Printing out of CS5 Problem Outside of Printable Area

[na goodman]

I'm on a Mac running 10.6.3 and CS5. I am using version 2 profiles but am experiencing a similar problem to the gray bands and V4 profiles. I get a very light blueish color out side of the printable area. Has anyone else seen this? Again, I am using V2 profiles.

[Farmer]

I'm on a Mac running 10.6.3 and CS5. I am using version 2 profiles but am experiencing a similar problem to the gray bands and V4 profiles. I get a very light blueish color out side of the printable area. Has anyone else seen this? Again, I am using V2 profiles.

Are you using custom profiles?  Are you using papers with OBs and are the profiles made with or without UV filtering and which printer are you printing to?

(We've seen something similar that basically came down to a custom profile that was made by doing some blending between UV and non-UV filtered readings - it took ages to track it down and point out to the company doing the profiling what was going on).  So, I'm wondering if you're seeing something similar?

[na goodman]

Thank you for your response. It is a custom profile (I make my own). I did switch to a paper canvas that does have OBA's. I did print out of CS4 just to see if it printed the same and it did. I think it probably always was doing it but since I usually set my canvas size to the roll size I was not noticing it before. The profile was made with a non UV i1 and color match. Which, supposedly has UV adjustiment built into the software.

[digitaldog]

You’re not using an Absolute Colorimetric intent by any chance? That would produce the results you describe.

[na goodman]

Nope, using relative.

[na goodman]

Are you using custom profiles?  Are you using papers with OBs and are the profiles made with or without UV filtering and which printer are you printing to?

(We've seen something similar that basically came down to a custom profile that was made by doing some blending between UV and non-UV filtered readings - it took ages to track it down and point out to the company doing the profiling what was going on).  So, I'm wondering if you're seeing something similar?

I'm printing to an Epson 9800.

[Farmer]

I'd be tempted to get a hold of another profile, made either as a canned profile by the canvas vendor or perhaps someone else's custom profile made with a UV filter and see what result you get.

In the issue that we investigated, it certainly pointed to the introduction of data from a substrate with OB taken without a UV filter, into the profile.

[na goodman]

I'd be tempted to get a hold of another profile, made either as a canned profile by the canvas vendor or perhaps someone else's custom profile made with a UV filter and see what result you get.

In the issue that we investigated, it certainly pointed to the introduction of data from a substrate with OB taken without a UV filter, into the profile.

So, do you think it is affecting the profile? Since I didn't notice it before, I'm wondering if I should just go back to having the canvas size the same as the roll size and that should take care of it, at least on the running length side. I did have someone make me a profile with an isis but I haven't used it yet because for some strange reason my profile has a much wider gamut. Why do you think it prints the faint blue outside of the printable area?

[Farmer]

Yes, I think it's affecting the profile, but if the image itself looks as you expect, it's only a problem if you're worried about the blue in the white surrounds.

I don't know why it's printing the blue for sure, but if it is the same issue that we came across, then it seems to be because the OBs are not being UV filtered and so white isn't white - it's moved towards UV.  That's a guess.  There are far more knowledgeable and qualified people on the subject on these boards (like Andrew), and I'm just assuming you're seeing the same thing we saw.  But as a general understanding of what's going on, that's mine :-)

At the end of the day, if the prints look right, then that's the most important thing.

[Ernst Dinkla]

Yes, I think it's affecting the profile, but if the image itself looks as you expect, it's only a problem if you're worried about the blue in the white surrounds.

I don't know why it's printing the blue for sure, but if it is the same issue that we came across, then it seems to be because the OBs are not being UV filtered and so white isn't white - it's moved towards UV.  That's a guess.  There are far more knowledgeable and qualified people on the subject on these boards (like Andrew), and I'm just assuming you're seeing the same thing we saw.  But as a general understanding of what's going on, that's mine :-)

At the end of the day, if the prints look right, then that's the most important thing.

Last weekend someone I know mentioned the same problem. He is using Fuji RC paper on an Epson 3880, Mac/Snow Leopard etc. Custom profiles he made with the Eye 1 Pro show the blue. Fuji profiles downloaded as well. Epson's own profiles, so not (custom) made for the paper, do not show the problem. Same conditions in the driver etc.

If it was OBA/UV reflectance simulation build in the profile for a non-UV measuring spectrometer it should only make a shift to more neutrality for the total if the objective of the profile creation was to create neutral greys etc on OBA paper. One wouldn't expect extra blue added but more an orange mix. If the objective was to follow the paper color for greys etc then it shouldn't change the paper color at all with ink additions.

I'm not familiar with the choices in the profile creation software used in this case but apart from the OBA simulation choices is there also a choice for neutral greys etc versus paper color greys etc in that software?


met vriendelijke groeten, Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

[Jack Flesher]

I'm guessing you used an i1 with iMatch software to create the profile?  Three of my group noticed the same behavior -- it appears to be an issue with the i1 software on the latest Mac OSX.  We all print on Epsons and you didn't mention your printer, but it could also be Epson/Mac/i1 specific issue. Epson canned profiles do not exhibit this behavior, only i1 generated profiles.  

It seems as though OSX>Epson is treating an i1 profile as an absolute colorimetric profile regardless of how you print with it. Note that if you build your profile by reading the target on a white base, it prints a lighter gray/cyan outside the print area than if you create the profile reading the target over a gray backing -- so it does appear to be an issue with how i1 is reading and then rendering the profile.  And FWIW this was with a version 2 profile out of i1 3.6.3, so I rebuilt the profile using i1 3.6.2 and had the same issue.

[na goodman]

I'm guessing you used an i1 with iMatch software to create the profile?  Three of my group noticed the same behavior -- it appears to be an issue with the i1 software on the latest Mac OSX.  We all print on Epsons and you didn't mention your printer, but it could also be Epson/Mac/i1 specific issue. Epson canned profiles do not exhibit this behavior, only i1 generated profiles.  

It seems as though OSX>Epson is treating an i1 profile as an absolute colorimetric profile regardless of how you print with it. Note that if you build your profile by reading the target on a white base, it prints a lighter gray/cyan outside the print area than if you create the profile reading the target over a gray backing -- so it does appear to be an issue with how i1 is reading and then rendering the profile.  And FWIW this was with a version 2 profile out of i1 3.6.3, so I rebuilt the profile using i1 3.6.2 and had the same issue.

Yes to all, and thank you all for I thought I was losing my mind. And yes Jack I have the exact same set up. MacPro and MacBook Pro running Snow Leopard 10.6.3, i1 Pro with i1 Match software, though this target was read through Measure Tool and then brought back into i1 Match but it sounds like exactly the same thing. So, at this point there is nothing I can do, other than have my printing width the same size as the roll and then I wouldn't see it. These are version 2 profiles I created because of the V4 and gray bars. That is why I created a new V2. The printer is an Epson 9800. I haven't tried anything on the 3800 but I'm assuming the same would happen. I think I will call XRite tomorrow. Thanks everyone for at least giving me some piece of mind that it isn't just me. And as far as the profile printing, we have had to warm it up because the images looked a bit too blue. I do have a datacolor device here, I wonder if I should try printing targets with that and see what happens. If anyone can think of anything else to try, I'm all ears.

I'm wondering if  shouldn't wipe the laptop and put Leopard back on it. I think CS5 will run in 64 bit on it. Does that sound like a viable solution?

[Doyle Yoder]

I may have missed it but has anyone tried the different Conversion Engines available in CS5?

Doyle

[Jack Flesher]

So, at this point there is nothing I can do, other than have my printing width the same size as the roll and then I wouldn't see it.

Unfortunately, this won't help -- it prints that cyan/gray on any section of printable paper white.  So if you have any paper-white specular highlights in your image, they will get this fine dot pattern too (yes, I confirmed that via a test print).  The only workaround I can think of with the i1 is to print borderless and pull your print whites down to something like 253 gray. You might want to try a profile built with different software.  Next thought is that before you go that route, you may want to send your spectro in for a recert.  I think they charge about $300 for this.  

Fortunately for me, I'm now on the 7900 and the Epson canned profiles for it are excellent -- good enough I have no desire for anything different.  Unfortunately for you, I cannot say the same about their x800 profiles back when I had my 7800 and 3800... Another interesting sidebar factoid: I pulled my i1 profiles into Colorsync and compared them directly to their Epson counterpart -- in all cases the Epson profile was physically larger, and in all cases the very tip of the i1 profile did NOT hit pure white on the 3D model. Which would support my suspicion that Colorsync is reading them as Absolute Colorimetric profiles.

Couple other thoughts...  If I had an x800 printer, I would probably set up a cheap print server. Any older Mac running Leopard or even an el-cheapo PC laptop running XP or higher that can be networked.  Just a thought and it would be easy to set up and cheap.  Alternatively, you could try running XP or higher under Parallels or Fusion on your existing Macs and print from that. I have not tried that, but suspect it would work fine.

[na goodman]

Unfortunately, this won't help -- it prints that cyan/gray on any section of printable paper white.  So if you have any paper-white specular highlights in your image, they will get this fine dot pattern too (yes, I confirmed that via a test print).  The only workaround I can think of with the i1 is to print borderless and pull your print whites down to something like 253 gray. You might want to try a profile built with different software.  Next thought is that before you go that route, you may want to send your spectro in for a recert.  I think they charge about $300 for this.  

Fortunately for me, I'm now on the 7900 and the Epson canned profiles for it are excellent -- good enough I have no desire for anything different.  Unfortunately for you, I cannot say the same about their x800 profiles back when I had my 7800 and 3800... Another interesting sidebar factoid: I pulled my i1 profiles into Colorsync and compared them directly to their Epson counterpart -- in all cases the Epson profile was physically larger, and in all cases the very tip of the i1 profile did NOT hit pure white on the 3D model. Which would support my suspicion that Colorsync is reading them as Absolute Colorimetric profiles.

Couple other thoughts...  If I had an x800 printer, I would probably set up a cheap print server. Any older Mac running Leopard or even an el-cheapo PC laptop running XP or higher that can be networked.  Just a thought and it would be easy to set up and cheap.  Alternatively, you could try running XP or higher under Parallels or Fusion on your existing Macs and print from that. I have not tried that, but suspect it would work fine.

Thanks Jack for the info. The spectro really isn't that old. I think I will call Xrite tomorrow but I guess I better think about setting up a print server or something else. I do have Parrells but took it off my Mac. I'm thinking since I just use the macbook pro for driving the printer that maybe I should just wipe it, put Leopard back on it and go from there. I could reinstall CS5 on it or an earlier version of Photoshop. All of the editing is done on my MacPro. I just wonder if this will ever be straightened out. I'm hoping Xrite will be able to tell me more tomorrow. Thanks for your input.

[Ernst Dinkla]

Any older Mac running Leopard or even an el-cheapo PC laptop running XP or higher that can be networked.  Just a thought and it would be easy to set up and cheap.  Alternatively, you could try running XP or higher under Parallels or Fusion on your existing Macs and print from that. I have not tried that, but suspect it would work fine.

Is the flaw happening at printing time or is there something going wrong in the profile creation on the latest Mac OSX that triggers that printing flaw with the latest OSX? Someone tried iMatch Windows created profiles on the latest OSX and observes the same blue background or not? Then I could help my friend with creating the profile and he doesn't need to switch to a PC or an older OSX system for printing. There is of course nothing wrong with printing from a PC/Qimage print server but I understand that it isn't considered a slick solution for an Apple fan.


met vriendelijke groeten, Ernst Dinkla

spectral plots of +100 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

[na goodman]

I think this quote from Jack really explains it

QUOTE (Jack Flesher @ May 31 2010, 07:46 AM) *
I'm guessing you used an i1 with iMatch software to create the profile? Three of my group noticed the same behavior -- it appears to be an issue with the i1 software on the latest Mac OSX. We all print on Epsons and you didn't mention your printer, but it could also be Epson/Mac/i1 specific issue. Epson canned profiles do not exhibit this behavior, only i1 generated profiles.

It seems as though OSX>Epson is treating an i1 profile as an absolute colorimetric profile regardless of how you print with it. Note that if you build your profile by reading the target on a white base, it prints a lighter gray/cyan outside the print area than if you create the profile reading the target over a gray backing -- so it does appear to be an issue with how i1 is reading and then rendering the profile. And FWIW this was with a version 2 profile out of i1 3.6.3, so I rebuilt the profile using i1 3.6.2 and had the same issue.

I'll talk to Xrite tomorrow but right now I'm thinking I'll downgrade and install Leopard on the laptop. It looks like CS5 will run in 65bit on Leopard. Maybe someone can confirm this. I just think that may be my quickest and best solution at this point.

[Jack Flesher]

Is the flaw happening at printing time or is there something going wrong in the profile creation on the latest Mac OSX that triggers that printing flaw with the latest OSX?

AFAIK, the issue is specific to i1 Match created paper profiles printing out of OSX 10.6.3 (Snow Leopard). Then in all the cases I have seen first-hand the printers have been Epson.  So it may be specific to i1 Match profiles to OSX 10.6.3 to all printers or only then to Epson.

I theorize the following, but do not know for certain: that it is a specific flaw in the i1 Match profile software that generates paper profiles that OSX 10.6.3 Colorsync reads as an Absolute Colorimetric rendering intent profile *regardless* of the rendering intent chosen, and thus attempts to render the included paper white point when printed.

[Ernst Dinkla]

AFAIK, the issue is specific to i1 Match created paper profiles printing out of OSX 10.6.3 (Snow Leopard). Then in all the cases I have seen first-hand the printers have been Epson.  So it may be specific to i1 Match profiles to OSX 10.6.3 to all printers or only then to Epson.

I theorize the following, but do not know for certain: that it is a specific flaw in the i1 Match profile software that generates paper profiles that OSX 10.6.3 Colorsync reads as an Absolute Colorimetric rendering intent profile *regardless* of the rendering intent chosen, and thus attempts to render the included paper white point when printed.

Based on that assumption the flaw isn't happening at the target printing stage (there have been/are other issues with an Apple/Adobe/Epson chain in target printing) so the target patches should be identical to PC created ones. The flaw isn't happening at the spectrometer measuring stage, data files are easy to compare I guess. It happens at the profile creation stage, a "tag" forces the rendering to absolute colormetric for the latest OSX 10.6.3 Colorsync. Isn't that a one day, one man job, to solve that at X-rite? A profile like that should work correcty on a PC for the same paper, driver setting combination?

met vriendelijke groeten, Ernst Dinkla

spectral plots of +100 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm

[Jack Flesher]

Isn't that a one day, one man job, to solve that at X-rite?

Seems like it could be. But then it would require somebody at X-Rite to care enough to look into it.  

Howeverbutt, it does not address why the i1 profile is so much smaller than the Epson factory profile.  I can only assume that i1 Match has some kind of profile limit imposed by the current (and like 5 years old now) 2-page target.  IOW even if they fixed the rendering issue, they also need a revised, broader gamut target that actually exceeds the bounds of current printers. And of course that is going to be a lot more than a one person, one day job...