Rendering deep shadow detail

[photopat]

I must say, though I've used IP successfully with my Epson 4000 for over a year, i find that rendering intents and shadow point compensation have been a source of misunderstanding and confusion. I guess i'm gonna have to dig through the ImagePrint application folder on my hard drive to find the documentation!

I suppose that with shadow point compensation set to zero, one would expect deeper blacks at the potential cost of lost shadow detail?
[a href=\"index.php?act=findpost&pid=52634\"][{POST_SNAPBACK}][/a]

Colorbyte don't really have any "real" documentation" about the shadow point compensation.
But thrue  trials/testings  and discussions with other IP(and Colorbyte when they answer....) users it has come to this.

Normally when using the shadow point compensation slider the "default" value is 0
Changing the value (black point compensation) to 100 will efectivly have about 4% lighter shadows.
This aplyes when using perceptual intent  as rendering intent within IP.

However...When using relative intent as rendering intent within IP  the value 0 will make the shadow to deep and you'll lose details in the deep shadows.
Using 100 will  be  more "correct" since this will print more or less "exactly" as your soft proof looked like in PS when chosing relative  as rendering intent.

This has been discussed at the Yahoo ImagePrint group some times.

Anyway.This is how I work with IP and my prints look"exactly" when printed(and wiewed under correct lightning)as they do in my softproof wiew within PS.

This is "on" my Eizo CG19 display hardware calibrated (using Eizio's ColorNavigator software  and Eye One display(hardware)to use the full 10bit lut of the display) and D65 and gamma 2,2.

Patrick.

[photopat]

Mark.
I have   not seen you mentioned(or anybody asking) what kind off display you are using(lcd  or crt)
How is it calibrated??(I', not talking hadware here)
D65 ??D50?? gamma? whitepoint(cd/m2)?

What's your "workingconditions"(dark room"cave" or bright light)
Edit: I just re read this =)

I forgot to mention that my whole room is illuminated with D-50 (when I turn the lights on to examine the prints). I bought a Solux track and fixture system from Tailored Lighting in Rochester NY (www.solux.net). This was much cheaper than a state-of-the-art viewing box, but no dimming - so I replace dimming by holding the print nearer or further from the ceiling

Did i understand you correct that you work with these lights turned off while working and only turn them on to evaluate your print???

Even though I don't belive this to be the case,but setting the wrong (to high) cd/m2 will fool you when it comes to shadowdetails etc.

I myself work in what I would call slightly light cave invirement (got some dim D50 ligthtning in the room)

I calibrate my display(Eizo CG19) to D65 gamma 2.2 and 100cd/m2 for that invirement.
I've read somewhere that the  Eye-One(which I know you don't use) default of 140 cd/m2 for LCDs(wich is pretty bright if you work in a dark invirement)

I'm pretty shure you know your calibration "stuff",but I just wanted to rule this option out before you feel the need to try out if IP is (better)for you.


Patrick.

[mposter]

Print a black stepped gradient with no color management on various papers with various Epson driver paper settings to see how different they can be. I've used Moab Entrada Natural for over a year. It's good stuff, but it separates dark tones poorly. As others have said, most matte fine art papers are similar but some work better than others.

I've recently done some testing of papers along with setting up an X-Rite Pulse profiling system. The best I found for black separation so far is Premier Hot Press. And surprisingly, with almost all the papers I tested, including PHP the "Smooth Fine Art" paper setting in the driver yielded the best separations. FWIW, PHP is highly rated by Wilhelm, has no OBAs, no flaking, and is similar in price to Moab Entrada.

I bought the Pulse after having profiles made by others on i1 Photo and the Pulse. I had been having two continuing problems for years: poorly separated dark tones and infrequent, but annoying poorly separated colors (some browns and greens particularly). I threw custom profiles made by others at my situation with no improvement until I used a profile made by the Pulse. I'm fairly certain now that the dark tone separation problem is due mostly to paper choice and paper setting choice though a good profile helps a bit. The color separation problem was solved with the Pulse profile.

Michael P.

[Mark D Segal]

Patrick, the monitor is a Sony Trinitron CRT calibrated at 6500K, gamma 2.2 and luminance between 99~102 depending on the calibration session. The latter is set by ColorEyes Display as part of the profiling process. I do the monitor work in a darkened environment - only one lamp with a 60W bulb burning in the background some eight feet away. During the day most daylight is blocked out using an opaque window shade. I'm glad you asked, because for sure this problem can be related either to the printing end or the monitor end, including their associated profiles. It is possible that my monitor, despite soft-proofing, is showing me dreams that cannot come true. I've had that discussion with Integrated Color Corp. (the ColorEyes Dsiplay people) and I shall do a sanity check by resetting the monitor to factory defauls, and redoing the who re-calibration re-profiling routine to make sure there isn't some glitch there - but I doubt it.

Michael P - that is a very interesting idea - testing a selection of different paper types in the printer driver to see which one performs best - that is indeed thinking out of the (Enhanced Matte) box! I shall try the "smooth fine art" and several other paper settings on Enhanced Matte and see what happens. Enhanced Matte has a satisfactory texture and it is inexpensive. (I know it can yellow over time if left exposed to the elements, but that is not what I do with it.)

Interestingly, I am able to test IP without buying it, so on Monday I'm bringing two of these challenging images into the retailer for a "this versus that" demo on their set-up (similar to mine) and we'll see what emerges. I'll report the outcomes here. Meanwhile if there are more ideas and suggestions, please keep them rolling.

[photopat]

Mark.
It might be that you have a "bit" to high contrast ratio with your setup(for that enviroment).
100 in luminance for a CRT in a dark envirement feels a bit high for me..(others might disagree)
But then again calibration and setup isn't just numbers....=)( I wich it was that simple)

I used to calibrate my LaCie CRT display to 0.3 blackpoint and about 85-90  luminance
This was done with Optical(colorvision) and a DTP-92

There are people that have better knowledge than me on these matters when it comes to  calibration/colormangement.(Even though I feel pretty confident in what I do  =)......)
They could probably say alot about this .


I can't really say(for sure) that lowering your luminance to≈85-90 would be the way to go.
But try it as a start(you can always recalibrate)
The next step would be to wiew these images on a system(other than yours) that you "know"
is calibrated and working,to see if those "dark" look the same or differ from your display.


Patrick

[Mark D Segal]

Thanks Patrick, you may well be on the right track about whether the monitor is providing me the correct information on luminosity. I shall have a better handle on this Monday when I take the problem files downtown for viewing and testing on a professional retailer's system. I shall be looking both at what their monitor shows the image to look like, what the Epson driver produces on paper and what IP produces on paper. From that point, I'll know whether my solution is to be monitor or paper oriented, or a combination of the two.

The monitor calibration package I am using is state-of-the-art: Integrated Color Corporation's ColorEyes Display with a Monaco Optix XR colorimeter. This is the same hardware as X-Rite DTP-94. The ColorEyes program determines optimal luminance automatically, and according to I.C.C. the luminance values I am getting are normal. That of course is not necessarily the end of the story, as I share your view that this calibration process is not simply a numbers game. I'll be playing with these variables over the coming several days.

[photopat]

Seems like you are close to a slution here Mark.
Let us know how your tests turned out.


Patrick.

[mposter]

Mark,

Theoretically Enhanced Matte will yellow behind glass since it's not acid free. My testing of EEM showed the wackiest gradient of all the papers I've tried. There was lots of separation, but some of the patches that should have printed darker actually printed lighter. That is baffling, but it's what happened.

There are cotton rag papers with archival ratings that are relatively inexpensive, show good results for color and tone that I'd test before tearing your hair out.

It seems to me you've covered your bases as far as monitor and lighting management is concerned. And if your screen shows all but the low end is the same as your prints then to me that means you're just seeing what I've been seeing all along and that's that matte papers are a bit low end challenged. I have had to boost some darker tones prior to printing and expect I will continue to have to do that from time to time to get subtleties to expand on matte papers. But if you choose a paper and settings with that in mind at least you know you've mitigated the problem.

I'd be interested in what IP can do with that end of the scale though. My guess is they've overcome some of the problems with the Epson driver and might very well extract more from these papers. Still, I am starting to believe we can get more from these papers within the driver with a little bit of cautious paper choices and setting selections and a good profile to boot.

Michael P.

[photopat]

Michael P.
I do agree that there are (relative inexpensive)cotton rag papers that will out perfome EEM(which I rarley use these days anyway). My personal choise is Innova photo smoot high white (much lager gamut).

And I do agree that some deep shadows that shows  subtle detail in soft proof might not print.
However in the first post Mark made he said that ,they vere MUCH more visible.

He has(and no one included me have not asked him )mentioned the valuse of those shadows.

I use IP and in general I would say that it handles shadows better than the "Epson" driver.
But no matter what I use (even with the epson driver) I am in what you would call WYSIWYG situation (even on EEM with the epson driver)

I will admit that I'll start "losing" details(hard to separate to it's nerest neighbour) when I come to 93-94% k(or in rgb 18,18,18-15,15,15) in my prints on EEM.
This is when printing a numeric grayRamp from 1k to 100k .(with no adjustments and using realtive intent and epson driver,printed from PS print with prewiew))
And  to be onest about these "observations", I know there should be separations  so I look for them(and find them) but it's wey subtle.. =)

But then again those shades starts to be really hard to separate in softproof also
I see the separation but it's kinda no suprise to me when they don't  show on print.

So if I would think they were "Important" to the print I would to do some adjustments to bring them up in PS soft proof.

But then again(my favorte expression =) )
Trying to understand (and judge) somebodys troubles only by the written word is sometimes wery difficult.

What someone (Mark in this case) refers to as MUCH, might be in some one elses mind SMALL.

This makes in my mind forums both a pleasure and frustrating,since it's all about  understanding what the other person really is saying.


In this case I might have (or might not) understod what Mark really meant(when he said MUCH).
Eighter way all answers are benefitial and no matter who's right
The only things that's important is that the person that needs help is helped =)
And the problem is solved.
Patrick.

[Mark D Segal]

Michael, for all the grayscale tests I've done on Epson EM, there has never, ever been a reversal on the luminosity scale. Perhaps you have a profile issue or a corrupted file somehow. From what I understand the yellowing is not caused by light but by atmospheric pollution (gases); if I understood Wilhelm's fine print. displayed behind sealed glass it should not yellow for many years. I think this is why they changed the name from Archival Matte to Enhanced Matte.

Patrick, yes I never mentioned any numbers - silly me - and if you don't have the numbers and you don't see the prints or the monitor image it is hard to know really what I am complaining about. Now when you hear this I expect to see the next posts telling me that I am "over-reaching" - but heck, why not - otherwise how to do we improve, eh? So the LAB tonal area of concern is much worse separation on paper (compared with the monitor) in the LAB tonal range of 3 to 16, for steps such as 3 to 8, 8 to 11, 11 to 16, etc. (The RGB numbers in ProPhoto colour space are below 18, but best focus on the LAB numbers - they always mean the same thing.

[mposter]

Not a reversal per se, but some tones misplaced for sure. Nothing to do with a profile since with no color management I'm not printing with a profile in any case. That's the point of testing without color adjustments or profile. The test will show you the behavior of the paper, ink and paper setting without influence of profile (or rendering intent.) It sounds like Patrick is printing with no color adjustments in the driver but _with_ a profile since he mentioned relative colormetric rendering. That is fine if you wish to test the paper, ink and profile, but I think what you want to do here is remove the profile from the equation at least until you've established how the paper handles ink.

Ethan Hansen at drycreekphoto.com offers the technique to customers wanting to buy his profiles. He has an evaluation file (DCP-PrintEval.tif) that he asks you to use before printing your targets. That file specifically tests how each paper setting in Epson's driver lays down ink. You're not so much selecting papers using this file (although it helps that process) as much as you're selecting the best paper setting to use when printing (targets and real-world work).

And using EEM is not inappropriate, but it is not acid free and from what I've seen, papers containing acid will yellow even in dark storage. I was only suggesting there are acid-free, archival alternatives that are not unduly expensive and behave better. I make photographs for sale and produce fine art prints for my artist-wife as well. I use 100% cotton rag papers since they have a long track record for longevity. But that doesn't rule out the use of EEM. You should use what you like.

All this is quite theoretical at this point. I'm only saying that matte papers have a tougher time separating the low end tones and that to find the ones that do it best I print a stepped gradient with no color management trying various paper settings. IMO, this at least demonstrates the best the driver has to offer. A good profile and, as a last resort, slight adjustments to the image do the rest. And again, IP may (and probably does) do better than Epson's driver in that regard, but after some testing and an investment in profiling tools I'm getting the results I want (at least for now ;o)

[photopat]

Mark.
Now that we have "all" the details,we might be closer to a "answer"
I've done some testing,but I would take these conclutions with some "marginals" since I'm doing these tests on a 2100(2200) printer .And it's how you "interp" my written word.

I did 4 blocks with the lab values 16/11/8/3 next to each other in that order .
I printed and compared the print with the softproof wiew  .
This was done in 2 ways
1. Printing with the epson EEM profile from PS (colormanaged of course)
2. Printing from IP using their profile for EEM. (also colormanaged)

The Epson profile  printed "darker" compared to the soft proof and in my mind printed to dark overall.
There's still separations but the soft proof wiew is a bit optimistic (Even for some one with experience
looking at a soft proof knowing that a print is still a print and it will never print exactly as the soft proof)

The print made from IP and their profile printed  lighter than the epson/PS combo(more "true") and with better  separation( not by far but  defenatly notisable)
The print from IP was much closer to the soft proof in PS(using the IP profile offcourse) and printed more or less how I would expect it to print.

I'll leave it to this until you've done your own tests with IP to compare.But I'll not be supriced if the IP print has better shadow details than the epson profile.

Michael P

It sounds like Patrick is printing with no color adjustments in the driver but _with_ a profile since he mentioned relative colormetric rendering. That is fine if you wish to test the paper, ink and profile, but I think what you want to do here is remove the profile from the equation at least until you've established how the paper handles ink.

In Marks first post he wrote

I am using Version 3.1.07 of ColorEyes Display for monitor calibration and the Epson print driver for the 4800 printer and Enhanced Matte paper with Matte Black Ink. In general, the screen-to-print matching is quite reliable, except for deep shadow detail which is MUCH more visible on the monitor with soft-proofing active than it is emerging in the prints. I am wondering whether this an Epson profile issue, whether anyone else using this combination of hardware and software has experienced similar issues, and whether ImagePrint 6.1 would improve it.

So it is one of the things to test.
Since you do your own profiles you know that a  printer profile  is a 2 way profile aPCS(profile connection space)-to-device vs device-to-PCS,which makes it possible to soft proof.
And it's possible to edit the "part" in the profile that makes it possible to soft proof ,if the soft proof differ to much from the print.(This is nothing I can do though)


I do however agree with you in what you are saying about matte papers  have a tougher time separating the low end tones compared to glossy papers
And there are ways to test what a paper is capable of "doing"

But  a correct(good) profile (when soft proofed) should be somewhat close to the print if wiewed in correct lightning.
This has nothing to do what the paper is capable of "doing" since the soft proof should  more or less"reveal" it's limits and show you what to expect.
And he's trying to achive a WYSIWYG situation here.

Patrick.

[Mark D Segal]

Patrick and Michael, thank you ever so much - and Patrick double thanks for taking the time to do those tests. I think they are very revealing from your description, so I am ever more looking forward to what I shall see when I do this with actual photographs tomorrow at the retailer's set-up. I'll report back what I find.

It is true compared with glossy, matte papers generally provide a bit less sub quarter-tone separation - so there is a paper factor. Perhaps it is the price one pays for preferring matte over coated media. Hence the challenge with matte is to find the combination of software and paper selection that does the best job of it. I think this is what our dicussion boils down to.

As for yellowing, if you look carefully at Wilhelm's results for the Epson 4800, particularly footnote 6 in conjunction with the characteristics of all the papers tested in that chart, you can interpret his information to mean that for all papers listed, dark storage yellowing is related to storage conditions and low level air pollutants - not UV brighteners. (Display is another matter - I know for FACT Enhanced Matte definitely yellows within a year stuck on the fridge with no protection.) Beyond Wilhelm ratings, the choice of paper depends on taste and budget. People who like paying less than half of what anything else costs and like the brightness will use Enhanced Matte, others will use other papers, and others still will use a mixture depending on the image.

[Eric Myrvaagnes]

As for yellowing, if you look carefully at Wilhelm's results for the Epson 4800, particularly footnote 6 in conjunction with the characteristics of all the papers tested in that chart, you can interpret his information to mean that for all papers listed, dark storage yellowing is related to storage conditions and low level air pollutants - not UV brighteners. (Display is another matter - I know for FACT Enhanced Matte definitely yellows within a year stuck on the fridge with no protection.) Beyond Wilhelm ratings, the choice of paper depends on taste and budget. People who like paying less than half of what anything else costs and like the brightness will use Enhanced Matte, others will use other papers, and others still will use a mixture depending on the image.
[a href=\"index.php?act=findpost&pid=52789\"][{POST_SNAPBACK}][/a]

Mark and MPoster:

Can you offer some specific suggestions for reasonably priced alternatives to EEM? You've made me nervous about EEM yellowing, since that has been my primary paper so far.

And Mark:

I look forward to the results of your IP versus Epson tests. I have an Epson 2200, and I gave up on the Epson driver a while ago and ended up with IP, which seems to work nicely for me.

-Eric

[mposter]

Patrick, all profiles are not created equal. Printing with no CM and no profile allows you to see how the ink lays on the paper, nothing more, with no influence from the (good or bad) profile. It's a small piece of the testing puzzle, but in my opinion an important one. You do this to find the best Epson paper setting only. Then you use that setting to make your real world prints with profile applied. You can skip this step entirely if you wish, but I've been using this test as a precursor to profile-making and it's shown me some interesting variability in Epson's paper settings.

And Mark, use EEM, it's great paper. I'm only suggesting that there is better. Galleries, museums and collectors are just beginning to accept ink-jet prints as a viable and worthy alternative to traditional media. They tend to apply the same standards to work created using our ink-jet methods to those of any artist using paper media and that means acid-free (and usually buffered) paper and acid-free anything that touches the paper, from the backer board to the tape holding the print to the mat. I show and sell my work and create fine art prints for my artist-wife and I've chosen to use acid free materials. Again, don't get me wrong, I'm not telling you your choice is wrong.

Eric, there are many third-party alternatives some of which are double sided allowing you to proof on your rejects to save money, but none quite as cheap as EEM. Some are: Moab Entrada, Premier Hot Press, Hahnemuhle Photo Rag. Inkjetart.com has a good selection, Atlex.com has a smaller selection but at lower cost for those papers that overlap. I've just tested the papers I mentioned and Epson Ultrasmooth Fine Art as well. The Premier produced results at least as good as the more expensive alternatives with my images.

Michael

[Chris_T]

Hi all, great thread. Let me share my experience with HWM on 1280 and EEM on 2200 with Epson profiles. Color prints.

The problems are with highly saturated colors and/or deep shadow details. Since these are subjective terms, I find that the images in Dale Cotton's great article can best demonstrate these problems.

http://daystarvisions.com/Docs/Rvws/EpsonPaper/pg1.html

Fig. 2 is what the image looks like ín PS RGB1998, i.e. before Soft Proofing. Fig. 3 is how Soft Proof with a matte paper profile and in print looks like. Notice that in Fig. 3, not only does the shadow details suffer, but also the more saturated colors on the boats at the upper right corner. When I encounter this kind of problem, I will Soft Proof with a luster profile, both the monitor image and the print on luster paper will be more like Fig. 2.

While the nature of matte paper plays a role in this, I think that the profiles themselves and the Epson driver are the more significant contributors to the problems. Here's an action that plots out a profile conversion. Plots of the matte paper profiles provided by media vendors (Epson, Ilford, Red River, etc.) all look about the same, implying that they won't make that much a difference. The same is true with luster paper profiles. However, when I plot Bill Atkinson's profiles, I notice that there is a big difference in the shadow end.

http://www.curvemeister.com/downloads/profileplotter/

Soft Proofing and printing with different rendering intents and with black point compensation turned on/off do make a difference. With some image, the difference can be significant. But iterating through the permutations of different profiles, intents and bpc is very tedious and time consuming, not to mention the amount of paper and ink wasted.

There's got to be a simpler solution.

[Mark D Segal]

Mark and MPoster:

Can you offer some specific suggestions for reasonably priced alternatives to EEM? You've made me nervous about EEM yellowing, since that has been my primary paper so far.

-Eric
[a href=\"index.php?act=findpost&pid=52797\"][{POST_SNAPBACK}][/a]

Eric, anything apart from EEM you use for good quality matte will cost AT LEAST twice and in some cases up to four times the cost of EEM. The only problem I've had with EEM so far is photos posted on the fridge door. I have prints from 2000 in dark storage whose white borders look the same as EEM bought yesterday, and I doubt they will change much any time soon. As far as I'm concerned there is nothing to be nervous about. It all goes back to issues of purpose and taste. I make prints to bind in books or keep in portfolio cases - i.e. dark storage. If I were making prints for sale that will be displayed, unless they will be framed according to Wilhelm test specifications, I would change to an acid-free paper - the cost of the paper becomes a non-issue if your prints sell.

In preparation for tomorrow's tests I just finished preparing and printing on EEM (Rel Col. intent with Photoshop Color Management active - to simulate what I do for normal photographs) a series of neutral grey test strips - LAB gradations of L from 0 to 22 in 2 point increments with 0 values for both "a" and "b" throughout. I can hardly see any difference between 0 and 22 (equal to RGB 0 to 40) whether I use the EEM paper setting or the Single Weight Matte setting, which in principle should lay-down less ink. (The Smooth Fine Art setting is not available - in fact the Epson driver won't display any media that is not meant to be used with the Matte Black cartridge - cutting out playing with anything "smooth".) Comparing my results with the monitor image, Soft-Proof active, gradations upward of LAB 4 are obviously differentiated on the monitor. The gap between the print and the monitor image is striking as one progresses upward of LAB 6.

[mposter]

Mark,

As far as I know the smooth fine art setting is not avilable for other reasons, not that it's not meant for matte black ink. It is available for use with the 4000 driver (and other larger formats) but only for manual feed with sheets and for rolls of course. My guess is that it's not available in the 2200 driver because the sizes offered are too large for that printer.

MP

[photopat]

Michael.
I see your point and that's  actually what I would do as a "linear" test to see what paper settings
would produce the best starting point before printing a test chart for "profilemaking"
If I used the epson driver.
And off course it does not cost more than a couple papers and some ink to include  it it the testing.
If I in  my posts have given the impression that this is of no value,I'm sorry.

It still bothers me though that this is all epson products(printer/driver/paper/ink)
So the soft proof  and print should in my mind not differ that much as they aparently do.
Espessially whith the improved factory pre calibration these printers(x800) aparently gets
and the "better" driver.

I have no problems whith these things in IP.
What I see when soft proofing is what I get when printing.
And this is when using canned IP profiles

Patrick.

[Mark D Segal]

Michael, I'm using a 4800 so it is not the size of the paper.

Patrick, Michael and others, what I reported above about the results of my test strips is garbage - block it from eyesight. Rewind. For some wicked reason - perhaps because I was messing around with rendering intent settings or whatever, Photoshop was on printer color management and the printer driver was on no color management, so I got what I should have got - garbage. That nonsense found and rectified, I've now produced two sets of these strips: one goes from LAB 0 to 22 (= RGB 0 to 40), and the second from LAB 24 to LAB 46 (= RGB 44 to 90).

When you add a curves adjustment layer to either of these test sets, the deep dark one at its brightest is about half way between bottom left and the quarter-tone boundary, and the second one is a bit more than half way between the quarter-tone and half tone boundary.

The soft-proof is a bit brighter accross the range than are the prints. Tonal differentiation is better in the softproof. In the deep dark set, the upshot is that differentiation is extremely poor accross LAB 0,2,4,6. It becomes a bit more obvious at 8 and is somewhat too subtle between 8 and 22 relative to what the soft-proof indicates it could be. I'm taking these files into the retailer tomorrow along with the problem prints, so I can see on their monitor what they look like, and if they are patient enough with me, do both the tests and the photos in Epson and IP. If I can take them through all that, it should be determinative.