Why not subscribe now?
Useful Link(s)...
Hosting by
Bluehost
April 2006
David Em davidem@earthlink.net
www.davidem.com
Copyright 2006 David Em.
The Media Lab is first and foremost a User's Column. I often write about blue sky technologies, but for the most part I concentrate on imaging hardware and software that my editorial associates and I have actually tested in a real world setting.
Of course readers are always interested in the latest of everything, so I try to report on new products as close to their ship dates as possible. This works better in some cases than others. For example, it's not too hard to analyze a new camera's capabilities, but a new version of a deep 3D animation program like Autodesk's 3ds max (1000 new features!) can take a while to wrap your head around.
Sometimes it takes a period of pounding away before a tool's hidden capabilities or deficiencies reveal themselves. This week I'll take a Second Look to two photographic products I've recently reviewed, Nikon's Super CoolScan 5000ED slide scanner and Epson's R2400 archival desktop printer. In addition, I'll give a First Look to PremierArt's inkjet spray fixative and Neat Image's impressive noise reduction software.
Based on the amount of mail I received on my review of Nikon's Super CoolScan 5000ED slide scanner a couple columns back, it appears an awful lot of people are very interested in digitizing their towering mountains of slides.
There's a lot to like about the 5000ED. Except for its funky software interface and overpriced options (such as a bulk loader that costs as much as some scanners and doesn't even do a very good job), I got excellent results. I was particularly impressed with its speed and color fidelity.
After I wrote the review, a photographer friend stopped by the lab with some slides he'd shot in Thailand twenty years ago using a somewhat exotic black and white 35mm film made by Polaroid. The film was designed for photographers who needed to produce "quick and dirty" tests on the spot. The film could be developed in the field with tap water. The result was a photograph with a very sharp fine-grain dither pattern.
The grain pattern gives the slides a unique quality, a rare case of grain being a feature rather than a drawback. My friend had printed the slides optically through an enlarger and some of the grain's sharpness was lost, so I suggested we try scanning them with the 5000ED. We did, and they came out sharp as a tack.
Then we tried printing a proof of the scan on Epson's R2400 printer. After a little jiggling with the driver settings, we were looking at a very nice 8-inch x 10-inch print. But when we blew it up to 13-inches x 19-inches, we discovered that the while the central portion of the image was ultra-crisp, the edges of the image were out of focus. This problem would not be noticeable on most photographic images, but in this case it was plain as the nose on your face.
A quick investigation revealed the original film was slightly curved. To its credit, the 5000ED lets you adjust film focus very precisely, but the field of focus is so narrow that eliminating the blur at the edges puts the center of the image out of focus. We attempted to flatten the film out with a number of industrial-strength slide mounts with no success. We even tried building one from heavy matte board and clamped it to a metal insert, to no avail.
I was reluctant to put the slide between two sheets of glass for fear of Newton ringing, but finally decided there was no other recourse. The good news was that the dreaded Newton rings didn't appear. The bad was that a subtle texture pattern resembling paper grain appeared in light areas of the image. In the end we scanned the slide both with and without glass and composited the best parts of the two scans in Adobe Photoshop CS 2. The final 13-inch x 19-inch print on Epson Velvet Fine Art paper was stunning.
Nevertheless, the process of scanning the image and getting it ready to print took a couple of hours instead of less than a minute. The solution to the problem may lie in a new generation of affordable desktop flatbed scanners (accent on flat!) such as Epson's upcoming V750 PRO that features 6400 optical dpi (compared to the 5000ED's 4000 dpi) and a DMax (tonal range) rating of 4, which is about the limit of human vision.
Another product I recently gave a fairly glowing review to Epson's R2400 printer, mentioned above. The R2400 delivers lush color prints up to 13-inch x 19-inch in size on a variety of media. It uses a new archival pigmented ink system called K3 that print-longevity expert Henry Wilhelm (http://www.wilhelm-research.com/) estimates produces prints that can last up to two hundred years. The printer also makes beautiful black and white prints that don't suffer from metamerism, a phenomenon that causes monochrome images to radically shift color under different illumination sources.
The R2400 has one significant design flaw. The K3 ink system employs two different blacks, depending on whether you're printing on traditional gloss photographic media, or on matte Fine Art paper. The slick Photo Black ink looks dull on matte paper, while the deep Matte Black doesn't mix well with the glossier papers.
This wouldn't be much of an issue if it wasn't for the fact that switching between the two blacks requires physically swapping out one cartridge for the other. That's not such a big deal either, in itself. The kicker is that when you perform this action, the printer recalibrates itself, a process that drains a good bit of ink out of all the cartridges.
Epson provides a utility that keeps track of ink levels. I kept it on while I tested a variety of papers to compare their surface qualities. Going back and forth between the two inks drained an entire set of eight ink cartridges worth over $100 with less than ten prints to show for my efforts. The same problem obtains with the R2400's bigger siblings in the new Epson K3 ink printer lineup.
The communication gap between Epson's marketing and engineering departments must be vast. I can't imagine how a completely fresh printing system design could have a flaw this big built in. Epson says the problem's not simply mechanical, which could be addressed by adding another slot to the cartridge area, a relatively simple fix.
It seems the design of the actual print engine is involved here. I've no clear idea of what it takes to redesign the core of a printer, but I'm sure it's not the least bit trivial, so there's probably no product refresh right around the corner.
On the up side, if you're not switching inks, you can get between twenty to thirty 13-inch x 19-inch prints per ink set, which is quite reasonable cost-wise, around three bucks a print. The bigger printers in the K3 line use bigger wells that are twice as cost-efficient (the inks cost half as much in the bigger cartridges).
Right now today Epson has no meaningful competition in the archival art-quality printing market. But midsummer both Hewlett-Packard and Canon will release directly competitive print systems that don't require switching inks. Epson had best be prepared to bite the bullet and refresh their line way ahead of schedule, or for the first time lose market- and mind-share in what's rapidly becoming a very significant sales segment, namely every person who's bought a good digital camera over the last couple years.
Much as I appreciate them, there's a big flaw to pigmented ink systems. The wonderful deep Matte Black that Epson, HP, and Canon use is a pure carbon pigment that, unlike all the other ink colors (including Photo Black), doesn't lend itself to encapsulation, so when it hits the surface of the print, it's pretty close to pure pigment applied to paper, similar to traditional art pastels made of pure chalk.
This makes the surface of matte prints far more fragile than gloss prints, which absorb the ink differently and are extremely durable. One solution to the problem is to immediately put prints on matte paper under glass. But this isn't always practical. For example, if you run an edition of thirty prints, you might put an exhibition print under glass, but the others will likely live in a box until they're sold, leaving plenty of opportunity over time for physical contact with studio assistants, framers, art dealers, and others (not to mention the occasional hurricane or other natural disaster).
Another problematic scenario involves digital images printed on canvas to simulate the look of a traditional painting. In this situation, the canvas is typically stretched on wood stretcher bars. Once up on the living room wall, the print's completely unprotected from dirt, moisture, fingerprints, smog, and smoke wafting in from the kitchen.
One solution is to spray your prints with a protective coating. There are several on the market. Wilhelm recommends PremierArt's Print Shield spray that's designed specifically for inkjet prints. I got a can of it and tested it out. The spray is composed of thin lacquer and a substance called isopropanol. Both are highly toxic. The combination's also highly flammable, so I advise not smoking or barbecuing while spraying.
Don't spray indoors unless you have access to a space with an industrial-strength ventilation system. Another strategy is to use a self-contained oxygen mask system. Sculptors who work with toxic polyester resins often resort to this method. Of course, the problem with using it outdoors is that airborne particulate matter can get stuck on or trapped under the freshly sprayed surface.
I conducted my tests on the front steps of our lab on a windless Sunday night when there wasn't much traffic. I used goggles and a filter mask. The instructions on the can recommend spraying the prints twice, at 90-degree angles. Treating ten 13-inch x 19-inch matte prints took a little over an hour. The smell of lacquer remained for some time, but eventually dissipated.
I detected no darkening or color shift on the prints. The surface was more water-resistant and less fragile than it had been, although I was still able to scrape the ink off the paper with my fingernail. According to Wilhelm, a UV blocker in the spray doubles the life of the prints before light fading occurs, a very Good Thing. The can was used up after the ten prints, so at fifteen dollars a can, the process adds roughly a buck and a half to the cost of each print.
Spraying matte papers and canvas is clearly good practice, but it's also a lot of effort and it can be injurious to your health. I haven't tried it, but spraying big pictures can't be easy. I'd be more inclined to pay a premium and job this out to a third party, but I've not yet found one that provides this service.
Digital prints on high-quality matte Fine Art papers are visually as good as any print ever made, but if Rembrandt had been a digital artist, I doubt many of his prints would exist undamaged today.
Unlike the Polaroid slide I scanned on the Super CoolScan 5000ED, grain and visual noise are qualities most photographers strive to get rid of. Film noise is usually a byproduct of poor lighting conditions, and the problem exists in the digital realm as well. As ISO light sensitivity numbers go up, so does the corresponding amount of noise introduced to the image. Different camera makers combat noise with a mixture of hardware and builtin software with varying results, but none have managed to eliminate the problem.
If you like to shoot with natural light as I do, you're no stranger to noisy pictures. Except in certain studio situations, I avoid flash lighting whenever and wherever possible. Many times the noise in a picture's just fine, contributing to the verite look and feel of the image. Other times it's just plain aggravating.
Photoshop comes with noise reduction filters, but they're not terribly good. The general idea is to blur away the problem, but this affects other parts of the picture as well, usually with less than desirable results.
I recently ran across a program from a company called Neat Image that takes a different approach. The software, also called Neat Image, analyzes a portion of a picture that's clear of visual information except for the noise pattern. It then analyzes the pattern, builds a profile, applies the profile to the entire image, and extracts the noise elements from the picture.
It works like a charm. I was very impressed with how grungy picture data became clear, sharp, and, for want of a better word, "coherent." Neat Image also does a good job with film grain and JPEG compression artifacts. It always creates a new image file, so the process is nondestructive. The latest version can handle 16-bit grayscale and 48-bit color images and does batch conversions.
Neat Image reads any file format supported by Photoshop and it works as a plugin to Photoshop and several other imaging programs. The company provides noise profiles on their site for many models of cameras and scanners, or you can create your own.
It has controls for Luminance and Color channels as well as Sharpness, with sliders for the amount of noise to reduce (eliminating all of it usually looks artificial). You can go pretty deep to fine tune a setting with the controls, but in my tests I found it wasn't really necessary most of the time. You can compare Before and After versions of your image in either a selected portion of a picture or the whole image.
Neat Image provides free demo version on their web site. There's a $29.90 Home version, a $74.90 Pro version with all the bells and whistles, and a couple variations with a mix of features in between. Recommended.