Why would factory profile be better than mine?

[Some Guy]

Got a roll of B.C. Silverada (Metallic) Canvas to play with (It smells bad so good to keep it in the plastic bag!).  Used the ColorMunki Photo to generate a profile for the Epson 3880 with "dye ink" loaded (for metallic shine-through over pigment.).

A NorthernLight Standard Test image (One with the 4 kids, Utah arches, strawberries, B&W graduated scale.) doesn't look so hot.  Strawberries "red" is sort of washed out and dull with the profile.  Kids look pale and cyanish.  The graduated B&W transition goes from grey to a black fast and not smooth at all (Shadows blocked up).  The one of two B&W scales, i.e. the black one that has several lighter black square within it, I cannot even pick out any of the other lighter black squares.

So I downloaded the B.C. ICC profile for that paper and printer.  Wow!  It does look far better than my own one.  The graduated B&W is very smooth, blacker blacks, kids skin colors look far better, strawberries are redder and richer.  This is with the "dye ink" too!

So where can this mistake be made at in my Colormunki Photo?  I'd expect it to be better than factory especially since I am using dye ink, but this is 180 opposite.

Same "Photo Black" dye was called for both the factory profile and my own.  Don't know where the mistake was made or how they got a much better profile.

I do have access to an i1 PhotoPro 2 so maybe it would make a difference?

Notes on paper (other than smell):  It is heavy (0.43 mm thick).  Seems to have some glossy varnish-like coating.  No loading (Skew errors) with the printer on any sheets I fed into it (Stuff is stiffer than most fabric canvases).  Curls from the edges inward rather than the length of the roll (maybe the coating?).  This "metallic canvas" is not like metallic aluminum-looking papers, looks more like a pearl or opalescent surface (yellowish) tint (No OBA's, or so they say.).

Tia.

SG

[hugowolf]

If you want to reliably scan textured gloss surfaces, than then you need something a little more industrial than the Color Munki Photo. You would need a multi-angle spectrophotometer.

Something like the X-rite MA98 and about US$20-30k.

http://www.xrite.com/ma98-portable-spectrophotometer
http://www.youtube.com/watch?v=l8VUgL92c0s

Brian A

[dgberg]

SG,
I gave up trying to profile metallics and metal as well. As Brian says the gloss surfaces do not scan well enough to get usable results with the Munki.

On the factory profiles you mention BC uses "Photo Black" dyes for the factory profile.
The 3880 is pigment ink and that is the profile I have and use from BC.
Where did you see that they have a Dye profile? Or did you mean to type pigment instead of dye?
The profile is printer,paper and ink specific.

My experience is that dye inks will look better on most gloss medias then pigments. That may be all there is to it.

[Some Guy]

SG,
I gave up trying to profile metallics and metal as well. As Brian says the gloss surfaces do not scan well enough to get usable results with the Munki.

On the factory profiles you mention BC uses "Photo Black" dyes for the factory profile.
The 3880 is pigment ink and that is the profile I have and use from BC.
Where did you see that they have a Dye profile? Or did you mean to type pigment instead of dye?
The profile is printer,paper and ink specific.

My experience is that dye inks will look better on most gloss medias then pigments. That may be all there is to it.

Thanks for the information Brian and Dan.

Sorry if I wasn't clear on the "dye" part.  It is what I use in the 3880 (Cone's "dye" ink) as I prefer the gamut over the Epson K3 pigment for metallics.  I'm not concerned about longevity since I toss the stuff once the club judging or competition is over, or I get tired looking at it.  Just a hobby.

I did run a profile making run with the i1 PhotoPro 2 on the BC Silverada Canvas and it produced a much smoother grey gradient than what happened with the ColorMunki Photo.  The Munki did a quick transition into the grey to black around a RGB of 40 and all shadow detail is covered and dark.  Still, the reds lag a bit over the factory profile which is much stronger red color.  This canvas seems to have a weak dMax too no matter the profile (Mine or the factory, at least with dye ink.)

I dug into the factory profile with Notepad++ and found "CreatorApp 1Profiler" at the end of the file so x-rite was used someplace to make theirs.  Whether they used the MA98 head or similar I didn't see in that file, but seems they did use the xrite generic(?) i1Profiler software.

I'm wondering if some sort of dulling spray on the surface might make for a better reading?  The stuff does have all those glossy peaks on the surface which might be the issue for the two spectrophotometers I used (i.e. Munki & i1 PhotoPro 2).

Whole thing sort of flies in the face of "I can make a better profile (with the above units) over the factory's generic profile, even though I also used a non-OEM ink as well." 

SG

[Ken Doo]

....

Whole thing sort of flies in the face of "I can make a better profile (with the above units) over the factory's generic profile, even though I also used a non-OEM ink as well." 

SG

Though the ColorMunki offers a lot for its price point, I know BC is using MUCH better color profiling equipment than a ColorMunki.  IMO, you'll have to use better than a ColorMunki to beat most factory profiles for current printer models.

Technology hasn't stood still for color management and profiling and it has gotten easier and better.  I think paper manufacturers/suppliers have benefited from that advance too.  With my old 9800, I could easily produce much better custom paper profiles over factory generic profiles.  With my 9900, factory supplied profiles are significantly better, and you need to do some work to make improvements.

ken

[digitaldog]

Might want to try the optional optimization process with the ColorMunki. May help.

[Justin B]

SG, which media type setting are you using? Any adjustments to Density setting during target print? Same question for profiled evaluation print.

[Some Guy]

SG, which media type setting are you using? Any adjustments to Density setting during target print? Same question for profiled evaluation print.

Spent day printing and reading calibration charts.  Looking to see if the blacks can be dMax'ed any better on this canvas.

All new charts made for "i1 PhotoPro 2" (I abandoned the ColorMunki temporary since I was going to waste a box of Costco's Kirkland Glossy $15 for 150 sheets stuff.).

Used both the Canon 9000 II, and the Epson 3880:  Canon has OCP "dye" inks loaded.  Epson has Cone's encapsulated "dye" ink loaded.

All images printed for Glossy Paper and via Qimage Ultimate.

There was a small square RGB 0,0,0 square on each print that I added to the Northern Lights Test TIF image so I could read that black with the i1 head in Colorport.

Results:
Canon 9000 II with OCP:  Black square visual density at 2.551 without OCP profile applied.  With the OCP profile added it is 2.534.

The Epson only reads 1.84.  I did add 10% on a second print and it went  up to 1.917.  Another density kick to 20% and it got up to 1.946, but the shadows are blocking up badly with more density applied which xrite warned me would happen by boosting the density.  I didn't do a OBC profile for the Epson as the numbers are already low.

Don't know why the OCP dye ink out of the Canon is that much blacker over the Cone dye from the Epson.  Sort of disappointing that the cheaper Canon makes for a far better black.  Tonality of the graduated strip was good from both, with the Canon a bit warmer than the neutral grey out of the Epson.  Epson printer seems sharper too.

Add:

I followed BC method to setup and print with the Canon 9000 II dye printer.  No explicit profiles for it on their website, but they suggest dialing back on the Intensity in the Canon 9000 II driver by -20 to keep the dye ink from "pooling on the surface."  Seems that happened on the Epson when I cranked its density up as well prior.  So I did that and still the blacks seem to suffer on this canvas.  Seems this canvas does not like to play well with dye ink.  Maybe the pigment inks are better suited to it.  Sort of defeats the purpose of a metallic paper shining through though.

I did find an old post on this site about it: http://www.luminous-landscape.com/forum/index.php?topic=83320.msg675490#msg675490  I'll guess he was using pigment ink, however even with dye the blacks on mine do seem muddy too.

SG

[PeterAit]

Why WOULDN'T the factory profile be better? You can bet they have better equipment than you and the people making the profile have more experience than you.

[Some Guy]

Why WOULDN'T the factory profile be better? You can bet they have better equipment than you and the people making the profile have more experience than you.

Simple answer is they are most likely NOT using the same non-OEM ink that I am using for their profile (i.e. OCP dye in one printer, and Cone's dye ink in the other.).

More likely, they are using the default OEM ink for their profiles.  Not perfect for dye ink on this paper though, which appears not to like dye ink as it puddles on the surface so they recommend pulling the density down in the Canon printer by -20 that uses dye ink (it puddled badly at 35% in the Epson with dye ink too.).  My feeling is this paper, although claimed as metallic-like canvas, is not suited for dye ink as its receptive layer doesn't absorb it and pulling back on the density screws with the color density/blacks too.

I noted too that their profile is made by x-rite's "Profiler" software too.  They might be using the x-rite MA68 spectrometer that sells for $25K too if it uses x-rite's "Profiler" software too.  Could also be the generic "i1 PhotoPro 2" also, just they are doing something that I cannot duplicate since I am not using pigment ink.  If I were using OEM pigment ink, maybe they would match.

SG