Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: nihil on January 09, 2012, 08:56:43 am
-
Hello,
has anyone compared the icc-profiles from Profilemaker and i1Profiler? I consistently get visibly lighter blacks with i1Profiler, across many different papers on Epson Pro 4900, 7900, 11880 and 3880 printers. I prefer the colors from i1Profiler but the poor blacks is a deal breaker for many images. PS: It's not THAT poor, I probably wouldn't have noticed unless I compared the same test page, in which the difference is clearly visible.
-
Have you measured the density of an all black patch to see what the % difference is? Theoretically a profiling program if properly implemented should not change the black point from what the native printer driver can deliver. I don't use either program, preferring instead to use ArgyllCMS and I know that the black point is not altered by the profile.
-
So far that's something I have NOT noticed in comparing side by side prints made with profiles from both apps. Using Relative Colorimetric on a good photo paper, you'd be hard pressed to see ANY difference between the two. If anything, the ProfileMaker profiles are actually better, and the difference is in blue skies, which are ever so slightly more magenta in i1Profiler. The Perceptual tables seem, so far to be better in Profiler, but I haven't tested them extensively yet, or on enough images on matte media, where it's likely to be more of a problem.
I'm printing my targets through the Adobe Color Printer Utility and measuring with a Spectroscan and a UV filter on both.
-
Here are some measurements made with BabelColor CT&A:
The paper is Permajet Portrait White, printed with Perceptual. I had two profiles made with each program, measured 5 times each and then averaged:
Profilemaker 1:
C: 1,56
M: 1,59
Y: 1,62
K: 1,57
Profilemaker 2:
C: 1,56
M: 1,59
Y: 1,62
K: 1,57
i1Profiler 1:
C: 1,49
M: 1,48
Y: 1,46
K: 1,49
i1Profiler 2:
C: 1,48
M: 1,48
Y: 1,46
K: 1,49
From these data it appears that i1Profiler tries to make a neutral black, sacrificing dmax, while Profilemaker simply gives the darkest black possible.
Judging by my eye (which is all that matters to me) - The Profilemaker ones seems to be neutral, or too black to really know what tint it is. (or by stretch ever so slighly warm) The i1Profiler blacks seems dark grey with a blue/cool tint. Tints aside, the Profilemaker blacks definitely looks nicer.
-
From these data it appears that i1Profiler tries to make a neutral black, sacrificing dmax, while Profilemaker simply gives the darkest black possible.
Yes, that's consistant with my findings as well. That's always been one of the differences between PMP and Monaco Profiler (MP) and now PMP and i1P. In the long run, I've seen several instances where I've visually preferred the results from MP and i1P profiles.
The other thing that might effect your results are your measurements. Have you compared results with and without UV filtration, or with other instruments?
Another thing that's hard to discuss with words is the "shadow contrast". I've seen instances where the shadow contrast was nicer with MP and i1P which lead to more visually pleasing results, more "punchy blacks" if you will, even though the Dmax values we're a little weaker. In the end I think it's important to let the visual results supersede the numbers.
-
The Profilemaker ones seems to be neutral, or too black to really know what tint it is. (or by stretch ever so slighly warm) The i1Profiler blacks seems dark grey with a blue/cool tint.
Of course your numbers suggest that the i1P blacks are neutral and that the PMP blacks are yellow. It's all relative right? Of course UV filtration can effect the entire gray axis is perceived and, in turn, how the shadow values are perceived relative to the rest of the gray axis.
Feel free to download my color evaluation image at http://www.on-sight.com/downloads/ and use the image in the lower left corner to evaluate the extreme shadows. How the lower densities are handled from 94%-100% are perhaps more important than the 100% black itself.
"Better DMax" could be defined a number of different ways. I prefer to define it, in part, by how the entire 94%-100% density range is handled, not just the raw numbers for 100%.
-
You may also want to try an optimization of the profile in i1P. The 2502 target at the URL works quite well (or email offline, I also have a 900 patch version).
http://www.i1upgrades.com/2011/08/how-to-use-the-tc-2502-gray-optimization-chart/
Be interesting if this makes a difference. It might point to the original target used (number of patches, where in color space the patches are generated). You do know you want to create the targets in i1P and then use it there AND ProfileMaker Pro right?
-
Of course your numbers suggest that the i1P blacks are neutral and that the PMP blacks are yellow. It's all relative right? Of course UV filtration can effect the entire gray axis is perceived and, in turn, how the shadow values are perceived relative to the rest of the gray axis.
I use an eye-One Pro. I don't know if it is UV-sensitive or not. Any insight about this would be appreciated! Is there any settings regarding this that might make i1Profiler behave more like PMP regarding shadows?
-
"Better DMax" could be defined a number of different ways. I prefer to define it, in part, by how the entire 94%-100% density range is handled, not just the raw numbers for 100%.
In terms of separation, smoothness and general quality of the shadows (except max black of course) I'd say the i1Profiler and PMP profiles are pretty comparable.. Which wins slightly seems to vary by printer and paper.
For most images I'd say deepest possible black is not that important. But have a look at mine and you'll understand how crucial deep blacks are to me .)
-
You may need then, to run a few tests with the ink density option in the Epson driver to see if running at +5 percent or +10 percent or something like that make a visible difference to your images. Some papers need a little bump to get the most out of them in terms of highest d-max. Print the target with, say, +5, +10 and +20 percent settings, read and profile all three, make test prints and then repeat with intermediate settings if desired. Just remember to make any subsequent prints with the same settings you printed the best target with.
You may very well find that that method gives you more pleasing prints for you individual style and preference. You can also try a modest edit in ProfileMaker's edit module to bump the blacks as well.
-
You may also want to try an optimization of the profile in i1P. The 2502 target at the URL works quite well (or email offline, I also have a 900 patch version).
http://www.i1upgrades.com/2011/08/how-to-use-the-tc-2502-gray-optimization-chart/
Be interesting if this makes a difference. It might point to the original target used (number of patches, where in color space the patches are generated). You do know you want to create the targets in i1P and then use it there AND ProfileMaker Pro right?
I will try your tip about the gray optimization chart soon, but something tells me that it will only optimize grays, and not really do anything about the pure black? For the record I made all my i1P-profiles with 1720-patch targets.
I didn't know you could use i1Profiler targets with Profilemaker, but I managed to figure it out. I exported the targets and measurements from i1P and created the icc-profile with PMP. The result is a profile almost exactly like the ones i made with PMP alone in every regard. Blacks as deep PMP too.
So the problem remains. i1Profiler for slightly better colors or PMP for much better blacks. As for i1Profiler trying to neutralize the black (into "neutral dark grey"), I experimented with the "Neutralize gray"-slider: With either minimum(0) or maximum(100) the results are identical to my eye. Perhaps because the paper (Permajet Portrait White) is already pretty white.
-
The target contains pure black, so it may.
-
The target contains pure black, so it may.
I measured the gray optimization charts and here are the results from BabelColor CT&A (again 5 times each and averaged)
Profilmaker TC9.18:
C: 1,66
M: 1,69
Y: 1,73
K: 1,68
i1Profiler 1720 patches
C: 1,59
M: 1,59
Y: 1,57
K: 1,60
i1Profiler 1720 patches + 2520 patches gray chart
C: 1,60
M: 1,62
Y: 1,62
K: 1,61
So there is a slight improvement. Now at least the blacks don't have that blue tint (to my eye). But that's a lot of work for not getting even half-way to the Profilemaker ones.. Thank you for your tip though.
-
not really this topic.. but i just read the 4096 patches target yesterday with my DTP 20 for harman baryta glossy.. and used the same data to build 2 profiles with PM (color and crome) and one with the APS of HP and get totally !!! different colors using perceptual.. even the extreme colors are viewable different!.. the shadow detail is visibly best with PM crome..
-
Hello,
has anyone compared the icc-profiles from Profilemaker and i1Profiler? I consistently get visibly lighter blacks with i1Profiler, across many different papers on Epson Pro 4900, 7900, 11880 and 3880 printers. I prefer the colors from i1Profiler but the poor blacks is a deal breaker for many images. PS: It's not THAT poor, I probably wouldn't have noticed unless I compared the same test page, in which the difference is clearly visible.
Sorry to be late on this subject. I recently purchased a new desktop printer (a Canon iP4800 series) and, because most pictures I will print with it does not justify paying OEM-ink extra costs (I don't sell prints and many of them are just to be shared with friends and relatives), I was waiting for some ordered non-OEM inks to arrive. After the initial profiling round, I unfortunately noticed that the i1Profiler generated profiles render too light blacks for most papers I use. For some of them (Ilford Galerie Smooth Pearl 290g, Hahnemuehle Photo Rag Duo 276g, plain offset paper) this is not a deal breaker problem, as black, while lighter than what I get with Canon OEM ink and i1Profiler combination (or with non-OEM ink and PM5 combination), is acceptably dark enough. But for the rest, the achieved blacks are too light to live with. The images become dull, almost no punch and the overall contrast (not only at the shadows) is unacceptable. I could simply go back to PM5 for those papers and no more questions about this subject. But as many of you may agree, it's easy to get used to the qualities of i1Profiler: gradient smoothness, gray axis neutrality and natural transitions and saturation. On these things, i1Profiler has a clear superiority over the the old PM5. Unfortunately, i1Profiler has also an exaggerated tendency to inclination to neutralize all the gray ramp, throwing out even those deep shadows with casts that would be unnoticeable by naked eyes.
Justice must be made, and according to some additional tests I have done, I discovered that it's not all fault of i1Profiler though. The OCP dye-based black ink (the IJBK-124) and Canon drivers for the iP4800 printer have parts of them. Canon iP4500 printer with CLI-8BK OEM cartridge produces a raw black patch of Lab = (3.97, -0.61, -2.12), while the newer iP4800, also with an OEM black cartridge, gets a black patch of Lab = (5.65, -1.22, -1.59). Also, the IJBK-124 OCP ink is not only lighter than CLI-226BK OEM ink but also less neutral than the latter. It gets a raw black patch of Lab = (9.92, 3.70, -10.25). These scores were achieved with the same media settings and the same paper. As I already said, I could live with these L/Dmax scores (I get even darker black Lab of (6.52, 7.85, -6.96) with Ilford GS Pearl) but i1Profiler simplies throw them out because of they're too much off the gray-axis, so it rises the rendered black point to an unacceptable L value for most papers I have. As for the small part of papers I achieved almost neutral raw black with non-OEM inks (Hahnemuehle Rag Duo and plain offset papers), i1Profiler profiles gave me almost as good/dark black as PM5 ones did.
I'm currently tempted to mix the Canon OEM dye-based black with the other OCP dye inks, or try to find another OCP compatible black darker and/or more neutral than the IJBK-124 one, or even other non-OEM brand of ink. Before going on with this, does anyone know how to edit i1Profiler profiles so they could render shadows nearer to a not so neutral raw black? Or is it feasable to mix parts of profiles tables? What tool would do it?
-
[....]I could live with these L/Dmax scores (I get even darker black Lab of (6.52, 7.85, -6.96) with Ilford GS Pearl) but i1Profiler simplies throw them out because of they're too much off the gray-axis, so it rises the rendered black point to an unacceptable L value for most papers I have[....]
It's sad i1Profiler doesn't offer us the choice to neutralize/lighten blacks or not. If you or anyone else figure out a solution some day, please let me know.
-
It's sad i1Profiler doesn't offer us the choice to neutralize/lighten blacks or not. If you or anyone else figure out a solution some day, please let me know.
Yesterday I printed some targets on a cheap RC paper with the combination of cyan, magenta and yellow OCP inks and black Canon OEM ink. Profiling it today, those deep blacks came back but, because gamut grows a bit decentralized at the Lab diagram (although black is as dark as L=7, 1st and 2nd quadrants are only populated at L>=17), dark reds and dark greens (L<17) are badly printed. But it was the best result after profiling that printer/paper combination with all Canon OEM cartridges. It's now clearer than ever to me that the black ink darkness is not the only factor of the light black problem, but also its neutrality. While the OCP black ink is only 4 L units lighter than the OEM one, i1Profiler produces with the former a profile 10 L units lighter than with the latter. Maybe a better solution (besides using PM5 Chroma+ ou Colorful Paper-colored Gray) is to purchase non-OEM inks with hues as similar as possible with the Canon OEM inks, even their black were not so dark. Or to use a RIP which would control, at ink level, the balance between those inks. I expected PrintFab would do this for the desktop printer I use, but unfortunately the only controls that work before applying a colorprofile to the driver are "type of paper", "print quality" and "ink saturation limit", second to the response to an inquiry I posted at their website.
-
I've done some additional tests with the mixing of OEM and non-OEM inks and should say it has some problems that forced me to abandon this attempt. Surely the OEM black gave me a darker and more neutral black. But, as the other inks do not converge into this "black", there are some smoothnessless problems into transitions between shadows and pure black. So, the best solution is profiling with PM5 and choose Colorful Paper-colored Gray rendering indeed. PM5 PcG solves another problem with Canon dual pigment/dye black cartridges. As i1Profiler throws the absolute black out, its profiles never lead to usage of pigment black when printing on plain paper. There's no black ink, even for text. PM5 profile, otherwise, triggers the pigment ink when printing texts or even part of pictures with black ou very-near-black tones.
Hope one day X-Rite implements an option in i1Profiler that would imitate the Paper-colored Gray we have in PM5.
Another update: I downloaded yesterday a trial version of basICColor Print 3, which let me experiment all its features for 14 days. I reprofiled two papers (among them the most problematic ones regarding to the black lightness) using the i1Profiler data, which was imported seamlessly. I was impressed with some extraordinary features, like the possibility to implement a rendering similar to the relative intent with black compensation in place of perceptual intent (nice to be used with software without native BC). Also, it has a variable setting for OBA compensation, from 0 to 100, that could be finetuned better than the PM5 on/off feature. And best of all, basICColor Print 3 renders colors and gradients something like i1Profiler. Unfortunately, for both papers I worked on, its profiles produced a subtle yellow cast and the grayscales had some problems between midtones and shadows. Maybe it would have better performance if I used its own targets. I was afraid basICColor Print could magically solve everything because I wouldn't have 1,200 EUR to spend on that beast! Until now, the king of not-so-good combinations of printer, ink and paper is Profilemaker indeed.
-
You may also want to try an optimization of the profile in i1P. The 2502 target at the URL works quite well (or email offline, I also have a 900 patch version).
http://www.i1upgrades.com/2011/08/how-to-use-the-tc-2502-gray-optimization-chart/
Be interesting if this makes a difference. It might point to the original target used (number of patches, where in color space the patches are generated). You do know you want to create the targets in i1P and then use it there AND ProfileMaker Pro right?
Interesting link, thanks Andrew.
But... back to the OP - I too have noticed far better DMax in all colors with older PM profiles than i1 profile. its dramatic and clearly visible to the untrained eye. Both in profiles made professionally (by 'experts') and by me with iSis XL and i1profiler, so its not my spectro being out.
Amazed this is not getting more commentary.
-
It's sad i1Profiler doesn't offer us the choice to neutralize/lighten blacks or not. If you or anyone else figure out a solution some day, please let me know.
Yesterday I have the idea of playing with CMY adjusters at Color Adjustment tab of printer driver settings. I do not know if those adjustments are applied at an ink level or just at RGB level (as the raw 0,0,0 black does not change), but I decided to go on with this experiment anyway. I printed raw (color unamanaged) sets of 100%, 90%, 70%, 50%, 30% and 10% gray patches with different CMY adjusting values until I find the numbers which produced the most neutral raw grayscale after a reasonable curing time. A kind of preprofiling calibration. For example, the settings found were CMY = (-12, -4, 26) for a given paper. So I reprinted the usual set of targets, waited overnight and then reprofiled it using i1Profiler. The result was a black almost as dark as the one I have got with ProfileMaker/Paper-colored Gray and default printing settings. PM5 profile has still a bit more punch, but now i1P one has enough contrast combined with its native smoothness, natural saturation and color precision qualities. Also, I have got bigger gamut through all L scale, mainly at shadows. I think additional playing with Brightness, Intensity and/or Contrast parameters before printing targets would produce even darker black and better looking profiles, but for now I have not enough time to test all those possibilities.
So, the problem with i1Profiler is, indeed, its insistence to neutralize the grayscale at all costs. When the paper/ink combination naturally prints not-too-high a/b values for gray raw patches, i1Profiler is ok. But when this is not the case, i1P absurdly rises, without limits, black point until it finds a neutral one. I wish i1Profiler neutral-gray dimmer for perceptual rendering could produce more visible difference between 0 and 100 values, not only for lighter shades of gray but also for those near-black ones -- something like the differences we get between PM5 Neutral Gray and Paper-colored Gray settings.
-
Hi guys,
I know this thread is old. I just tried i1profiler on the weekend and I must say, I am surprised, how light the blacks are compared to PMP5.
I observed that setting the profile generation options to "Colorful" improves this situation somewhat but also boosts the contrast.
It would be great if someone could give me some input how to fix this problem.
I read this thread here (http://www.luminous-landscape.com/forum/index.php?topic=72499.0) but I cannot derive any actions to be taken that could help in solving the problem.
Is everybody just tolerating this lower DMax now? Boosting the contrast cannot be the final solution because it alters the image...
Cheers,
Jani
-
Adjusting the contrast does nothing for dmax. Also, it affects only the perceptual tables only. Either live with it or use PM5. A lot of things are wrong with i1Profiler.
-
i1Profiler stopped crashing blacks since v1.4 or v1.5 - what revision are you using?
-
i1Profiler stopped crashing blacks since v1.4 or v1.5 - what revision are you using?
Really? Where did you get this information from, can you provide any links to Release Notes/Bugereports or any threads in this or other forums that confirm this. I am using 1.5.0.
-
Just adding a side comment that might help for those exploring all of this...
I found the Saturation rendering intent w/ an increase of ink density on my 9900 produced some wonderful color and Dmax, however, it's smoothness
is quite harsh. This using i1Pro built profiles.
Speaking with DataColor at a Tradeshow, and via emails, here's a bit of information that might help:
"Datacolor does offer an exceptionally good Saturation Intent in our printer profiles, and a good alternative for where that is not ideal: our 'Perceptual' Intent is actually not Perceptual (meaning it does not blindly shrink the gamut), instead it is halfway between our Colorimetric Intent, and our Saturation Intent. This allows max color saturation with Sat, a step back from that with Percep, and then the most literal color with Colorimetric. Its a good system. "
I'm going to pull the trigger on their SpyderPrint, and see how this compares to i1Pro
-
Ja i i do not know your print environment, nor do i have pmp5. I use iproviler 1. 5 witb i1pro2 and have little issue with max black. Depends on papers but on a gloss pe photopaper around 2.5 and on a fine art mat around 1.6 to 1.7 . This with a smooth tone curve.
-
"Datacolor does offer an exceptionally good Saturation Intent in our printer profiles, and a good alternative for where that is not ideal: our 'Perceptual' Intent is actually not Perceptual (meaning it does not blindly shrink the gamut), instead it is halfway between our Colorimetric Intent, and our Saturation Intent. This allows max color saturation with Sat, a step back from that with Percep, and then the most literal color with Colorimetric. Its a good system. "
Sounds like a lot of marketing speak to me. There are no rules in how one creates a perceptual rendering intent. Just like there are no rules in how a film manufacturer could build a rendering into their film stocks. So the: does not blindly shrink the gamut seems a bit disingenuous, implying everyone does this incorrectly but they have some magic process. You should test this but not based on what they wrote above.
-
Really? Where did you get this information from, can you provide any links to Release Notes/Bugereports or any threads in this or other forums that confirm this. I am using 1.5.0.
I had horrible problems with crashed shadows on c-prints - profiles generated by earlier i1Profiler revisions had distinctly crashed blacks. With i1Profiler 1.5.0 the problem is virtually eliminated, RGB 0,0,0 is only slightly neutralised, so the resulting RGB values after conversion are usual 0 for one channel, and 1-2-3 for other two channels. It's still not perfect, but much better than before.
-
I found using i1 Profiler software with the i1 PhotoPro 2 head today (Windows 8.1-64), if I use Qimage as the pre-print tuning software, I can set a boost to the black part of the curve (In Qimage's "+Color" tab) to regain shadows lost or crunched up in a 21 step gray tablet. If I don't, often it will bunch the blacks up around the 70% part of the 21 step scale and anything beyond 70% of the scale goes dark. I saved it as a filter to apply to other normal images and it did help with regaining shadow detail.
Been monkeying with it today on a new glossy paper and it needed a good kick in the black and Profiler's Optimization didn't address it either. Oddly, letting PhotoShop Manage Colors produced a smoother gray scale too that the one the i1 produced that bunched up the blacks. Saw zero affect to applying Perceptual, Absoute Colormetric, Relative Colormetric, or Saturation in Qimage either from 4 prints made with each.
SG
-
however, it's smoothness is quite harsh. This using i1Pro built profiles.
Setting the saturation slider in i1Profiler to zero from the default of 40 does a wonderful thing for smoothness.
-
V1.5 renders the grayscale slightly darker towards the black point than earlier versions of i1Profiler for me too. This is by lowering the separation between the darkest tones, albeit slightly, and it does nothing to better the dmax for what I profile. My profiling experience is limited to relatively neutral native output, so I don't face any issues with dmax. Like Jan, I've been happy with the quality of the profiles after quite a bit of tweaking and investigation. V1.5 compared to earlier versions of i1Profiler has smoother gradations in dark blues - that the main tweak they have made in the profiling engine. It does sacrifice a bit of blue edge gamut separation.
But V1.5 is very buggy for me. Lost of connectivity with i1 Pro 2 at random (halfway during a large patch measurement process is not fun), increased errors in measurements and a strange bug when measuring light patches or paper white, very slow profile generation (sometimes very very slow at random) and more resource hungry than V1.3.2. Crashes without warning frequently.
-
i1Profiler, like Monaco Profiler before it neutralizes the grey axis all the way tot the black point, whereas PMP leaves the black point alone. So PMP will leave the black point at RGB=0,0,0 whereas MP and i1P might set the black point to RGB=0,23,8 to try and neutralize a magenta, yellowish cast. I've suggested that i1P give us the ability to choose between the two behaviors but I think that's wishful thinking. Neutralizing the black point is probably the right approach and a colorful black point represents a failure in the calibration process prior to profiling. All said I'll take the i1P results and exceptional Perceptual rendering anyway. Perceptually it is fantastic. If you have a beef with it the correct approach is to improve the calibration prior to profiling.
-
So re-reading this old thread, what is the recommended path for getting best out of i1 profiler now - use Sat or Colorful intent? increase ink density +5/10/15% on the Epson panel? reduce or increase the smoothness slider?
It is very confusing - but I, like others here, found that i1 Profiler reduced the DMax compared to profiles made on older software. Generic profiles made by the paper manufacturer 3-5 years ago on PMP often give far better max density in all colors than i1 profiler, but do not have the smooth gradations Profiler does. You win something, you loose something - it is very smooth, with no uneven color steps, but at a real hit to max density of each 100% CMYK squares. not good. The explanations of i1 attempting to neutralise the black at cost to density make some sense, but do not explain all.
I was lucky enough to download the optimizer chart & a page grab when it was released (the blog post and link has subsequently been removed by x-rite) and that helps a bit with greyscale smoothness, but I can't notice any meaningful change to max density. (I suspect that grey scale optimizer chart will be added back to Proilfer 2.0 as a 'feature' when it is released, and we'll all have to shell out a $500 upgrade fee to get it!)
-
It is very confusing -
It is and you can thank X-rite for putting in so many sliders that in some respects do little or nothing or turn around and produce visible differences. All without a lick of documentation! Best you can do is test and come up with a rough set of settings. Or move to another solution (expensive I know). Or use legacy software (the reason I spent all of $160 on eBay for an old MacBook that can even boot OS9). Or continue to put pressure on X-rite to get their act together.
-
So re-reading this old thread, what is the recommended path for getting best out of i1 profiler now
Start with an excellent calibration (ink densities set optimally for chroma peaking, and perfect linearization) and you'll get perfect profiles from i1P. Calibration and profiling are two different pieces to the puzzle that must fit together to get optimal results.
Now, if you're using a RIP there are ways an analyzing the ink densities and developing optimal linearization curves externally and then using them in the RIP but we won't go go down that rabbit hole here. If you're using the driver you may need to do some experimentation to determine what the optimal settings are for a given media. Again, too much to go into here.
i1Profiler makes awesome profiles. If you're having problems with Dmax then that's a calibration problem, IMO, not a profiling problem. Don't blame the profile if it's the calibration (driver) settings that aren't right. Yes, PMP may have delivered better Dmax from a poor calibration but there are problems with PMP's approach too, namely colorful blacks. The bottom line for me is that if you're using optimal driver and or RIP settings you'll love the blacks you'll get from i1Profiler profiles, especially with Perceptual.
use Sat or Colorful intent?
Perceptual.
increase ink density +5/10/15% on the Epson panel?
Again, you may need to do some testing to determine the optimal settings for a given media. More is not necessarily better! If you have the tools, chart the chroma of the CMYK channels and look for the peaks. This is an advanced topic and a rabbit hole... If you don't have the tools, I wrote an article years ago about a beginner's approach to determining optimal media settings in a printer driver with the need for a spectro here: http://www.on-sight.com/how-to-determine-the-optimal-media-selection-for-any-paper/ It doesn't mention the ink density slider in the Epson driver because it didn't exist back then, but you get the idea and can include this slider in your testing.
reduce or increase the smoothness slider?
I'll save you the hassle of a lot of testing and encourage you to just set it to 70.
-
Start with an excellent calibration (ink densities set optimally for chroma peaking, and perfect linearization) and you'll get perfect profiles from i1P. Calibration and profiling are two different pieces to the puzzle that must fit together to get optimal results.
Thanks Scott - I have to be honest here and declare that a profile I paid to have custom made by you ~4 years ago (before I bought my own iSiS, etc) on i1 profiler was not as good as the generic PMP one from the paper company in the DMax. Your one certainly had far better smoothness but significantly reduced DMax. Same media settings. You remade it twice with boosted Color and Sat, but it changed nothing in the max densities reached. I used the old generic one for my prints. I do not say this to embarrass you, far from it, I have learned a lot from your helpful posts, but simply to point out that Color Pros like yourself and Andrew here, could not get results from X-rite latest and greatest that equaled the old profiles from Monaco etc in a very very important area - maximum color density. It is a remarkable deficiency in an expensive program.
-
Interestingly, I have managed to produce a profile with better dmax in i1Profiler than i1Match (which is basically the PMP profiling engine at default settings) with the same measurement file.
I'll save you the hassle of a lot of testing and encourage you to just set it to 70.
I have done a lot of testing too and I humbly disagree. Setting it to 70 will reduce overall gamut slightly and will also reduce the profile's black point slightly. Keeping it to 65 or less will not do either but is still effective enough to smooth any bumps in the CLUTs. If possible, just leave it at zero.
Also, I think narikin was referring to the two default choices for profile settings in i1Profiler, colorful or saturation, not the rendering intent. Neither will do anything to affect dmax or the overall gamut.
-
Thanks Scott - I have to be honest here and declare that a profile I paid to have custom made by you ~4 years ago (before I bought my own iSiS, etc) on i1 profiler was not as good as the generic PMP one from the paper company in the DMax.
What's your real world name? I'd like to look back at those profiles. If it was 4 years ago I was using Monaco Profiler back then... I'd like to know what process we're talking about...
Also, how are you determining the Dmax differences? Are you comparing color gamut renderings or taking measurements off of the prints? I would argue that the later is the only fair way to do such a comparison.
Your one certainly had far better smoothness but significantly reduced DMax. Same media settings. You remade it twice with boosted Color and Sat, but it changed nothing in the max densities reached. I used the old generic one for my prints.
I see quell the appropriate thing would be to have let me know at the time so we could discuss and address it. Bringing it up in public years later is poor form. Perhaps we did discuss it before - let me know your real world name and I'll take a look.
...simply to point out that Color Pros like yourself and Andrew here, could not get results from X-rite latest and greatest that equaled the old profiles from Monaco etc in a very very important area - maximum color density.
I've yet to see i1Profiler produce a profile that's inferior to Monaco Profiler. When I compare PMP, MP and i1P profiles on a variety of printing processes I rarely see what you're talking about. Would love to take a look at what you're seeing. Tell us more could you!
-
I have learned a lot from your helpful posts, but simply to point out that Color Pros like yourself and Andrew here, could not get results from X-rite latest and greatest that equaled the old profiles from Monaco etc in a very very important area - maximum color density. It is a remarkable deficiency in an expensive program.
I'd agree! I recently rebuilt some profiles for a client where we had the time and energies to send the same averaged data (a huge amount of it) through various profile products including i1Profiler, PMP, MonacoProfile and Copra. We had various output technologies (CMYK) on differing substrates. We sent a lot of actual images and synthetic's though each and what we found were instances where the Copra profile was better (in some areas), i1P was better in others, PROFILER better yet in other areas. Each has certain strengths and weaknesses. You have observed one area where you are not as happy with i1P. And that's no surprise. The kinds of images you send are telling to! Some synthetic image (like a Granger Rainbow but better) tells you one thing about the color engine. The Roman 16's tell you another. In the end, you have to send a lot of such tests through the profile and look at the areas that are most important to you to select which product will be used to build the final profile.
-
I've yet to see i1Profiler produce a profile that's inferior to Monaco Profiler.
Based on my work, that indicates you need to send a lot more differing images, real and synthetic through both because depending on a lot of factors, MonacoPROFILER can indeed show better results.
-
Based on my work, that indicates you need to send a lot more differing images, real and synthetic through both because depending on a lot of factors, MonacoPROFILER can indeed show better results.
I've printed ALOT of images, on a lot of processes. I do agree that one might like features in one profiling solution over another in different situations. There are advantages sand disadvantages to easy system. But do tell me more about your observations, particularly regarding Dmax on RGB inkjet profiles (since that's what I think we're talking about here)!
-
I've printed ALOT of images, on a lot of processes. Tell me more about your observations, particularly regarding Dmax!
It isn't solely about Dmax! The tests were designed to look at as much of what's happening within color space as possible.
I can't go into detail for the following reasons:
1. The project is under NDA.
2. Many of the images used to exmaine the results of the profile were developed by Karl Lang (I name I suspect you recognize) and they are his properly. But they are designed specifically to look at many areas of color space and detect issues with the color engines and they do an amazing job of this task.
3. This was CMYK output on very specific kinds of presses and substrates so YMMV depending on what you are profiling.
I can say that it was easy to see some issues with blue mapping from Copra profiles where it won in all other areas. I'll point out that sending reports to them resulted in quick fixes to improve this mapping although not enough to our full satisfaction. But it was really nice to deal with a small company that both recognizes they can improve their profiles and did so quickly enough that we could use them for this job.
For reasons I can't discuss, all the testing was done with Perceptual rendering. That opens a lot of potential can of worms as you know.
None of the various profile packages were perfect, none didn't show some flaws. In the end, it was about selecting which deficiencies we could live with and again, in some cases i1P profiles were used in process A while Copra was used in process B. PROFILER in process C, etc. Again, you have to spend a lot of time and with the right tests to determine this but I'll say, based on this work, i1P was not the winner in all cases, hence my suggestion you throw a lot of images and output into the testing before picking!
-
Andrew, yes, I'm not surprised that you're seeing better results elsewhere for CMYK profiles and of course I know Karl, but what I'm trying to get to the bottom of here is Narikin's assertion (as I understand it) that i1P's RGB profiles have poor DMax. I'd love to hear from him, get his name (so that I can look up the profiles I did for him that's he's saying were poor) and hear what printing process he's talking about. Something just doesn't sound right there...
-
So re-reading this old thread, what is the recommended path for getting best out of i1 profiler now - use Sat or Colorful intent? increase ink density +5/10/15% on the Epson panel? reduce or increase the smoothness slider?
It is very confusing - but I, like others here, found that i1 Profiler reduced the DMax compared to profiles made on older software. Generic profiles made by the paper manufacturer 3-5 years ago on PMP often give far better max density in all colors than i1 profiler, but do not have the smooth gradations Profiler does. You win something, you loose something - it is very smooth, with no uneven color steps, but at a real hit to max density of each 100% CMYK squares. not good. The explanations of i1 attempting to neutralise the black at cost to density make some sense, but do not explain all.
I was lucky enough to download the optimizer chart & a page grab when it was released (the blog post and link has subsequently been removed by x-rite) and that helps a bit with greyscale smoothness, but I can't notice any meaningful change to max density. (I suspect that grey scale optimizer chart will be added back to Proilfer 2.0 as a 'feature' when it is released, and we'll all have to shell out a $500 upgrade fee to get it!)
Would like to see some measurement of Dmax and how it was measured. Until then this is about blind guessing which shade of grey each of us talking about, yet assume were are all talking about the same shade. Also Dmax is just 1 (one) point in a gamut that one observes in its total behaviour (very large number of point if you like) when looking at a print.
-
Maybe a database with Dmax numbers for different papers could shed some light here or should I say make things less gray. There is still my plan to add Dmax / L-min numbers to SpectrumViz for up to 4 matte black and 4 gloss black inks. Two Epson desktop printers with 6 channels each are cleaned and prepared for that job. I made a conversion of the QTR partitioning target for this purpose. All I asked was to send me some near empty PK and MK carts of Canon and Epson wide formats, HP Vivera I have here. MIS Eboni etc and possibly Cone could be added. All I got was 1 near empty 3880 cart, thank you Peter.
In practice I have seen that profiling software usually cuts at Dmax compared to the best black sample in the calibration patches. To get the tone range correct in B&W with QTR tools for example has less impact.
--
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
April 2014, 600+ inkjet media white spectral plots.
-
i1Profiler, like Monaco Profiler before it neutralizes the grey axis all the way tot the black point, whereas PMP leaves the black point alone. So PMP will leave the black point at RGB=0,0,0 whereas MP and i1P might set the black point to RGB=0,23,8 to try and neutralize a magenta, yellowish cast. I've suggested that i1P give us the ability to choose between the two behaviors but I think that's wishful thinking. Neutralizing the black point is probably the right approach and a colorful black point represents a failure in the calibration process prior to profiling. All said I'll take the i1P results and exceptional Perceptual rendering anyway. Perceptually it is fantastic. If you have a beef with it the correct approach is to improve the calibration prior to profiling.
I doubt neutralizing the black point can actually be done as it will take way more color ink (cyan mainly) to get there than the coating can handle while Dmax then really would go down the drain. It is always a compromise and the odd thing is that the higher the Dmax is there is less need to neutralize it, the color cast loses effect with less reflection. The compromise is where to let go of neutralization and at what rate. Best solution of course is more neutral gray-black inks. The HP Zs have an advantage there, at least the PK and Gray inks, the MK is not neutral either though but in the media presets that use MK, PK, Gray, Light Gray the PK can at least go a long way in the shadows. MIS (Image Specialist's) Eboni is actually the most neutral carbon matte black ink, still a bit warm. I do not think that the underlying paper white is a factor on the neutrality at good Dmax which is something else than observing the color contrast between white and black in a checkerboard image on OBA papers.
--
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
April 2014, 600+ inkjet media white spectral plots.
-
Ernst , near empty? I can send you my replaced epson 4900 mk and pk cartridges. What address?
-
Ernst , near empty? I can send you my replaced epson 4900 mk and pk cartridges. What address?
Is doorgegeven via je website. Merci.
--
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
April 2014, 600+ inkjet media white spectral plots.
-
Just came across this old thread. I have a Epson 1400 (dye ink) but have bought OCP 3rd party inks. They black is more bluish than the Epson OEM, and when I made my own profiles I lost a lot of depth, never got any real blacks. Very neutral though (Still going to optimize it). My profile was made in i1Profiler 1.5.6 with an i1Pro2. After reading this I exported my measurements and imported them into PM5, and made 2 profiles. One with neutral grey, and one with paper grey.
Measured the D-max on the i1P and the 2 different PM5 profiles I got the following:
i1P: 1,67 Dmax
PM Neutral grey: 1.53 Dmax
PM5 paper grey: 1.99 Dmax.
Since the OCP black has a colorcast, I am guessing using neutral grey in PM5, and i1P tries to neutralize it using different inks, and when going lower than this It can't correct it, and thus caps of as the blackest (neutral) black it can create. But using paper grey on an Epson lustre paper with and L104.1,a-0.2,b-4.1 it can create a deeper black since the paper is somewhat bluish also? Could this be correct?
Update: Measured the blackest black as lab 9.9, -0.7 -11.4.
On a real print the paper grey gives a better impression, with a better bottom and you can't really notice the not so neutral black.