Requirements: Mac OS X 10.2.8, 1.25 GHz G4, 512 of RAM, USB. Universal.
Recommended: 1.6 GHz G5, USB 2.0.
Like many serious computer users of all platforms, the concept of a paperless office has often occurred to me as something of a pipe dream. As much as I love the idea, I had been fairly convinced that anything close to a realization was a long way off.
When I first heard about Fujitsu’s ScanSnap, however, the brass ring of a (nearly) paperless office seemed within reach for the first time. As I have two four-drawer filing cabinets (both about three-quarters full) sitting beside my desk, plus another four drawers’ worth of files in other places around the house, I would love to see these scale down a bit (or completely).
Where the ScanSnap is different is in the purposeful design. Most scanners are either flatbed scanners, designed primarily for photos, and able to handle documents primarily for duplication purposes—not for archives or searchable replicants; or they are scanners built into a multi-function printer, and the scanner is again geared toward copying or faxing a document. A dedicated document scanner that easily handles any size paper up to 8.5 inches wide is a completely different tool.
What’s in the Box
The ScanSnap comes ready to use: of course the scanner itself and a power cable are included, and it also includes a six-foot USB cable and bundled software.
The ScanSnap atop one of my (soon to be vanquished) file cabinets.
Though the price is well above most home/consumer scanners, the return on investment is immediate: ScanSnap is bundled with Adobe Acrobat Standard (though version 7.0 came with mine, not the latest version 8.0) and ReadIris Pro 11. With Acrobat selling for $299 and ReadIris Pro for $129, you could say the scanner itself only costs $65—though that would imply a value far below its worth. The real return on investment will come on the hours saved by what this scanner offers.
When it comes to a scanner of this sort, the usefulness is limited only by the imagination of the user. This could be the solution to a household that has trouble maintaining adequate records for tax preparation, a way to archive all of the family recipes, or finally getting the business cards you keep in a Rolodex entered into Address Book or Entourage. It could be a tool for academic research or data mining. What papers do you have stacks of at your house? You could probably figure out a way to use the ScanSnap to get them quickly stored on your computer.
As for the included software—this is a true bargain. To begin with, Adobe Acrobat is a great tool to have on-hand for many reasons. Combined with the ScanSnap, any paper form—an application, a job information form, a report—now becomes an editable document while preserving all graphical and layout elements. With Acrobat, it’s easy to combine PDFs, make comments or add elements, or apply watermarks. A lot of these could be done with good OCR and a robust word processor or layout tool, but Acrobat makes it easy and a lot less painful.
Those early adopters, whose introduction to Optical Character Recognition (OCR) technology coincided with the rise in availability of scanners in the late 1990s, will likely be predisposed skeptics. Consumer-level OCR used to outright stink, and many of us who tried it then made up our minds for a while. Let this serve as the official announcement that OCR has matured to a very usable level. If you get the right OCR engine—and ReadIris is the right OCR engine—accuracy is simply not a factor. There are still plenty of clunkers around, but ReadIris nails it every time, in my experience.
When my ScanSnap arrived, I was so excited it was almost like opening a Mac. It was well-packed and easy to set up—complete with a quick-start setup guide—and after a few minutes installing software I was on my way.
For my very first scan, I inserted two single-sided pages into the “hopper” and pressed the large, prominent Scan button. The first time through, the staple holes of the second page caught the trailing edge of the first page, and a page-size error showed up. While this took a little of the magic away, I remained undaunted. I re-inserted the pages and pressed Scan again—this time it took, scanning both pages quickly, and saved them as a PDF on the desktop (where I had specified to save it). The PDF was then opened in Acrobat—when Acrobat was installed, the ScanSnap Manager software set this as the default application to open scanned files. I quit Acrobat and opened ReadIris, dragged the PDF into it, and it was instantly recognized. I asked it to complete the OCR process, and it simply asked me for a file name. The file was saved as an RTF file, again where I specified, and opened flawlessly in TextEdit—verifying that all of the OCR had successfully converted the image to text. Great!
For my second scan I inserted two double-sided pages (putting the duplex scanning to the test) and pressed Scan. This time, however, I had directed the ScanSnap Manager to open them directly in ReadIris (I had to show it where to go; ReadIris, though shipped as a bundled version, was not a default option for the ScanSnap Manager). Once again, ReadIris did a beautiful job of OCR conversion. (Note: because I didn’t know to delete the previous pages, it actually added the new pages to them; I had to delete them manually before conversion), including bolded and italicized text.
My third scan was the “money shot” for my purposes. Inserting nine double-sided pages (17 total), I again pressed Scan. This time, the default application was DEVON Technologies’ DEVONthink Pro Office, which also incorporates the Iris OCR engine. The scanner handled all of the pages well, and DT Pro Office opened the document without trouble. It wasn’t surprising that the OCR built into DEVONthink was a bit slower in converting the scans, but on the other hand it was fully automated—I didn’t need to touch a thing for the scan to appear in my DT database as an editable, searchable PDF document. When the conversion was done, DT opened a dialog window that invited me to change the creation date, title, and author (default to creation date: today; title: scan date; and author: me), as well as as add keywords (which DT uses heavily for its artificial intelligence engine), and a subject line. Very smooth—and everything I could possibly want.
After these initial tests, I began to really put the ScanSnap through its paces. Scan after scan, this machine took everything I threw at it—including pages that were stapled, wrinkled, and even torn. Multi-column documents and documents containing tables and/or images were also no problem, and maintained formatting nicely. Regardless of whether the orientation is portrait or landscape, the scanner recognizes which and automatically switches formats. Even newspaper clippings were a success.
Pages that had been heavily curled needed to be flattened as much as possible—and even then, they would occasionally catch on each other and cause a scanning error. (This cannot be said to be a fault of the scanner, however, as it certainly did a better job at handling these than I expected.) Thus, I learned that, when scanning a set of well-used pages, it’s not a bad idea to watch the pages being fed through—or count them after scanning—to verify that all pages made it in.
I have had one or two really bad snags—one 38-page essay that was especially dog-eared jammed up pretty badly. While the ScanSnap made quite a disturbing noise, there was no real problem in clearing the jam—the unit is well-designed in this way, as well—and starting over, after smoothing the corners a little more.
By the end of the first day, I had more than 50 academic articles and essays—representing more than 600 pages—put through the ScanSnap and imported into DEVONthink. I’d barely begun to knock out the contents of one file drawer, but I realized that with a few more evenings’ work I can put a file cabinet or two up on Craig’s List.
The fruit of my labors.
After using it for a week, I’ve emptied a file cabinet with almost 2,000 pages in it. Wow! I’m excited about the possibilities, and have already begun to brainstorm how useful this can be. For example, I’m planning to set up a local Web-server (already possible with DEVONthink Pro Office) on one of our home computers that has all of our manuals and documentation, receipts, medical records, insurance information, and other documents that fill a four-drawer cabinet, binders, or are scattered about the house; my wife and I can access this stuff through a simple search, rather than tracking them down and digging through them. And my writing research will be so much simpler from now on.
About the Scanner Itself
Fujitsu has been making document scanners for years. Every doctor’s office or hospital that I’ve been to has one on every desk for scanning insurance cards and IDs. What once was available mainly to offices and professional services is now becoming a consumer-level success.
As a unit, the scanner “closes” into a nicely compact “blobject,” with nothing protruding to get caught on (it even turns itself off and on automatically when closed or opened). This is nice to keep the footprint of the unit small when it’s not in use. A carrying case is even available through Fujitsu for a reasonable price (around $40). I can see how this would be nice if a trip to the library were planned for major research work.
All closed up—a compact “blobject.”
You can also order cleaning supplies and other consumables for it, since toner and paper dust will eventually make the platen get hazy, and certain parts (such as rollers) will wear out over time. Fujitsu has thought this through, and it is ready for consumer use.
Obviously I’m an even bigger fan of the ScanSnap than I thought I would be—and it has earned my favor in every step. I think the ScanSnap represents a great addition to any household or office, and could substantially improve the way that records and file archives are kept.
I would even go so far as to say that the ScanSnap/DEVONthink combination could be an essential tool for moderate to heavy academic work. I wish these had been available when I started graduate school—my papers would have improved drastically from the way that research is made more efficient.
As I write this, I’ve literally sat for several minutes trying to think of something to say that isn’t favorable—surely there is something to critique? I haven’t tried the ScanSnap for photos, but I would imagine that it doesn’t excel in that area as much as, say, Epson or Canon does. But so what? That’s not what it is trying to offer. (By the way, the ScanSnap can handle images and even color on documents it scans.)
We’re on our way to a truly all-digital, paperless society—and the ScanSnap is a big step in making the journey a lot less painful.