, etc. etc. I can't speak for Bruce, Andrew or Jeff but I lack Spocks ability to mind meld. Help us help YOU!




Andrew Rodney
Posted on Fri Aug 22, 2003 11:28 AM

on 8/22/03 9:50 AM, IR - Allen Pacheco wrote:

> I remember reading somewhere that you could use a very wide space like
> ProPhoto as the output destination for Camera Raw when processing a camera
> profiling target. Then you'd develop a profile and and in order to use it
> you'd continue to use ProPhoto and simply apply the custom profile "over" it
> in PhotoShop. Is this far-fetched or does it carry any weight?

But why? I'm not sure what the advantage would be to take all the time an energy to do this when ACR doesn't really require it.

> I have the stand-alone rather than the plug-in so I am assuming that the
> profile construction routines are very much alike.

The newer versions of the plug-in are significantly better than the older stand-alone. Not that you'll get perfect results 100% of the time but I'd see what it takes to move from the stand-alone to the plug-in.


Andrew Rodney
Posted on Fri Aug 22, 2003 11:40 AM

on 8/22/03 9:50 AM, IR - Allen Pacheco wrote:

> I remember reading somewhere that you could use a very wide space like
> ProPhoto as the output destination for Camera Raw when processing a camera
> profiling target. Then you'd develop a profile and and in order to use it
> you'd continue to use ProPhoto and simply apply the custom profile "over" it
> in PhotoShop. Is this far-fetched or does it carry any weight?

But why? I'm not sure what the advantage would be to take all the time an
energy to do this when ACR doesn't really require it.

> I have the stand-alone rather than the plug-in so I am assuming that the
> profile construction routines are very much alike.

The newer versions of the plug-in are significantly better than the older
stand-alone. Not that you'll get perfect results 100% of the time but I'd
see what it takes to move from the stand-alone to the plug-in.


Allen Pacheco
Posted on Fri Aug 22, 2003 2:17 PM

Ian,

Discussing inCamera...

First, I'm going to refer to the 24-tile chart, not the ColorChecker DC.
And for additional reference I use Bruce Lindbloom's computer generated file of the ColorChecker chart based upon the Lab values of the chart....
http://www.brucelindbloom.com/index.html?ColorCheckerRGB.html

If you open the reference file that comes with inCamera you will see similar values.

So far so good.
The Pictographics profiling engine is going to use these values as a basis for computing the profile for the target image provided.

Assuming you use even lighting and well-controlled neutral balance then the target exposure now becomes the main variable, correct?

And the key values to determine exposure are the six neutral tiles on the target.

So, if you make a series of exposures, gross over and under exposures are easy to dismiss by just viewing the results. You end up with a range of exposures which could be considered quite good. So I tried to zero in on an image that satisfied the profile requirements.

After studying the results and making comparisons to the "desired" values in the reference file, it's clear to me that the best results are obtained when using an exposure that matches as close as possible as many of the neutral tiles as possible.
In my case, using Canon cameras, I get the best results when I match tile #20 (neutral 8) which is approximately L81 (Lab). However, the resultant profile wants the bright white tile to be L96. And when you apply this profile to an image the highlights are dealt a pronounced clipping. This soft shoulder of the processed data is straightened right out.

Deep shadows and three-quarter tones are also lifted and the contrast in these areas reduced but that can be addressed in an editor and may or may not be a bad thing.

But the highlights are gone. And to make things worse the clipping is not equal across all three channels...the blue channel is worse than the others and the result is a greenish off-white effect in the highlights.

I experimented using target images with more exposure to compensate and the resultant profile curve pulls and tugs at the image data. In other words, if the data in the target image deviates by a large amount from the reference values the resultant profile will end up distorting the image. In my mind I want the least amount of distortion possible. A profile without contrast bias.


As I mentioned, I have version 3.1.0.0 of inCamera. Perhaps, as Andrew mentions, the plug-in produces different results.
I spent some time talking with a rep at this year's PMA and he felt the best use of the product was for low contrast scenes with controlled lighting so as not to end up with highlight data too far beyond L90 (Lab). Is this practical for some? Perhaps. But it is still a limitation. Unless you have a way around it.

If I'm off in left field then show me the error in my method and I'll praise this product loud and clear.



Allen Pacheco
Posted on Fri Aug 22, 2003 2:52 PM

Hi Andrew,
You have been a staunch supporter of custom printer and monitor profiles. To deal with the unique output characteristics of a printer or monitor via generic profiles is almost always shunned, correct?

So why the reversal with using the generic T.Knoll profiling in Camera Raw? True, it delivers the output to a standardized color space but doesn't this fly straight into the face of precise color-management?


Bruce Fraser
Posted on Fri Aug 22, 2003 3:19 PM

Andrew I'm sure has his own opinion. I'll simply point out that profiling digital cameras is fundamentally different from profiling other devices.

Monitors and printers have a fixed color gamut. Capture devices do not, but in the case of scanners it's rendered moot because the film that we scan does have a fixed color gamut.

Digital cameras have something much more slippery�they have a color mixing function. They DO have a somewhat fixed dynamic range, modulated by the exposure controls, but they don't have a fixed color gamut.

ACR uses not one but two 'generic' profiles for each camera, one created under tungsten, the other under D65, but more importantly, it allows interpolation between, and extrapolation beyond these profiles.

In a fixed lighting situation with locked exposure controls, a conventional profile makes sense. Outside of these conditions it becomes progressively less reliable the further away the conditions are from the ones in which the target was captured.

So it makes sense to recognize cameras as a special case, and deal with them accordingly. Thomas' approach in ACR is one that makes a lot of sense, but more importantly, it works rather well when the camera's behavior bears a reasonable resemblance to the one he profiled.


Jeff Schewe
Posted on Fri Aug 22, 2003 4:10 PM

Allen Pacheco wrote:

> So why the reversal with using the generic T.Knoll profiling in Camera Raw?
> True, it delivers the output to a standardized color space but doesn't this
> fly straight into the face of precise color-management?

There is no such things as a "precise color-management" ability when dealing
with a device that always changes. . .Yes, one CAN profile a camera. . .and
that profile will be "useful" in direct proportion to the exact nature of
the state of the camera when the profile was made. Change _ANYTHING_
relating to the camera such as exposure, lighting, light balance, lens and
even exposure time, and the usefulness of the profile diminishes.

If you have a "standard" studio set up with a typical lens, a typical light
source and typical exposure, a profile can be quite useful. However, change
anything and it's a crap shoot.

This is akin to difficulties in profiling various scanners over the years
that contained a non-changeable "auto-setting" that made consistent and
repeatable scans impossible. Those scanners, one of the Polaroids was
particularly vexing, in effect were NOT "profilable" because the scanner
changed based upon the images scanned.

Another analogy is that while you CAN scan an E-6 target and create a
profile for E-6 film, there are no targets for color neg film. Why? Because
each film and each process batch can vary. . .making profiling impossible.

The geeks & gurus who have advocated color management practices have long
lamented the limitations and difficulties of camera profiling. But the
simple truth is, unless there are no variables, color management simply will
not work.

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





Andrew Rodney
Posted on Fri Aug 22, 2003 4:13 PM

> Hi Andrew,
> You have been a staunch supporter of custom printer and monitor profiles. To deal with the unique output characteristics of a printer or monitor via generic profiles is almost always shunned, correct?

Not always. I've found the canned profiles in ImagePrint are quite good and while I can (and do) make custom profiles for it, I could see where a large majority of users would be quite happy with them. But that's an exception to the rule.

> So why the reversal with using the generic T.Knoll profiling in Camera Raw? True, it delivers the output to a standardized color space but doesn't this fly straight into the face of precise color-management?

You didn't read my "Point-Counter Point" on ACR in PEI mag ("Jane, you ignorant slut" <g>).

What ACR does, on a calibrated display is give you virtually all the controls to render an image as you wish, and then bring that into Photoshop with a Working Space you pick and honors that appearance and the numbers you saw in ACR. That's all I really need.

Without this kind of workflow, the numbers you get from a camera are undefined (without RAW, you can kind of believe that you're getting sRGB or Adobe RGB but I don't really buy into that). So with this RAW file you have numbers with no meaning and Photoshop hasn't a clue how to preview or convert that data. So you need a profile. But you don't with ACR. So in a nutshell, there's no reason with ACR to worry about building a profile. What you see is what you get. What more do you want?


Bruce Fraser
Posted on Fri Aug 22, 2003 4:52 PM

>But the
>simple truth is, unless there are no variables, color management simply will
>not work.

Make that uncontrolled variables. There are always variables....


Jeff Schewe
Posted on Fri Aug 22, 2003 5:40 PM

Bruce Fraser wrote:

> Make that uncontrolled variables. There are always variables....

Yeah. . .what Bruce said. . .

:-)


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





Herbert Gibson
Posted on Fri Aug 22, 2003 6:09 PM

Unless I have lost my marbles (not out of the question) the geeks and gurus BCR (Before Camera Raw) used to advocate camera profiling. There was a minor argument, for instance, that the EOS D30 did not produce EXACTLY sRGB but the major selling point made by these folk for profiling hardware and software was variation across individual items of equipment.

As soon as Adobe launched ACR, the field seemed to split into Pals of Phase One and Mates Of Adobe. At that same time ACR advocates started to downplay the variation argument.

Sorry; but if the variation argument was valid before, it remains valid as an argument against Camera Raw. If it was not valid, then some folk were urging the purchase of expensive profiling gear where it was not necessary.

I realise that in some cases (reputedly the 2100 printer) variation may have been reduced, but that case has not been made for cameras.

There is an argument that ACR tries to balance profiles between 2 color temp end points whereas a user profile is only good for one point. But there is no reason why the user cannot make a number of profiles for different conditions and call them up as required. No great hardship.

Unfortunately, it appears that those resident in the Adobe camp see their role as cheerleaders for the Camera Raw implementation rather than presenting a balanced view. Ironically, that has only served to weaken your argument.


Andrew Rodney
Posted on Fri Aug 22, 2003 6:21 PM

There was nothing like ACR before ACR! We're dealing with RAW data here (which is a plus). Other than RAW, the camera spits out what it thinks is "right" and often, the numbers are undefined (or not defined correctly). So you had to have some profile to describe the numbers to Photoshop so it could preview and convert that data correctly.

That's just not necessary with ACR. Think iof RAW and ACR as having a color neg and you get to decide what filter pack you want.

Think of non RAW files as chromes that came out of the soup with the color and density fixed. You got what you got (but imagine if you will that the light in which you viewed the chrome was so undefined and could be virtually any color until you told the light source "I want daylight" (the profile).

If you want to use ACR, you pretty much forget using profiles for the portion of the workflow where you bring the numbers into Photohop. If you want to use another method of producing the color you'll likely need a profile or at the very least the right meaning of the numbers for Photoshop (which is all a profile does anyway).

Why do I like ACR? Well if I don't have to spend $500 on a good target and then a few hundred for software, then spend the time to make a profile (which is hit or miss half the time).

As I've said elsewhere; people keep asking for easier/quicker/simplier solutions for color management. When we get one, a group complains that we color geeks are happy there' less work to produce better results. I don't get it.

If someone came out with a self calibrating and profiling printer, I'd be super happy. You think I LIKE measuring patches?


Jeff Schewe
Posted on Fri Aug 22, 2003 6:25 PM

Herbert Gibson wrote:

> Unfortunately, it appears that those resident in the Adobe camp see their role
> as cheerleaders for the Camera Raw implementation rather than presenting a
> balanced view. Ironically, that has only served to weaken your argument.

You never saw me advocate profiling a camera. . .I tried it with several
packages. . .and to my eyes and way of working, no profile has ever worked.
. .I advocate 16 bit color correction, in large part BECAUSE camera
profiling sucks. . .

Quite honestly, I don't give a rat's ass about a "balanced view". . .I've
tried C1, and it sucks for usability and UI on the Mac. I've tried the Canon
shit and that sucks. . .I'm a photographer and user of this stuff. . .I say
what works and doesn't work for me.

Camera Raw suck a lot less than any of the other applications out there if
you have a Photoshop centric workflow. . .one that I've ALWAYS indicated
that I use. . .

I think Mr. Gibson, that your anti-Adobe slant is peaking through?

Use whatever works for you. . .or quite working.


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




Bruce Fraser
Posted on Fri Aug 22, 2003 6:30 PM

Herbert,

I don't think you'll find any published piece or documented conversation where I have advocated camera profiling with any greater force than saying "if you shoot in the studio under tightly-controlled lighting, it may be worthwhile." (Which in my book falls substantially short of advocacy...)

Camera-to-camera variation isn't even slightly the point here. The point is that a single camera's response varies quite dramatically under different lighting conditions. It's a fundamentally different problem from those we face in profiling printers, displays, or even scanners.

Yes, you could make profiles for a camera under different lighting conditions. Conceivably, you could even write actions that would let you blend between renditions offered by the different profiles. But the profiles in ACR aren't endpoints�ACR can extrapolate beyond them. I can't think of a way to do that with static profiles and images in Photoshop. ACR offers a lot of functionality for $99. Its weakness, which I've pointed out on many occasions, is that a good many people out there seem to own cameras that don't behave the same way as the sample that Thomas profiled�particularly Nikons. That's being looked at by the people who may be able to do something about it.

I'm not in any particular camp, and I outgrew my cheerleader outfit decades ago. Not am I making any particular argument. I just explain how the damn things work, and let people draw their own conclusions. I personally find Camera RAW much more useful than Phase One Capture, but I only shoot under available light, and don't even pretend to be a photographer.

I think the Camera RAW approach is a conceptual breakthrough in addressing the central problem of characterizing the color produced by digital cameras, which is that they have a very variable response depending on the scene dynamic range and color temperature. Is it the be-all and end-all? Hell, no. But it seems to make more sense in more situations than the alternative, which is to try to lock down all the lighting and exposure variables, then profile the result. That's the philosophical argument, such as it is.

But the only reason I use ACR is because it works better than anything else I've tried.


Jeff Schewe
Posted on Fri Aug 22, 2003 7:55 PM

Bruce Fraser wrote:

> I'm not in any particular camp, and I outgrew my cheerleader outfit decades
> ago.

You got a picture of you in your cheerleading outfut? I would pay good money
to see that. . .I've seen you in a skirt, er, kilt, but you in a
cheerleader, that WOULD be rich. . .

Jeff


Bruce Fraser
Posted on Fri Aug 22, 2003 8:19 PM

We should do it together...


Ian Lyons
Posted on Sat Aug 23, 2003 9:55 AM

Allen,


>
> Assuming you use even lighting and well-controlled neutral balance then the target exposure now becomes the main variable, correct?

To a large extent exposure is the main variable, but not the only one.

Building a profile for the non linear images converted via Canons FVU or Phase One is relatively easy. So I assume that you're trying to build a profile based on the linear raw data. In such circumstances and without some aid to get the max white close to optimum then you'll likely struggle with every digicam profiling app currently available. If the conversion is to a gamma corrected (some prefer to call it toned) value and it results in highlight end spikes then the chances are that you will either have blown highlights or will cause the profiling app to produce less than ideal profiles. For such images you've got to find a compromise between linear and the gamma/tone corrected image from Canon FVU or Phase One. I described how I went about it in the following tutorial http://www.computer-darkroom.com/d30-profiling/d30_1.htm


With the Canon D30/60 my compromise meant that the camera profile tends not to be totally accurate. Simply applying a standard gamma curve to to the linear image appears to produce an image with its own set of problems; one of which is the green cast in the highlights that you mention. Since I work on the principle of holding the highlights as best as posible I use my own compensation curve. I build the profile based on this curve and apply it to those images what will benefit from it. The image will tend to be low in contrast but I hold everything that was capatured. I then fix the remaining in Photoshop. If the image isn't going to benefit from this curve/profile I'll now use Adobe Camera Raw and save myself a headache.

I don't have a clue about highlight characteristics of other cameras so the curve and how to determine the optium exposure to build it isn't something I can help you with.

If the image doesn't have one or two channels hitting level 255 or blown highlights then I'l just use Adobe Camera Raw.

Unfortunately there are currently no conversion apps that allow me to totally avoid building profiles. However, I do find that there is lot more flexibility in my approac



>
> And the key values to determine exposure are the six neutral tiles on the target.
>

Max white and lightest grey (patches 19 and 20 <neutral 8>) are the only ones I look at nowadays. The other 4 patches aren't really worth the card they're printed on for determing the optimum exposure for building a profile.


> So, if you make a series of exposures, gross over and under exposures are easy to dismiss by just viewing the results. You end up with a range of exposures which could be considered quite good. So I tried to zero in on an image that satisfied the profile requirements.
>
> After studying the results and making comparisons to the "desired" values in the reference file, it's clear to me that the best results are obtained when using an exposure that matches as close as possible as many of the neutral tiles as possible.

The gross over exposure theory doesn't work as it screws up everything else on the target. I simply choose the one that places patch 19 between level 240 and 250. I'll use patch 20 to determine/adjust the white balance. Patch 19 is used to choose the image with the optimum exposure. Once I capture the colorchecker with patch 19 lying between level 240 and 250 I'm done buggering around. If it falls a tad short or over this value I'll cheat.

> In my case, using Canon cameras, I get the best results when I match tile #20 (neutral 8) which is approximately L81 (Lab). However, the resultant profile wants the bright white tile to be L96. And when you apply this profile to an image the highlights are dealt a pronounced clipping. This soft shoulder of the processed data is straightened right out.
>

That suggest that you are profiling the raw linear image. A linear raw image doesn't have a shoulder. There are no soft landings with digital capture. Basing your exposure on patch 20 means that you will b leaving a lot up to chance. If necessary forget tonal accuracy - just get the damned highlights right.


> Deep shadows and three-quarter tones are also lifted and the contrast in these areas reduced but that can be addressed in an editor and may or may not be a bad thing.
>

You're expecting miracles from the current targets. Max white is at or around level 242 and max black at or around level 50. Everything above and below are pure invention on the part of the profiling apps.


> But the highlights are gone. And to make things worse the clipping is not equal across all three channels...the blue channel is worse than the others and the result is a greenish off-white effect in the highlights.
>

I only see this when building profiles for the for linear raw images. Other vendors have a diferent colour cast or worse - total reversal (i.e. near white turns black!) Again, I'm talking about the Canon D30, D60 and 10D!

I'm told that if you want into profiling nirvana you should speak to Jack Bingham or get his ColoreEyes 20/20. The top secret manual is supposed to be the answer to all our profiling problems. Me? I'll make do with inCamera and Adobe Camera Raw







Allen Pacheco
Posted on Sat Aug 23, 2003 5:48 PM

Ian,

> Building a profile for the non linear images converted via Canons FVU or Phase One >is relatively easy. So I assume that you're trying to build a profile based on the >linear raw data.

Actually, I was not using linear data. I do have a D30 on a microsope that I use in linear mode but I use ImagePro+ to work the data, not PhotoShop, and I have no need for profiling it.

Surprisingly, I see the problems with highlights in non-linear processed images. Two 10Ds..processing via Canon SDK (BreezeBrowser) or Capture One...depending upon how I feel about C1 that particular day. I did try a modification of your curve method only I used it on the non-linear image data. I applied curves based upon what the reference file wanted to see. Results were encouraging but the highlights were still off.

> You're expecting miracles from the current targets. Max white is at or around level >242 and max black at or around level 50. Everything above and below are pure >invention on the part of the profiling apps.

And perhaps you're right....I was hoping to push the envelope a bit. Seems that just a little work by the programmers could alleviate the issue.


> I'm told that if you want into profiling nirvana you should speak to Jack Bingham or get his ColoreEyes 20/20. The top secret manual is supposed to be the answer to all our profiling problems. Me? I'll make do with inCamera and Adobe Camera Raw

Well, you could be right. I had no intention of mentioning J.B and ColorEyes. Although, I did read his Thumbnail faceoff in PEI with Andrew. Good points by both parties. The original poster (poor guy, remember him?) was solely interested in using inCamera with ACR.
BTW, I have corresponded with a user of 20/20 who indeed had a few issues with highlights, also. But they were resolved with some editing.
Which begs the question...is there a profile editor that can modify inCamera's LUT-based profiles with a fine degree of control? The included editor does more damage than good.

Thanks for taking the time for responding.


Andrew Rodney
Posted on Sat Aug 23, 2003 7:13 PM

-->Which begs the question...is there a profile editor that can modify inCamera's LUT-based profiles with a fine degree of control?

ProfileEditor from GretagMacbeth should be able to. I'll have to check if the new OSX version of ColorBlind Edit will but that product as it stands is so screwed up it's probably not worth considering.


Steve Bohne
Posted on Sat Aug 23, 2003 10:54 PM

> Jeff Schewe wrote:

> Quite honestly, I don't give a rat's ass about a "balanced view". . .I've
> tried C1, and it sucks for usability and UI on the Mac. I've tried the Canon
> shit and that sucks. . .I'm a photographer and user of this stuff. . .I say
> what works and doesn't work for me.
>
> Camera Raw suck a lot less than any of the other applications out there if
> you have a Photoshop centric workflow. . .one that I've ALWAYS indicated
> that I use. . .

I started with a Fuji S1 (well, actually with a Kodak DC120, then a Coolpix 990, but we won't go there), then moved up to an S2, getting one of the first cameras released. While I loved the image quality (have a 30x40 in my reception room that even the lab thought was film scanned into a digital file), I became unhappy with the tendancy of the Nikon AF engine to backfocus.

I moved up/over to a Canon 10D (coveting the 1Ds, but lousy year this year forces prudence in the fiscal responsibility area). Liked it, so I bought another, and sold the S2 (still have one to sell).

Many are touting the ability to capture RAW. I have always worked with a jpeg workflow, therefore always shot for good capture. I am confused by what I read on this and other forums about RAW capture:

1) some say Capture One is the only way to go;

2) others (e.g., Mr Schewe) say the Adobe Plug in (which I hate to pay for since they'll include it in PS8 (which forces me, kicking and screaming, to "upgrade" to OS X, and most likely, a newer computer--bye bye bux) while others say it blows;

3) while Jeff states the Canon software is feces, I have others say it is much better than the ARP or C1, although file updating is admittedly slow.

I have a medium volume business. I photograph 100+ high school seniors, and I have enough work in sizing and rotating files for on screen viewing. I can't imagine adding yet MORE work in my Photoshop workflow. What is the advantage of shooting RAW if my exposures are always good (well, ok, at least in the neighborhood)? I normally have to boost the contrast a little, and maybe adjust the highlights slider a bit.

WHY does the Canon software suck (I've not used it)?
WHY should I use the Adobe Raw Plugin...and does it work with 10D files?
IS CaptureOne worth $500?

Inquiring minds want to know...


Jeff Schewe
Posted on Sun Aug 24, 2003 12:45 AM

Steve Bohne wrote:

> I have a medium volume business. I photograph 100+ high school seniors, and I
> have enough work in sizing and rotating files for on screen viewing. I can't
> imagine adding yet MORE work in my Photoshop workflow. What is the advantage
> of shooting RAW if my exposures are always good (well, ok, at least in the
> neighborhood)? I normally have to boost the contrast a little, and maybe
> adjust the highlights slider a bit.

Steve. . .

To be honest, you may actually be in a situation where a raw workflow is
just not gonna help you.

Seth Resnick and I have, for a couple of years, been arguing with Jay Maisel
about the benefits of raw and the weakness of jpg. Jay's argument against a
raw workflow is pretty much this, "look at my prints", do they look like
they were shot with jpg or raw?

I've seen Jay's work up close, and you can't tell. Oh, sure, on a 30x40, I
can spot jpg artefacts. . .if I move in close to inspect the prints. But
they don't get in the way of the image. . .

You have to understand where I come from. . .I want total, absolute control
over every aspect of my image. I want the total quality I can get out of my
equipment in pursuit of the perfect image. . .and I can get it. But, my
experience may not be relevant for everybody, nor my way of working.

In your case, dealing with raw captures is gonna be a royal pain in the ass.
You will always be fighting time. . .you will always be fighting getting the
image done vs getting the image perfect.

I do NOT mean this to be in ANY WAY a show of disrespect, but some venues of
photography are just not meant for raw captures. When time, workflow or
meeting a deadline matter, what's more important has to take precedence. Get
it done.

Screw raw capture if it makes your job impossible. If you have a
controllable shooting situation, can add fill or bounce light, can adjust
and pretty much nail exposures, you won't really see much benefit tp raw and
you WILL see a lot of time and effort get in your way.

Save raw capture for times where you CAN'T control the lighting, or where
the needs of the image require that absolute best image quality.

Yes, you'll have to live with a limited palette of color, you'll have to
live with a limited bit depth and you'll have to live with a limited
interpretation of your image. But, the upside is you may actually get to bed
at a reasonable hour.

:-)


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




Chuck Jones
Posted on Sun Aug 24, 2003 12:59 AM

Steve,
I'll toss my couple of thoughts in here, as I've been going through this very evaluation myself since changing over a lot of my own work to the 1Ds.

> > Jeff Schewe wrote:
>
> > Quite honestly, I don't give a rat's ass about a "balanced view". . .I've
> > tried C1, and it sucks for usability and UI on the Mac. I've tried the Canon
> > shit and that sucks. . .I'm a photographer and user of this stuff. . .I say
> > what works and doesn't work for me.

Steve, I have to agree with Jeff on Capture One. C1 may be the holy grail on a PC, but their Mac version bites. Guess it also may work ok for studio shooters, but I am out and about constantly, and always in very different mixed lighting conditions. Even when I can grab a Macbeth shot, most often the actual shots are going to be in slightly different light. Every stinking one of the 500 or so images. After reading all the "rave" reviews, and user posts, I downloaded the 30 day demo. Does PhaseOne pay these guys to write this stuff or what? I've yet to get a decent color balanced image out of C1 that didn't require extensive further color correction in Photoshop! And believe me, I have really tried, figuring that I just hadn't learned the software well enough. Well, to make a long story short, after untold hours of screwing around with it, I have reached the conclusion that (1) Schewe isn't off his rocker, and (2) he has probably the best single line statement I have read yet to discribe C1.

I bought ACR when it first came out to use with my D1X. I felt it was a little off for that camera. I tried it with my 1Ds, and woah, a tweak here, tweak there, couple of slider adjustments and wham, an almost perfect rendered image inside Photoshop, right where I want it. For me, very much worth the $99 price tag. Not worth it for the Nikon, but what I think is a must have for the Canon.

I've also used the stock Canon stuff and found that pretty crude, and very limited. About what you come to expect from all the digital camera makers to include free for the point and shoot croud. Except for Kodak that is. The Kodak Photodesk software they ship with the 14n is excellent. And it is free! They even release regular updates to keep it current and improve the feature set.

> Many are touting the ability to capture RAW. I have always worked with a jpeg workflow, therefore always shot for good capture. I am confused by what I read on this and other forums about RAW capture:
>
> 1) some say Capture One is the only way to go;

Someone should tell THEM where to go!

> 2) others (e.g., Mr Schewe) say the Adobe Plug in (which I hate to pay for since they'll include it in PS8 (which forces me, kicking and screaming, to "upgrade" to OS X, and most likely, a newer computer--bye bye bux) while others say it blows;

It works, works pretty darn well, and also has a hidden side benefit of giving you probably the best software available today to resize your images without much loss.

> 3) while Jeff states the Canon software is feces, I have others say it is much better than the ARP or C1, although file updating is admittedly slow.

Those folks are probably from the point and shoot croud who most likely think that color temperature is some kind of sunburn from lying around on the beach all day.

> I have a medium volume business. I photograph 100+ high school seniors, and I have enough work in sizing and rotating files for on screen viewing. I can't imagine adding yet MORE work in my Photoshop workflow. What is the advantage of shooting RAW if my exposures are always good (well, ok, at least in the neighborhood)? I normally have to boost the contrast a little, and maybe adjust the highlights slider a bit.

Try working in RAW. I think you would find using ACR for your tweaks would be even faster than having to click the menu items to adjust the things you discribe above. ACR presents all of those as sliders right on the same screen. Should save you even more time. With a well shot RAW image to start with, I can get it exactly where I want it using ACR in ten to fifteen seconds. Four finished images in under a minute is pretty darn good productivity, and impossible to match using JPGs.

> WHY does the Canon software suck (I've not used it)?

How high is up? Who is burried in the tomb of the unknown? Just another of life's little mysteries. I'd just be guessing here, but probably because it is designed by programmers who make their living writing software, not from taking pictures for clients.

> WHY should I use the Adobe Raw Plugin...and does it work with 10D files?

Not sure on the 10D question. Someone else will have to answer that part for you.

> IS CaptureOne worth $500?

Let me put it this way, if it was on sale for $49, I mite have to consider buying it. Until then, when the demo expires it will be wiped from my editing system.

Warmest Regards,
Chuck Jones
www.chuckjonesphotography.com


Seth Resnick
Posted on Mon Aug 25, 2003 7:00 AM




> Steve Bohne wrote:
>
>> I have a medium volume business. I photograph 100+ high school
>> seniors, and I
>> have enough work in sizing and rotating files for on screen viewing.
>> I can't
>> imagine adding yet MORE work in my Photoshop workflow. What is the
>> advantage
>> of shooting RAW if my exposures are always good (well, ok, at least
>> in the
>> neighborhood)? I normally have to boost the contrast a little, and
>> maybe
>> adjust the highlights slider a bit.
>

On Sunday, August 24, 2003, at 01:45 AM, IR - Jeff Schewe wrote:


> Steve. . .
>
> To be honest, you may actually be in a situation where a raw workflow
> is
> just not gonna help you.
>
> Seth Resnick and I have, for a couple of years, been arguing with Jay
> Maisel
> about the benefits of raw and the weakness of jpg. Jay's argument
> against a
> raw workflow is pretty much this, "look at my prints", do they look
> like
> they were shot with jpg or raw?
>
> I've seen Jay's work up close, and you can't tell. Oh, sure, on a
> 30x40, I
> can spot jpg artefacts. . .if I move in close to inspect the prints.
> But
> they don't get in the way of the image. . .
>
> You have to understand where I come from. . .I want total, absolute
> control
> over every aspect of my image. I want the total quality I can get out
> of my
> equipment in pursuit of the perfect image. . .and I can get it. But, my
> experience may not be relevant for everybody, nor my way of working.
>
> In your case, dealing with raw captures is gonna be a royal pain in
> the ass.
> You will always be fighting time. . .you will always be fighting
> getting the
> image done vs getting the image perfect.
>
>
> Screw raw capture if it makes your job impossible. If you have a
> controllable shooting situation, can add fill or bounce light, can
> adjust
> and pretty much nail exposures, you won't really see much benefit tp
> raw and
> you WILL see a lot of time and effort get in your way.
>
> Save raw capture for times where you CAN'T control the lighting, or
> where
> the needs of the image require that absolute best image quality.
>


Jeff raises an interesting point but and a big but I would respectfully
submit that if one is truly concerned with quality that RAW is the only
option. Steve, I too mean no disrespect and in your case you may not
really care about improving quality and you may not have any need to go
back and work on an image from a previous shoot but I will bet you
could still improve your quality by shooting RAW. The RAW format is all
of the information captured from the image sensor without first
processing it. The RAW format records color and other information that
is applied during processing to enhance color accuracy and other
aspects of image quality. When it comes to digital RAW is the golden
rule. It is the only way to gain the full quality and control to
produce the best final image. Using a raw file converter like ACR
(Adobe Camera Raw) allows color space and exposure to be accurately
controlled. Since one has the original file with the raw file, it is
also always possible to go back and reprocess with different set of
standards. For example an original raw file could be processed as if it
were shot with tungsten lighting or with daylight lighting. This is
unlike a JPEG image where data is permanently changed or deleted during
processing in the camera and can never be recovered.

Steve, way back when I worked for the Syracuse Newspapers and we had a
black and white processor called a Kodak Versamat. We called it the
Vershitscratch. If you had a job that you really didn't care about the
Vershitscratch was wet to dry in about 5 minutes or less. The film was
ok and produced satisfactory prints. The highlights were typically
blown out and the shadows were blocked and the midtones lacked contrast
but the prints looked fine in the newspaper. That said, if we ever shot
anything that was something we really cared about we hand processed the
film. There was no comparison between the hand processing and the
machine processing. What looked ok from the machine looked awesome if
it was hand processed and wet printed. The argument for RAW is quite
similar.

Further, when I first went digital and was stuck using Alpha versions
of Canon software which barely worked I didn't save my raw files and
went directly to PSD. I could shoot myself today because with what I
know now, I would love to be able to go back and use a good RAW
converter like ACR. I know that images which were portfolio images even
then could be better today.

Best,

Seth


Andrew Rodney
Posted on Mon Aug 25, 2003 8:15 AM

on 8/25/03 6:00 AM, IR - Seth Resnick wrote:

> Jeff raises an interesting point but and a big but I would respectfully
> submit that if one is truly concerned with quality that RAW is the only
> option.

I have to side with Jeff (and I'm actually proud of the big guy for brining
up what I've felt for awhile). Quality is in the eye of the beholder. It's
not like JPEG is pure shit that a client will object to IF the alternative
is not getting the job done on time. For say an event/wedding/news
photographer who may never output the file to close to max size and does
have a huge time constraint, the difference in time between the two
processes may make or break the entire job. Yes it's possible a wedding
photographer will be asked to blow up one or more images to the point that
JPEG may degrade the quality a lot more than if the image was shot RAW. But
it's a crap shoot. If the possibility most of the time is the job simply
can't get done in time, JPEG is the only alternative the photographer has.

What the issue is here is that shooting RAW is NOT productive enough! When a
business has it's back to the wall to output their product, the key is
getting the product out the door. We need better/faster and cheaper tools to
allow anyone wanting to work with RAW to do so as quickly (or with a far
lower speed hit) than we have today.


Seth Resnick
Posted on Mon Aug 25, 2003 8:50 AM


On Monday, August 25, 2003, at 09:15 AM, IR - Andrew Rodney wrote:

> on 8/25/03 6:00 AM, IR - Seth Resnick wrote:
>
>> Jeff raises an interesting point but and a big but I would
>> respectfully
>> submit that if one is truly concerned with quality that RAW is the
>> only
>> option.
>
> I have to side with Jeff (and I'm actually proud of the big guy for
> brining
> up what I've felt for awhile). Quality is in the eye of the beholder.
> It's
> not like JPEG is pure shit that a client will object to IF the
> alternative
> is not getting the job done on time. For say an event/wedding/news
> photographer who may never output the file to close to max size and
> does
> have a huge time constraint, the difference in time between the two
> processes may make or break the entire job. Yes it's possible a wedding
> photographer will be asked to blow up one or more images to the point
> that
> JPEG may degrade the quality a lot more than if the image was shot
> RAW. But
> it's a crap shoot. If the possibility most of the time is the job
> simply
> can't get done in time, JPEG is the only alternative the photographer
> has.
>
> What the issue is here is that shooting RAW is NOT productive enough!
> When a
> business has it's back to the wall to output their product, the key is
> getting the product out the door. We need better/faster and cheaper
> tools to
> allow anyone wanting to work with RAW to do so as quickly (or with a
> far
> lower speed hit) than we have today.
>

Andrew,

As someone who regularly shoots for magazines on deadline with
sometimes only hours to get files in, I still only shoot in RAW. The
key is workflow. JPEG's still need metadata inserted and the files
still need to be edited for the client. With proper actions one can
still shoot RAW and edit the files very rapidly. The time for
processing RAW files if one has a good solid workflow is fairly
minimal. There are many times where the original client may have been
satisfied with a jpeg but after the fact the image is going to be used
for something else and the jpeg was not sufficient. Almost every file I
have in WorkbookStock is from a magazine assignment. Had I shot jpegs
none of them would have been accepted. It is not always just the
immediate job that matters but rather the future of an image. My
retirement account is in part my image collection and I want that to be
the best it can be. I do deliver images to some client as jpegs but
they are made as an action from the RAW. I admit that certain jobs like
a highschool portrait may be just fine for jpegs but for the most part
with a good workflow shooting RAW is very feasible.

Seth


Andrew Rodney
Posted on Mon Aug 25, 2003 9:10 AM

on 8/25/03 7:50 AM, IR - Seth Resnick wrote:

> As someone who regularly shoots for magazines on deadline with
> sometimes only hours to get files in, I still only shoot in RAW.

And that works for you based on your skills, the actions you've created and
more importantly the number of images you need to process. It's certainly
possible isn't it that someone might have to shoot a lot more images and
process them under a tighter deadline no? It's possible that with your
assistance, they could get that work done. It's possible with better tools
or a faster machine they can get it done. But if the reality is that TODAY
they have to get the job finished and they can't with RAW and can with JPEG,
I would suggest they get the job done. When they have the time to
investigate RAW and prove they can get the job done, fine. But I don't like
to tell people not to shoot one format verses the other without knowing the
specifics. Nothing need be written in stone. For a great deal of work,
you'll never see the difference. For some work you will.

4x5 will always produce better image quality than 35mm but we can't expect
every photographer to shoot 4x5 for every assignment. It's simply not
practical.

> JPEG's still need metadata inserted and the files
> still need to be edited for the client.

Some photographers don't need this nor do their clients. Yes, again I think
having proper Meta data in files is very important. But not in lieu of not
getting the job done and getting paid if that's a possibility.

> There are many times where the original client may have been
> satisfied with a jpeg but after the fact the image is going to be used
> for something else and the jpeg was not sufficient.

Yes and I said that's a crap shoot. If 98% of the time shooting RAW produces
a bottleneck that causes non payment due to non delivery of on time work and
of that, 10% might go to an output that might be insufficient, what choice
does the shooter have?

> Almost every file I
> have in WorkbookStock is from a magazine assignment. Had I shot jpegs
> none of them would have been accepted.

Because the client saw it was JPEG? That's easy to fix. And again, in your
situation, you know your clients have a certain quality standard that would
cause real issues if you didn't shoot RAW. That's not the case with all
photographers.

This JPEG thing kind of reminds me of the sRGB discussions where some people
would give you the idea that if your file was in sRGB, it was crap and would
look awful no matter where output. That's simply not the case. Yes I would
never recommend someone go out of their way to produce sRGB. And while the
gamut is going to clip colors to a lot of devices, when rubber meets the
road, the output is acceptable to a very large number of people. I think
that's the same with JPEG. I recommend people shoot RAW for all reasons we
know about. But I also know that some folks simply can't effectively. And I
think the solution should be more people asking for better and faster RAW
tools than telling people that if they don't shoot RAW until those solutions
become available, they are in dire straights. I really don't believe that
for ALL photographers.


Allen Pacheco
Posted on Mon Aug 25, 2003 9:21 AM

It has been mentioned several times in this thread that, to paraphrase, camera profiles are basically suited for only those conditions that match those at the time of the target capture. I assume the color temperature of the ambient or strobe lighting is considered to be the chief culprit. Correct?
Well, after doing my own tests using both 3400K photofloods and Calumet travelites in the studio (not in the same image) and also both bright sunlight and open shade outdoors, I'm convinced that with careful custom neutral balancing I can null out the impact of the color temperature issue. Sure, there may be issues with mixed lighting and perhaps some other extremes like event lighting but I am in agreement with the concept that one profile can be made to cover a lot of bases. And as long as you don't go blindly using the profile with every frame you click off you should be OK.

It does depend upon having a good neutral balance routine in whichever flavor of converter you use.


Bruce Fraser
Posted on Mon Aug 25, 2003 11:17 AM

Allen,

The issue isn't just the color temperature, but also the camera's response under the different color temperatures. SOme camera responses change in a more or less linear manner, and a single profile may work for them. Others change in a distinctly nonlinear manner (though I suspect that few current cameras are as bad as the DCS 460 on which I first noticed the issue), and with those, a single profile probably won't work all that well.


Ian Lyons
Posted on Mon Aug 25, 2003 12:43 PM

Allen,

> I'm told that if you want into profiling nirvana you should speak to Jack Bingham or get his ColoreEyes > 20/20. The top secret manual is supposed to be the answer to all our profiling problems. Me? I'll make > do with inCamera and Adobe Camera Raw

>Well, you could be right. I had no intention of mentioning J.B and ColorEyes. Although, I did read his >Thumbnail faceoff in PEI with Andrew. Good points by both parties.

I'm not familiar with the article. PEI doesn't trvel too well ;-) BTW: There was no intent to criticise ColorEyes I was simply making the point that the secret of it's success appars to be largely related to the methodology decribed in the manual


>The original poster (poor guy, >remember him?) was solely interested in using inCamera with ACR.
>BTW, I have corresponded with a user of 20/20 who indeed had a few issues with highlights, also. But >they were resolved with some editing.
>Which begs the question...is there a profile editor that can modify inCamera's LUT-based profiles with a >fine degree of control? The included editor does more damage than good.

ProfileMaker Pro probably does abetter job than most but it isn't going to fix the problem you describe (sadly).


Andrew Rodney
Posted on Mon Aug 25, 2003 1:15 PM

on 8/25/03 11:45 AM, IR - Ian Lyons wrote:

> I was simply making the point that the secret of it's success appars to be
> largely related to the methodology decribed in the manual

It's really all about shooting the target. A good deal is common sense and a
few pointers where to place your white point. Shooting the target in a copy
light environment (but using one light), keeping flare from the gloss
patches to nil and placing some black paper with a hole near the lens etc.
They feel (and to some degree I agree) that having super flat even light on
the target is super important. How anal you really need to be to insure all
four white corner patches are within a 1/10 stop is probably up to debate
but the more even, the better.


Herbert Gibson
Posted on Fri Aug 29, 2003 12:44 PM

Andrew Rodney wrote:

>It's really all about shooting the target.

Andrew,

Would you care to advise on how best the target should be illuminated?

Flat illumination should result from a small source close to the lens axis or from bouncing a head backwards inside a white painted box/room. Any preference?

Is a single source really necessary. To boost the levels in a large white room, could you fire a couple of heads backwards towards the same spot behind the target? Surely the light bouncing around the room to the target would be an even mixture?

Some weeks ago you were seeking shots of targets for Pictographics assessment. Did you draw any conclusions from that exercise?

The new Lasersoft converter has profiling built in. Doesn't that make it a strong contender?


Andrew Rodney
Posted on Fri Aug 29, 2003 1:05 PM

> Would you care to advise on how best the target should be illuminated?

Real flat with no light bouncing back into the lens and no glare on the gloss patches.

> Flat illumination should result from a small source close to the lens axis or from bouncing a head backwards inside a white painted box/room. Any preference?

I've always done copy work like this with two lights but I think the issue may be different light temps per light and hence it's might be easier to white balance precisely using one. I don't know that two would hose the issue. You want as even as humanly possible on all four corners and center.

> Some weeks ago you were seeking shots of targets for Pictographics assessment. Did you draw any conclusions from that exercise?

I'm still testing some software which I can't speak about (NDA). I could use more files to test but I'm using a special target now which makes it even harder to get results.


Herbert Gibson
Posted on Sat Aug 30, 2003 7:55 AM

"> Andrew, A while back you mentioned on one of the forums about a larger sized
> card of just the #20 - Neutral 8 patch color, or something similar. Is there
> such a beast?

Yes but no one outside of GretagMacbeth and a few testes have one but it's
coming... "


What is this beast? Is it a 'light' grey card?

Does this tie in with Jeff's suggestion that it is important to 'white balance' rather than 'grey balance'?

What is the reasoning behind Jeff's suggestion?

Is this a quirk of Camera Raw or something fundamental?

Presumably, since we don't have perfect sensors, it is preferable to shoot a light grey card rather than overexpose a mid grey card?

What Photoshop RGB values should these grey tones present in a perfect exposure? Bruce Lindbloom measurements?




Andrew Rodney
Posted on Sat Aug 30, 2003 8:50 AM

on 8/30/03 7:00 AM, IR - Herbert Gibson wrote:

> What is this beast? Is it a 'light' grey card?

No but that is almost done and will be out around PhotoExpo in November (the
Gray/White/Black card). And it will help with the suggestions Jeff has made


Bruce Fraser
Posted on Sat Aug 30, 2003 12:44 PM

>What Photoshop RGB values should these grey tones present in a perfect >exposure?

That would depend on

a) the scene

and

b) your chosen working space.

You may have scenes that contain something much brighter than the target white patch and something much darker than the target black patch. You just want the gray tones to be neutral�it's counterproductive to aim for a specific tonal rendition without taking the rest of the scene into account.


Jeff Schewe
Posted on Sat Aug 30, 2003 1:00 PM

Herbert Gibson wrote:

> What is the reasoning behind Jeff's suggestion?
>
> Is this a quirk of Camera Raw or something fundamental?

It all depends on what sort of algorithm is being used for adjusting the
white balance. I don't think it's a Camera Raw quirk, but on the other hand,
not all raw converters actually do a white balance in software.

The reasons for using a 3/4 tone vs. mid-grey is that the calculations are
working primarily with the white point. If you set the target too far down
the curve (say at a middle grey) you may be over-correcting the white
balance. Grey balance is really a function of gamma adjustments on the red,
green & blue channels. This works fine on gamma encoded images. . .but since
the raw captures are still in gamma 1 (linear) the post adjusted mid point
is way far towards the deep end.

I can't tell you what exact tone settings work best since any RGB values
would be post converted readings. . .but Thomas has indicated that
non-specular textural white is fine.


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




Allen Pacheco
Posted on Sat Aug 30, 2003 3:26 PM


Jeff, not to be picky but I think a lot of folks are used to 3/4 tones meaning the area between the midtones and shadows. Quarter tones are uaually between highlights and midtones.

(Maybe it's a RIT thing?)


Steve Bohne
Posted on Sun Aug 31, 2003 1:34 PM

Andrew Rodney wrote:

>This JPEG thing kind of reminds me of the sRGB discussions where some people
>would give you the idea that if your file was in sRGB, it was crap and would
>look awful no matter where output. That's simply not the case. Yes I would
>never recommend someone go out of their way to produce sRGB. And while the
>gamut is going to clip colors to a lot of devices, when rubber meets the
>road, the output is acceptable to a very large number of people.

Gee, I'm one of those. Shoot in JPEG, output in sRGB. My color lab wants sRGB, so I acquiese. Prints look good, though. At least people trade me their money for my pictures (or is it my money for their pictures?).

Anyway, it works.


Andrew Rodney
Posted on Sun Aug 31, 2003 1:42 PM

Problem becomes when you use or need a lab or output device who's gamut far exceeds sRGB (and there are quite a lot). In that case, the wider gamut is long gone and you can't get it back.

Having a file in a wider gamut (say Adobe RGB) means you can convert to sRGB (better an output space) AND anything that needs a wider space. It's all about flexibility and options. Using just sRGB can certainly work but you can paint yourself into a corner too.


Jeff Schewe
Posted on Sun Aug 31, 2003 2:10 PM

Andrew Rodney wrote:

> Having a file in a wider gamut (say Adobe RGB) means you can convert to sRGB
> (better an output space) AND anything that needs a wider space. It's all about
> flexibility and options. Using just sRGB can certainly work but you can paint
> yourself into a corner too.

The other advantage of bringing images in as Adobe RGB and using Adobe RGB
as your working space instead of sRGB is that then, sRGB is just your
"output space". In addition to the ability to repurpose, you also get one
additional big carrot-you can softproof your sRGB output and determine how
best to correct for the reduced gamut.

If all you had to start with is sRGB, your ability to softproof and correct
for sRGB is eliminated.

This can be a big deal and improve on the canned sRGB formula of digital
cameras.


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





Ken Frazer
Posted on Wed Oct 15, 2003 9:17 PM

My question is about camera profiling vs. ACR. ACR is using two profiles and allowing you to balance between them, but doesn't every camera chip have it's own "fingerprint"? I have not used ACR extensively and it looks like a great tool on the Canon line of cameras, but my Kodak 660 files are very difficult to manage with it in my experience. I have used Jack B's 20/20 for about a year now and it has been a great timesaver with my workflow which almost all revolves around a more or less daylight/ cool light/ flash color temp. I daresay that it would fall flat with straight tungsten. I have found with my carefully made profile, and a careful neutralizing custom white balance, color is far more accurate than I can achieve any other way. Am I out to lunch here? I think ACR is a great product and worth more than it sells for, but custom allows me a better more accurate and smoother transition of tonal values, especially in skin tones. Any thoughts would be appreciated.


Jeff Schewe
Posted on Wed Oct 15, 2003 9:35 PM

Ken Frazer wrote:

> My question is about camera profiling vs. ACR. ACR is using two profiles and
> allowing you to balance between them, but doesn't every camera chip have it's
> own "fingerprint"?

Ask the question again after you try Camera Raw II in Photoshop CS. CR I had
a shortcoming if the camera you used varied greatly from samples tested. CR
II has a calibrate function that allows for finetuning the color rendering
and correcting for hue torque where some colors came out off.

Add chromatic aberration correction and vignetting correction and the
greatly enhanced File Browser and the ability to batch apply CR II settings
to multiple files, then the CR II story becomes more compelling. . .

Of course, you still have to have one thing. . .patience since Photoshop CS
& CR II won't ship till sometime in Nov.

:-)



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





Bruce Fraser
Posted on Wed Oct 15, 2003 9:44 PM

Ken, I don't think there's a single answer to the question other than that if it works for you, it works.

I do suspect that Canon's chips are a great deal more consistent than Kodak's�Canon is basically a chip manufacturer, while Kodak is flailing around looking for a business to be in. The Canon's I've looked at seem remarkably consistent from unit to unit.

I'd say that if each chip in a particular camera line has its own fingerprint, the vendor needs to improve their process control and QA. The lens should be a far bigger source of variability than the chip.

20/20 comes highly regarded. It's the only custom camera profiling tool I'd recommend. But if the camera doesn't respond linearly to changes in color temperature, no single profile will handle all lighting conditions. Since that's not a problem for you, use it in good health with a clear conscience.

But do check out ACR 2 when it ships. I think it solves a lot of problems.


Ken Frazer
Posted on Wed Oct 15, 2003 9:56 PM

Thanks all of you for the excellent dialogue! I have learned more in the last 24 hours on this professional forum than in all the other time I have spent in the others. I greatly appreciate the opportunity to be in touch with such an incredible group of experts!!!





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