Need help: color spaces & color management workflow

[qveda]

Hi Colin, as you can observe the next image, in which I'm comparing a sRGB gamut with the profile of my iMac, which is not a Pro monitor, you can observe that the colour range of my monitor is wider than sRGB.

@ Qveda, I think that your workflow is ok, the only recommendation I could give you is to have a colour managed environment, that includes your printer/ink/paper set, otherwise you'll never get the perfect match between what you capture/see/print.

You also mentioned that you'll get Spyder to calibrate your monitor. I recommend you to get the Spyder3Studio. I know that it's more expensive, however, believe me that it's cheaper than getting first the one to calibrate your monitor and then the one to profile your printer.

Hi Daniel,
I only have an HP photojet printer at home - HP doesn't offer an ICC for it. I only use it for rough proofs, so not worth the extra expense of Spyder3Studio. But the Spyder3Pro, or the X-rite Display2 should be fine. If they ask for a Gamma preference during the calibration, would it matter whether I choose Gamma1.8 for my Mac vs 2.2 which is more "common" ? As long as you have target ICC's . would it make any difference ?

[Colin Southern]

The reason I personally dislike ie goes way back and is nothing to do with colour profiling. It's based on pure technical reasons, mainly security but it's rendering isn't as good as other alternatives, I often use none MS OSs and hence none ie capable machines so that is deciding factor but the biggest is it's super insecure.

I hear this quite often, but to be honest, if just doesn't make much sense to me.

In terms of operating systems, take a look at a graph that lists operating vulnerabilites in the first 90 days since release ...

http://blogs.csoonline.com/node/218

Something like about 5 for Vista - about 15 for XP - about 100 for Unbuntu and about 200 for Red Hat Linux ...

In a 6th month analysis of severe vulnerabilities only we get ...

http://blogs.csoonline.com/windows_v...ability_report

about 11 for Vista - about 25 for XP - about 30 for Ubuntu and about 73 for Red Hat Linux.

In terms of browsers (IE -v- Fire Fox) ...

[COLOR=#008000]blogs.technet.com/security/attachment/2594822.ashx


To late 2007 we have 193 for Fire Fox -v- 65 for Internet Explorer.

[Davey]

Hehehe yeah I know ff had more number wise last year but it gives the wrong impression. There are 2 important factors to consider a) how serious an exploit it is b) how long before it's patched. Running a slack box behind pro box router, with iptables etc etc and most apps mandatory access controlled and regular updates and patches means I don't suffer from ubuntu exploits usually nor do i suffer from xp or vistas ones. My home box is far from water tight (but it doesn't need to be, trade off between userbility and set up and basic security). I'll agree with many points but ie being half decent is one I'll will duel to the death on hehehehe.

As for hosting images with none sRGB profiles I'll definately agree it's not ideal thing to do since for other people it's better to put in sRGB, and if it's for yourself why upload them when you can view locally? Obviously people not realising the implications it has on others viewing their image is the main issue since many wont thing "washed out..........hmm profile issue?" but rather "hmmm washed out image, bad photo".

[Colin Southern]

Hehehe yeah I know ff had more number wise last year but it gives the wrong impression. There are 2 important factors to consider a) how serious an exploit it is b) how long before it's patched.

Agreed - unfortunately, Firefox also had far more serious ones (in fact Fire Fox has had more vulnerabilities than IE, Safari, and Opera combined - how secure does that make you feel?).

In terms of "time to patch" yes, I agree that that is important - but you also have to factor in the number of vulnerabilities as well ... if we multiply the number of vulnerabilites by the number of days people were exposed to these vulnerabilities then you still end up with significantly less exposure using IE

I'll agree with many points but ie being half decent is one I'll will duel to the death on hehehehe.

Well call me old - call me a fool (or call me an old fool) but I find that when I click on a link or type in a URL good old IE takes me to the site - and that's about all I ever ask of a browser. Bit like my cell phone - did you realise that you can actually use them to make and receive voice calls? (you know, where you can talk to the other party without having to use text?)

As for hosting images with none sRGB profiles I'll definately agree it's not ideal thing to do since for other people it's better to put in sRGB

In a more ideal world we'd shoot Adobe RGB as a minimum - all monitors would be able to display the full gamut - and all printers would be able to print it, but unfortunately, we're not there yet

[Davey]

I still find ie is more wide open since you cannot close holes and they remain open until ms patch them (can't hold my breath that long hehehe), granted ff has serious exploits along with others apps but if you stay on top it's much much more secure than ie and doesn't suffer the same exploits as ie. Incidentally many of the ff exploits I have seen affect ubuntu, a distro I hate more for the way it's promoted than technically (even though I never did like debian) because it's touted as secure and user friendly. Unix OSs are not user friendly, I don't care what anyone says put quite simply Linux and others are a right PITA to set up properly but once you do most holes are covered and ones that open you can close quickly.

I do own vista too btw and I quite like it I never recommend xp over vista because it's wide open and far less secure than vista and TBH on new machines vistas underperformance is severly overdone by bandwagon jumpers. Windows is known to be insecure though, it's the highest % target (so most viable to exploit and ie is the main browser of choice) and you're at the mercy of MS to patch you up. Granted however using ubuntu or similar with a million services running out the box and using stock everything so you don't need to learn how it works is not how you get security nor performance in my book and I've caused a right poo-storm suggesting such users are prob more secure on windows on unix boards (it doesn't go down well ).

As for the phone thing hahahaha don't get me started! Funny I was just discussing my mobile technical prowess ( or more specifically my total and complete lack off!) today with a friend who was demonstrating what their blackberry can do. Impressive but I can just about text and it takes me aeons and even then no-body can understand my texts , I prefer calls usually via wires when possible as i still don't trust low microwaves that close to my head. Call me old fashioned or ludite but it seems smart thing to do, I use my mobile phone to "phone" when I need to and not for extended periods, the overuse trend is a bit like driving to a store 3 minutes walk away to me. People laugh especially since I like technology so seems strange viewpoint for me to take but I'll be the one laughing when their over reliance on and excessive use of modern gadjets fails them (cue maniacal laughter).

[Colin Southern]

I still find ie is more wide open since you cannot close holes and they remain open until ms patch them

I hear what you're saying, but I still think it's false logic - for sure, from when a vulnerability is published to when a patch is available you're vulnerable - and for sure, the shorter that time the better - but - if MS take twice as long as the FF people, but the FF people have 4 times as many holes to plug (both "reasonably accurate" figures), then your overall exposure is doubled on FF compared to IE.

Personally, I just don't go anywhere near the types of sites that are most likely to try to exploit vulnerabilities ...

Windows is known to be insecure though, it's the highest % target (so most viable to exploit and ie is the main browser of choice) and you're at the mercy of MS to patch you up.

It's a lot more secure than a lot of other platforms, despite what others might want you to believe. Looking at the 2007 stats for example, Vista had a total of 13 Highly Critical and Extremely Critical vulnerabilities discovered - the Apple Mac in the same period had a whopping 234.

http://blogs.zdnet.com/security/?p=758

In the real world however, my personal experience both with my own networks and those of my customers is that it's "pie in the sky" stuff. We keep our systems patched - we run AV software - we sit behind a firewall - and we have a policy that regulates the types of sites that our users are allowed to frequent (or more specifically we warn them against sex sites and "bad taste" sites) ... and in the last 15 years - with the exception of the odd virus that sneaks in - I really can't say that we've had a problem; no "machines taken over" - no "data spread all over the net".

[Davey]

I must admit most of the scaremongering is done as a marketing trick to get people to buy into stuff. Most pay soft firewall and av are a waste of time because they won't protect you without input. Automated tools are no match for people fact, if it's been made it can be cracked (it's just time to crack vs gain, along with a little how much noisy it makes in the process). The prob with ie is if you get it to execute code that's a big problem due to close ties to OS it can run some pretty heavy functions so ie has inherent flaw. On ff under slack running MAC that isn't an issue at all, someone can crack it sure but the most that's easily grabbed is my bookmarks or the like and believe me they aren't that interesting hehehe. I've never had a problem ever but I'd never boast I'm invinsible because it would be easy to get into my home box with time and effort but pointless. I just use common sense which is the first line and perhaps strongest defense.

The problem isn't so much what you see on the machine but what you're not seeing. Some of the nastier stuff doesn't trip many alarms but unless you are completely out of date and wide open (as many are) it's unlikely to just happen unprovoked so to speak to an open connection or unless you visit a dodgy site/open dodgy content. Targeted attacks are another matter and you need sec admin with extensive pen testing and frequent audits, cycle keys, packet sniffing etc etc but only applies to business.

[Daniel Salazar]

Hi Daniel,
If they ask for a Gamma preference during the calibration, would it matter whether I choose Gamma1.8 for my Mac vs 2.2 which is more "common" ? As long as you have target ICC's . would it make any difference ?

I recommend you using the Gamma 2.2 which is the common one and the one recommended to be used by Datacolor, the company that produces the Spyder.

I only have an HP photojet printer at home - HP doesn't offer an ICC for it. I only use it for rough proofs, so not worth the extra expense of Spyder3Studio.

Regarding this, I could tell you that if you get the Studio version, you could crate by your self your icc profile for your printer. As you stated before, you live far from any printer service, so if you could print by yourself you could get really good results, of course with a cheap printer won't work and the investment won't be worth it.

Let me tell you that for a couple of months I had problems when printing using a combination of HP ink/HP paper, because the printer was applying a colour cast when selecting the HP paper. Since I create my icc profiles by myself, I'm getting perfect prints.

[Daniel Salazar]

Thanks Daniel - as I suspected, very little in it. Out of interest, do you have a graphic that maps your monitors gamut against the Adobe RGB colourspace?

Hi Colin, here you can see a comparison between the Adobe RGB colourspace and my profile created with a Spyder3 calibrator.

Hi Chris,

I'm not familiar with the gamut of Mac 'pro' monitors, so it is possible that they're capable of some colours outside of the normal "sRGB Gamut" (I put that in quotes because sRGB isn't a gamut definition, but my understanding is that what most monitors typically display anyway), but what I suspect you're seeing is probably more of a spread of tones than additional colours (perceptual rendering).

As per the comparison between sRGB and my profile you could observe that the Mac displays also go beyond sRGB gamut, is not a spread of tones, and even a "little" the important thing is that it goes beyond.

[Colin Southern]

Hi Colin, here you can see a comparison between the Adobe RGB colourspace and my profile created with a Spyder3 calibrator.

Hi Daniel,

Thanks for that - it's always nice to have the graphs to backup the theory.

On the face of it, it appears as if your display is pretty well "aligned" with sRGB (as expected) - it certainly exceeds it slightly in some areas (no doubt about it), but I think it's also worth noting that it also falls well short of Adobe RGB in other areas (as do 99.999 percent of monitors on the market). So yes - a small advantage to be gained in using Adobe RGB colourspaces (and assuming of course that the capture device also exceeds the sRGB gamut in the same areas, unless you're manipulating the image into those areas) - otherwise it's going to be a case of "two ships passing in the night".

[crisscross]

Hi Colin, here you can see a comparison between the Adobe RGB colourspace and my profile created with a Spyder3 calibrator.

As per the comparison between sRGB and my profile you could observe that the Mac displays also go beyond sRGB gamut, is not a spread of tones, and even a "little" the important thing is that it goes beyond.

I may be interpreting Daniel's image incorrectly and not sure how the super-imposition has been done, but from the profiles I have, I am now coming round to Colin's exposition - at least in theory. Later on (when it clouds over!) I will do some experiments in practice.

For now, I can see that starting with my normal screen calibration and turning to sRGB, the kite (or is it a gamut map?) enlarges in all directions. Then even more for Adobe RGB, but I can see that is academic unless of value when printing....on the other hand it could be that that is what causes trouble printing if what I am actually doing is compensate visually/intuitively the image as I see it on screen to overcome what the OS/colorsync had done in the 1st place to squeeze Adobe RGB into the kite area available.

[Daniel Salazar]

Hi Chris, you have to use the ColorSync Utility, then you have to select the colourspace you would like to compare, e.g. Adobe RGB and then on the right side of the ColorSync Utility, where the colour Gamut it's shown, you have to click on the triangle located on the top left side of the screen and then select something called like "Hold for Comparison" (I'm not at home so I don't remember exactly how is called) and then on the left side of the utility you have to select the other profile you would like to compare.

When both are selected you will get something like what I have.

@ Chris, could you show a picture of the profile comparison you meant? in another thread once I compared the http:////www.cambridgeincolour.com/forums/thread765.htm colour Gamut of my screen profile and one printer profile set. On this comparison you were able to see that the printer in some areas exceed a lot the colourspace of my printer, therefore I think that showing a comparison could clarify your comment.

[crisscross]

Hi Chris, you have to use the ColorSync Utility, then you have to select the colourspace you would like to compare, e.g. Adobe RGB and then on the right side of the ColorSync Utility, where the colour Gamut it's shown, you have to click on the triangle located on the top left side of the screen and then select something called like "Hold for Comparison" (I'm not at home so I don't remember exactly how is called) and then on the left side of the utility you have to select the other profile you would like to compare.

When both are selected you will get something like what I have.

Thanks Daniel, got that far. Only on your image the 2nd comparison profile appears to be entirely within the 1st. Are you saying the 1st is the monitor profile and the 2nd is the sRGB? With mine it would be reversed.

(However on my 1st trial image, an 'ex-camera' one, converted to sRGB Firefox appears to have lost its colour management capability and only Safari is showing anything like the original image (in pbase original) and pbase reduction still knackered. So recant may be postponed)

[Daniel Salazar]

Thanks Daniel, got that far. Only on your image the 2nd comparison profile appears to be entirely within the 1st. Are you saying the 1st is the monitor profile and the 2nd is the sRGB? With mine it would be reversed.

it depends which one you choose first to hold for comparison.

I made already two:
1. sRGB vs. my screen profile
2. Adobe RGB vs. my screen profile

in both my screen profile is the one shown with colour.

[Colin Southern]

Hi Chris,

I've been trying all day to think of a good analogy for monitors and colourspaces - and I think that I finally have one ...

... Back in the "good old days" when I was a kid we used to have a black and white TV (colour TV and colour transmission hadn't been invented yet). The black and white TV is like a typical computer monitor, and black and white transmission is like the sRGB colourspace ...

... as the years rolled by colour transmission started (analogous to Adobe RGB) - BUT - the black and white TVs didn't suddenly start displaying images in colour because it is incapable of displaying them (like a monitor that's only capable of displaying the sRGB colourspace) - so the TV converts the colour signal to a monochrome one, which is what we see.

Now that we have colour broadcasts in high definition (perhaps the ProPhoto colourspace) - great if you have a high-definition capable TV, but if you still only have the black and white TV then that's still all it can display.

It's one of those horses for courses types of things - sRGB is based around a "lowest common denominator" approach - covers most colours pretty well - and makes it harder to mess things up; Adobe RGB is more a superset of typical printer and monitor gamuts (although some colours will be mutually exclusive as the practical application of it is based around the physical capabilities of the respective devices) - whereas with the "all encompasing" colourspaces like LAB it's very much a case of "with great power comes great responsibility" as it's possible to stuff things up something chronic, but not realise why because most monitors can't display the errors of ones ways (but the printer may well).

And sometimes I talk too much!

[Davey]

You mention about lab colour, I've seen pop up in conversation but never used it. As you say many of it's colours fall outside the range of our vision and monitors can't display them so that brings me to the question why use lab? I presume there is some benefit to working in lab (hence it exists and is an option in ps) but guess it's not to do with it's larger gamut which is surplus to requirements.

[crisscross]

I think I am gravitating back to heresy, but have to quell a couple of rebellions first (a) lost the colour management on Firefox, but it will give me a chance to look at things on a PC (b) I have noticed that NX2 changes 'Color Mode' at will between Nikon I,II,III, ie sRGB and Adobe RGB, BUT if the image was shot with the camera in I or III the 'colour space' stays sRGB irrespective.

Modes I & III jazz things up redwards, a bit like Canon 'portrait' & 'landscape' in the 'picture style' menu.

But, accepting that the monitor can't show a lot of the colours in Adobe RGB, it can still use them for PP work. A bit like working on a tif file rather than a jpg (ignoring the later progressive loss of quality) even if you can't see any difference between them on screen??

Hence perhaps the Nikon suggestion that Modes I & III are used for files pretty near ready to print with no PP and Mode II for ones requiring more work. Virtually all my work is on landscape and nature subjects with very few of the bright colours you get on man-made/urban stuff and especially short on red (outside the poppy flowering season) - so it does usually need a lot of PP work. It does vary a lot with the image content; sometimes I have to look quite hard to see any difference, sometimes it is quite startling and not pleasantly so.

Having done all the PP work, why then mess it up again changing the profile at the last minute before generating the forum jpg? I ran up the Dell Dimension 2400 and was relieved to see my pbase images looking fairly colourful irrespective of profile and browser so it would not be for casual viewers.(Can't find a colour profile to feed to Firefox there).

Its over to Safari for me, and making sure I never accidentally change the colour mode on the camera again. I know I can globally change colour profiles in other software, but still see no point.

[qveda]

I recommend you using the Gamma 2.2 which is the common one and the one recommended to be used by Datacolor, the company that produces the Spyder.



Regarding this, I could tell you that if you get the Studio version, you could crate by your self your icc profile for your printer. As you stated before, you live far from any printer service, so if you could print by yourself you could get really good results, of course with a cheap printer won't work and the investment won't be worth it.

Let me tell you that for a couple of months I had problems when printing using a combination of HP ink/HP paper, because the printer was applying a colour cast when selecting the HP paper. Since I create my icc profiles by myself, I'm getting perfect prints.

I'm happy to report that even though I don't have my monitor calibrator (Sypder or Display2) yet, I successfully ordered prints online and they came out very close to how they looked on my monitor when I used the remote printer's ICC and adjusted accordingly .

[Colin Southern]

You mention about lab colour, I've seen pop up in conversation but never used it. As you say many of it's colours fall outside the range of our vision and monitors can't display them so that brings me to the question why use lab? I presume there is some benefit to working in lab (hence it exists and is an option in ps) but guess it's not to do with it's larger gamut which is surplus to requirements.

LAB colour is wonderful - I use it a lot

In the LAB (or more correctly C.I.E.L.A.B) colourspace luminocity (the "L") is kept seperate from the colour information (the "A" and "B" channels), which makes it easier to do many things without luminocity affecting the colour, and vice-versa.

The colour channels are 2 pairs of 2 colours; "A" ranges from magenta to green and "B" ranges from blue to yellow - it might sound confusing at first, but it works pretty much the same way our brains do when processing the RGB information from our eyes. Once you get used to it it's much easier to do many things than it is in RGB mode (some say that there are things you can do in LAB colour that just can't be done in RGB modes; other RGB diehard fans dispute this, but in my experience, the LAB way can certainly be a lot easier in many cases).

It's also a true colour standard (infact THE colour standard in many cases) - my understanding is that photoshop uses it INTERNALLY for all processing regardless of the mode selected by / presented to the user.

It's a huge space that's used as a reference for containing all other spaces - and can specify colours that can only exist in theory (eg yellow, with 100% luminance).

Some of the benefits ...

- You can sharpen just the L channel so that you sharpen the image, but don't accentuate chromatic noise.

- You can use it to drive subtle colours apart in shots with little colour varience.

- You can use it to remove colour casts with ease

- You can use it to accentuate certain colours without affecting others, without affecting their luminocity.

Scott Kelby went to far as to say that "it will forever change the way we do colour correction" (which hasn't happened yet of course!).

Personally I just find it more intuative - and I get better results when I use it on difficult images.

[Colin Southern]

But, accepting that the monitor can't show a lot of the colours in Adobe RGB, it can still use them for PP work. A bit like working on a tif file rather than a jpg (ignoring the later progressive loss of quality) even if you can't see any difference between them on screen??

Hi Chris,

The problem isn't so much that monitors "can't display most of the extra colours of the Adobe colourspace" as it is one where those extra colours are rendered into SOMETHING that the monitor CAN display. People then adjust the image so that WHAT THEY SEE on the screen looks right - but then get bitten in the bum come print time because the colours that they thought they were changing the image to weren't really what they were changing them too; they were really being changed to something else, but the monitor couldn't display them. Good example of this is when I had a photo of fishing boats with bright red paint on the hulls - looked great on screen (because the monitor has a dedicated RED channel), but come print time the hue shifted considerably to something far more orangy because it was out of gamut for the printers inkset. (OK perhaps not a great example, but hopefully you get what I mean).

Having done all the PP work, why then mess it up again changing the profile at the last minute before generating the forum jpg?

For two good reasons ...

1. The main one being that 97% percent of the population won't be able to view your images correctly unless they're tagged for sRGB (a) the images will look horrible and (b) even if they could use larger colourspaces like Adobe RGB or ProPhoto their monitors are incapable of displaying most of the extra colours anyway (similar to the broadcasting of high-definition colour TV signals if 99% of the population only had black and white TVs).

2. The other consideration is that images put up for display are usually severely down-sampled anyway. Down-sampling will degrade image detail far more than using sRGB over Adobe RGB so for most, the uploaded image will only ever be a low-grade copy of the original (my originals are always 16 bit PSD files in large colourspaces - usually over 5000 pixels with and over 3000 pixels high, whereas what gets uploaded is 8 bit - sRGB and only 1024 x 512 pixels).