Epson SureColor P800

[digitaldog]

Okay, agreed. No more rancor.

Now you're moving away from the childish.

[digitaldog]

Great, but that's NOT what Bob originally said.

He said nothing about ultimate printed output, nothing about accuracy. The reality is that many care more about pleasing color than they do about accurate color. You may prefer Adobe's pipeline, Bob prefers Epson's ONLY as far as the P800 / Epson Ultra Premium Lustre Photo Paper goes - now who's to say he's wrong and you're right ?

I and other's have no beef with that opinion. Read the posts again! I've outlined specifically what technical points he's made that is unsound and incorrect from the perspective of color management. See the text with the specific bullet points that he made that are either totally and technically wrong or use language that is unclear and sounds made up (the bit about S/N and his ideas about algorithms is pure speculation).

Did anyone here say that PMC is poor or people should not use it? No.

As for being obnoxious, I really don't care what you think of my attitude. I do care that technically correct and factual text about color management and digital imaging are expressed here for those who need to learn the facts. Much of what Bob wrote doesn't fall into that camp. He's entitled to believe that science fiction but other's who know as little as he does on this subject need to get peer review comments that ARE accurate or they end up as confused and posting more technically incorrect text for others to read and unfortunately accept.


Do you really believe Bob has Adobe and Epson's source code, understands it if he did (he doesn't) and can actually post one is superior? If you do, you're in the same unreality bubble and none of us can help you. 

[Alan Goldhammer]

Bob is happy with his workflow for that printer/paper combination.  this really should be the end of the story but of course egos are too big to let it alone.

[Schewe]

Bob is happy with his workflow for that printer/paper combination.  this really should be the end of the story but of course egos are too big to let it alone.

Which would be fine if he didn't advocate a suboptimal CM workflow...sorry in this day and age, I don't accept the concept of "alternative facts"...

:-)

[digitaldog]

Which would be fine if he didn't advocate a suboptimal CM workflow...sorry in this day and age, I don't accept the concept of "alternative facts"...

:-)

Yup, it's the alternative 'facts' that were easily dismissed that some of us keep pointing out, not his use of PMC. Got zero issue with how he prints whatever he prints. 

[BobDavid]

I just finished reading Mark Segal's impressive article on various MOAB papers and output from the Epson P800 and Canon Pro-1000. His methodology is tight, and no doubt the data has integrity.

I'm sticking to my guns: The  extremely specific combo of letting the P800 manage output on Epson Ultra Premium Luster Photo Paper, is fine for pleasing color.

Incidentally, using a closed-loop and inflexible "printer managed" workflow, is an asset when running limited editions.

Finally, if my prints were worth $$$$, I wouldn't be using a prosumer  printer and pedestrian media.

[digitaldog]

I'm sticking to my guns: The  extremely specific combo of letting the P800 manage output on Epson Ultra Premium Luster Photo Paper, is fine for pleasing color. 

Do so, that's fine as no one has said you shouldn't use PMC or not stick to your guns about what print path you desire.


What you should not stick to is the following (again) which is mostly wrong or if not wrong, ambiguous and without a lick of proof:

• No need to soft proof, the monitor is setup for aRGB.
• Try "absolute
• Rendering intents are disabled when the printer manages output.
• Higher S/N ratio: Adobe's algorithms are one step removed from the printer's
• Truth is, I think the Epson algorithms are superior to Photoshop's. (Truth: You Bob of course didn't examine (understand?) the source code from those two companies).

[BobDavid]

Schewe: 'Which would be fine if he didn't advocate a suboptimal CM workflow...sorry in this day and age, I don't accept the concept of "alternative facts"...

:-)'

I'm not advocating or recommending anything to anyone. Here's my original post:


'I've been using Epson Stylus Pro wide format printers since 2006. I've owned a 9800, 7800, 9900, and a 7890. I loved the 7890 for its reliability and output. When my wife and I decided to move from Florida to Colorado, I sold the 7890. I didn't want to risk shipping it 1,800 miles. ... And at this point, a 17" printer is sufficient for my needs. I ultimately decided on the SureColor P800.

Going from a pro printer to a prosumer printer didn't thrill me. Epson's professional 17" printers are over twice the price of the P800. When I owned and operated a studio, the P800 would not have been up to the task, neither would any 17" printer.

There are really only two options for a 17" prosumer printer: 1) Canon imagePROGRAF PRO-1000; 2) Epson P800. 

The P800 has exceeded my expectations: It makes stunning prints. The coolest thing about the printer is that I let it manage output rather than Photoshop. Who would have ever thunk that that would work. Truth is, I think the Epson algorithms are superior to Photoshop's.'

It seems all the flap/arguments center around the comment 'I think the Epson algorithms are superior to Photoshop's.' That that observation has rankled so many is interesting. Of course I qualified the observation by discussing one paper and one printer, in one particular situation.

Merely noting an observation-based opinion is not the same as pushing "alternative facts," or bending the truth to promote an agenda.

It is true that when the printer assumes management, it overrides rendering intents specified in the Photoshop printer dialogue box.
 
Unlike Rodney, I have not looked at Adobe's source codes. I suspect no one on this thread has recently, if ever,  been allowed to delve into Epson's proprietary hard code (burned into the chip set) and every line of software code.

[digitaldog]

Truth is, I think the Epson algorithms are superior to Photoshop's.'

Speculation. One may indeed produce a better print. There's a lot more going on just in the OS to make a print than the two other components you think you know is producing a 'better' algorithm!

It seems all the flap/arguments center around the comment 'I think the Epson algorithms are superior to Photoshop's.' That that observation has rankled so many is interesting. Of course I qualified the observation by discussing one paper and one printer, in one particular situation.

No, the flap is far more and has been shown to you and others multiple times here.

Merely noting an observation-based opinion is not the same as pushing "alternative facts," or bending the truth to promote an agenda.

It is an alternative fact when you bring algorithm's from differing companies you have no idea about into the discussion.

It is true that when the printer assumes management, it overrides rendering intents specified in the Photoshop printer dialogue box.

It is true expect when it's not. Because it depends. Even Adobe points that out!

Unlike Rodney, I have not looked at Adobe's source codes. I suspect no one on this thread has recently, if ever,  been allowed to delve into Epson's proprietary hard code (burned into the chip set) and every line of software code.

I've never looked at the code and clearly we know you haven't either. You don't have to suspect further as you've done originally; no one here has been allowed to delve into any proprietary code or if they have, they work for the companies or have an NDA. Further, you will never see them making made up statements about the code, let alone statements based on fact concerning these algorithms. Those who know, don't talk. Those who don't know, talk a lot as we've seen here about who's algorithms are superior. It's simple a figment of the imagination of the writer based on speculation. Enough said. 

[BobDavid]

Bob is happy with his workflow for that printer/paper combination.  this really should be the end of the story but of course egos are too big to let it alone.

Alan, you're right. It's time to let it alone.

[digitaldog]

Alan, you're right. It's time to let it alone.

Yup, just ignore the bits that are factually wrong about color. After that, print any way you desire.

[BobDavid]

Yup, just ignore the bits that are factually wrong about color. After that, print any way you desire.

No more rancor from me.

[Doug Gray]

I'm sticking to my guns: The  extremely specific combo of letting the P800 manage output on Epson Ultra Premium Luster Photo Paper, is fine for pleasing color.

I'm curious about two things. Since the P800 Printer Manages results are quite pleasing to you, the gamut mapping they use could be interesting and I'm curious as to how much it differs from using canned profiles and Perceptual Intent.

Gamut mapping, that is not image dependent, could just as easily be included in a supplied profile Perceptual BtoA0 table but, for some reason, were not in the two printers I have. Perhaps the motivation is market segmentation and addressing customer expectation differences in them.

However, by not doing so the printer companies have removed the ability to soft proof against the printer driver gamut mapped prints and that's quite unfortunate and completely unnecessary.

In any case, if you have the tools to measure color charts it would be interesting to see how these colors are mapped. The following post has an attachment, LAB10Ref.tif, and if you print it using your preferred "Printer Manages Color" settings it will tell quite a bit about the gamut mapping the P800 defaults produce. If you don't have the tools to measure them I would be happy to scan them and post the results. I can send you a mailing address by PM.

http://forum.luminous-landscape.com/index.php?topic=120679.msg1000735#msg1000735

[digitaldog]

Gamut mapping, that is not image dependent, could just as easily be included in a supplied profile Perceptual BtoA0 table but, for some reason, were not in the two printers I have. Perhaps the motivation is market segmentation and addressing customer expectation differences in them.

Image dependent how? You mean picking the RI by each image? 
Yeah, the lack of soft proofing and selecting an RI isn't ideal.

[digitaldog]

On some images (many?) in some color spaces, going out to print, the visual differences between RelCol, Perceptual and even Saturation (Absolute, no) is significant. I can't see spending the ink and paper and money to make even an 8x10 without viewing those options before printing. Put output specific edits and paper/ink simulation aside and consider just those three options.

[Doug Gray]

Image dependent how? You mean picking the RI by each image? 
Yeah, the lack of soft proofing and selecting an RI isn't ideal.

Gamut mapping, which is what the Perceptual Intent does, is not image dependent. It's purely a function of each pixel's color. Here's a description of the terminology.

https://www.researchgate.net/profile/Jan_Morovic/publication/236121201_The_Fundamentals_of_Gamut_Mapping_A_Survey/links/560511e108ae8e08c08adae2/The-Fundamentals-of-Gamut-Mapping-A-Survey.pdf

First, one of the most noticeable trends in the reviewed gamut mapping work is the agreement among different studies that image-dependent methods are preferred over medium-dependent methods, which is in some sense supported by a number of sources. At the same time, however, there is some work that suggests that while determining gamut mapping on an image-dependent basis gives better results, it is not an image’s gamut that determines how it ought to be gamut mapped.

Second, there is significant number of studies where clipping is given preference over compression

Perceptual intent can be and is used for gamut mapping though I1PRofiler's is rather minimalist.

[digitaldog]

Gamut mapping, which is what the Perceptual Intent does, is not image dependent. It's purely a function of each pixel's color. Here's a description of the terminology.

I would think the same is true of any RI. How does an ICC Profile know anything about an image? I'm I thinking of a different concept of image dependent?

[Doug Gray]

I would think the same is true of any RI. How does an ICC Profile know anything about an image? I'm I thinking of a different concept of image dependent?

It's true but what's allowed varies with the RI.

Relative Intent is specified for all PCS colors not outside the device gamut so there is no flexibility there except for how out of gamut colors are mapped to the gamut surface. Perceptual and Saturation Intent allow for arbitrary gamut mapping though Saturation intent is supposed to favor preserving hue rather than clipping to the nearest gamut surface

ICC profiles in any RI can only do image independent gamut mapping*. Image dependent mapping can include things like comping for the dynamic range of a scene by decreasing or increasing contrast. But this really is subjective and people are better at knowing what they like in a picture than neural nets. So far.

I'm pretty sure I've seen you point that out several times but perhaps used different terms than "independent gamut mapping" as you have a very good understanding of how ICC profiles work in practice.

[Schewe]

Gamut mapping, which is what the Perceptual Intent does, is not image dependent.

While the mapping isn't image dependent, the success of that mapping is...as is the starting color space. The only way to know what intent os best is either do a sample print or use soft proofing...

[Doug Gray]

While the mapping isn't image dependent, the success of that mapping is...as is the starting color space. The only way to know what intent os best is either do a sample print or use soft proofing...

Indeed. And that's why people will often try Relative Intent w/BPC as well as Perceptual or even Saturation and pick what looks best. I've had a few cases where relative Intent w/o BPC is best.  And that can only be done by soft proofing or killing more trees.