Imageprint - worth the $ ? - compare to QImage? other views please!

[PeterAit]

In the world of art, does "technically superior" even matter? There's way too much emphasis on the technical, and let's face it, the technical aspects of photography are neither particularly interesting nor really challenging.

I had a refreshing experience the other day - I was showing some photos to friend who is an excellent painter. She had some very helpful and interesting things to say about the images, their composition and lighting. Never once did she ask about my equipment or software, nor did she press her nose to the print to squint at the sharpness.

[Sven W]

Metamerism, while in the past it was a major concern, is a non issue in the inks today with yellow ink..... Eleanor

That's true, but the yellow ink will lower the display rating for your b/w prints.

/Sven

[eleanorbrown]

Svan, I simply don't care if my toned prints using Adv. B&W in the Epson driver have a "slightly" lower display rating  than those toned prints made through Imageprint.   It's a non issue for me. Eleanor

That's true, but the yellow ink will lower the display rating for your b/w prints.

/Sven

[eronald]

 I get good results from Gutenprint, except for some remaining gamut issues. I really don't know how they get those reds out of the native 3880 drivers.

Edmund

The edge of a custom RIP (e.g., ImagePrint) generally comes not from the ICC color profiles, but from the substitute ink mixing algorithms and the way the ink droplets are placed (in other words, the mechanics of the process). Of course, the profiles have to be solid to realize the benefits.

To be really clear: There are some (many) types of images where you're going to see minimal difference. Once you start pushing gamut boundaries on a particular surface type, that's when differences (e.g., graininess or lack thereof as you progress from one hue to another) tend to be more apparent, and even then sometimes only on close inspection. I think this is an indication of how good the makers' own drivers have become.

[Gemmtech]

When will people learn that what they see with their eyes isn't necessarily what another person sees with his/her?
Telling a person what they see with product X is better than product Y is ridiculous, it's like telling me which car seat
feels better to my ass! 

"Beauty is in the EYE of the BEHOLDER" 

[alfin]

It’s true that people don’t always see the same, but when an app like ImagePrint is considered inferior to the standard print driver, especially with canned profiles, one has to start wondering whether they actually have used Imageprint and compared for themselves or if they just reject the RIP because it costs too much.

[Czornyj]

It’s true that people don’t always see the same, but when an app like ImagePrint is considered inferior to the standard print driver, especially with canned profiles, one has to start wondering whether they actually have used Imageprint and compared for themselves or if they just reject the RIP because it costs too much.

So what's wrong with canned profiles - especially when the Epson is linearized with ColorBase?

[Sven W]

When will people learn that what they see with their eyes isn't necessarily what another person sees with his/her?
Telling a person what they see with product X is better than product Y is ridiculous, it's like telling me which car seat
feels better to my ass!  

"Beauty is in the EYE of the BEHOLDER"  

Wow....who dares to continue now?
I thought this was a forum to discuss and solve printing problems.
/Sven

[alfin]

So what's wrong with canned profiles - especially when the Epson is linearized with ColorBase?

There is nothing wrong with canned profiles per se, but lots of them are still not as good as custom made profiles. For someone who is making their own custom profiles this is not even a question. ImagePrint profiles are also “canned”, but they are really good for all the different papers I have tried, and that’s more than I can say of most other canned profiles, even though I acknowledge that Epson’s canned profiles are good nowadays.

[eleanorbrown]

Agree, there are some really really off the wall bad canned profiles from different paper manufacturers but these days I only use a very small number of papers and the Epson profiles are really good and many times they will work exceptionally well on my non epson papers. eleanor

There is nothing wrong with canned profiles per se, but lots of them are still not as good as custom made profiles. For someone who is making their own custom profiles this is not even a question. ImagePrint profiles are also “canned”, but they are really good for all the different papers I have tried, and that’s more than I can say of most other canned profiles, even though I acknowledge that Epson’s canned profiles are good nowadays.

[digitaldog]

The edge of a custom RIP (e.g., ImagePrint) generally comes not from the ICC color profiles, but from the substitute ink mixing algorithms and the way the ink droplets are placed (in other words, the mechanics of the process). Of course, the profiles have to be solid to realize the benefits.

Absolutely! In fact, when I was using ImagePrint, I preferred profiles I built compared to profiles they built, especially with respect to blues. I don’t know if things have changed in the past couple years, but the IP profiles were exhibiting a tendency for blues to shift magenta, a somewhat common issue. In just this regard, the color output I got from the Epson driver using a even a canned profile was better than the IP canned profiles. The big difference was the ink delivery, dither etc.

[Czornyj]

I don’t know if things have changed in the past couple years, but the IP profiles were exhibiting a tendency for blues to shift magenta, a somewhat common issue.

That's exactly the same thing I observed when I had tested the latest IP. Blues shift to violet, and reds shit to orange, so appearently nothing changed.

[alfin]

Well, everything is relative according to Einstein. If we all agree that ImagePrint profiles are canned, how come it’s a surprise that custom made ImagePrint profiles are even better? It just proves what I said before, that custom made profiles are better than canned profiles in general!

[Sven W]

That's exactly the same thing I observed when I had tested the latest IP. Blues shift to violet, and reds shit to orange, so appearently nothing changed.

If you mean the famous "blue-turns-purple,red-turns-orange"-problem? Which in fact depends on the curved radius in the CIE
color model. Nowadays I only see that in some cmyk-workflows, e.g. in magazines. Never in my IP workflow.

So my last reply in this topic;
I suggest that we all gather together and print the same file (Ansel Adams?) and send it to a jury to judge 
/Sven

[digitaldog]

If we all agree that ImagePrint profiles are canned, how come it’s a surprise that custom made ImagePrint profiles are even better? It just proves what I said before, that custom made profiles are better than canned profiles in general!

But one canned profile using one print “engine” is superior to the other IMHO and experience and the superior canned profile wasn’t IPs.

[Czornyj]

Well, everything is relative according to Einstein. If we all agree that ImagePrint profiles are canned, how come it’s a surprise that custom made ImagePrint profiles are even better? It just proves what I said before, that custom made profiles are better than canned profiles in general!

If you mean the famous "blue-turns-purple,red-turns-orange"-problem? Which in fact depends on the curved radius in the CIE
color model. Nowadays I only see that in some cmyk-workflows, e.g. in magazines. Never in my IP workflow.

I did a simple test and printed Bill Atkinson's test image on my 7880 using IP (OS-X) and Qimage+Epson canned profiles (Windows). So it's not a matter of canned vs custom profiles, but rather a matter of L*a*b vs CIECAM02. The effect is quite obvious.

[alfin]

But one canned profile using one print “engine” is superior to the other IMHO and experience and the superior canned profile wasn’t IPs.

I’m not sure I understand what you mean here Andrew, do you mean that you consider Epson’s print driver with Epson canned profiles, superior to ImagePrint and their supplied profiles?

[alfin]

Oh my gosh! After more than six years, I have been upgraded from Newbie to Jr member! Great! Another six years and I will hopefully become a real member! 

[digitaldog]

I’m not sure I understand what you mean here Andrew, do you mean that you consider Epson’s print driver with Epson canned profiles, superior to ImagePrint and their supplied profiles?

In terms of solely color, yes. That is, the Epson canned profiles don’t exhibit this blue to magenta shift that I see in the IP canned profiles. I don’t like my skies magenta <g>

[Sven W]

Couldn't let the topic fade away.
I did a hardcore test on a Epson canned profile vs the same for ImagePrint.
And I presume that QImage would behave similar to the Epson profile test.

Printer Epson 9900
Paper Epson SemiGloss
Absolute Colorimetric rendering intent for both.

Printed then a 16 patch strip, which is built in Measure Tool.
The criteria is set like this:
DeltaE Average
Under 3, good. Under 6 acceptable
DeltaE Worst
Under 6, good. Under 9 acceptable

And the winner is..........both.
The results are excellent. It became the apple & pears decision.

Epson profile:
DeltaE Average 1.43
DeltaE Worst 2.57
(worst, a yellow patch)

ImagePrint profile:
DeltaE Average 1.33
DeltaE Worst 2.78
(worst patch: a black, which is blacker(!) than the actual L value)

A red patch was equal on both, no orange.
A blue was DeltaE 0.8 more correct in the Epson profile compared to the ImagePrint profile, and I would absolutely not call that "a purple shift"

Tomorrow I will do the same test with custom made profiles.
/Sven