If you are using an Adobe program for post processing, wouldn't it make sense to take your photos in Adobe-RGB as opposed to S-RGB.
I'm sure that this has been discussed somewhere on the internet over the past few years. But, since I'm lazy, I thought I would ask for opinions here and see what everyone else does.
__________________
K20D/BG-2, FA 35 f/2, FA 50mm f/1.4, Tamron 90mm f/2.8 Macro, DA 16-45 f/4, DA 18-55 f/3.5 - 5.6 II, DA 55-300 f/4 - 5.8, AF540FGZ
Wildly over-simplified and terribly dated! Gotta love the internet.
You shoot RAW; no matter weather dSLR or P&S; HAAA. Biggest difference wrt outputs: in a dSLR you can actually retrieve the RAW data--it ain't an image! Not even 1/2 an image--it's a third of the photo!
AdobeRGB is proprietary and made for POSTSCRIPT devices and CERTAIN PRE-PRESS machines. You may get some richer colors as that article details---then again you may get absolutely nothing! OR you may get a terrible mess---we could discuss statistics, but...
AdobeRGB and sRGB have EXACTLY THE SAME NUMBER OF COLORS:16,777,216 in 8bit per color, 3 colors per pixel mode. Not one more or less different. They are distributed differently. For the different distribution you gotta give up something---yep, blues, dark reds, true black. Much of the colors making skin tone-----Eismann has a discussion, badly watered-down in Restoration and Retouching book.
Best bets: Shoot sRGB for web and snapshot printing done at commercial outlets like Walgreens, WalMart, Costco, etc. Especially useful is Dry-Creek Photo--they have some sRGB based printer profiles for some of these outlets.
Shoot aRGB if you plan on publication in a magazine and don't have good skills with PhotoShop.
BEST BEST: Shoot RAW. Keep the RAW files; customize output files for specific tasks (as above). If printing at home (or with any printer under your direct control) use the "largest' evenly distributed well behaved, non-experimental color space available: ProPhoto or ColorMatch. Acquire something Like MonacoEZColor with Optrix: calibrate and profile the monitor, profile the printer, profile scanners. For absolutely accurate color use something like Tom Fors' ACR calibrator and Macbeth Colorchecker or PictoGraphs inCamera and one or more standard targetss to profile camera output on location. Workflow: Raw from camera, convert RAW to editor native format (in Photoshop its PSD) with Prophoto/Colorsync colorspace and file tagging, allow Photoshop to manage color when printing (turn all print driver enhancements OFF) and tell it the name of the printer profile just calculated.
The collective 'Human vision' colorspace is assumed, by definition to be the largest-essentially a large continuous volume when expressed in tristimulus notations, i.e. RGB, Lab, HUV etc. Your cameras colorspace is nearly this large! Next might be the computer itself especially when using high bit counts per color, but the computer has no native output device.
Monitors have relatively small native colorspaces--and it's a range or distribution not a count of crayons. Exactly one expensive monitor makes a claim for full sRGB capability range. All others are less capable. sRGB has a smaller range, a smaller subset of human vision than aRGB (which is also much smaller than the human range; sRGB about 25-35% and aRGB about 40-45% of human vision). ProPhoto and ColorSync come surprisingly close to the human range-estimated about 90-95%. See note at end.
Printers are a poor last place finisher. No expansion of the range of color has occurred in --well, never (we still use the same paper and the same color dyes and pigments as originally discovered and developed! All expansion is in the ability of the printer to spray finer and finer droplets, thereby filling-in and making smoother gradients. It's a chemistry thing, the major ink improvement of the digital age is the addition of something called soap.
Note at end. Computer color models, and monitor and printer models by extension are discrete;they use only a small set of positive integer numbers. Each color represented is a single point. The human color model is continuous-a volume in 3-space---i.e. real numbers-infinite fractions. There are infinite human shades of color between every pair of computer model color point pairs. The percentages given in some fashion represent the 'volume occupied' of the smaller color space extracted from the human model. It's very difficult to compare small finite counts to a universe!
Beyond this it gets very complicated--a place where no photographer might wish to go...
Wildly over-simplified and terribly dated! Gotta love the internet.
You shoot RAW; no matter weather dSLR or P&S; HAAA. Biggest difference wrt outputs: in a dSLR you can actually retrieve the RAW data--it ain't an image! Not even 1/2 an image--it's a third of the photo!
AdobeRGB is proprietary and made for POSTSCRIPT devices and CERTAIN PRE-PRESS machines. You may get some richer colors as that article details---then again you may get absolutely nothing! OR you may get a terrible mess---we could discuss statistics, but...
AdobeRGB and sRGB have EXACTLY THE SAME NUMBER OF COLORS:16,777,216 in 8bit per color, 3 colors per pixel mode. Not one more or less different. They are distributed differently. For the different distribution you gotta give up something---yep, blues, dark reds, true black. Much of the colors making skin tone-----Eismann has a discussion, badly watered-down in Restoration and Retouching book.
Best bets: Shoot sRGB for web and snapshot printing done at commercial outlets like Walgreens, WalMart, Costco, etc. Especially useful is Dry-Creek Photo--they have some sRGB based printer profiles for some of these outlets.
Shoot aRGB if you plan on publication in a magazine and don't have good skills with PhotoShop.
BEST BEST: Shoot RAW. Keep the RAW files; customize output files for specific tasks (as above). If printing at home (or with any printer under your direct control) use the "largest' evenly distributed well behaved, non-experimental color space available: ProPhoto or ColorMatch. Acquire something Like MonacoEZColor with Optrix: calibrate and profile the monitor, profile the printer, profile scanners. For absolutely accurate color use something like Tom Fors' ACR calibrator and Macbeth Colorchecker or PictoGraphs inCamera and one or more standard targetss to profile camera output on location. Workflow: Raw from camera, convert RAW to editor native format (in Photoshop its PSD) with Prophoto/Colorsync colorspace and file tagging, allow Photoshop to manage color when printing (turn all print driver enhancements OFF) and tell it the name of the printer profile just calculated.
The collective 'Human vision' colorspace is assumed, by definition to be the largest-essentially a large continuous volume when expressed in tristimulus notations, i.e. RGB, Lab, HUV etc. Your cameras colorspace is nearly this large! Next might be the computer itself especially when using high bit counts per color, but the computer has no native output device.
Monitors have relatively small native colorspaces--and it's a range or distribution not a count of crayons. Exactly one expensive monitor makes a claim for full sRGB capability range. All others are less capable. sRGB has a smaller range, a smaller subset of human vision than aRGB (which is also much smaller than the human range; sRGB about 25-35% and aRGB about 40-45% of human vision). ProPhoto and ColorSync come surprisingly close to the human range-estimated about 90-95%. See note at end.
Printers are a poor last place finisher. No expansion of the range of color has occurred in --well, never (we still use the same paper and the same color dyes and pigments as originally discovered and developed! All expansion is in the ability of the printer to spray finer and finer droplets, thereby filling-in and making smoother gradients. It's a chemistry thing, the major ink improvement of the digital age is the addition of something called soap.
Note at end. Computer color models, and monitor and printer models by extension are discrete;they use only a small set of positive integer numbers. Each color represented is a single point. The human color model is continuous-a volume in 3-space---i.e. real numbers-infinite fractions. There are infinite human shades of color between every pair of computer model color point pairs. The percentages given in some fashion represent the 'volume occupied' of the smaller color space extracted from the human model. It's very difficult to compare small finite counts to a universe!
Beyond this it gets very complicated--a place where no photographer might wish to go...
boy I think something is wrong with me.... I got most of this! at least the important parts
my one question is..... since sRGB2.1 is a smaller color space then Adobe RGB 1998, with the most popular intent being relative colometric (sp?) wouldn't more colors outside the garmet be converted to the closest one then Adobe RGB 1998 (being that the color garment is larger)
so to sum that up, wouldn't the colors be more accuarate in a larger color space then the sRGB2.1?
I knew that Prophoto was close to the human eye, but I did not know color match was. good to know.
wouldn't the colors be more accuarate in a larger color space then the sRGB2.1?
randy
Yes but its horses for courses..
There is no point having more colours than you can use and most monitors cannot display the full gamut of Adobe RGB. Such images need to be mapped anyway to a more suitable colourspace for displaying on screens.
Since a lot of apps/browsers do not support colour management, it is useful to use sRGB for images destined for the web or on-screen display. Most decent printers can use a larger gamut than this so Adobe RGB is a reasonable choice.
Personally, I always use RAW.
I output to sRGB for screen display/web.
For prints I output to Adobe RGB. It's not perfect, as editing using a monitor means that you can't actually see some of the colours that you hope to print, but it works for me.
For prints I output to Adobe RGB. It's not perfect, as editing using a monitor means that you can't actually see some of the colours that you hope to print, but it works for me.
dave
that's exactly what I have been doing for a few years now
At the moment I would ask that you 'stand by'; I got some household chores to finish and then I'll show you some interesting stuff. You're the first person in a long time who seems to appreciate the situation and to have formed some 'considered' questions--very refreshing that post of yours about rendering intent. The answer is complex and I'm working on it for you--need a few hours more. I also have some corrections to my first post. I may bump this whole thread to new thread if no one seriously objects--it's an important subject.
At the moment I would ask that you 'stand by'; I got some household chores to finish and then I'll show you some interesting stuff. You're the first person in a long time who seems to appreciate the situation and to have formed some 'considered' questions--very refreshing that post of yours about rendering intent. The answer is complex and I'm working on it for you--need a few hours more. I also have some corrections to my first post. I may bump this whole thread to new thread if no one seriously objects--it's an important subject.