LR, CS and 3800 blues

[Robert Campbell]

I took the plunge and got an Epson 3800 as my first proper printer. I'm generally very pleased with the output...but I'm having problems with blues.

I have a calibrated monitor, the canned profiles, and a small selection of papers. Generally the colours are very accurate, except for 'rich, deep blues' which come out more as 'slate grey'.

So, I looked at soft proofing in CS3; most profiles show this slate blue-grey instead of rich blue -- though Epson's Luster paper looks quite accurate -- but I don't have this paper. Changing from perceptual to relative colorimetric or vice-versa doesn't change colours that much.

I normally print from LR where there's no soft-proofing [yet].

So, am I doing something wrong? Do I have to soft-proof to see if the particular paper will give me the blues I expect? --is this something to do with the gamut of the paper? I'm not looking for total accuracy, just 'acceptable' would be good enough.

Any thoughts/comments?

TIA

 

[Eoin]

Can't say I've noticed the same issues, I presume since it's a new printer and this may seem very basic,
(1) that you have shaken the ink carts before installing them. Some settling may have taken place during storage.
(2) You have checked to make sure the ink carts are in the correct slots, cyan is not mixed up with light cyan and so on.
(3) You have run a cleaning & alignment routine from the printers control panel rather than from the driver?.
(4) you are using the correct epson paper type for the (I presume) 3rd party paper you're using and have applied the correct paper manufacturers ICC profile in the LR or CS in the printing screen and subsequently switched off colour management on the printer?. I suspect you're well aware of double profiling and correct colour management routines, so please excuse the basics.
(5) You're not running on a Mac with Snow Leopard installed :deadhorse:

Other than that, which is very basic colour management I can't help you.

 

[Terry]

I found you still need to use custom profiles not simply the manufacturers. For instance with Lightroom and my 3800 the Epson premium Luster paper looked awful with the Epson profile and spot on perfect with the profile I got from Jack at the printing workshop. It was specially tuned to his printer but it still looked great on mine and worlds different to the base Epson profile.

What papers are you using?

 

[ddk]

Its impossible to tell without actually seeing and troubleshooting the entire chain, but the main difference between 3800 and 3880 is a new ink set that will supposedly give you more accurate blues and yellows. Personally, I never experienced anything like what you mention, the 3800 is capable of very deep, rich blues, so I wonder if the problem is one or more of the following.

- Monitor not properly calibrated.
- Working with Adobe RGB files on sRGB monitor.
- Out of gamut blues.
- Paper type and/or paper's ICC profile.

As far as canned profiles go, people's experiences will vary depending on their machines and papers, hence the need for custom profiles.

Paper and ink combination can be the limiting factor too. I tried many types and brands of paper and for my money the Ilford Gallery Gold Silk is the most accurate paper that I found for my 3800, even with their icc profile. I can say with confidence that I regularly get prints that match the monitor at 90% or better with this combination, I wish I could find a gloss paper this good.

 

[Robert Campbell]

Thanks all for the input.

I've re-calibrated the monitor using the suggested 80 instead of 100 for luminance; there is now little difference between what I see and what CS soft-proofs.

However, the dark blues, though rather better, are still not right; and in other pix, the colours are now much more saturated.

I'l try fiddling a bit more.

 

[ddk]

Thanks all for the input.

I've re-calibrated the monitor using the suggested 80 instead of 100 for luminance; there is now little difference between what I see and what CS soft-proofs.

However, the dark blues, though rather better, are still not right; and in other pix, the colours are now much more saturated.

I'l try fiddling a bit more.

Which monitor and what calibrating software/hardware?

 

[Robert Campbell]

Can't say I've noticed the same issues, I presume since it's a new printer and this may seem very basic,
(1) that you have shaken the ink carts before installing them. Some settling may have taken place during storage.
(2) You have checked to make sure the ink carts are in the correct slots, cyan is not mixed up with light cyan and so on.
(3) You have run a cleaning & alignment routine from the printers control panel rather than from the driver?.
(4) you are using the correct epson paper type for the (I presume) 3rd party paper you're using and have applied the correct paper manufacturers ICC profile in the LR or CS in the printing screen and subsequently switched off colour management on the printer?. I suspect you're well aware of double profiling and correct colour management routines, so please excuse the basics.
(5) You're not running on a Mac with Snow Leopard installed :deadhorse:

Other than that, which is very basic colour management I can't help you.

Thanks;
1. shook the carts
2. they are where they should be
3. Haven't run any cleaning
4. Yes - check every time that the [cryptic] Epson code is correct; have ICM set to off
5. No - windoze!

 

[Robert Campbell]

I found you still need to use custom profiles not simply the manufacturers. For instance with Lightroom and my 3800 the Epson premium Luster paper looked awful with the Epson profile and spot on perfect with the profile I got from Jack at the printing workshop. It was specially tuned to his printer but it still looked great on mine and worlds different to the base Epson profile.

What papers are you using?

German Etching and Photo Rag from Hahnemuehle
Epson Premium Semigloss and Archival Matte, Premium Glossy
Ilford Galerie gold fibre silk
Some from Calumet [Brilliant]

The shops here in NornIrland don't have much to offer in the way of paper selection -- I'm still trying to discover what I like.

 

[Robert Campbell]

Which monitor and what calibrating software/hardware?

David,

It's an Eizo CG241W; it comes with it's own profiling software. The calibrator is an x-rite eye-one display2.

I'd understood that 100 was the 'correct' luminance for photo editing, but the Eizo says that 80 is the default, to I've chosen this -- not that I can see much difference.

There really isn't a problem with 'the prints are too dark/light'.

 

[Robert Campbell]

Thanks again to all of you -- I've responded to your comments above.

I've tried a few more things; I found that out-of-camera jpgs will print +/- exactly as I see them on screen -- and that the softproofing in CS is accurate for them -- in fact, the colours hardly change.

I print from Lightroom 2.5.

I found a jpg with the problem 'rich deep blue' -- in this case it's called Oxford Blue as painted on our front door. The print matched the monitor almost exactly -- my viewing light isn't perhaps the best. So it is possible to print these colours, but there are images that don't work.

The image that concerned me was in fact a tiff that had been through PhotoAcute -- using 3 bracketed exposures -- a snow scene in the Alps.

I've tried several out-of-camera jpgs -- camera model seems irrelevant -- and they are all excellent matches [they are in aRGB]. jpgs that have been modified a bit -- clarity, vibrance -- sometimes come out a bit too saturated, but are generally accurate enough [for me].

A pseudo HDR tif from Photomatix [two raw developments] has printed accurately.

Some -- but not all -- raw developments seem to print accurately; so far I can't determine any common factors -- still looking.

I do have some difficulty with the blue end of the spectrum --I can't see much difference between blue, indigo and violet in rainbows -- but the 'deep rich blue' turning to 'slate blue' was very obvious.

Thanks again.

 

[Jack]

German Etching and Photo Rag from Hahnemuehle
Epson Premium Semigloss and Archival Matte, Premium Glossy
Ilford Galerie gold fibre silk
Some from Calumet [Brilliant]

The shops here in NornIrland don't have much to offer in the way of paper selection -- I'm still trying to discover what I like.

Any of the matte papers will not reach saturated blues. Your best paper in that list should be Epson Glossy, but even it won't render as good of blues as Epson PPP Luster.

 

[ddk]

David,

It's an Eizo CG241W; it comes with it's own profiling software. The calibrator is an x-rite eye-one display2.

I'd understood that 100 was the 'correct' luminance for photo editing, but the Eizo says that 80 is the default, to I've chosen this -- not that I can see much difference.

There really isn't a problem with 'the prints are too dark/light'.

I have the same monitor so we can be sure that the calibration is correct. At this point it can be only one of two things, either your blues are out of gamut or the ICC profiles are not working for your machine. The other thing that you should check is the print driver, make sure that the right type of paper is selected and that the other printer controls are off!

German Etching and Photo Rag from Hahnemuehle
Epson Premium Semigloss and Archival Matte, Premium Glossy
Ilford Galerie gold fibre silk
Some from Calumet [Brilliant]

You're not going to get those deep rich blues with either of the Hahnemuehles, nor with Epson's Matte and Semigloss, but you should get close with the Premium glossy. The Ilford should give you an exact or a very close match, if your blues aren't out of gamut. I printed a whole series with this kind of blues without any problems, are your blues deeper and more brilliant than this?

 

[Robert Campbell]

Any of the matte papers will not reach saturated blues. Your best paper in that list should be Epson Glossy, but even it won't render as good of blues as Epson PPP Luster.

Thanks, yes I did appreciate this. No local store has Epson PPP Luster [and the postmen are striking ]. I didn't expect quite so much change.

 

[Robert Campbell]

David,

thanks; the problem pic has rather darker blues; but I can get dark blues accurately from an out-of-camera jpg. I'm beginning to think that I'm doing something wrong in processing. Some, but not all, of the blue can be out of gamut in PS.

 

[robertwright]

I'll take a swing: Lr has it's own internal colour space, melissa-rgb, which is the profoto space with a linear tone curve. It's linear because raw data is linear. They chose profoto because it is larger than what cameras can capture currently and can contain all the colour data.

The adobe engineers in their podcasts talked about how Lr was meant to purpose images for many destinations, web, print, inkjet. So the conversion to those destinations is held off until the very end, as is the representation of the data-for example, - the percentage numbers of rgb will not relate to any destination specifically, they are not the same as the rgb numbers in Ps. The reason this is is that first you would have to have a known destination, a gamma and a rendering intent. Since Lr designed for multiple outputs, this is unknown while you are doing edits. What the gamut warnings are showing you is clipping for profoto with gamma 2.2, which is larger than aRGB and sRGB.

You mentioned that camera baked jpegs are printing fine- this is very interesting because they are doing a colour space conversion in camera from the larger capture gamut to sRGB--how they do this is secret sauce--they use memory colours and other tweaks to get an ideal conversion that preserves a lot of the saturation of colours with non of the posterization you are getting from Lr.

You can't expect to print highly saturated colours on matt paper. Epson's lustre is about the best for this kind of thing. And probably their canned profile is going to be close-it is for me on the 3800.

I don' think you are doing anything wrong, I just think you are bumping up against the limits of colour management in Lr. You will have to export a 16 bit tiff in profotorgb into Ps and do a soft proof against you paper profile, and then tweak that to really handle difficult cases. Most of the time Lr will work with the defaults. If your print driver has a 16 bit option you can tick that also. Probably won't make much of a difference.

 

[jjlphoto]

I suppose a real test would be to run the same RAW fine in ACR, your camera's RAW processor, and CO(if you have it), process them all into sRGB TIFF files. Then run some prints.

 

[Robert Campbell]

Thanks for the further comments.

I downloaded a couple of test pix designed for colour proofing, and they printed properly. So at least part of my workflow is OK.

The images giving trouble were all processed 'there' -- in Switzerland -- now that I come to think about it; the set up 'there' should be the same as 'here' -- Ireland --, but perhaps it ain't.

So I reprocessed a couple of the troublesome images 'here' and they now print +/- correctly -- certainly the soft proof in CS3 with Simulate paper color checked looks very accurate compared to the print. I don't really follow how this change of location could alter things so much; the monitor 'there' is the same as here, but the luminosity may differ -- I'll check it shortly. The images and the LR catalog file are on a portable HDD that goes to and fro.

It's been mostly the blues that have been troublesome -- but here, the weather is uniform 18% grey which isn't very helpful for comparing jpgs and raws [I don't have ooc jpgs for the problem pix].

If this is the answer, then another strange problem has arisen; a tiff [with no LR adjustments] sent from LR to CS3 and opened in Viveza looks +/- identical to the LR view; but the same tiff sent from LR directly to Viveza appears much more saturated. I haven't found an answer to this yet.

 

[Robert Campbell]

I printed a whole series with this kind of blues without any problems, are your blues deeper and more brilliant than this?

David,

the blues are more or less these -- on reviewing, in this post they look a bit light on my monitor

I know these are chocolate box pix, but I was thinking of them for the Xmas cards; that day, the sky at about 90 degrees to the sun was the deep blue/black that you get high in the Alps -- the prints were more cyan/green -- ugh! IN CS, there is no gamut clipping.

 

[docmoore]

Thanks for the further comments.

I downloaded a couple of test pix designed for colour proofing, and they printed properly. So at least part of my workflow is OK.

The images giving trouble were all processed 'there' -- in Switzerland -- now that I come to think about it; the set up 'there' should be the same as 'here' -- Ireland --, but perhaps it ain't.

So I reprocessed a couple of the troublesome images 'here' and they now print +/- correctly -- certainly the soft proof in CS3 with Simulate paper color checked looks very accurate compared to the print. I don't really follow how this change of location could alter things so much; the monitor 'there' is the same as here, but the luminosity may differ -- I'll check it shortly. The images and the LR catalog file are on a portable HDD that goes to and fro.

It's been mostly the blues that have been troublesome -- but here, the weather is uniform 18% grey which isn't very helpful for comparing jpgs and raws [I don't have ooc jpgs for the problem pix].

If this is the answer, then another strange problem has arisen; a tiff [with no LR adjustments] sent from LR to CS3 and opened in Viveza looks +/- identical to the LR view; but the same tiff sent from LR directly to Viveza appears much more saturated. I haven't found an answer to this yet.

LR export edit has three options two are edit copy and edit copy with LR adjustments...perhaps your LR/PS export differed from LR/Viveza?

Bob

 

[Robert Campbell]

LR export edit has three options two are edit copy and edit copy with LR adjustments...perhaps your LR/PS export differed from LR/Viveza?

Bob

Perhaps, but I deliberately chose files with no adjustments, and chose edit original.