Hot,
I’m going to assume the color management experts (many of whom hang out at the Color Management forum <
http://www.adobeforums.com/cgi-bin/webx?14@@.eea5b31>, where you might also post your questions) will jump in to help you.
In the meantime, you might read this article by Bruce Fraser from the now-defunct Adobe Magazine:
<
http://www.adobe.com/products/adobemag/archive/pdfs/98auhtbf .pdf>
It’s from the Photoshop 5 era, but PS 5 was the first version to implement the color architecture scheme we have today, and the information is still quite valid. It certainly helped me understand the basic concepts, and maybe it’ll help you.
I will answer your question #4. CM as it exists today isn’t perfect, but both my photography and prepress work would be MUCH more primitive and difficult without it.
=-= Harron =-=
Lot to be confused about , <G>
1) Once you have a calibrated display, the working space profile interprets the RGB numbers of the image and displays them accurately on the display. The profile provides a reference for how to interpret those RGB numbers. Without a profile, the display would be based on the famous "monitor RGB" unless the application had a different default.
2) "Assigning" a profile to an image DOES NOT alter the image numbers at all. It simply tags the image with a reference color space so that the image can be correctly interpretted by a color managed application.
3) There is no conversion associated with question 2 and the "Assign" function . Clearly however , IF you do use the "convert to profile" several times to different colorspaces ( especially those with wildly differing gamuts ) you will do a lot of damage to the color of the image, compared with the original image .
4) Works just fine as long as you have accurate profiles, and remember the limitations of all the many devices you have to deal with potentially. I have no problems with color management printing in my basement. <G>
5) I’m confused on this one. I’m confused about why you want to use your monitor’s colorspace for either softproofing, or as a working space. The Photoshop levels commands don’t have anything to do with just how it looks on your particular monitor, they affect the image itself and how the image should look on every calibrated monitor.
Let’s say that you are working in the AdobeRGB colorspace. What you see on your display is the image accurately displayed according to the AdobeRGB space ( thanks to the wonderfully accurate monitor profile you have built <G> ) .
🙂
Brent
H_D,
If it helps you in any way here is my short "idiot’s guide to CM" <g>
The numbers in an image file do not represent specific colours. For example, 100R, 0G, 0B just means "make the brightest, most saturated red you can". It doesn’t mean a specific SHADE of red, and the red that will result will depend upon the capabilities of the device the data is being sent to (usually either a monitor or printer of some sort).
Only when coupled with an ICC profile that describes the ‘colour space’ do those numbers represent a specific shade.
There is data in a file. That data doesn’t represent specific colours UNTIL the colour space is stated (embedded profile or assign profile).
The working profile sets up the colour space you are working in, and the embedded profile allows the CM engine to convert the file data values to your working space values so those values still represent the same colour as originally indicated by the file data and embedded profile.
The monitor profile alters that data from your working space on the fly to allow the monitor to display the colours represented by the data within the working space profile.
When you print, the output profile alters the data to allow the printer to reproduce the colours represented by the data and the image profile.
The monitor profile effectively drops out of the equation when you print.
Provided the ‘translation’ from working space to monitor is correct AND the translation from working space to output device is correct, the print will match the monitor.
However, only if the translation from embedded profile to working space is also correct will the monitor and print also match the original file intentions.
For a fuller explanation, spend some 15 minutes or so over at
http://www.computer-darkroom.com
That’s an excellent précis, Len.
=-= Harron =-=
Len
Can you also explain the ASSIGN and CONVERT profile functions in photoshop (yes I read the manual)… but im puzzled in what they do physically etc..etc..
Len,
A. So e.g. I scan a file with a scanner and CHOOSE not to attach the scanner colour profile when I save it via the scanners scanning software. Now I load it in photoshop. What does photshop do with this untagged scan ?, and what translations in colours take place ?, and are they just visual or RGB pixel values altered
B. If I however scan and include the scanners profile via the scanners software on saving a scan; NOw if I load it in photoshop and the RGB working space is of my actual monitor, then…what is happening ? Is
C. What is the difference in the PIXELS rgb values in a file output with a tagged colour space at scanning time and without a colour space at scanning time.
D. IF I scan a file and save it without any attached scanner profile, but in photshop ASSIGN a profile, the original scanners profile, what have I effectively done ?, is it the same as saving it with a profile at scanning time or are the results different ?.
Brent,
yes read the computer darkroom stuff, and it was similar to the photshop manual so it did not help much.
When you ASSIGN a profile, no changes are made to the image data whatsoever. The numbers in the file remain exactly the same, but you ARE telling Photoshop (and other colour-managed applications) what those numbers represent in terms of real colours.
This is synonymous with taking a page written in a foreign language, and adding a note saying "This page is written in Gujarati "
Now if you assign a profile that does NOT match the colour space in which that file was created, that profile will ‘lie’ about what colours those unaltered numbers represent. So if you ASSIGN an incorrect profile to a scanned image, the appearance on screen and when printed will not match the original, which is synonymous with adding the note saying "This page is written in Gujarati " when in fact it is written in Hindi or Urdu, and expecting a machine translation of the page to be correct.
When you CONVERT to a profile, the numbers in the file are changed, because in this case you need DIFFERENT NUMBERS to represent the SAME COLOURS under the target profile as they did under the original profile. Synonymous with translating from one language to another.
To ASSIGN an accurate profile, you need to know the current colour space of the file.
To CONVERT to a profile, the file needs an existing profile – otherwise the application does not know to convert from what space.
This is synonymous with being given a page written in a foreign language, but not knowing WHICH foreign language! How can you hope to do an accurate translation under those circumstances?
Len
Is Conversion a lossy (pixel values) process ?, and in which circumstances ?. and is using the CONVERT TO PROFILE command the only way that data is modified ?, Also when you load a file with a different profile, photoshop asks to CONVERT to the current working space ?
Am i correct thinking, I should scan and save with profile or ASSIGN the scanner profile manually. THEN, work with the file in photoshop with the VIEW PROOF option ON and set to the MONITOR or WINDOWS (average windows monitor), in order to Colour correct , adjust corectly, and view the image as it is intended on the monitor. Also if I do this is Photoshop DYMNAICALLY convering the original Pixel values to the monitor profile and showing them ‘on the fly’ or have any modifications been made ?.
What should I do with graphics that I come across that have no profile (off web pages etc) and I want to work with them on the monitor ?, Should I attach my CUSTOM monitor profile or WINDOWS average monitor profile (sRGB) then, ALSO preform colour correction to make sure the image looks right, good ?. Or not attach anything and JUST work in PROOF view with CUSTOM monitor or WINDOWS monitor selected ?.
H_D,
Is Conversion a lossy (pixel values) process ?, <<
When converting to a smaller colour space, yes, but not otherwise.
is using the CONVERT TO PROFILE command the only way that data is modified
?<<
As far as changing the data in memory, Yes, although values will be converted on_the_fly for the screen representation, when viewing proof colours or when printing with a target profile.
when you load a file with a different profile, photoshop asks to CONVERT
to the current working space ?<<
It gives that option, yes. Sorry, but I don’t understand the implied question there.
I should scan and save with profile or ASSIGN the scanner profile
manually. <<
Correct.
THEN, work with the file in photoshop with the VIEW PROOF option ON and
set to the MONITOR or WINDOWS (average windows monitor), in order to Colour correct , adjust corectly, and view the image as it is intended on the monitor<<
NO! If you use Proof colours, the proof profile should be set to the intended output device. If that is a monitor, and you are not embedding a profile when saving, THEN monitor RGB would be appropriate. If intended final use was printing, then the profile for the output device would be appropriate.
However, for working with your images, it is more appropriate to do that in your working space (which is designed to be a linear editing space), and then soft-proof using Proof colours.
What should I do with graphics …that have no profile (off web pages etc)
<<
Web images essentially will be sRGB. Either convert to your editing space or, if re-use on the web is intended, keep it in sRGB by assigning the sRGB profile.
What is diffrence in sRGB and e.g. Generic 2.2 Gamma Monitor ?.
So I have a file with no profile, so I attach the sRGB . It is for the web. Am I correct in saying that I PROOF VIEW option as Windows Monitor ?, and also make any colour adjustments acording to THIS Proof view ?.
Continuing from above, I have the final image ready for the web. Should I save it as (via SAVE FOR WEB) is (sRGB) or use CONVERT TO PROFILE to convert from sRGB to Windows Monitor ?, or Generic 2.2 Gamma Monitor (and then SAVE FOR WEB) ?? which ?, why ?
If I do PROOF VIEW as Windows Monitor (paragraph 1 above), then Would I be one step better in Proof viewing in my monitors custom Profile (created via adobe Gamma, or manufacturer), and making colour adjustments in this Proof view,……But then FINALY CONVERT TO PROFILE of Windows Monitor or Generic 2.2 Gamma Monitor ?.
Why Would I want to (In what circumstances) CONVERT TO PROFILE (and save the file?) of a printer e.g. my inkJet Printer’s profile if when the file is printed it is ACTUALLY converted automatically on the fly (by the printer driver ?)
So I have a file with no profile, so I attach the sRGB . It is for the web. Am I correct in saying that I PROOF VIEW option as Windows Monitor ?, and also make any colour adjustments acording to THIS Proof view ?.
You _assign_ sRGB. You can proof with Windows monitor if you want, but remember most screens used for browsing aren’t calibrated, so it’s a crapshoot.
Continuing from above, I have the final image ready for the web…
for the web I would keep the file in sRGB
Why Would I want to (In what circumstances) CONVERT TO PROFILE (and save the file?) of a printer e.g. my inkJet Printer’s profile if when the file is printed it is ACTUALLY converted automatically on the fly (by the printer driver ?)
Some print processes need the file to be converted beforehand
H_D,
I have the final image ready for the web. ….<
The web is not a colour managed environment. There is no point in saving files destined for web use with an embedded profile, and Save for Web strips all meta data (unless you tell it to do otherwise) giving you the smallest possible file-size.
The vast majority of web users have uncalibrated and badly adjusted monitors. Consequently there is absolutely no way that you can control or prophesy what any image will look like to any particular viewer. The best you can do is to ensure that it looks as good as possible in the browser on YOUR machine – and Save for Web will give you that view, as will viewing Proof colours with a proof profile of Monitor RGB
Why Would I want to (In what circumstances) CONVERT TO PROFILE (and save
the file?) of a printer e.g. my inkJet Printer’s profile <<
With your workflow as outlined, you wouldn’t. You would keep the file in your working RGB colour space. However, if you were having your image output by an outside source on, say, a Fujix Frontier, then converting to that output device profile would be appropriate.
Len
Can you also answer the questions that you have not yet in my last post. e.g Generic 2.2 Gamma Monitor related etc…
Fujix Fontier / Outside source…., why ? convert for this, why is it not possible/sensible to send the file to the printer people and thier equipment uses thier printers profile to do ‘on the fly’ conversion at printing time ?, just like my inkjet would ??
H_D,
Fujix Fontier / Outside source…., why ? convert for this<<
You cannot be sure they will do anything but dump the file to the machine. If it’s in the correct colour space, it is harder to muck up.
As to the sRGB/Generic monitor, I’ve never compared profile specifications (and don’t intend to start now!<g>)
Yes, this is safe for offset printing – convert every-
thing to CMYK.
No need for embedding any profile (once the designer
and the printer agreed on a process).
But for printing on large format inkjets it´s different. The RIP should offer:
Input profile for CMYK Vector (including text)
Input profile for CMYK Raster
Input profile for RGB Vector
Input profile for RGB Raster
Output profile for machine
-Rendering intent for Vector
-Rendering intent for Raster
An esoteric extension:
Input profile Lab.
Best regards –Gernot Hoffmann
Len
NO! If you use Proof colours, the proof profile should be set to the intended output device. If that is a monitor, and you are not >embedding a profile when saving, THEN monitor RGB would be appropriate. If >intended final use was printing, then the profile for the output device would >be appropriate.
"and you are not embedding a profile when saving…>", and if I was embedding a progile when saving (and which profile did you have in mind and why ?)
When would I/Should I, find myself embedding my Monitors custom/manufacturer profile in an image (what situations).
OK, so I am going to preform colours adjustments/brightness contrast etc… , Do I do it so the image looks right on my RGB working space , i.e. the monitor’s working space, or the Proof working space if it is not the MY monitor or e.g. standard WIndows monitor or if it is a printer.
Len,
When I assign my scanner profile to a Scanned image in photoshop, the colours on the screen of the scan change , like it has been lightened,…….can you explain what is happing and why (I am taking it that when I save the file the RGB values are the same as before assignmnet). Also which representation is the representation of the actual paper print that I scanned, the modified display when a scanner profile is attached, or when proof viewed in MONITOR profile ?
Now carrying on from above, if I PROOF>VIEW to MONITOR profile, I get an image that looks like BEFORE the assignment. I take this is Photoshop taking the RGB values and also the scanner profile, then converting them to some intermediate space and then into the MONITOR colour space, all on the fly ?.
Fujix Fontier / Outside source…., why ? convert for this, why is it not possible/sensible to send the file to the printer people and thier equipment uses thier printers profile to do ‘on the fly’ conversion at printing time ?, just like my inkjet would ??
Because Fuji Frontiers operate in an entirely different way than your inkjet, for one, their primary use is as a minilab, to print from film or digital cameras, digital file outputs is secondary (at least for most people who own them). Second, the software that comes with it assumes your file is in sRGB mode, period, it does not do any profile to profile conversion.
When I assign my scanner profile to a Scanned image in photoshop, the colours on the screen of the scan change , like it has been lightened
This is because when assigning a new profile, you are changing the interpretation of the colour numbers, thus changing what the same numbers look like.
Also which representation is the representation of the actual paper print that I scanned. When I assign my scanner profile to a Scanned image in photoshop, the colours on the screen of the scan change , like it has been lightened
The accurate representation of the colours is when you have assigned the scanner profile.
When would I/Should I, find myself embedding my Monitors custom/manufacturer profile in an image (what situations).
Never.
H_D,
can you explain what is happing and why <<
Yes. Once you have assigned the scanner profile, the numbers in the file will represent specific colours. Those colours will then be displayed on your screen. Before you assign the profile, Photoshop is assuming what colours those numbers represent.
if I PROOF>VIEW to MONITOR profile, I get an image that looks like BEFORE
the assignment. <<
Proving that before assignment, Photoshop was assuming a colour space somewhere near to your monitor profile, which suggests that your RGB working space is set to sRGB (or God forbid, Monitor RGB!)
sRGB (or God forbid, Monitor RGB!)
What bad about Monitor RGB ? (its display is quite similar to my custom monitor profile). what is sRGB , and what is it used for used for ?.
What other RGB working spaces exist ?…
I scanned this document with huge patches of BLaCK areas. On assignment of the scanner profilw the black area’s brighten considerably and the image looks awful. It does not look like the piece of paper that I scanned. What do I make of this behaviour ?? (Looks OK when I proof view in Monitor RGB/Custom Monitor profile)
What other RGB working spaces exist ?…
All display profiles, most scanner profiles and some output profiles are RGB. In addition, device independant profiles are usually RGB profiles.
Some device independant profiles are sRGB, AdobeRGB, ProPhotoRGB
BTW, how about letting us know what your colour setup is (exactly)? We may be better able to help you if we knew.
Rene
Agfa scanner, and just a IBM Monitor (both with profiles). I am not intending to print. But I did want to know overall colour management.
Its become very clear now…..just a few more things to think ago.
HD,
You just told us what equipment you have, what I was asking for was you colour settings in Photoshop
H_D,
What bad about Monitor RGB ? <<
It is NOT an editing space profile. Editing (or working) space profiles need to be as linear as possible. Monitor profiles are designed to take account of Monitor non-linearities.
what is sRGB , and what is it used for used for ?.<<
sRGB is a ‘standard’ colour space. See
http://www.w3.org/Graphics/Color/sRGB.html for details.
You may also wish to look at
http://www.cie.co.at/cie/
"sRGB is a standard promoted primarily by Hewlett-Packard and Microsoft. It reflects the standards for HDTV broadcast (Gamma=2.2, Primaries=HDTV, White Point=6500K).
One of the reasons to use sRGB is that Hewlett-Packard is promoting a workflow in which hardware devices such as scanners, non-PostScript printers, and Web browsers will be optimized for RGB data in the sRGB space. If you are using such devices, sRGB will provide the simplest workflow.
sRGB reflects the characteristics of the average PC monitor. If you are producing graphics to be viewed on the Web, sRGB will reflect what most viewers see. The downside to sRGB is that it has a limited color gamut and cannot represent as many colors as other color spaces. It is not a good choice for professional prepress users since too much of the CMYK gamut lies outside of it. "
This is one of my many clipboard references, I believe stolen from an Adobe Techdoc.
🙂
Brent
Hot,
What other RGB working spaces exist ?…
Did you read the Bruce Fraser article I pointed you to in post #1 of this discussion? It provides an excellent summary of pre-defined ("standard") RGB workspaces, including sRGB, AppleRGB, CIE RGB, ColorMatch RGB, NTSC, PAL/SECAM, SMPTE-240M, SMPTE-C, and Wide Gamut RGB. It even explains how you can define your own custom RGB workspace and why you might want to do so.
=-= Harron =-=
Len
OK, so I am going to preform colours adjustments/brightness contrast etc… , Do I do it so the image looks right on my RGB working space or when PROOF VIEWed using the Windows RGB or Monitor RGB, or Printer working space if intened for printout ??
I scanned this document with huge patches of BLaCK areas. On assignment of the scanner profile the black area’s brighten considerably and the image looks awful. It does not look like the piece of paper that I scanned. What do I make of this behaviour ?? (Looks OK when I proof view in Monitor RGB/Custom Monitor profile)
sRGB, When I open a file that has no colour space attached…eg. off the web, or scanned from a scanner but no profile, and the RGB working space is set to sRGB, what is happening to the image data ?. Also I want as much colours as possible (ALL if possible), no loss of data, so since sRGB has a gamut range, am I not loosing colours or the pixels values beign modified ?. What RGB colour space will preserve all the colours in the file without modificaiton, if you know what I mean ?….
H_D,
On assignment of the scanner profile the black area’s brighten
considerably and the image looks awful. It does not look like the piece of paper that I scanned. What do I make of this behaviour ?? <<
Having assigned a profile you should then convert it to your working profile.
what is happening to the image data ?<<
Absolutely nothing until you convert the image to a particular profile.
What RGB colour space will preserve all the colours in the file without
modificaiton, if you know what I mean ?….<<
That’s the ‘wrong’ question but I’ll answer it anyway
Wide Gamut RGB but you really don’t want to go there! The result is that you can make changes THAT YOU CANNOT SEE ON SCREEN and therefore have absolutely no idea what changes you have made!
Whenever you convert to a different profile, you will modify the data, but that doesn’t necessarily mean you are losing data.
For those who want a really big drink of water, may I suggest the book, "Real World Color Management" by Bruce Fraser, Chris Murphy and Fred Bunting. It covers the whole subject of color management thoroughly. A big read but worth it.
Erv
Hot_Denim,
To convince myself about how all this works, I downloaded a great picture from Phil Askey (www.dpreview.com) of his wife, Joanna, holding a GretaMacbeth color chart. It’s a 1.7MB JPG and does NOT have an embedded color profile. If you want it, it’s on my website:
<
http://www.stanford.edu/~mauryg/pcc.jpg>
I made several duplicate copies without changing the color profile. Then I opened each copy,converted it to a different color profile (sRGB, AdobeRGB, MonitorRGB) and resaved the file with a new name. Finally I opened all the copies and tiled them in the PS window to compare them side by side with different working space settings.
To all you other members of the forum….THANK YOU for some very interesting and helpful explanations.
Maury
You can get an Electronic version (photographs are a waste of time, especially when they’ve been processed) of the Gretag Chart at the following link:
<
http://www.drycreekphoto.com/Learn/monitor_calibration.htm>
You want the the CIE Lab version – then you can convert to whatever colour space you like.
Len
Again,
OK, so I am going to preform colours adjustments/brightness contrast etc… , Do I do it so the image looks right on my RGB working space or when PROOF VIEWed using the Windows RGB or Monitor RGB, or Printer working space if intened for printout ??
Having assigned a profile you should then convert it to your working profile.
Why CONVERT why not just PROOF VIEW in your working space when you feel like it ?, converting would mean possible data degradation ?
Again, I meant can you explain the visual changes …i.e. whats happening ?? :
"I scanned this document with huge patches of BLaCK areas. On assignment of the scanner profile the black area’s brighten considerably and the image looks awful. It does not look like the piece of paper that I scanned. What do I make of this behaviour ?? (Looks OK when I proof view in Monitor RGB/Custom Monitor profile"
Wide gamut RGB; But is easily possible to make changes to a file (via color correction) that the eye cannot spot on screen even standard RGB working spaces….and especially in 16-bit mode
Out of all the RGB working spaces, sRGB, adove(RGB) etc… can you tell me the good and the bad of each so that I can know why I choose one over another.
H-D,
Why CONVERT why not just PROOF VIEW <<
Because you want to edit within as linear editing space as possible.
But is easily possible to make changes to a file<<
With Wide Gamut as a working space you could move levels sliders a considerable distance without seeing any changes on screen. The same applies to other controls. If you want to wreck a few images, try it.
can you tell me the good and the bad of each so that I can know why I
choose one over another. <<
Depends entirely on intended use. You have already been pointed towards Bruce Fraser’s book on colour management – read that if you want to get to grips with the nitty-gritty
Why CONVERT why not just PROOF VIEW in your working space when you feel like it ?, converting would mean possible data degradation ?
Because if you do not convert from the scanner profile to your working space, you are limiting yourself to the colours that the scanner can "see". Device independant profiles usually have a wider gammut (with a better distrivbbution too)
The scanner converts colours into numbers. The numbers represent the colours that the scanner sees, not the ones in the image. The scanner profile describes how the scanner sees colour, and a comparison is made with a known standard with a much wider range than the scanner, and the numbers are converted.
The numbers are changed in a look up table before they are sent to the monitor. They are then displayed using a monitor profile which describes how the monitor displays colour.
The working space refers to the colour range used for editing images in. It is useful to have a wide range of colours to edit it as the prepress process reduces the range considereably. Many device spaces are not based on pure primaries and so any neutral colours will have colour casts. A device indepenant working space is purer and yeilds better results.
You can have say 3 documents open in PS at the same time and each one can be in it’s own colour space. You can chose to stay with the document space or convert it to the working space.
The numbers in the file are changed, but if profiles are not used the numbers do not have any colour meaning. So they are changed in order to maintain consistant colour independant of the devices used to scan, display or print them.
Len,
Should I find myself in need of new knowledge of complexity equal to that of color management, and should I further find myself in need of your help after having exhausted other avenues of study, I can only hope that you will be there to assist me as you have so many others here. And after you have guided me out of the darkness, I shall do the unthinkable: thank you publicly.
Heck… I might even fly across the pond and buy you a pint of your favorite.
=-= Harron =-=
Well, thank you, Harron.
It’s good to know that this topic has been useful to you.
Len
Because you want to edit within as linear editing space as possible
I know that monitors are not 100% linear, i.e. the gamma adjustment. Tell me a bit about others e.g.. scanners (input) and printers (output) so I can get the whole picture
You have already been pointed towards Bruce Fraser’s book on colour management – read that if you want to get to grips with the nitty-gritty
Cant afford a book for this question. Are there any web resources you know that will answer this question if you wish not to.
color profiles affect things like default foreground and background color picker tool thingy.
any kind of pixel mask.
defines a source or destination color space.
and pretty much controls the behavior of the file.
H_D,
Tell me a bit about others e.g.. scanners (input) and printers (output) so
I can get the whole picture<<
Every device will have it’s own characteristic curve.
In the case of a scanner, CCD’s do not have a linear relationship between input intensity and output voltage, and neither do they have a linear response over the whole visible spectrum. The same intensity at different frequencies will produce different output voltages, and to make matters worse the light source will not be linear either.
With printers, applying the same signal to the cyan printhead and the yellow printhead will produce distinctly different reflectance values – again it is not linear. Even when considering a single printhead, the visual difference (and difference in measured reflectance values) between a 2% ink coverage and a 7% ink coverage will be vastly different than that between a 50% and a 55% ink coverage or between a 95% and 100% coverage.
Len
Did you do a lot of reading about colour management to get to know it like you do ?. I consider myself kinda good at ‘reading the manual’ but photoshop puzzeled me in this area (read Photoshop manual), did it puzzle you tooo ?, anyone else ?……[i feel like a stranger]
H_D,
Did you do a lot of reading about colour management to get to know it like
you do ?. <<
Not that much at all. Most of it is a mélange of standard colour theory and logic.
LEN
With Wide Gamut as a working space you could move levels sliders a considerable distance without seeing any changes on screen. The same applies to other controls. If you want to wreck a few images, try it.
I did try on my own image with the Levels sliders, but they were working as good/natural/sensitive as ever.
Can you point me to a good example image on the web and give me step by step details of how to experience this……
H_D,
they were working as good/natural/sensitive as ever.<<
What you now have is a screen image that no longer represents the data in the file in any meaningful way.
So I CONVERT to sRGB, or ASSIGN a sRGB to images from an unknown source.
But when I try on GREYSCALE images to ASSIGN, I am not given sRGB as a choice, what should I do (ASSIGN) and also WHY ?…
If I scan A image (Black and white photo in 24-bit scanner mode), then convert to greyscale the Present SCANNER profile is lost, so if I them assign a GREYSCALE profile is everything still consistant etc..or does this not make sense ?>
H_D,
Think about it. Greyscale really doesn’t need to worry about a COLOUR space, now does it? You only have ONE channel. All that can alter are the levels.
OK, I understand why im not getting the sRGB profile for selection , But im getting some ones titled INK GAIN and GAMMA, I guess from knowing WINDOWS monitors are gamma 2.2, I should choose this.
Is my deduction correct for my problem or is the answer different ?
H_D,
‘Greyscale Profiles’ are essentially one channel of a CMYK profile.
CMYK profiles carry information for press dot-gain (that’s the ratio of the size of the dot on paper and the size of the dot in the file, and is due to a variety of effects such as image-spread at the film stage and when making the plates and even more so due to ink-spread on the paper)
LEN :
So what do I use ?, the GRAYSCALE equivelent of sRGB ?
LEN,
I scan an image, In photoshop I open it up, the colours look good. I then ASSIGNa profile which I believe to be the scanner profile due to similar name, when I do this the colours lighten considrable and it does not represent the paper orignal as much as the unASSIGNed version, you have told me WHY previously. If I PROOF VIEW in Monitor/Windows view the image looks good again, like before the ASSIGNment.
But then I CONVERT to sRGB, and the image looks near identical to the previous scanner ASSIGNed profile version. I.e. no chaange. NOW when I proof view in MONIOR/WINDOWS view there is NO change in the displayed image, why not now ?………., [I am confused at this point onwards]
I.e. I have converteed it to a WORKING sRGB space profile (because I do not want to edit in the n on-linear smaller scanner gamut), but it looks worse now, What have I done wrong, and what am I to do extra ?… I have a bad looking image now.
HD
Try and think in the simplest terms.
I scan an image, In photoshop I open it up, the colours look good. I then ASSIGNa profile which I believe to be the scanner profile due to similar name, when I do this the colours lighten considrable and it does not represent the paper orignal as much as the unASSIGNed version,
What does that tell ya? The canned profile may be crap.
It tells you that the scan that you just did, (the color space in which the scanner is seeing, ((IS)) not the same as the profile that you are assigning it.
It’s all about pairing the image with the correct color space. It’s that simple.
If the profile that you assign to the image is not describing the capture, "scanner or digital camera", than image color, will not represent the original.
ie. art or camera scene.
It’s all about preserving or "copying" the original with color mgmt.
Color mgmt. tries to preserve the original as best as possible during color space conversions.
Look at it this way…..
When you scan an image, it’s going thought CCD chips, with color look up tables and all that junk out to an RGB file. How and what the process is, who knows, but when you profile that scanner or camera, you are giving meaning to the numbers you captured. Once you describe the numbers correctly with a profile, you then "preserve the color appearance" or, you want to, when you do color space conversions. aka. image reproduction.
Editing in scanner space is usually a really bad idea because it’s more times than not, a non-linear color space. ie. The values or meaning that describes the scanner’s numbers are all over the place through the entire tonal and saturation range. aka. color space.
The correct workflow is to create a custom profile, use a canned scanner profile, or assign a working space that "BEST" describes the numbers of the scanner. Quality between these options can vary dramatically.
Again, what the scanner see’s and what the profile is saying what it is.
Even when you have everything lined up, things can go sideways and bring you to false conclusions so keep it simple / stupid when trying to track down the bless "it" trinity for your workflow.
mo
H_D,
In photoshop I open it up, the colours look good<<
And what do you have set for your CM policy? How is Photoshop set to handle missing profiles?
. I then ASSIGNa profile <<
As Mike says, the profile you are assigning is incorrect.
So what do I use ?, the GRAYSCALE equivelent …<<
For what end-use? If screen viewing Gamma 2.2, if for printing via separations then a suitable dot-gain for the press
And what do you have set for your CM policy? How is Photoshop set to handle missing profiles?
sRGB
As Mike says, the profile you are assigning is incorrect I have about 4 profiles that seem related to the sanner
Agfa : SnapScan 1212U :Agfa paper : Rx
Agfa : SnapScan 1212U_2 :Agfa paper : Rx
Agfa : VirtualMonitor
Agfa : VirtualMonitor22
I used "Agfa : SnapScan 1212U :Agfa paper : Rx";
If I try all of them one by one, what result should the correct one give me ???
The one that describes the scanner.
The one that describes the scanner.
I dont know , because I no longer have the scanner…. so I cannot scan and EMBED a profile via scanner software, then open in photoshop to see the name of the embedded profile…..
Thanks to the patience of the people who are answering the questions here (particualrly Len and Brent), I think I am gradually getting the hang of of colour management issues. I’m in the process of transitioning from Elements, which has extraordinarily basic colour management to CS (I’m currently using the trial version – I’m hoping that before it expires I can find the funds to pay for the license!)
Have I got this right: – I generally want to work in sRGB – most of my images are either for web or screen display, or either printed on a low end photoprinter or sent out for printing on a Frontier or other automated machine. My digital camera produces untagged images, which are close to, but not exactly sRGB as far as I can tell.
So what I should do is set the working space to sRGB, open the files and tick the box to assign working RGB (given that I can’t find any better profile to attach to the camera images).
If for some reason I wanted to work in Adobe RGB (say I’m sending an image out to someone and they want it in that colour space for print purposes): set working space to AdobeRGB, open the untagged image from the camera and then tick the check box that says "assign profile and convert to working colour space", and make sure that I assign the sRGB profile.
Susan S
H_D,
If you have your CMS policy set to convert files with missing or mis-matched profiles to sRGB, then there is no point in THEN assigning a profile – the file has been converted (though probably not accurately in the case of a missing profile), and the only relevant profile is sRGB.
IF you are going to ASSIGN a profile you MUST set your policy Preserve Embedded Profiles and to "Ask When Opening" for missing and mismatched profiles.
If you have your CMS policy set to convert files with missing or mis-matched profiles to sRGB, then there is no point in THEN assigning a profile – the
file has been converted (though probably not accurately in the case of a
missing profile), and the only relevant profile is sRGB.
I dont know as I saved the files several times on Photoshop over a long duration of time so it may be possible that it was. BUT i however remeber that WHEN assigning profile THE ONLY profile and NO COVERTion inbetween that the STATUS bar showed UNTAGGED RGB ?, So am I safe to assume due to no CONVERTion / ASSIGNmnet in between that I did infact not AUTO-convert the file ?. I.e. Does AUTO-convert assign a profile that shows up on the title bar.
Susan,
don’t confuse yourself by changing the working space to Adobe RGB from sRGB.
Use the assign profile or convert to profile features.
and no, you don’t need to use Adobe RGB for print.
What you do need is a full understanding between assign and convert and when to and when not to.
Susan,
No, assigning the AdobeRGB profile would merely be lying about the colour space of the file. You would need to use Image>Mode>Convert to profile
H_D,
Does AUTO-convert assign a profile that shows up on the title bar.<<
No, because until you save the file with an embedded profile the image is untagged.
Len
Sorry i meant STATUS BAR
Thanks Mike and Len – I’ll go away and read the material again, but I think I’ve got it clearer – is this OK?: Never assign a profile that doesn’t match the original colour space of the file, you need to convert if the colour space of the image and the working space are mismatched.
So with my untagged images from my camera (which are close to sRGB) I should assign the sRGB tag, but in the unlikely event that I wanted to use a different working space I need to convert them to that different working space. And if I understand Mike correctly, I am probably better off sticking to sRGB as a working space for my strictly amateur purposes anyway, rather than mucking around converting to AdobeRGB.
And I presume I am better off having my preferences set up so that CS asks what to do if an image is unprofiled or there is a profile mismatch, so that I can make the choice, rather than relying on having rememebered to set up correct automatic choices.
Thanks for the help.
susan S
Susan,
Yes, you have now got it spot-on!
H_D,
Yes, I know you meant the status bar. The answer I gave stands.
Never assign a profile that doesn’t match the original colour space of the file, you need to convert if the colour space of the image and the working space are mismatched.
well……
um…
yea…
the problem is that it’s assumed that the profile in the image is an accurate description of the files capture or current working space (converted state were as the image has been rendered correctly to preserve the color appearance), but there is no way to guarantee this, so……
for example.
If an image looks super saturated in flesh tones and reds, then you know that the image was captured with Apple RGB or sRGB, but assigned Adobe RGB.
The big picture , you’d assign a profile that gives you pleasing color, then convert to your output space.
It’s a pretty simple concept.
Practice may be another issue.
Mike – in an ideal world I would undoubtedly have an accurate colour profile provided by the camera manufacturer which would mean I would not have to play "guess the profile" to see what works best.
Another wrinkle on this. Up to now I’ve been shooting jpegs which come out in the colour space that the camera’s processor chooses for them (approximately equal to sRGB, but not quite – the reds are still a tiny bit oversaturated I think). But now I’ve got Adobe RAW converter to actually allow me to convert RAW images in a practical fashion it looks like I’ve got more precise control over the colour space that the image is created in. It seems to me that I should have better control over the image profile by shooting RAW and then choosing either sRGB or Adobe RGB in the plug in dialogue when I convert from RAW, (plus of course I can correct any white balance or exposure inaccuracies to a much greater exten tand get 16 bit non-lossy images)
Susan S
Yes,
ACR is the drug of choice if your camera is supported.
ACR is an interpetor for RAW files, which by the way, is in a flux color space until you render it out to a working space.
To get the best, accurate results, render the file from RAW to Prophoto 16 bit.
jpeg sRGB is a huge step down from ACR and RAW file processing.
