This content has been marked as final. Show 60 replies
I empathize. I own a couple of Nikon DSLRs-- but in my line of business I work with raw brands from all kinds of camera makers. Like you, like many others, I therefore prefer ACR for my workflow-- but accurate colo(u)r processing can be as elusive as riding atop a bucking bull. You can only stay on the saddle as long as you accurately predict, within narrow tolerances, the next wild swing.
From my humble experiences, calibration works best for daylight and tungsten-ish. I assume, and the more technically informed members are free to correct me, that's because those are the two fixed targets third party solutions will profile for. Most any deviation to the known profiles therefore becomes a moving target. Fortunately, in my case, most of the photos I deal with are more or less true to daylight. However for mixed lighting, the camera makers packages have shown to demonstrate their own advantages. I'll frequently disagree with some arguments Mr. Borg puts forth in the link you provided--but in this case I believe his reservations are accurate.
From what I have learned, here's why: Adobe and most third party alternatives don't have access to all the colo(u)r information your camera produces. Most, if not all, camera makers won't publish or document that data. Therefore Adobe, and other raw processors, must take the time to generate new profiles for every camera once it's released. Those new profiles, won't be anywhere near as detailed as the camera makers have avaialble to them. I believe this is the case because:
1. There isn't enough time to generate more than a couple of profiles (probably daylight and tungsten) before the market HOWwwwwels for their favorite new camera to be supported.
2. The ISO sensitivity of each sensor is not an absolute. The camera makers know much more about camera to camera deviation than the entire networked raw processing community.
3. There could be other proprietary channel data that only the camera makers are privy to.
So what does this mean? My interpretation of this is that if you use the Fors script to recalibrate ACR for a specific camera's deviation, you will have the best results when shooting at temperatures the camera was only initially profiled at. Anything else is a moving target and requires further futzing. Nikon Capture can lookup nikon's full library of profiles, and therefore provide you with more accurate results the first time around. This can be advantageous if you're shooting under deviant lighting and need bulk processing. Bruce, or others, are free to weigh in and dispute any of these interpretations if I'm not correctly describing them.
So what can be done? If you shoot under deviant conditions, you can either futz in ACR or grow your beard waiting for NC to process your images. Freshly enlightened, you could also join the growing movement to ask camera makers to document their formats so that everyone has access to the same colo(u)r information. In DPreview you'll read that the Open RAW foundation and others are active proponents of this solution. As an alternate, but I believe ultimately more effective solution, is one that you'll find myself, as well as some other forum members putting forth. That would be insisting camera makers --such as Nikon--first support the *option* of saving your raw files to DNG over NEF.
Why DNG? DNG places the burden of providing the correct spectral response with the file's maker. Therefore, Nikon, as the camera maker, would need to embed the correct, or at least a sanctioned spectral response into any in-camera DNG --thus making it self contained and independent of third party profiles. It's my assumption, but I'm not yet clear if this is absolute, that Nikon could do this *without* the need to openly document what they are doing. Furthermore, I believe that while camera makers will need their arms twisted by their customers to offer either:
A. full and open documentation or
B. DNG support
it will be far easier to leverage market support for a universal raw container--such as DNG before camera makers succomb to openly publishing the ingredients to their secret sauce.
In conclusion, while your title suggests NEF+ACR is hopeless, it's my belief that DNG+ACR could be very hopeful to your cause. :-)
Thanks for your thoughful (and detailed) reply. A few quick points:
* I'm a subscriber to Juergen Sprecht's list, and have signed on to and agree with the Open Raw petition, but it seems that it's momentum (and impact?) is slowing
* Julia Borg is one of the rare female regulars on the D1/D2 forum
* Given that tweaking color in ACR is probably unavoidable, maybe there needs to be a simpler way to do it. I always feel as if I push on one problem only to have another pop up somewhere else - there's almost no end to it. In NC it's a couple of quick slider movements and you're done.
* I agree that Nikon should be more open about the NEF format. I't seems their reticence is self-defeating. After all, how many Nikon shooters chose the brand for color rendition? Or image quality in general? I think their advantage is in the camera-as-tool factors, but that's for another thread.
Any thoughts on Capture NX?
>Julia Borg is one of the rare female regulars on the D1/D2 forum
it's OT--but so he would like you to believe. there are reasons which i don't need to go into and are really irrelevant. :-)
>After all, how many Nikon shooters chose the brand for color rendition? Or image quality in general?
to the victor goes the spoils. nikon, as well as canon, have legitimate commercial reasons why they would like their formats to remain closed. i believe DNG could offer a path out from their dilemna. i suspect it will not be OPEN RAW petitions that sway the manufacturers as much as industry compliance to an open standard container. the demand for that container lies not with photographers, but with cost-efficiences inherent to archiving, printing, and publishing . eventually that part of the ecosystem will put pressure on photographers to supply them in some pre-defined template.
>Any thoughts on Capture NX?
a few. last week i had a chance to work with a beta copy. as such it was still prone to crashing, and performance wasn't finalized--but it did appear to offer some significant improvements. the biggest one is speed. NIK replaced many of the older ARCSOFT routines and have made the entire package run much more efficiently--but don't look for ACR efficiency. also, the NIK interface is very sweet. they use some of the same colo(u)r selection and feathering routines found in Nik Sharpener Pro. selecting an area to work on and interfacing with is much more precise than anything currently offered by ACR. some other observations:
NX isn't a substitute for PS. there are no layering tools. the tools available are limited, and the beta didn't support third party plug-ins--including NIKs. i'm told that won't change in the final version.
NX introduces yet another non-standard NEF container. NX adds a set of edit decision lists to not only RAW files, but to JPEGs and TIFs which it then wraps into a NEF. there is enough confusion about NEF scan files and NEF raw files. now we will have NEF files containing JPEG and TIF data with appended edit lists that no one other than nikon can support or open. you can imagine how confusing this will be when third party programs that support some NEF files (mostly raw processors) cannot open or process these other NEF files. it'll be another exception that needs to be added to the ACR FAQs.
although i don't support nikon's NEF strategies, i will probably still purchase NX as a tool for my digital darkroom. the final price, was not set as of last week and was still being determined by Nikon. it will probably cost more than NC because of the additional licensing structures with NIK.
hope that was helpful. i'm gunning for the curvemeister family seal of approval ... :-)
Greg, et al,
I have been following your DPR forum and this one. I am very frustrated too. It does almost seem that converting to Canon or film (no more Velvia 50??!!) might be the only alternatives to the scenario outlined herein by nunatak.
Bruce Fraser says that ACR can duplicate anything that Capture can do. I'm just afraid that you may have to actually be Bruce Fraser - or Thomas Knoll - to do that.
At some point Nikon is going to have to cater to mere mortals.
Thanks again, Nunatak. You're quite the fountain of knowledge!
Interesting about Mr/s Borg. My curiosity is officially piqued.
Interesting also about Capture NX. It sounds like editing with NX is a dance with devil. But if it get's me from point A to point B substantially faster than today then I'm cool with it (so to speak).
God forbid - I'm having evil film thoughts myself. I wonder how many Hail Marys that is... I'll certainly wait for Capture NX before doing anything drastic.
I'm curious what Mr Knoll's and/or Mr Fraser's take on all this would be?
Ya know. . .it really just shouldn't be that hard bud.
I watched the thread on the DPReview site with amusement because for whatever reason it seems Nikon shooters seem unable to get Camera Raw colors to match Nikon Capture. I have no problem adjusting Canon raws to match DPP (although there seems to be plenty of Canon shooters who can't).
Doing the auto calibration script is a useful time saver but unless you have half a clue what it's doing, it's just a black box result. I really DO think you SHOULD actually go through the manual calibration routine to learn what the controls in Calibrate are doing so you can have a degree of control over the process instead of flailing about as you have done. Don't get me wrong, it's an impressive amount of flailing...but as you've indicated, you seem at your wits end.
The main difference between the camera company's software and ANY 3rd party software (that is NOT using the camera SDK for processing) is that the 3rd party software is dealing with a wide variety of camera formats and models-none of which are really fine-tuned on a camera by camera basis. That's why Capture One relies so heavily on camera profiles and why Camera Raw has the calibrate function-to allow the user to adjust the rendering of the resulting process.
The camera company's software is designed to replicate the in-camera JPG rendering...and that seems to be what many digital photographers seem to fixate on and seem to end up trying to replicate in the 3rd party software. Well, it's pretty easy for Nikon and Canon to duplicate THIER in camera jpgs in their SDK and software. Others who don't use the SDK will NEVER match the camera jpgs by default.
So, that means users of Camera Raw or other 3rd party raw processors will need to learn how to use the color controls of the processor to adjust the rendering of the resulting processed raws. It's entirely possible, if you know how to use the controls, to infinately adjust the renderings even to the point of matching the in-camera jpgs...but it does require a bit of learning and experience-which you really don't get if you are trying to rely upon an automatic calibration routine.
I guess ACR is just beyond me. I suppose that given enough time and effort I could tame it, but I'm afraid I don't have the inclination to make that kind of investment with little guarantee of success. Show me one high-volume Nikon shooter who's using it successfully and I might be persuaded to try.
If a person with my background (30 years of heavy computer use, 8 years of Photoshop, close to a hundred thousand digital images captured) can invest the time I have into learning ACR only to still be "flailing" (however impressively) then I wonder what chance the average Nikon user has.
The last Nikon camera I actually shot was an 8008s, so my experience has thus far been limited to problem NEFs sent to me by others. Absent irreperable exposure errors, I've generally managed to replicate the appearance of NC in ACR. Take that for whatever it's worth to you.
But leaving that aside, there seems to be some confusion about why the Calibrate controls are there and what they actually do.
When Thomas Knoll creates support for a new camera, he generally only gets one unit. He profiles that unit, and builds Illuminant A and D65 profiles for that camera into Camera Raw. The built-in profiles in Camera Raw are all camera model-specific. But they're generally based on a single unit. If that unit was atypical, Camera Raw's built-in support for that camera model may be less than ideal.
The profiles are simple matrix profiles with a white point, primary chromaticities, and a tone reproduction curve. ACR's white balance controls interpolate between the profiles.
The Calibrate controls adjust the primary chromaticities. That's ALL they do. The reason they're there is to account for one specific type of unit-to-unit variation, which is small differences in the thickness of the color filters.
How common, and how great are these differences? Nobody knows except the camera vendors, who aren't providing any data.
Jeff Schewe and I did a comparison of 3 Canon 300Ds a couple weeks ago. After wading through a very large data set, we found we were absolutely certain that the three chips had significantly different sensitivities. After compensating as best we can for those, we're pretty sure that there are also smaller variances in the color primaries, that Tom Fors' script irons out pretty well.
Moving the primaries to account for lens differences seems like a bad idea. They should mostly be handled by the white balance controls. Lenses certainly make a difference to overall color, contrast, and even saturation, but they don't affect the chromaticities of the filters in the color filter array, and that's what you're adjusting when you adjust the Calibrate settings.
Some people are snotty about wedding photographers. Personally, I think what they do is heroic. (Shooting a woman in white lace next to a guy wearing black under noon sunlight is about as hard as it gets.) One problem I've seen is that our eyes see a woman wearing white getting into a white car, while the camera sees a woman wearing pale blue getting into a pale yellow one when the image is adjusted to produce reasonable skin tones. That's observer metamerism in actionthe color filters in the camera are fairly unlike the cone responses in the human eye, and I don't have a solution for it.
That said, a lot of problems can be avoided by determining the real (as opposed to nominal) ISO sensitivity of the camera, so that the internal meter becomes somewhat reliable as opposed to flat-out lying.
Note that if the camera vendors adopted DNG, one of the things they could do is to embed unit-specific color matrices, thereby making much of the problem go away.
I'm curious about Mr. Knoll's take on this too, but since most of what I know about it was learned from him, my hope is that it would be somewhat like mine....
well as long as we're confessing our curiousities, I wouldn't mind confirming if nikon were to save their RAW data as DNGs, then:
1. would the colo(u)rs more accurately reflect what nikon, photographers predict?
2. and would nikon be able to protect this response in the private maker tags?
Greg and Jeff have very different workflows and if ACR, as a solution, is to work there needs to be accomodation.
Thank you for the post. I'm impressed by the seriousness and depth of the replies I'm getting here, and it's especially gratifying to hear from the guy who literally wrote the book on the subject, a book I clearly need to read more of.
A few points:
* I badly misunderstood the purpose of calibration. So have a lot of other people, apparently.
* I've been meaning to determine the real ISO of my camera bodies; that would be time very well spent.
* I hear what you're saying about our eyes reacting differently than a camera. But I'm not asking ACR to do the impossible or even the somewhat unreasonable. At least I don't think I am. All I really want to do is to be able to get an NC-quality jpeg in a reasonable amount of time.
You know what would be cool? If you (or someone else who knows what they're doing) could take a few nice NC-produced jpegs and their corresponding unedited NEFs and show the process of working the NEFs in ACR with a Russell Brown style video. Now that's something I'd pay to see.
-->1. would the colo(u)rs more accurately reflect what nikon, photographers predict?
-->2. and would nikon be able to protect this response in the private maker tags?
No. Well, yes, they could populate the DNG matrices with some bogus information and then use private makernotes to decode it. But I'd say that if they did that, they weren't supporting DNG....
-->You know what would be cool?
I'd really rather leave that to a Nikon shooter. I already went through the agony of swwitching from Nikon to Canon. I couldn't face doing it again!
>All I really want to do is to be able to get an NC-quality jpeg in a reasonable amount of time.
if that's the case Greg, all you need to do is shoot NEF+JPEG and you're set. the real benefit of raw is for custom processing--outside and beyond "nikon's default look".
>But I'd say that if they did that, they weren't supporting DNG....
eye don't anticipate nikon to support DNG willingly. more likely they'll be dragged screaming and kicking into it only if industry standard templates require photographers to supply their raw files as DNG. yet for nikon, this option has to be better than fully documenting and distributing the full cartography for NEF.
imo, it's important to offer a practical alternative to the open raw initiative and eutopia. :-)
>All I really want to do is to be able to get an NC-quality jpeg in a reasonable amount of time.
>if that's the case Greg, all you need to do is shoot NEF+JPEG and you're set. the real benefit of raw is for custom processing--outside and beyond "nikon's default look".
I'd actually use few of the camera-generated jpegs. Maybe I'm a bit obsessive, but I tweak the majority of my keepers. Even aside from those images that are over- or under-exposed or have color balance issues, sometimes (frequently, actually) I just want to change some aspect of the image (such as the contrast or what Nikon refers to as the color mode) in a non-destructive way. And shooting NEF+JPEG would mean almost double the card space (and card changes).
I need a jpeg as a final product because that's what ends up in my online galleries and it's what I send to my lab for printing. I also use jpegs as input to PS for those images that need more work but will not be extensively retouched (16-bit tiffs otherwise).
Greg, Jeff, Bruce, nunatak, et al,
Many thanks for the replies that help us mortals try to understand this subject. Bruce, I'd like to second Greg's motion to have you do an instructional video on ACR calibration for Nikon folks. Would it help if you could consider it as loyalty to your enthusiastic readers? I'm sure that you and Jeff probably get chuckles from how slow some of us are, but we nevertheless appreciate your teaching efforts.
John Shaw taught me a valuable lesson about wedding photography. Say no. Thanks, John. And salutes to Greg who does the white/black/emotion laden task for a living.
>John Shaw taught me a valuable lesson about wedding photography. Say no.
yup. he also gave me invaluable advice on how to best photograph birds--take lot's of photographs. :-)
The Nikon guru Thom Hogan made some interesting observations vis a vis Nikon, Canon, and professional photographers. Nikon continues to lose market share among pros because it does not give them what they want while Canon does:
Certainly, pros do not want to be locked into a proprietary raw format that limits their processing options now and in the future. While Canon has not yet embraced open RAW, at least they haven't encrypted their WB.If I were in charge at Nikon, I certainly would do everything possible to meet the needs of pros still using Nikon, and embracing open raw would be a good start.
BTW, have you tried Bibble Pro? They have a PS plugin and were among the first to circumvent the WB encryptation issue. I don't know if they have better WB than ACR, but it might be worth a look.
I think Greg Zillgitt is absolutely on target. Many of us are simply up to here with endless "CSI" exercises in camera or on the computer.
"a woman wearing white getting into a white car, while the camera sees a woman wearing pale blue getting into a pale yellow one when the image is adjusted to produce reasonable skin tones"
Bruce seems to be right about this, but does it help? Jeff & Bruce got dissimilar results from several same model cameras, not what I wanted to hear either.
One well-known dSLR user recommends setting separate color calibrations for each lighting situation before shooting. I wouldn't eagerly anticipate doing this every five minutes during a six or ten hour wedding. Or any other assignment. What do you do?
How many dSLR's allow you set individual color calibrations as conveniently as f/stops or shutter speeds and at what cost? A dSLR with a function knob with mucho color calibration clicks on it could be the next big thing. -Ed
Correct to read:
AN AFFORDABLE dSLR with a function knob with mucho color calibration clicks on it could be the next big thing. -Ed
I think there is a LOT of disconnect between the concept of doing a White Balance and doing a calibration of a camera. There is absolutely no reason to do separate "calibrations" for anything other than "Daylight" and "Tungsten". Aside from the metamerism of the chip under substantially different spectral light sources, daylight, cloudy and other "above" tungsten light will pretty much calibrate similar if not the same.
Calibrate adjusts the 3 color primaries of a sensor's responce to more accurately capture color hue & sat...this has nothing to do with white balance-except for the period of time that you need to white balance in order to calibrate.
I kinda started this whole calibrate thingie when I found Bruce Lindbloom's synthetic ColorChecker file in Lab. Under guidance from Bruce, we adapted the visual approach I was using to a more accurate measured approach. Once Bruce came up withe the "numbers" Tom Fors followed with an automated script.
But, the bottom line is to get the color rendering the way you want.It doesn't nessicarily follow that it has to be technically absolutely accurate, just visually accurate so the the colors red, green and blue (and various mixes) are rendered the way you want.
But...then you gotta learn how to adjust white balance.
Back in the old days of film, I almost NEVER shot chrome without some sort of filtration-either CC or LB filters, to control the balance of the film and processing. If the scene was outdoors and was anything other than warm sunlight, I would almost ALWAYS shoot with a Wratten 81EF warming filter. And let me tell you, that's warm-but does wonders for overcast or cloudy light. I've even gone to the extent of taking spray paint to the insides of my strobe light boxes to adjust the color temp of my lighting.
I will also say that one factor that is often left out of the picture (so to speak) the the major impact that exposure has over color rendering. Back when I shot 8 x 10 film, I would test film emulsions and order film by the case. It was not at all unusual for me to do 1/6 of a stop brackets on 8 x 10. Why? Because I could see a difference (that and I charged $50/sheet so it was worth it to me).
These days with auto everthing, I think a lot of digital photographers have really lost the ability to judge and measure light. They just shoot it and chimp the LCD on the back of the camera to see what they just got. That's NOT the same as measuring highlights, shadows and determining the contrast ratio and manually selecting the optimum exposure.
The problem with digital is there is no "right" rendering...the power of digital is there is no "right" rendering...you can make images look anyway you want. And, I love that about digital. To0 many digital photographers seem to want digital to look like film or even worse, want their digital to look like their camera's jpgs. Making a raw capture look exactly like some other rendering isn't easy...I can do it because I'm pretty good at Camera Raw's controls. Most people don't seem to be.
I'm also big on automating as much of everything I can to make my life easier...I often alter or change the "Camera Defults" settings in Camera Raw. . .I'm always making saved settings for specific processing parameters so I can easily apply them to masses of images. And, as much as I like Camera Raw and Photoshop, I do want a life not staring at a computer display (sometimes I want to stare at prints-don't ya know).
I see a whole bunch of hand wringing and flailing about by a lot of photographers-particularly if they have gone digital only recently. It's to be expected...digital is new and a completely different medium than film. It behaves differently, has different strenths and weaknesses and basically needs to be thought of differently than film-whether neg or chrome (which is ALSO radically different).
Everybody is looking for a magic bullet...the magic bullet is really learning how to control the medium to get what you want-whatever it takes!
Hey Jeff -
I very much regret coming to this forum for help. It's really not about helping people, apparently. It seems to be more of a mutual admiration society for condescending snobs.
Thanks for nothing. You can get off your high horse now. Or not.
Today's magic bullet is this:
Shooting Menu > Hue Adjustment > +9
Use this one setting under all lighting conditions.
If it works, great, if not, let's do this:
Tomorrow's magic bullet is a Sony DSC-R1 with a Real Time sensor, a built-in eyedropper with which to set unlimited in-camera WB settings. It would include a Color Temp knob on camera & Tint knob on the camera if you want to eyeball a change without the eyedropper option. Otherwise, select a neutral area, apply eyedropper, Save As, Enter, shoot the picture.
In film technology are big excesses of UV or florescent not attenuated by the spectral fabric of the film itself? Is it accurate to state that one RGB color can just rip away from the rest of the spectral fabric?
I have no idea what options are available for all dSLR users to calibrate their sensors; meantime, you noodle WB, Sat, Hue, etc., and then de-code and de-befuddle a series of images which look different as thumbnails, as Browser images, and as .psd's. Several gigs at a time. Or you might just get smart and upscale to black and white to your customers. - Ed
C'mon Greg, that was uncalled for. I appreciate your diligence in pursuing this issue, frankly color management sems like an ocean of ways to spend your money and time, and its an unmatured technology.
But Jeff didn't deserve that broadside hit. I might mention that I read many of your posts on dpreview.com concerning your calibrations, others admire your posts.
I am also a professional photographer. I might mention that my wife has adamately (sp?) cordoned off (for now) any use of digital for wedding photography, as she is a superb color darkroom printer. We are all digital except for weddings btw, so I know where you are coming from with much gigs of who knows what in ACR.
Take a nap we'll talk later ok?
I'd like to commend your efforts, and at the same time say that this forum is what it is.
By coming here, you bow before the princes of light on bended knee. If fortune smiles on you, you receive a pittance of largess along with the debris and vegetables that are tossed in your general direction.
>> By coming here, you bow before the princes of light on bended knee. <<
Could it be that the (excellent) architecture of Camera Raw, the sense & simplicity of ACR calibration and the basic photo math was NOT communicated carefully enough TO the princes of light BY the Adobe engineering team?
But who am I, just a hobbyist,
Several decades ago film color management was already old news.
Jeeesh Greg...I was responding to Ed's question about multiple calabrations for different lighting...I wasn't even writting to you dooode. But, if the shoe fits....
Jeff I read and re-read your original post, the need to differentiate sensor calibration and white balance got me to thinking. The D100's Sony sensor reminds me of a then new generation of Fuji/Kodak film several years ago: dichroic heads set at old settings suddenly began to output color prints with very magenta skin tones.
My non-calibrated D100 output similar skin tones. With this in mind, I just stood at the front door shooting outside then inside, with and without flash, surfing WB and combining it with Hue until I got a useable set and forget combo. The combo was darned handy regardless of light source or combination with the D100's built in flash. The deeper shadow .NEF's were a tad yellow but overall not bad. In particular I noticed a more realistic green in less illuminated areas. In ACR the needed tweaks require nowhere near the demands I dealt with before. - Ed
Despite my aptitude for just about anything "...for Dummies", I sincerely appreciate these (and this ACR) forums. And depsite the fact that Jeff, Bruce, Thomas, et al, are intimidatingly brilliant with respect to the physics - and metaphysics - of pixels, I find that these forums add a fabulous learning dimension without which I would stumble much more severely. Yes, digital stuff is frustrating...and sometimes overwhelming. But this forum is helping me eat the digital elephant one bite at the time.
Greg, you started a very cool thread in DPReview. Then continued the subject here. Thanks to all who have contributed. Unless, of course, this is just a conspiracy to get guys to spend big bucks on those ColorChecker cards.
There is no convenient way to just pop a gray card in front of your camera and calibrate out of spec camera sensors. Digital photography also is an almost fifty year old technology.
I am challenged by digital and it is a highly creative way of thinking, but you have to tiptoe around its logical inequities at this point. For example does anyone remember when color negative film whipped the UV/florescent problem years ago? Fast forward to the 21st century: why do dSLR's include UV and florescent icons as WB options?
Its the output that matters, like Jeff Schewe said, you want to look at prints. Until you get to that point you're tied down to a monitor, and that's not where your picture is.
>Its the output that matters, like Jeff Schewe said, you want to look at prints.
Of course output matters. Some are happy with standard lab style prints, while others want custom printing. What I believe Jeff is advocating, and ACR is particularly suitable for, is the latter.
Many others are just looking for a reasonable amount of consistancy, continuity, and predictability from their output. They want that in a way the integrates reasonably well into their workflow. Calibration is one way to mitigate the extremes of ACR, but it's certainly not perfect. My understanding is it can't match the camera maker's response over every scenario unless more openess is forthcoming from the camera makers.
Which brings me to openRAW. I fully embrace openRAW as an initiative, but I don't see them as offering any real leverage for change. Right now it's more of a wish list held up by in-fighting -- kind of like how peace talks during the vietnam war were held up while participants fought over the shape of the table.
So one can either live with the differences in output, fight them, or learn to compensate with the tools we have at our disposal. If at this moment I had to pick and choose just one battle, it would be for Nikon to offer me the *option* to save in-camera as a DNG. If that were to happen right out of the box, Nikon images would output with greater consistancy, continuity, and predictability--and many of us could instead turn our attention back to the art rather than the technology. Through osmosis, such a change would rapidly affect all camera makers and software developers-- in the process making life easier for photographers and consumers of photography.
Until consumers demand consistant solutions from their primary parties, they will continue to be distracted by a mixed array of arguments and solutions -- and pay dearly for that privilege.
Its not just the output from digital cameras, what about the input into digital cameras? Digital is film without UV and florescent shock absorbers.
The sensor and the light source are the issue. The sensor cannot be calibrated and the lightsource has all kind of band-aids after the fact such as WB and Hue and Saturation. Its not a Real Time experience free of chimping which is the antithesis of taking the picture.
Maybe a color meter option can be combined with the spot meter in camera? I predict Olympus and perhaps Sony/Minolta will address this issue. The methods we now use for color management are just overly engineered nonsense and absurd.
Digital has a very sensitive take on colors at capture, perhaps too sensitive. So upstream strategies to me are the answer, not another set of PS actions or workarounds downstream.
>Its not just the output from digital cameras, what about the input into digital cameras?
that's a good question. films would, for their part, mostly react in a predictable manner. the responsibility for standardizing input characteristics fell to the manufacturer of film-- not the camera.
>The methods we now use for color management are just overly engineered nonsense and absurd.
welcome to digital. it's a process, not a standard. we all welcome predictable controls so that every image doesn't become another science experiment. however, the precision we enjoyed with film has been traded for proprietary colo(u)r solutions. unfortunately this means colo(u)r may vary per manufacturer, camera model, and sometimes camera to camera. we've traded predictability for convenience--leading to the possibility that each camera can be it's own film varient.
as camera makers know their input, conditioning methodologies, and varience, they can at the very least take responsibility for globally predicting their colo(u)r output and not leaving it to third parties to interpret. that's what can be done now. imho, it'll be left to some future evolution, if ever, before input and conditioning characteristics are normalized to a global colo(u)r targeting scheme. i hope it happens sooner.
"we all welcome predictable controls so that every image doesn't become another science experiment"
Well, like Groucho said, I wish I'd said that earlier, everybody seems to be repeating it around the club.
For retouching I'll stick with PS, for color correction I am looking elsewhere, especially in May 2006.
If you're referring to NX-- it still ain't no panacea of predictability. it's just another tool. :-)
I'm not trying to be combative, but I've been pondering your post for 24 hours in an attempt to extract meaning, and other than the fact that you seem to have some kind of color problem, I'm failing.
-->Digital is film without UV and florescent shock absorbers.
Digital is pretty much impervious to UV, it has a hard enough time simply capturing blue. It does require IR filtration, but most current cameras have that built in. I have no idea what a florescent shock absorber is, but there are many different types of fluorecent light sources, and film never handled any of them particularly well, though it did handle some of them spectacularly badly.
-->The sensor and the light source are the issue.
Yes, this is photography....
-->The sensor cannot be calibrated
As has always been the case in photography. However, the sensor can be characterized somewhat more easily than could varying lots of the same emulsion.
-->and the lightsource has all kind of band-aids after the fact such as WB and Hue and Saturation.
Those are not bandaids one can apply to a light source. They are image editing routines that you can either have the camera apply automatically by shooting JPEG, or can take control of yourself by shooting raw.
-->Its not a Real Time experience free of chimping which is the antithesis of taking the picture.
Even when I shot polaroids, I distinctly remember having to wait some 90 seconds or so. When I shot film, the experience could never remotely be described as real time, especially when I shot and scanned color neg. DSLRs make taking the picture closer to a real time experience than anything I'd experienced previously.
-->Maybe a color meter option can be combined with the spot meter in camera?
The sensor in a digital camera is at least as accurate as any of the emission colorimeters sold as "color meters." What would adding a $30 part achieve? A case could be made for adding an emissive 16- or 32-band spectrophotometer to characterize the light source, but do you really want to add $500 to the cost of the camera to obtain a benefit that can be achieved equally well by click-balancing on a known reference?
-->I predict Olympus and perhaps Sony/Minolta will address this issue. The methods we now use for color management are just overly engineered nonsense and absurd.
Who is "we," kimo sabe? You may be using overengineered nonsense, which I imagine wouldn't work too well. I'm using open-standard color management tools, most of which are built into Photoshop, and I'm getting digital images that correspond well with my memory of the scene, and prints that correspond well with the soft proofs I see on the monitor. At this point, the only third-party tools I'm using to do so are a Macbeth Color Checker, and a Gretag colorimeter for display calibration and characterization. Epson's profiles for the K3 inkset are good enough that I haven't bothered to make custom printer profiles even though I have everything I need to do so.
-->Digital has a very sensitive take on colors at capture, perhaps too sensitive.
If you're complaining that digital captures colors more accurately than film, I'd have to agree with the observation, but disagree strenuously with the notion that it is in any way a bad thing.
-->So upstream strategies to me are the answer, not another set of PS actions or workarounds downstream.
I have absolutely no idea what this means. That may well be my fault. I beg enlightenment.
Bruce - I will continue to use Photoshop's strengths in my workflow, perhaps try Lightroom and/or other solutions and see how that works out. Thank you for taking the time to post a response which I printed out and will be studying.