Michael Harvey wrote: > I have an image which we have printed on our profiled (Gretag 9.18 TC) Epson > 2100. > > The image is of a night sky and trees at dusk. It is predominantly blue and > shows star trails in the sky. I have used this image quite a bit to > demonstrate a magenta metameric cast under indoor lighting.
I'd believe that. > However, even > under D50 light the prints from the profiled Epson 2100 are still not > displayed correctly (too magenta). OK, how do you know it's too Magenta? Do you compare with a calibrated screen, or a profile Cromalin or what. And how are these calibrations verified? (I'm not kidding, I've seen ton's of bad profiling on screens that were used to deem verified printer profiles as "off". Same with impressive sounding proofing systems that just happened to be off. I have an entire folder of horrors (same image on a variety of reputable proofers). So how did you build the profile? What measurement instrument? Did you take into account the viewing light source? What grey axis did you use? What profile size? Also, if you really want superb results you should probably use a RIP and limit the inks to avoid the curving in of the chroma near the end of the gamut. (See a hex plot of you profile). > > In order to substantiate your claim about the Epson 2100 and its ability to > print numeric values correctly, I didn't say it would print numeric values correctly, although it's not too bad at that either. I said I'd deem it suitable as a proofer - most proofs are used as a visual reference under a controlled (often D50) lightsource. If profiled correctly you can have it simulate an ISO 12647-2 Fogra 1 dataset within an Delta E of under 2.5 (average). > I would be willing to send the file to you > and have you print it out on your Epson 2100. I wouldn't mind but I don't have a 2100... I've done comprehensive tests' on many 2100 though and on several 7600/9600 with RIP's. These use the same inks as you know. > I could then compare it to our > output on a profiled Epson 5500 and a digital cromalin. The 5500 suffers somewhat more from metamerism than the 2100, so that would be a very subjective test. And the digital cromalin is extremely dependant on both calibration and what it is set up to simulate. I've compared quite a few digital cromalins with the Fogra dataset, and they usually end up with a dE between 5-9 (on average) which isn't that impressive IMHO. Analogue Cromalins and MatchPrint's are even worse though, they just seem to be accepted as people "know them". > > Let me know if this is ok. No hurry, I'm just interested in getting to the > bottom of the problems with the Epson 2100. NP, send the image, I'll try and make a print next time I have the chance on a profiled Ultracrome printer with a suitable media when time permits. Best Regards Thomas Holm / Pixl ApS - Photographer & Colour Management Expert - Adobe Certified Training Provider in Photoshop� - Imacon Authorized Scanner Training Facility - Remote Profiling Service (Output ICC profiles) - Seminars speaker and tutor on CM and Digital Imaging etc. - Home Page: www.pixl.dk � Email: th[AT]pixl.dk -- =============================================================== GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE
