MediaStreet G-Chrome on an Epson 1280

[IGExpandingPanda]

I'm currently playing with different profiles color adjustments in order to match OEM more closely, but this should give someone a reasonable idea what this pigment ink looks like in an Epson 1270/80/90/890.





[source image]
http://i294.photobucket.com/albums/mm83/IGExpandPanda/test1-source.jpg
[OEM ink before the Switch]
http://i294.photobucket.com/albums/mm83/IGExpandPanda/test1-OEM.jpg
[Mediastreet's profile]
http://i294.photobucket.com/albums/mm83/IGExpandPanda/test1-MediaStreet-MIS-profile.jpg
[MIS's profile]
http://i294.photobucket.com/albums/mm83/IGExpandPanda/test1-MediaStreet-Epson-Profile.jpg

 

[mikling]

Is your scanner itself properly profiled?

I ask this because none of the profiles would be satisfactory to me especially if I am using expensive pigment inks. If your scanner color accuracy is close then I will suggest that either you get a custom profile or find out what papers and exactly which inks each profile was meant for. There is no such thing as a universal profile.

Going to the trouble of using pigment inks without the specific information could be wasteful. I would imagine your example is good to show why using generic canned profiles can be frustrating for someone wanting accurate archival prints not knowing the details of the profile.

Remember if you're going to the trouble of using profiles, you must ensure that you use the specific, ink, paper and print driver settings the profile was created for or get a custom printer profile. Other than that, it's hit and miss.


Here's an example of something of what you should expect of a proper profile. Both images were scanned at the same time.

The upper image is that of an Epson R220 using Precision Colors inks on Dollar Store paper.
The lower image is that of an Epson R260 using Precision Colors inks on the same Dollar Store paper.
Two different profiles on different printers and inks but using the same papers producing essentially the same image within gamut limitations.
There's no longer any guessing and hoping what the output will be and essentially the same output regardless of printer and inks used.
With my monitor profiled, the output is exactly what I see on my monitor.

 

[IGExpandingPanda]

Is your scanner itself properly profiled?

I ask this because none of the profiles would be satisfactory to me especially if I am using expensive pigment inks. If your scanner color accuracy is close then I will suggest that either you get a custom profile or find out what papers and exactly which inks each profile was meant for. There is no such thing as a universal profile.

Going to the trouble of using pigment inks without the specific information could be wasteful. I would imagine your example is good to show why using generic canned profiles can be frustrating for someone wanting accurate archival prints not knowing the details of the profile.

This is a first series of prints, without the proper profile, to serve a base point for adjustments. It was an out of the box type test. Given I'm running essentially 2200 inks in the 1280, a base point is required.

The scanner isn't calibrated, but does provide consistent results.

 

[websnail]

I think part of mikling's point is that the original source image is going to be an "unfair" comparison in that it's not been reproduced and then scanned so the output is going to have the variable of the scanner configuration/profile to cope with. Given that the OEM output looks very different to the source when it's obviously meant to look much closer to the source I'd say your results aren't going to be overly objective, or dare I say it, helpful.

That's just an opinion but surely a fully profiled screen, and scanner is going to help enormously along with properly profiled inks.

 

[IGExpandingPanda]

I think part of mikling's point is that the original source image is going to be an "unfair" comparison in that it's not been reproduced and then scanned so the output is going to have the variable of the scanner configuration/profile to cope with. Given that the OEM output looks very different to the source when it's obviously meant to look much closer to the source I'd say your results aren't going to be overly objective, or dare I say it, helpful.

That's just an opinion but surely a fully profiled screen, and scanner is going to help enormously along with properly profiled inks.

I'm sort of seeing your point. In the past I just used my Microtek 600Z, which was replaced by an Epson 1240. I made an adjustment curve based on a borrowed Q60 Kodak target. Then from there is was a simple matter of print, scan, evaluate, adjust the nobs a few points.

Straight output without any ICC profile using G6 ink is so totally off, which is to be expected considering it's ink geared for the 2200.

Ok, so how do I go about scanner calibration? I imagine there is no way without a target, as in I can't example print something from the Canon or Epson with OEM ink on a specific paper and get something adequate. Can I for example just pickup a standard IT8.7 target
http://www.targets.coloraid.de, or get one printed and use photoshop's assign profile?

 

[mikling]

One economical way is to get the it8 target that you mention then use something like Vuescan. That way the complete workflow is color managed. Once you start color management, each step needs calibration for it to pay off. The inkset that is geared for the 2200 "should" give a very nice gamut range when properly profiled for the 1280. I've used a similar strategy by using inksets for the 2100/2200 in the R200 series with phenomenal results....

 

[IGExpandingPanda]

One economical way is to get the it8 target that you mention then use something like Vuescan. That way the complete workflow is color managed. Once you start color management, each step needs calibration for it to pay off. The inkset that is geared for the 2200 "should" give a very nice gamut range when properly profiled for the 1280. I've used a similar strategy by using inksets for the 2100/2200 in the R200 series with phenomenal results....

Yes, this is where I think you were misunderstanding. If I run the printer without a profile, on MediaStreet g6 ink, geared for the 2200, I don't get a picture, well, not something that resembles a picture. If this was a case of minor differences, I could measure the relative difference between two prints on a non-calibrated scanner, and get "some" indication where corrections need to be made. Uncalibrated, it's not even in the ballpark, any ballpark.

I presumed Media Street just buys Image-Specialist ink. The only reason I went with them was their claim to generate ICC profiles free of charge, which is a service I have yet to try.

So, on to the IT8 cards. Do you know of a good source for these? I know Adorama camera carries the 5x7 q-60r2 on Endura paper for $35. I presume this would be adequate. I presume the data files I would need, unless I wanted to do a curve by hand again, are on ftp://ftp.kodak.com/gastds/Q60DATA.

There is also the ones by Wolf Faust, starting at $8 + $10 shipping from Germany http://www.targets.coloraid.de/ but to be honest, I don't want to wait for something from Germany.

And there are the Lasersoft ones. The price seems reasonable, but the shipping cost is $20.

Unless you have any reason not to, or know of a better deal, I think I'll go for the the Kodak q60R2 from Adorama.

 

[pharmacist]

Mikling,

Can you tell me what your experience is with Vuescan and IT-8 cards for printer calibrating ? I have a MP780 CCD scanner but want to make printer profiles with my Epson Pro 3800 and Canon i9950 printer. Are the results similar to those obtained with a spectophotometer like the Spyder3 ?

 

[mikling]

I'd say stick with your spectrophotometer, I use that and it is far more reliable. I use an inexpensive scanner an Epson V350 and it fills my needs. My scans are not at all critical so I really can't say if it is better because I haven't gone that route. What the scanner profiling to me is good for is at least getting close colors from a normal scan.

Yes, in theory you could create printer profiles after profiling the scanner with an IT8, and Vuescan is supposed to allow that but I haven't bothered to travel that road as I am perfectly happy with my Spyder and it works very well.

 

[pharmacist]

Mikling,

thanks for your anwer: however I do not have any profiling hardware/software for this moment. I want to purchase either the Spyder3 Studio or the Colormunki Design package to have a reasonable priced all-in-one profiling solution. I do not need very accurate and precise colours obtained by 3 or 4 digit costing systems. If the Vuescan and the IT-8 test charts can give satisfactory results I will go this way.