What is a true giclée fine art print?In the beginning only an image printed from an IrisGprint machine was accepted as a giclée print. But then people found out that this machine was just a slightly modified proofer (the Iris 3047), simply with the ability to accept thick fine art paper. So basically it was a regular Iris 3047 model proofer with a different label. About the same time people finally realized that the inks for Iris printers were only dye inks and hence subject to fading. Pigmented inks will not work in the Iris version of continuous ink flow (Herz) technology. In other words, Iris giclée prints are subject to fading rather quickly in a bright room with windows which allow in sunlight. On top of this people found out that the Iris giclée printer had a host of systematic inherent mechanical defects (listed and described in full detail in a FLAAR report). During these years of increased awareness of reality, Scitex stopped making the giclée version of the Iris printer (for many of the above reasons). Improved Technologies took the original design, re-engineered the printer to make it mechanically more reliable, and offered this as the Ixia printer. We have visited IT and found the output very impressive. But more importantly, all kinds of advances in inkjet technology continued to overwhelm the now somewhat outdated Iris system. Roland offered variable drop; ColorSpan offered considerably increased speed; Hewlett-Packard offered a completely user-friendly printer; even Encad developed a printer that could reproduce art. Today (going into 2007) Canon actually offers one of the more promising wide format inkjet printers: the iPF9000. During these recent years thousands of artists all across America began to buy their own printers to reproduce their paintings. In other words individual artists and photographers learned that they could often get more reliable output by buying their own printers rather than having unacceptable and costly prints attempted at a commercial location. In the final analysis, remember that the pseudo French word giclée means merely ejaculated (substance), as in "squirted ink" of an inkjet printer. Yes, a true giclée print should have a quality standard, but it's presumptuous to assume that only piezo technology can work. Some of the worst printheads in the industry are piezo (such as the Xaar printheads used on many solvent ink grand format printers). FLAAR has access to a several good piezo printers. However when it came time to print an exhibit for the art museum on our campus, we selected our thermal inkjet printer. It was easier to use and much faster. The exhibit was so well received that it is now traveling to other venues. Thus it is self-defeating to attempt to claim that only a piezo printer is "true giclée." Any such movement will simply result in end users of thermal printhead technology responding by pointing out the embarrassing defects of Iris giclée printers, the well documented propensity of most piezo printers to ruin images with horizontal banding defects, and so on. Indeed the printer favored by many elite giclée studios is precisely the printer that exhibits the most inherent generic defects of piezo printhead technology (virtually none of these defects is present in thermal printhead technology). Thus delta-E measurement is only one aspect of print quality. A low delta-E with banding defects is not a good advertisement for the world of giclée printing. Summary: you can get a beautiful giclée print from continuous (Hertz) technology; you can get a really attractive museum-quality print from a piezo printer; you can get photo-realistic exhibit quality from a Canon, ColorSpan and Hewlett-Packard thermal printer as well. Indeed one of the three most attractive exhibits at Seybold was accomplished 100% with thermal printhead technology (on an HP 5000ps). Today the leading printers are all thermal printhead technology: Canon iPF5000, iPF8000, HP Z2100 and HP Z3100 photo printers. Other exhibits were done with Epson printers. So both piezo and thermal can produce great images. This fact has been demonstrated at countless tradeshows. So if you need more details on which of these brands to actually purchase, this we have provided in an entire FLAAR Series on giclée printers. I have been to an art gallery in Carmel where the output was all done with a Roland printer. The images were of top professional quality (as you would expect in a haven of artists and old wealth such as Carmel ). So the question is no longer "which printer should I select," but "which system should I chose?" What makes a good image is a combination of your scanner, your originals, your color management capabilities, your selection of media, and only then your printer. In the FLAAR facility we can print exhibits with a ColorSpan Displaymaker XII if we want optimal depth of color. These prints evoke the most "ohh and ahhh" exclamations of viewers. We can print with an Epson but may face a problem with metamerism (though the newer models are getting better). When we need a quick print on a machine that never requires a technician, no cleaning, and which all our art and photography students can learn to use in half an hour, then we print our exhibits with the Hewlett-Packard 5000ps. Final observations: when I visited the facilities of Improved Technologies their sample images from Iris printers were awesome. If I collected art I would definitely prefer an image from an Iris or Ixia giclée printer. However at tradeshows in Germany and throughout the USA , I have seen outstanding images from Canon, ColorSpan, Mutoh, Roland, Hewlett-Packard, Epson, and Mimaki. Prints from any piezo and also any thermal printer will last many more years than a dye-based print from an Iris, so there is an immediate tradeoff there. Of courseEpson's exaggerated claims of "200 years longevity" are not taken seriously by knowledgeable people such as ink chemists. But today the pigmented inks will definitely last several decades. The UV pigmented ink of Hewlett-Packard is also among the first pigmented inks to have extended gamut (in comparison to pigmented inks, for the Roland, which have a notoriously restricted gamut). Note: FLAAR reviews are based primarily on what real end-users actually experience in their photo labs, fine art studios. So if we make a comment that a particular printer-ink combination has a restricted gamut, that's because we have a continuous rain of e-mail from end-users who indicate this is their primarily problem specifically with that printer. Our statements about banding defects are likewise based on end-user reports as well as what we have recorded at the printer manufacturer's booths at tradeshows. If they can't eliminate banding at a tradeshow with their own technicians, why should an end user trust the glowing advertising claims which somehow blissfully fail to admit that feature of piezo printheads. Of course thermal printheads also may band but it is easier to replace these heads or use back-up nozzles to overcome the banding. Epson does not offer either of these technologies with their piezo heads. At FLAAR we use all three brands and both printhead technologies. FLAAR existed decades before giclee printers became popular, so we are not owned or operated by any printer manufacturer. Indeed we use Epson printers every day ourselves. It is precisely because we are cognizant of all three printer brands and models that we are called as consultants worldwide, to help giclee ateliers and photo studios. Some of the giclee businesses we consult for are in Istanbul , Thessaloniki ( Greece ) and Hawaii , among others. If you wish to obtain assistance from Dr Nicholas Hellmuth to get your business going, and if you seek color management help as well, contact us via ReaderService@FLAAR.org, but be sure to obtain the FLAAR Report series on giclee first. Key parts of these reports are being updated for December 2006 and January 2007. When you obtain the present versions the updates are available at no cost as long as the base rate has not changed. And if it has, you pay only the difference in cost if you wish a fresh update.
Most recently updated Dec. 4, 2006. |
|||