Printing - sRGB and Adobe questions

[Andrew76]

Monitors start with a black screen and add red, green, and blue - so no problem reproducing these colours.

About 3 weeks ago, my 3 year old added cyan, magenta, and yellow to my computer monitor. I have the best of both worlds, screen, and printing!

[Mike Buckley]

About 3 weeks ago, my 3 year old added cyan, magenta, and yellow to my computer monitor.

I'm not a parent, so I wonder if this is something that children use one color for every year of their age. If so, your problem will only get worse.

[ajohnw]

Funny cyan was mentioned - an early calibration attempt. I'm no sure that the software that came with the calibrator would even notice this one. There were a couple of others too.



John
-

[xpatUSA]

Except where the bride's mother was wearing a deep cyan jacket that was rendered blue by sRGB. That cost. Never again.

Tim

Always sad to see money "going South". How would you have recovered that deep cyan in sRGB? Is it possible?

I once had a long battle with a yellow flower image that ACR 5.4 blatantly messed up. Only after much research into gamuts and rendering intent did I finally end up with an acceptable image. More importantly, I learned why ACR messed when SPP didn't.



I too use sRGB all the way thru normally (I don't print and I use a sRGB gamut monitor) and, like Colin's 'sheeple', rarely notice a difference but every so often . . .

Bluebottle fly shot and rendering intent.

(my recent post of an insect with an iridescent body).

[xpatUSA]

Funny cyan was mentioned - an early calibration attempt. I'm no sure that the software that came with the calibrator would even notice this one.

Yes, I expect we all know that, on The Card (don't leave home without it), cyan patch 18 is out-of-gamut in sRGB D65. Lindbloom describes it well:

http://www.brucelindbloom.com/index....heckerRGB.html

cheers,

[Macmahon]

... How would you have recovered that deep cyan in sRGB? Is it possible?

...

Ted

Not possible, I believe. But I am most interested about getting good colour representation in print. (I seldom share images via web or projection because they so seldom appear as I intended.*)

Just for interest i used the colorthink to compare sRGB, the gamut of my printer (Epson 3880, Ilford Gallerie Smooth Pearl paper), and the colours from the wedding photograph with THAT jacket.



You can see how far beyond sRGB are those saturated cyans and blues. Yet they're well within the gamut of the printer.


Cheers

Tim


*PS Only partly due to gamut restrictions. Most desktop monitors around the world and almost every projector I've ever come across are not well calibrated and not at all colour profiled. The results are random, at best, and absolutely awful at worst.

[Richard Lundberg]

That's the question. How do you get from RAW t print and capture those cyans and blues. I am thinking that you print from a Pro Photo image in Photoshop to your Epson using Pshop manages colors, but I'm not sure what Epson profile you pick or whether you you should select Pro Photo from the list of profiles. I have used that profile and it seems to work well but I'm not sure what it means or why it works.

[Colin Southern]

That's the question. How do you get from RAW t print and capture those cyans and blues. I am thinking that you print from a Pro Photo image in Photoshop to your Epson using Pshop manages colors, but I'm not sure what Epson profile you pick or whether you you should select Pro Photo from the list of profiles. I have used that profile and it seems to work well but I'm not sure what it means or why it works.

If you're shooting RAW then all that can be captured has been captured (a RAW capture won't discard anything, unlike a JPEG capture) - from the RAW converter the image needs to be passed through to Photoshop as in a wide-gamut space (Preferably ProPhoto @ 16 bit, but Adobe RGB will still be better than sRGB).

Once in Photoshop you need to be careful as chances are your monitor won't be able to display the "problem colours" accurately; usually, all you can do is "do it by the numbers" -- white balance the shot correctly in ACR using a spectrally neutral reference, and in Photoshop, use the soft-proofing mechanism with gamut warning turned on to see where the problem areas are.

IF problem areas are an issue (and I say IF) then if you're already doing all you can do (as in shooting raw - working in a wide space - working to an accurate printer profile) then about the only other thing you can do is decrease the levels ("exposure") to bring the colour back into the gamut that can be printed; it's definitely a juggling act.

In terms of printing and profiles (in this often confusing area) the correct profile to use is the profile for that particular printer + inkset + paper + media settings combination. No exceptions. In practice what we do is start by printing targets (hundreds of patches of all types of colours) and then read these into the computer with a photospectrometer. The program knows what colours they SHOULD be, and the photospectrometer knows what colours they ARE and it uses both sets of information to create a profile that (in theory) translates "what you're asking for" into "what you wanted". It sounds hard, but it's not -- the software does the hard stuff; the rest is pure boredom as you hand scan 700+ patches

Having just said all that, the theory and practice don't always align that well; canvas is a good example. If you consider a scale of 0 to 100 where 0 is "the black of space" (no light reflected) and 100 being "the reflection of the media with no ink on it" then unsprayed canvas with matt ink scores about 23 (more like a dark gray) - and that's a problem because the software thinks "if 23 is the bottom end and 100 is the top, then I'd better fit everything else between those two numbers" - and the net result is the whole print lacks contrast, so in reality, for media like canvas the targets need to be printed - then over-sprayed (which lowers the 23 to about 17) - then scanned - and then the resultant profile adjusted to shift the low tones down down down.

After you've read Light, Science and Magic", grab a copy of Real World Color Management by Fraser, Murphy, and Bunting (2nd edition) (available on Kindle) - it's the universally acknowledged bible on colour management.

[Macmahon]

That's the question. How do you get from RAW t print and capture those cyans and blues. I am thinking that you print from a Pro Photo image in Photoshop to your Epson using Pshop manages colors, but I'm not sure what Epson profile you pick or whether you you should select Pro Photo from the list of profiles. I have used that profile and it seems to work well but I'm not sure what it means or why it works.

Richard.
I made my own profile for the Epson + IGSP, although on their website, Ilford supplies a reasonably good profile for this printer which would enable the saturated blues and cyans to be printed accurately.
I print from Lightroom, but Photoshop (albeit with a little more user concentration!) would do the same: Open the raw file in ProPhoto RGB, make whatever ever editing adjustments seem necessary, then print using 'Photoshop Manages colours' and select the print profile for the paper: in my case you can see on the picture my profile is called "i1P 3880 IGSP ... 110113.icc" (my naming convention tells me I made the profile with an i1Pro, for the 3880 printer, and the IGSP paper, and the date i made the profile)

Cheers

Tim

[Colin Southern]

Open the raw file in ProPhoto RGB, make whatever ever editing adjustments seem necessary,

It's the editing bit you have to be careful of though because you won't be able to see the out of gamut colours accurately and if you see that they're wrong - try to edit them - then you'll have a lottery come print time (unless of course you have a wide-gamut monitor that's correctly profiled) (Like an $8000 Eizo)

[Richard Lundberg]

I have Real World 2nd Ed. None of these books completely answer dilemmas, you have to do it yourself from a position of some knowledge and healthy skepticism.
Here's a Pro Photo test image: http://www.outbackphoto.com/printing...048/essay.html
I don't think Photoshop soft proofing is of much use in banging on Pro Photo.
Because of my problem camera, I'm just going to open in ACR as PP 16 bit do white balance and exposure, save as tiff, do lens in Photoshop until I get a PTLens profile, correct perspective in DxO Perspective, make edits in something easy, print in six inks using both the Epson profile and Pro Photo on matte and see if I like it.
In searching for the test image I dl some articles that may also help.

[Colin Southern]

I have Real World 2nd Ed. None of these books completely answer dilemmas, you have to do it yourself from a position of some knowledge and healthy skepticism.

The book provides the foundation - but after that, real world experience is the next step

I don't think Photoshop soft proofing is of much use in banging on Pro Photo.

It's just a tool - up to the 'tog to use it or abuse it. In the case of a wide-gamut image and the correct printer profile, it'll tell you what's in gamut and what's out of gamut before you print. No more, no less.

Because of my problem camera, I'm just going to open in ACR as PP 16 bit do white balance and exposure, save as tiff, do lens in Photoshop until I get a PTLens profile, correct perspective in DxO Perspective, make edits in something easy, print in six inks using both the Epson profile and Pro Photo on matte and see if I like it.

Sounds OK. The "trick" with wide-gamut images is to just "work by numbers" ie "don't try to adjust colours visually".

[ajohnw]

RGB Led monitors are likely to get cheaper and that allows Adobe RGB. In fact some more recent Dell's use them and even have hardware calibration built in. It will be interesting to see if this has an impact on all monitors at some point and actually upgrades the web. It's been the way it currently is for a long long time now and to a certain extent deeper RGB can be compatible with sRGB except for purists. There is also the fact that current sRGB was originally aimed at cathode ray type monitors not lcd variants.

There is also a lot of talk about different colour gamuts but I can't see that happening overnight as lots of equipment would have to be ideally replaced - brilliant for sales figures though. Almost over night obsolescence. Now they wouldn't do that would they? There are also other variants of LED about now such as RGGB. Mostly aimed at improving colour rendition for illumination purposes but ........

John
-

[Richard Lundberg]

I've printed the Pro Photo test image http://www.outbackphoto.com/printing...048/essay.html
from Photoshop on my six color Epson using the correct settings for printer and matte paper profile and the matte setting for media. I also printed it from Preview for comparison on copy paper with the media set to copy paper bright white.
The matte print has noticeably richer colors especially in red and blue with some improvement in yellow and orange (not as much as I expected). I wish I had a spectrophotometer.
The copy paper version was generally faded in appearance compared to the matte, but that is normal. I also printed to matte using the device independent Pro Photo profile for the printer. I have tried this on sRGB prints and it seemed not to make much difference. Not this time. It appeared to add black to the pictures and desaturated the color swatches a bit.
I think this proofs the path for printing raw to my satisfaction.
I'm still discussing raw processing with ACDSee, who have been prompt and helpful, but have not given me a definitive statement on support for my camera. I sent them a file to work with but have not heard about it yet. In another thread I mentioned that ACDSee Pro 3 View mode shows my image with lens correction, but Develop mode will not open it. It won't save from View. I might try to scan the three prints to see if the comparison survives the transition to scanner space and sRGB monitors.

[Richard Lundberg]

pro photo on copy paper

pro photo on presentation matte

pro photo with pro photo profile on matte


Amazing, some of the differences survived the scanner and jpg and aRGB. Not so much the red of the strawberries. Pretty obvious the pro photo profile experiment was a botch.

[Macmahon]

Amazing, some of the differences survived the scanner and jpg and aRGB. Not so much the red of the strawberries. Pretty obvious the pro photo profile experiment was a botch.

Richard
ProPhoto is not an output profile. It is, as you have named, a device independent space, invented to give the widest gamut margin internally when editing and saving. I am not at all surprised that it gives a cr*p result when used as a device dependent print profile: that is, one made especially to account for the colour reproduction characteristics of the output device (printer).
The copy paper and matte paper prints are interesting. I know this test image well, as I'm sure others do too. On my wide gamut monitor the colours in your images look 'muted' and, well, wrong: yellows of the aspens going to orange: the orange of the arches tending to red, while the red of the strawberries is quite underdone (but you noted that may be due to scanning etc)
What print profile did you use when printing to 'copy' paper? (I confess I've never put copy paper through my printer)
Tim

[Richard Lundberg]

I print to copy paper from Mac Preview not Photoshop, with the appropriate Print settings. The reason I printed from Photoshop with a device independent Pro Photo for the printer was that I had obtained decent results doing that with a sRGB image or so I thought at the time. I was just checking my work. I too am not surprised it was crap, but for a different reason. It indicates a different GCR was used, not only different but way wrong. I believe ink jets normally use Heavy GCR, but this was obviously Enormously Heavy GCR.

[Macmahon]

I print to copy paper from Mac Preview not Photoshop, with the appropriate Print settings. The reason I printed from Photoshop with a device independent Pro Photo for the printer was that I had obtained decent results doing that with a sRGB image or so I thought at the time. I was just checking my work. I too am not surprised it was crap, but for a different reason. It indicates a different GCR was used, not only different but way wrong. I believe ink jets normally use Heavy GCR, but this was obviously Enormously Heavy GCR.

Hi Richard

You simply can't send any image that's encoded in the ProPhoto colour space to an inkjet printer and expect a usable outcome. The colour numbers won't mean anything sensible at all: every colour will at best be totally undersaturated even if the hues are not all screwed up. (see the flower example here which shows what a ProPhoto-encoded image looks like when interpreted in a 'smaller' colour space)

You must either convert the image data from ProPhoto to the printer's colour space in the application (photoshop, preview, Lightroom , whatever) or request the printer driver to do it for you ("printer manages colours"). In the latter case the driver will select the manufacturer's generic profile for the media type you choose and assume the manufacturer's inks.

Even though I'm sure this isn't the issue with your ProPhoto encoded image, I am interested to see your comments on the Grey Component Replacement. Are you using a 3rd party RIP of some kind? My Epson 3880 driver doesn't give me any options for alterations of GCR. The appropriate settings are, to the best of my knowledge, optimally defined at the time that I create a profile for the media I'm using.

(BTW, I'm pretty sure that when you print to plain paper from Preview, the Epson printer driver applies a default profile called "Epson ..... Standard" unless you select a particular profile in Preview's print dialogue.)

Cheers

Tim

[Richard Lundberg]

The print from Preview had the appropriate settings for the Epson. The second example did exactly what you said, sent the ProPhoto to the printer/paper space using Photoshop manages colors. In the third case something about using Pro Photo as a printer profile added way too much black to everything but the color squares, as if very heavy GCR had been enabled. Strange that the color squares were not affected the same way.
The actual print to matte paper of the second image looks very nice in reality and that's the way I will print my raw conversions if they're keepers.
Posting an image scanned from a print that will be viewed on a wide gamut monitor is pretty dicey I suppose. Not sure I get why g ballard is so dead set against wide gamut. If the wg monitor is calibrated it should show more saturated colors if they are in the image.
Next year's budget is in flux, but a wg monitor is in the running.