Photoshop save for web color shift

245

Comments

  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    jfriend wrote:
    This one has me confused.

    Can you please explain why there isn't a single right answer for doing digital photography? When would the answer be 5000 and when would it be 6500 and when would it be something else? What do you choose and why?

    Does the display of an image in Photoshop vary depending upon what color temperature setting in your calibration/profiling software you choose?

    The white point is based on viewing conditions of a print. Are you viewing them at D50? And even if you do, does D50 provide a good screen to print matching? Its a bit too warm? So you calibrate to D55.

    Gamma is dumb, you should just use Native 99% of the time. I'd like the color management software companies to just do this automatically unless some expert preference is selected. But say you're on a Windows machine, using non ICC aware applications and you want to calibrate to a TRC gamma of 1.8 so it "looks" like a Mac. You're now given the opportunity to do so.

    This goes back to Baldy's idea that somehow, out of the box, display’s should somehow conform to a fixed behavior, one as poor as sRGB so web users can sort of, maybe see the same color. That's removing all kinds of useful functionality that many users want to fix an issue that can be fixed by simply using the right browser.

    Its like the guy who buys a $200,000 sports car capable of going 200mph but there's a speed limiter that only lets it go to 130mph.

    We could enforce the speed limits in the US by simply forcing all auto's to only go a max of 75mph. How well do you think that would go over to the buying public?
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • jfriendjfriend Scripting dude-volunteer Registered Users Posts: 8,097 Major grins
    edited January 23, 2008
    arodney wrote:
    The white point is based on viewing conditions of a print. Are you viewing them at D50? And even if you do, does D50 provide a good screen to print matching? Its a bit too warm? So you calibrate to D55.

    Gamma is dumb, you should just use Native 99% of the time. I'd like the color management software companies to just do this automatically unless some expert preference is selected. But say you're on a Windows machine, using non ICC aware applications and you want to calibrate to a TRC gamma of 1.8 so it "looks" like a Mac. You're now given the opportunity to do so.

    This goes back to Baldy's idea that somehow, out of the box, display’s should somehow conform to a fixed behavior, one as poor as sRGB so web users can sort of, maybe see the same color. That's removing all kinds of useful functionality that many users want to fix an issue that can be fixed by simply using the right browser.

    Its like the guy who buys a $200,000 sports car capable of going 200mph but there's a speed limiter that only lets it go to 130mph.

    We could enforce the speed limits in the US by simply forcing all auto's to only go a max of 75mph. How well do you think that would go over to the buying public?

    So, does the choice of color temperature influence how things look in Photoshop (e.g. in color-managed applications) too?

    Can you point me to any scientific method (e.g. step by step method) for figuring out what color temperature is the right choice for a specific office setting? Is this covered in your book? Any online references? The EyeOne Display 2 software just assume you know the answer to this question and doesn't offer any help at all.
    --John
    HomepagePopular
    JFriend's javascript customizationsSecrets for getting fast answers on Dgrin
    Always include a link to your site when posting a question
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    jfriend wrote:
    So, does the choice of color temperature influence how things look in Photoshop (e.g. in color-managed applications) too?

    Yes.
    Can you point me to any scientific method (e.g. step by step method) for figuring out what color temperature is the right choice for a specific office setting? Is this covered in your book?

    Measure the color with a Spectrophotometer like an EyeOne Pro. You need to measure the area or viewing box where you'll view the prints. Then control other lighting so its not affecting this area. This is why companies like GTI make Fluorescent boxes for viewing transparencies and prints as does Solux.

    Note that even a so called "daylight" Fluorescent box like those by GTI are not producing true, full spectrum daylight (impossible with a Fluorescent). And you may need to use a different correlated color temperature to produce a color match. For example, two weeks ago I was at photographer Art Wolfe's studio in Seattle setting them up. They have a GTI lightbox and a NEC 2690 display. Once we got the luminance of the two to visually match, we ended up calibrating the white point of the display to D55 to get a very close screen to print match. D50 was too warm, D65 too cool.
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • jfriendjfriend Scripting dude-volunteer Registered Users Posts: 8,097 Major grins
    edited January 23, 2008
    arodney wrote:
    Yes.



    Measure the color with a Spectrophotometer like an EyeOne Pro. You need to measure the area or viewing box where you'll view the prints. Then control other lighting so its not affecting this area. This is why companies like GTI make Fluorescent boxes for viewing transparencies and prints as does Solux.

    Note that even a so called "daylight" Fluorescent box like those by GTI are not producing true, full spectrum daylight (impossible with a Fluorescent). And you may need to use a different correlated color temperature to produce a color match. For example, two weeks ago I was at photographer Art Wolfe's studio in Seattle setting them up. They have a GTI lightbox and a NEC 2690 display. Once we got the luminance of the two to visually match, we ended up calibrating the white point of the display to D55 to get a very close screen to print match. D50 was too warm, D65 too cool.

    OK, I understand what you're saying now. But, I'm amazed that this is the desirable way to calibrate a screen for general purpose use.

    You are saying that you have to decide what print viewing conditions you want your screen to match and you have to calibrate your screen for that specific condition which is unique to your own little world. I get how there are some specific cases where this is the goal, but most people use their screen for a much wider variety of things than print viewing in one specific type of lighting.

    What if I'm calibrating my screen primarily for web display and I want my screen to show the same thing that other people with calibrated/profiled displays will see when we all use color-managed tools? Can't I just ask for "accurate" color reproduction on my screen that matches a color standard?

    What if I'm primarily using my screen to prepare prints for people that will view them in unknown lighting? What would I choose then?
    --John
    HomepagePopular
    JFriend's javascript customizationsSecrets for getting fast answers on Dgrin
    Always include a link to your site when posting a question
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    jfriend wrote:
    OK, I understand what you're saying now. But, I'm amazed that this is the desirable way to calibrate a screen for general purpose use.

    What's general purpose work?
    You are saying that you have to decide what print viewing conditions you want your screen to match and you have to calibrate your screen for that specific condition which is unique to your own little world.

    Well if your goal is to view a print and an emissive display and get them to match, yes. If you view a print alone (no display) outside in morning light, afternoon light and sunset, plus under a desk lamp, they don't look the same!
    What if I'm calibrating my screen primarily for web display and I want my screen to show the same thing that other people with calibrated/profiled displays will see when we all use color-managed tools? Can't I just ask for "accurate" color reproduction on my screen that matches a color standard?

    You can ask what everyone else is using and match that. Now consider that most users don't calibrate the display at all. And consider that the display is changing its behavior over time, as little as monthly. Then consider that you're asking that what you see and everyone of those users sees matches. See the problem? Worse, these people are not only not calibrating their displays the same way, they don't use an ICC aware web browser.
    What if I'm primarily using my screen to prepare prints for people that will view them in unknown lighting? What would I choose then?

    Guess. Good luck. This is why we try to limit the variables as much as possible. Calibration, profiling and using ICC aware applications make this far less variable but without communication with others, you're left guessing.
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • jfriendjfriend Scripting dude-volunteer Registered Users Posts: 8,097 Major grins
    edited January 23, 2008
    arodney wrote:
    What's general purpose work?



    Well if your goal is to view a print and an emissive display and get them to match, yes. If you view a print alone (no display) outside in morning light, afternoon light and sunset, plus under a desk lamp, they don't look the same!



    You can ask what everyone else is using and match that. Now consider that most users don't calibrate the display at all. And consider that the display is changing its behavior over time, as little as monthly. Then consider that you're asking that what you see and everyone of those users sees matches. See the problem? Worse, these people are not only not calibrating their displays the same way, they don't use an ICC aware web browser.



    Guess. Good luck. This is why we try to limit the variables as much as possible. Calibration, profiling and using ICC aware applications make this far less variable but without communication with others, you're left guessing.

    Wow, I'm surprised at the state of this technology. This whole process looks like it's designed for the professional who's producing a specific type of locally printed result viewed under controlled circumstances and will optimize their environment for that specific result. I expect that was the predominant marketplace for color calibration equipment 5 years ago, but I don't think that is the largest part of the marketplace any more.

    There is a new breed of folks who are not professionals, do not have a color controlled viewing station for printers, but they are now trying to calibrate their displays to improve their color accuracy. Many of this new breed may not even print important prints themselves. Rather, they want to:
    • Produce images for the web that will display as good as possible for all viewers (at least no obvious color casts).
    • Produce images for the web that will display the same on their screen as for other people who have calibrated their screens and are using color-managed viewers.
    • Produce images that will be printed on ICC profiled printers at various printing services (either the local Costco/Wal-mart or one of the online Labs like MPIX or EzPrints).
    • Show people images or slideshows on their own computer.
    I'm coming to the conclusion that today's tools are not designed for this. They are designed to match print view in a particular lighting when you know the color temperature of that lighting.

    It seems like the market would be ripe for a new set of calibration tools that would calibrate an emmissive display to a known color standard such that everyone who uses the tool on their emmisive display would see the same colors when using color managed applications. These new tools would not ask you a color temperature question because the goal is to match a color standard, not to match the lighting used for print viewing.
    --John
    HomepagePopular
    JFriend's javascript customizationsSecrets for getting fast answers on Dgrin
    Always include a link to your site when posting a question
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    Yes and it would be great if we had star trek transporters

    What you're asking for is that no matter what we view, where we view it, everything matches. And yet, we can't break the law of physics. If you agree that no matter what illuminant we use, it affects the perception of color on a print, how do we handle this automatic calibration? If we agree that everyone's display is different out of the box, let alone over time (or based on how they might adjust it), how does everyone sync up a new LCD producing 300cd/m2 in a dim room and someone with a 10 year old CRT sitting beside a pink wall and a large window?

    What you're saying is, you'd like a prefect world and so would I. Now back to reality. We have tools today to produce pretty good print to screen matching considering one's reflective and the other emissive and using quite different technologies just to make the light. It requires tools and techniques (and some money).

    The technology is just as much designed for pros as it's not. A $300 LCD and a $5000 LCD ain't the same in terms of who they are designed for and what they can provide.
    It seems like the market would be ripe for a new set of calibration tools that would calibrate an emmissive display to a known color standard such that everyone who uses the tool on their emmisive display would see the same colors when using color managed applications. These new tools would not ask you a color temperature question because the goal is to match a color standard, not to match the lighting used for print viewing.

    And it should be inexpensive and come with a brain probe to read the users mind because they can't possibly tell the product what ambient light conditions they are under, what products they are using and what their goals are.

    Really, you're asking for something that today isn't possible at any cost. The market is ripe for a car that gets 99mpg and runs on water but it don't exist. So instead of writing about science fiction here, lets get back to what IS possible today with current technology.

    IF you want your print to match your display and get a web page to match what you see in Photoshop, that's possible. I've described how. Now what?
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • jdryan3jdryan3 tao te grin Registered Users Posts: 1,353 Major grins
    edited January 23, 2008
    jfriend wrote:
    There is a new breed of folks who are not professionals, do not have a color controlled viewing station for printers, but they are now trying to calibrate their displays to improve their color accuracy. Many of this new breed may not even print important prints themselves. Rather, they want to:
    • Produce images for the web that will display as good as possible for all viewers (at least no obvious color casts).
    • Produce images for the web that will display the same on their screen as for other people who have calibrated their screens and are using color-managed viewers.
    • Produce images that will be printed on ICC profiled printers at various printing services (either the local Costco/Wal-mart or one of the online Labs like MPIX or EzPrints).
    • Show people images or slideshows on their own computer.
    I'm coming to the conclusion that today's tools are not designed for this. They are designed to match print view in a particular lighting when you know the color temperature of that lighting.

    It is typical of technology - 2 steps forward, 3, uh, 1 step back. I think there are 3 groups: those that care and are willing to do almost anything; those that care and are willing to do something, but not much ( look in the hlep, downlaod a patch, flip a setting); and those who not only don't care, they don't even notice. Absolutely indifferent.

    Putting almost anyone in this forum in group one, I think the B*tching by group 2 is what will drive change. That and just the typical maturity of a technology. And I don't mean digital photography. I mean browsers as a primary means of communication. It is still in its infancy. More people do it but most aren't yet demanding a 'quality' experience.

    Add in digital photography, HD DVD quality games, and HD TV... those thing will simply drive the end users to demand a quality experience across media. Only then will it happen.
    jfriend wrote:
    It seems like the market would be ripe for a new set of calibration tools that would calibrate an emmissive display to a known color standard such that everyone who uses the tool on their emmisive display would see the same colors when using color managed applications. These new tools would not ask you a color temperature question because the goal is to match a color standard, not to match the lighting used for print viewing.
    I think this plays right into Andrew's statements re: a lack of standards. If there were standards, you could eliminate the need for the ad hoc adaptive tools you mention. But as mentioned in the IEblog posting, MS wants to move that way, but leave no non-compliant enduser or their equipment behind, "like his mom."

    So if you could even enable in the options to turn on/off the ability (default being ON) IE8 to use ICC profiles, it would go a long way to heading down that path. Those that even slightly care, will notice. Those that don't, well ...

    :soapbox
    "Don't ask me what I think of you, I might not give the answer that you want me to. Oh well."
    -Fleetwood Mac
  • jfriendjfriend Scripting dude-volunteer Registered Users Posts: 8,097 Major grins
    edited January 23, 2008
    arodney wrote:
    Yes and it would be great if we had star trek transporters

    What you're asking for is that no matter what we view, where we view it, everything matches. And yet, we can't break the law of physics. If you agree that no matter what illuminant we use, it affects the perception of color on a print, how do we handle this automatic calibration? If we agree that everyone's display is different out of the box, let alone over time (or based on how they might adjust it), how does everyone sync up a new LCD producing 300cd/m2 in a dim room and someone with a 10 year old CRT sitting beside a pink wall and a large window?

    What you're saying is, you'd like a prefect world and so would I. Now back to reality. We have tools today to produce pretty good print to screen matching considering one's reflective and the other emissive and using quite different technologies just to make the light. It requires tools and techniques (and some money).

    The technology is just as much designed for pros as it's not. A $300 LCD and a $5000 LCD ain't the same in terms of who they are designed for and what they can provide.



    And it should be inexpensive and come with a brain probe to read the users mind because they can't possibly tell the product what ambient light conditions they are under, what products they are using and what their goals are.

    Really, you're asking for something that today isn't possible at any cost. The market is ripe for a car that gets 99mpg and runs on water but it don't exist. So instead of writing about science fiction here, lets get back to what IS possible today with current technology.

    IF you want your print to match your display and get a web page to match what you see in Photoshop, that's possible. I've described how. Now what?
    Andrew, I don't know why you decided to stop being helpful and go off into Star Trek land and science fiction. I'm working really hard here trying to keep the conversation with you on a productive level and I'm trying to add a perspective that you may not live every day of the photographic prosumer who uses the web a lot and doesn't do their own printing (which happens to represent a lot of the users here on dgrin and customers of Smugmug).

    To that end, I don't think I asked for nuclear fusion here. All I asked for was the ability to push a button (with a hardware puck attached and with no question asked about color temperature) and have an emissive display calibrated to produce a standard set of colors such that if a non-color-professional on the other side of the world does the same thing with their display, both our displays will look the same when using color-managed apps.

    I explicitly excluded anything to do with prints or print lighting. If you think that's not possible, I'd like to know why. If we need to turn off all ambient lighting so that all light is coming from the emissive display, I'm even willing to stipulate to that.

    Is this technically impossible? If it is, please explain why you think it's impossible rather than just telling me it's an unreachable thing to wish for with 99 mpg car analogies.

    Also, I do not believe you have described how one goes about setting up a display so that my view of a web page matches what another color-calibrated professional sees in their view of a web page. As you've described it so far, if the two of us choose different color temperatures, our views of that web page will be different even if we do everything else you've recommended. So, the only way Baldy and I can calibrate our displays to match is to communicate on picking the same color temperature? Or if we don't communicate and I pick 5000 and Baldy picks 6500, then he will view my images a lot cooler than I do (given all apps are color-managed)?
    --John
    HomepagePopular
    JFriend's javascript customizationsSecrets for getting fast answers on Dgrin
    Always include a link to your site when posting a question
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    jfriend wrote:
    Andrew, I don't know why you decided to stop being helpful and go off into Star Trek land and science fiction.

    Because what you're asking for is highly unrealistic. There's what we'd like to have and what we currently have. What we currently have isn't perfect but it works, it has for over a decade. It does require certain work and an expense by the user but so does a display, an ISP and something to digitize images to place on the web. I've stated how we can produce matching color appearance in all ICC aware applications. Wishing that some display out of the box that conforms to some as yet undefined "standard" will make the web appear to everyone the same way isn't going to happen any time soon and its currently not necessary.
    I'm working really hard here trying to keep the conversation with you on a productive level and I'm trying to add a perspective that you may not live every day of the photographic prosumer who uses the web a lot and doesn't do their own printing (which happens to represent a lot of the users here on dgrin and customers of Smugmug).

    I don't pretend to know what all consumers or Pro's want or need. I don't know that anyone can. I know what I need in terms of color matching and getting it isn't really difficult and by 1998 standards, pretty inexpensive.

    As I've said here in the past (and beautifully expressed by jdryan3 above), there are three basic types of users and two are not a problem (one doesn't care, the other does whatever they have to do to make things work). Group #2 is the PITA.
    To that end, I don't think I asked for nuclear fusion here. All I asked for was the ability to push a button (with a hardware puck attached and with no question asked about color temperature) and have an emissive display calibrated to produce a standard set of colors such that if a non-color-professional on the other side of the world does the same thing with their display, both our displays will look the same when using color-managed apps.

    We kind of have this but the problem is, there's no guarantee its going to work enough for camp #1 because we don't have any idea how they will view the prints. If all you want to do is calibrate a display inexpensively with the least amount of color geek speak, there's the Pantone huey. It works, it asks as few questions of the user as possible and doesn't try to deal with terms like "gamma" and "White Point". Its a step in the right direction. Its not going to fly for camp #1, its pretty close to making camp #2 happy (the jury is out here as there's no way we'll please all the people in camp #2 all the time).
    Also, I do not believe you have described how one goes about setting up a display so that my view of a web page matches what another color-calibrated professional sees in their view of a web page.

    In the context of camp #2, you get a huey and you set it for Web viewing (that's what they call it) then you hope everyone else you deal with does the same.
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • jfriendjfriend Scripting dude-volunteer Registered Users Posts: 8,097 Major grins
    edited January 23, 2008
    arodney wrote:
    Because what you're asking for is highly unrealistic. There's what we'd like to have and what we currently have. What we currently have isn't perfect but it works, it has for over a decade. It does require certain work and an expense by the user but so does a display, an ISP and something to digitize images to place on the web. I've stated how we can produce matching color appearance in all ICC aware applications. Wishing that some display out of the box that conforms to some as yet undefined "standard" will make the web appear to everyone the same way isn't going to happen any time soon and its currently not necessary.



    I don't pretend to know what all consumers or Pro's want or need. I don't know that anyone can. I know what I need in terms of color matching and getting it isn't really difficult and by 1998 standards, pretty inexpensive.

    As I've said here in the past (and beautifully expressed by jdryan3 above), there are three basic types of users and two are not a problem (one doesn't care, the other does whatever they have to do to make things work). Group #2 is the PITA.



    We kind of have this but the problem is, there's no guarantee its going to work enough for camp #1 because we don't have any idea how they will view the prints. If all you want to do is calibrate a display inexpensively with the least amount of color geek speak, there's the Pantone huey. It works, it asks as few questions of the user as possible and doesn't try to deal with terms like "gamma" and "White Point". Its a step in the right direction. Its not going to fly for camp #1, its pretty close to making camp #2 happy (the jury is out here as there's no way we'll please all the people in camp #2 all the time).



    In the context of camp #2, you get a huey and you set it for Web viewing (that's what they call it) then you hope everyone else you deal with does the same.

    Thanks. If you had some other calibration brand (besides a Huey) that asks for a white point temperature and you are most interested in web viewing, do you have any recommendation for what temperature setting to set it to? Do we know what the Huey decides to use in that case?
    --John
    HomepagePopular
    JFriend's javascript customizationsSecrets for getting fast answers on Dgrin
    Always include a link to your site when posting a question
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    jfriend wrote:
    Thanks. If you had some other calibration brand (besides a Huey) that asks for a white point temperature and you are most interested in web viewing, do you have any recommendation for what temperature setting to set it to? Do we know what the Huey decides to use in that case?

    D50.

    As for the huey:

    http://www.ppmag.com/reviews/200605_pantonehuey.pdf
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • BaldyBaldy SmugMug co-founder Mountain View, CaRegistered Users, Super Moderators Posts: 2,853 moderator
    edited January 23, 2008
    jfriend wrote:
    If you have an LCD display, you may NOT be calibrating your display to a particular color standard.

    You may be just profiling your display (measuring it's current color performance and characteristics) and dropping that profile in a standard location for color aware apps to find it. Firefox may not display any differently before or after profiling your display.
    John, you're the master of clarity. Thanks!

    I haven't read the links you referenced yet but will, hopefully tonight. One hypothesis Andy and I talked about was that, depending on your monitor/graphics card combo, the OS has limited control over how much color change it can make via controlling the hardware, whereas Photoshop has more latitude via software.

    The thing that's throwing us is when you change the monitor profile, Firefox's display changes (as does Photoshop's). On my Mac, I can flip between the sRGB profile Apple includes, the sRGB profile Adobe includes, the profile Dell includes, and the Eye-One profile from the hardware calibration/profiling--and FF renders the same photo quite differently for all 4 (the Eye-One profile makes web pages look too red and too saturated).

    And Andrew has said on our forums that the browser uses the monitor profile. In his book he says, "Most products ensure the profile is placed in the proper location on your hard drive and set this profile for use by the operating system (unless told not to)."

    The other color management books say, "As the OS is booting you'll see a shift in color. That's the OS putting the monitor profile to work."

    One confusing point to me is Andrew's book also says, "Apple's Safari running under OSX also is able to recognize embedded profiles and/or assume sRGB so it should preview images correctly." I'm assuming that's old/bad info from Apple.

    Yeah, I know I'm daft, but getting great clarity on this will do the world a big favor. I'll write it up as a help section if we can nail it. You see photographers everywhere on the forums struggling with this (and blaming SmugMug for making photos from their newly-calibrated monitors look like bletch).
  • BaldyBaldy SmugMug co-founder Mountain View, CaRegistered Users, Super Moderators Posts: 2,853 moderator
    edited January 23, 2008
    arodney wrote:
    No, no no! Repeat after me: Safari is correct, Photoshop is correct.
    Sorry to be daft, but can you show me some example web pages that Safari displays correctly? I get the feeling you're going to show us pages that are jpeg-only with ICC profiles attached. I can see how that satisfies photographers who are only focused on photos, but I think I speak for all web designers when I say we think the rest of the page should display correctly too.
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    Baldy wrote:
    The thing that's throwing us is when you change the monitor profile, Firefox's display changes (as does Photoshop's).

    Of course it does, as I explained to Andy. When you load the profile, a LUT is loaded into the graphic card (if a LUT exists, not all profiles do this). Now go into Photoshop and if you are quick, you'll see yet another redraw. That's Photoshop using Display Using Monitor Compensation based on the new profile. You will NOT see this in non ICC aware applications, they don't know the profile exits.
    And Andrew has said on our forums that the browser uses the monitor profile.

    ICC aware browsers do, non ICC aware browsers do not.
    The other color management books say, "As the OS is booting you'll see a shift in color. That's the OS putting the monitor profile to work."

    No, that's the LUT being loaded.
    One confusing point to me is Andrew's book also says, "Apple's Safari running under OSX also is able to recognize embedded profiles and/or assume sRGB so it should preview images correctly." I'm assuming that's old/bad info from Apple.

    OS X used to allow you to pick a profile for untagged documents OR assume sRGB. That was if memory serves me, the OS before Tiger. Tiger forward now assumes Monitor RGB (which is dumb). The functionality described on page 277-279 in my book is no longer available, you can't select what profile to use for untagged docs (step in the wrong direction). Look at figure 8.1 in my book and what you now have in OSX and you'll see that the Preference button and those options are now gone.
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    Baldy wrote:
    Sorry to be daft, but can you show me some example web pages that Safari displays correctly? I get the feeling you're going to show us pages that are jpeg-only with ICC profiles attached. I can see how that satisfies photographers who are only focused on photos, but I think I speak for all web designers when I say we think the rest of the page should display correctly too.

    Every web gallery I've got matches what I see in Photoshop and Lightroom:
    http://digitaldog.net/ARsAmazonPicks/
    http://www.digitaldog.net/GormanMendo07/index.html

    And so on...

    And then there's this:
    http://www.color.org/version4html.xalter

    By all means, speak for web designers, I'm not one. I'm also not viewing the boarders and stuff around images of web pages in Photoshop (nor do I really care).
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • AndyAndy Bicameral New YorkRegistered Users Posts: 50,154 Major grins
    edited January 23, 2008
    arodney wrote:
    nor do I really care).

    I think this is the bit that is making some of us bristle at your comments and line of reasoning.

    You are talking about a world where everyone has Safari - or goes and gets it ... but in reality, 90+% of the viewers of photos ( made by digital photographers who use Dgrin, SmugMug, Flickr, et. al., or their own websites or blogs) are in a world based on reality, that reality being a browser that is not Safari.

    And we (photographers who display our photos on the web, and order / sell prints from them, said prints made from a lab affiliated with our website) are looking for ways to mitigate the huge difference in display.
  • bwgbwg SmugMug Sorcerer Northern VARegistered Users, Retired Mod Posts: 2,119 SmugMug Employee
    edited January 23, 2008
    arodney wrote:
    Every web gallery I've got matches what I see in Photoshop and Lightroom:
    http://digitaldog.net/ARsAmazonPicks/
    http://www.digitaldog.net/GormanMendo07/index.html

    so, i've been lurking trying to soak all this in.

    The galleries you linked are in Flash, a non-color managed application.

    Based on what i've understood so far, those photos may look completely different to you and me even though I'm using safari and have a calibrated/profiled monitor. Correct?
    Pedal faster
  • BaldyBaldy SmugMug co-founder Mountain View, CaRegistered Users, Super Moderators Posts: 2,853 moderator
    edited January 23, 2008
    arodney wrote:
    Every web gallery I've got matches what I see in Photoshop and Lightroom:
    http://digitaldog.net/ARsAmazonPicks/
    I saw those. They're beautiful, Andrew. I dunno what shade of gray you're using and it doesn't sound like it's important to you. I know you know this, but who knows what that gray is gonna be on different monitors?
    Yikes, that's really scary. I'm assuming you used that as an ultra-ironic example of how bad it can get.

    The International Color Consortium doesn't care enough about color on its pages to use an ICC profile on the jpeg of crayons. Their logo, CMYK color swatches, and the color checker on the bottom are all mystery gifs.

    Not even Apple attaches an ICC profile to its banner for Aperture on its own pages:

    index00.jpg

    If even the ICC and Apple won't use ICC profiles in their own pages, web designers have to find a better way. I think John's suggestion of tags makes a lot of sense.
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    Andy wrote:
    I think this is the bit that is making some of us bristle at your comments and line of reasoning.

    You are talking about a world where everyone has Safari - or goes and gets it ... but in reality, 90+% of the viewers of photos

    I'm talking about anyone today who has the need for a color managed browser. Hopefully other's will follow. But what you say is moot since you either care what you view is correct or you don't. If you do, its not brain surgery to download Safari, it cost nothing.

    If of the 90% of those you specify, 80% don't care, fine. You only need to deal with the 10% that do.

    This really is a straw man argument on your part Andy. We have products that don't work properly and those that do. Those that care will use the appropriate products.

    Something like 33% of Americans are over weight or obese and the masses love MacDonalds. With your reasoning, the rest of the population is doomed because you assume they are not intelligent enough to eat well and exercise. That 90% of your users don't work with Safari may be factual. But the question is, how many care about the images they preview and how many know that the browser they use is the cause of the mismatch? Instead of crying about how many don't use Safari, why not spend some energies educating those that DO care but don't know why the current browser they use is providing incorrect previews?
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    Baldy wrote:
    I saw those. They're beautiful, Andrew. I dunno what shade of gray you're using and it doesn't sound like it's important to you. I know you know this, but who knows what that gray is gonna be on different monitors?

    Thanks. Maybe its luck or coincidence but the gray bkdng on the Amazon web galleries and the gray in LR.'s web module (Lightroom Flash Gallery) are the same. But you're right, I don't care all that much. If it were pink, I'd care!
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • AndyAndy Bicameral New YorkRegistered Users Posts: 50,154 Major grins
    edited January 23, 2008
    arodney wrote:
    I'm talking about anyone today who has the need for a color managed browser. Hopefully other's will follow. But what you say is moot since you either care what you view is correct or you don't. If you do, its not brain surgery to download Safari, it cost nothing.
    Ahh actuall it would cost a fortune :D How much do you think Microsoft has spent to keep it's dominant browser market share? The sad truth is, the majority of people out there don't even know there are other browsers, and don't know you can have more than one. They are afraid of anything different.

    If of the 90% of those you specify, 80% don't care, fine. You only need to deal with the 10% that do.
    Not really wise when you have millions of visitors and customers hitting the site. Do you see where we are coming from now?
    This really is a straw man argument on your part Andy. We have products that don't work properly and those that do. Those that care will use the appropriate products.

    Something like 33% of Americans are over weight or obese and the masses love MacDonalds. With your reasoning, the rest of the population is doomed because you assume they are not intelligent enough to eat well and exercise. That 90% of your users don't work with Safari may be factual. But the question is, how many care about the images they preview and how many know that the browser they use is the cause of the mismatch? Instead of crying about how many don't use Safari, why not spend some energies educating those that DO care but don't know why the current browser they use is providing incorrect previews?
    See above - it's a monumental task :)
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 23, 2008
    Andy wrote:
    Ahh actuall it would cost a fortune :D How much do you think Microsoft has spent to keep it's dominant browser market share? The sad truth is, the majority of people out there don't even know there are other browsers, and don't know you can have more than one.

    Tell them. Show them the differences. Or do you guys work for MS? I haven't used an MS browser on the Mac for years (and for good reason) but if memory serves me, it was free then and I'd have to assume its free now.

    But lets say people who us MS browsers, as you say, will go kicking and screaming before they switch. We have to assume then that the color appearance of images on the web is less important. Let em have the non color managed browser. That's not a hill worth dying on.

    Again, I keep saying this over and over, and the replies are (forgive me) lame. You have people who care about color consistency and those that don't. I'd like a dollar for all the posts I've made here and elsewhere from Photographers who want to know why images they see in their browsers don't match what they see in Photoshop (or Lightroom, Aperture, Bibble, Elements etc). I have to assume they care. They did notice and they do want the two to match. The solution is really damn simple. What isn't, is having to spend far too much time writing all this crap and convincing people that their beloved browser is totally brain dead about previewing images. Of course that's not the case with Mac users because thankfully (despite criticism of how they may upload images to the web), the browsers are color managed as a default install. So at least those that "think different" can be kept out of this goofy equation.
    Not really wise when you have millions of visitors and customers hitting the site. Do you see where we are coming from now?

    No I don't! You apparently host a site where images and how they preview are important? Have you gone to similar pages where someone has a link that shows a gray ramp in a somewhat ineffective effort to convince the viewer that the display isn't correct? You simply need a banner "do you want to see these images correctly?" or something like that. "No, I could care less, lets get to the photo's" one group will say. The other will say "yes, damn straight I want to view the images correctly" then you tell them how.

    Of you can assume they are mind readers. Or dismiss education because someone might have to switch to a new browser.

    Seriously, I get the idea that you'd like to bitch about his issue but nothing more. What are you going to do to educate your clientele about this issue of mismatching previews? You take ads? How about asking X-rite or Color Vision to pop up one for their colorimeter? Whatever. But to say "Its too hard" or "Too many users are on MS browsers, we're doomed" is not acceptable. Personally I don't care what you do. But I suspect YOU do.

    Or we can wait on that star trek perfect world. But I wouldn’t hold my breath.
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • pathfinderpathfinder Drive By Digital Shooter western IndianaSuper Moderators Posts: 14,649 moderator
    edited January 23, 2008
    Not only are there viewers who do not know the different way browsers display colors, there are lots of folks who view the web at work through the only browser they are allowed to use IE 6 or IE 7. DAMHIK:D

    Needless to say, I work with a Mac and Safari on a calibrated monitor when allowed. Like at home.
    Pathfinder - www.pathfinder.smugmug.com

    Moderator of the Technique Forum and Finishing School on Dgrin
  • BaldyBaldy SmugMug co-founder Mountain View, CaRegistered Users, Super Moderators Posts: 2,853 moderator
    edited January 24, 2008
    bigwebguy wrote:
    Your galleries are in Flash, a non-color managed application.
    He's right, Andrew. Your galleries are not color managed. They look the same in Safari, IE and Firefox, and different than your photos do in Photoshop. Safari has no choice but to ignore the ICC profile you're attaching, so it's only serving to slow down image loads.
  • jdryan3jdryan3 tao te grin Registered Users Posts: 1,353 Major grins
    edited January 24, 2008
    arodney wrote:
    As I've said here in the past (and beautifully expressed by jdryan3 above), there are three basic types of users and two are not a problem (one doesn't care, the other does whatever they have to do to make things work). Group #2 is the PITA.

    My groupings, his opinion. But not mine. I love your passion and have learned a lot from your postings. Your advocacy is needed. You are not tilting against windmills.

    But Group #2 isn't just Baldy & Andy's concern. They are the people who buy images from the photographers who are SmugMug account holders -regular up to Pro accounts.

    I'll agree most people who use SmugMug to host their images need to be educated. But based upon the online help SmugMug provides and the many threads in Dgrin about Soft proofing, calibration, and related topics, I think they (we?) have the photographers covered. Yeah, there are those who ask the question, get the answer and say "Oh, I'm not laying out $79 for Huey. I'll just eyeball it/compare to some print/ blah blah blah". But those folks in reality are Group 3's. Indifferent.

    The harder one to deal with is those photgrapher's customers. Who is responsible for them? Remember: a lot of those people are going to mypictures.smugmug.com, not dgrin. I do like the idea of a graybar in the footer of the page - I think I'll add one to mine. And maybe Baldy/Andy will consider having it load as part of the website template default.

    But we do live in an IE world. Period. End of conversation. Netscape lost, the Phoenix of Mozilla/Firefox not withstanding. Oops! Firefox is not ICC aware. Apple only last year came out with a version of Safari for the PC. Are you positive it is ICC aware? Or is only the Mac version? Apple has <10% of PC market worldwide. Do the SmugMug account holders and stakeholders just ignore them, send them off to Fl*ckr?

    Baldy is trying to find a way to resolve the issue of printed product differing from what is online. And I doubt we are talking slight color shifts. Someone on another thread was talking about green hair recently. Between my camera, post processing, loading to SmugMug and some person who I never met but liked my work at a benefit show ordering prints, there are at least 4 touch points where color management comes into play. I can control 3, some with SmugMug's help. But that last one, the customer ordering the print, I can't. Not gonna happen. But, unlike you, I do care. And I'm glad Baldy does too, enough that he puts it out there to try and find out what HE can do. Or at least find out enough to be able to tell those folks/ post on the order form/ include in the online help for buyers what the issues may be.

    :whew
    "Don't ask me what I think of you, I might not give the answer that you want me to. Oh well."
    -Fleetwood Mac
  • SloYerRollSloYerRoll Major grins Registered Users Posts: 2,788 Major grins
    edited January 24, 2008
    Here's a link for the rest of the world that's reading this thread and wonders what the real pecentages of web browser use is:

    http://www.w3schools.com/browsers/browsers_stats.asp
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 24, 2008
    Baldy wrote:
    He's right, Andrew. Your galleries are not color managed. They look the same in Safari, IE and Firefox, and different than your photos do in Photoshop. Safari has no choice but to ignore the ICC profile you're attaching, so it's only serving to slow down image loads.

    As far as I know, LR is embedding sRGB in the documents. And the previews perfectly match what I see in Safari on my machine to LR (and Photoshop).

    I may upload a gallery not using Flash (I don't really care if its a Flash gallery or not, that's just the LR default I've used a few times) to see if there's a difference.

    Color managing Flash IS something Adobe is working on. More I can't say due to NDAs.
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • arodneyarodney Major grins Registered Users Posts: 2,005 Major grins
    edited January 24, 2008
    jdryan3 wrote:
    But Group #2 isn't just Baldy & Andy's concern. They are the people who buy images from the photographers who are SmugMug account holders -regular up to Pro accounts.

    We could say the same thing about the masses who purchase stuff on the web, pick colors and either keep what they purchased or don't (and complain). You'd think that if MS thought this was a huge issue for the masses, they'd fix the browser. So maybe this isn't a big deal. Well it is a big deal for camp #1 or anyone who notices and makes a stink. And those are the people we can address.
    Yeah, there are those who ask the question, get the answer and say "Oh, I'm not laying out $79 for Huey. I'll just eyeball it/compare to some print/ blah blah blah". But those folks in reality are Group 3's. Indifferent.

    We can't dictate to others what is or isn't an important product to use to get results we think are more important than they do. If they don't want to calibrate their displays even with a huey, fine. Let them eyeball it. If they have a problem with color matching after that experience, they have to decide if the product is now a good expenditure or not. I'm not suggesting we force anyone to do anything. We do however need to have a simple, well written reply to those who have problems. Its called tech support (something I do every day for my company). I have people that are really smart and those that sometimes I feel shouldn't own a computer. But they have a problem and I have to tell them how to solve it as simply as possible. If they don't like the way Photoshop runs because they don't have enough ram and don't want to buy more, well we're done once I tell them what's absolutely required to speed up Photoshop. But its their money. Its their time.

    What I don't accept is the idea that there's no solution to the problem because people will not switch browsers or calibrate their display. That is primarily the fix. After I tell them how to fix the problem, I'm done, they can decide to take the advise or not.
    The harder one to deal with is those photgrapher's customers. Who is responsible for them? Remember: a lot of those people are going to mypictures.smugmug.com, not dgrin.

    The people selling them something are responsible.
    I do like the idea of a graybar in the footer of the page - I think I'll add one to mine. And maybe Baldy/Andy will consider having it load as part of the website template default.

    So when the user see's it looks like crap, now what? Alter the crude controls on the display? Showing them a mismatch is only the first start. Now do they fix the issue (correctly)? There's the link to the ICC I posted that shows a mismatch on non ICC aware browsers. Its too geeky a page for this group but someone could build a similar presentation and explain how to fix the issue.
    But we do live in an IE world. Period. End of conversation.

    Just as we live in a MacDonalds (Fast Food Nation). I don't have to eat there. That's not an excuse not to at least try to educate users who care about what they see on their browsers. It doesn't matter how many people use IE. It matters how many care about what they view and will do something about it. I'm no more likely to run into people's homes and install Safari as I am about to run into MacDonalds and pull Big Mac's out of people's mouths.
    Baldy is trying to find a way to resolve the issue of printed product differing from what is online.

    I told him how.
    And I doubt we are talking slight color shifts. Someone on another thread was talking about green hair recently.

    We can only speculate. Brown hair that looks green is a major screw up, its more than a browser or color management issue. The link Andy sent me is a good illustration. On the phone he told me there was big differences in the previews (how do we define big differences without measuring and using a metric like DeltaE?). I saw the link and wrote back that there were differences, I didn't think they were huge and what I expected to see with a ICC and non ICC aware browser. But nothing like green hair.
    Between my camera, post processing, loading to SmugMug and some person who I never met but liked my work at a benefit show ordering prints, there are at least 4 touch points where color management comes into play. I can control 3, some with SmugMug's help. But that last one, the customer ordering the print, I can't. Not gonna happen.

    Correct. We have no idea where they will view the print and under what illuminant or surround (pink walls?).
    Andrew Rodney
    Author "Color Management for Photographers"
    http://www.digitaldog.net/
  • BaldyBaldy SmugMug co-founder Mountain View, CaRegistered Users, Super Moderators Posts: 2,853 moderator
    edited January 24, 2008
    arodney wrote:
    As far as I know, LR is embedding sRGB in the documents. And the previews perfectly match what I see in Safari on my machine to LR (and Photoshop).
    I think what that says is you have a good monitor calibration that makes Photoshop match the web. Shipping versions of Flash ignore ICC profiles. And your gallery looks identical in Safari, IE and FF.
Sign In or Register to comment.