Several months ago, I ran across a review in MacWorld entitled The real paperless office, a brief explanation of setting up a paperless office. Quoth Wikipedia:
The paperless office is now considered to be a philosophy to work with minimal paper and convert all forms of documentation to a digital form. The ideal is driven by a number of motivators including productivity gains, costs savings, space saving, the need to share information and reduced environmental impact.
Although a paperless office is perhaps most useful within an office (even a small office or home office) environment, going paperless offers obvious benefits even for personal use. Whilst a single user may not benefit from the sharing of digital documents on the network, maintaining documents in digital form does allow the single user to reduce clutter and benefit from vastly improved desktop search tools (e.g. Google Desktop, Apple’s Spotlight, etc.).
Shopping for the right scanner
The article at MacWorld was interesting, but I didn’t have the equipment or time to devote to the project of moving paperless. One of the most important missing components for such a project was the lack of a scanner with an automated document feeder (ADF). My scanner at that time was a relatively cheap and feature-poor flatbed Canon LiDE 30. Having now had the benefit of a scanner with an ADF, I can confidently state that attempting this project with a typical flatbed scanner would be an exercise in frustration. Flatbed scanners may be very good for certain tasks, but helping you move to a paperless office is not among them.
Fortunately, a few coincident circumstances gave me the kick in the pants I needed to take on this project: First, I finished my degree recently, which gave me a bit of free time while I search for a job. (Anyone hiring? My degree is in Mathematics and Classics. That means I can do arithmetic and string together words into complete sentences.) Second, my Canon scanner decided it didn’t want to turn on for no good reason. The Canon has been accumulating other drawbacks recently, primarily that its driver is TWAIN-like, but not well supported. Third, the (somewhat poor) software that originally shipped with the device was OS 9 software that ran under Classic mode in OS X. After I upgraded to Leopard, even that mixed blessing (at least the scanner always worked with its own software) was a thing of the past. Finally, when the scanner mysteriously decided to stop working at all, I pretty much just threw up my hands in disgust and decided to get a new scanner.
The original article at MacWorld had recommended a Fujitsu ScanSnap scanner. Online reviews for the Fujitsu products have generally been very good, but a near universal negative that is listed for these scanners is that they are expensive. That’s certainly a fair criticism, if price is your only consideration. I managed to find an online retailer selling the S300M for about $50 under MSRP (my price about $250 before S&H). Even still, with multi-function devices that integrate printer, scanner, fax, and copier (many with an ADF) retailing for $50-$100 less than the price I ultimately paid, the premium will need to be evaluated.
To be fair to Fujitsu, similar products from competitors also carry a premium. The main competitor is Visioneer (makers of the PaperPort line, who also develop a line of scanners for Xerox). Unfortunately, the Visioneer scanners do not appear to include Mac-compatible software, so they were not serious contenders for my money.
The criteria that ultimately pushed me to pick the Fujitsu despite the price were fourfold:
- First, I had a space consideration. My “office area” is a small-ish desk in my bedroom. I needed a scanner that I could easily fit on my already crowded desk, and I don?t have the room for a separate printer/scanner stand.
- Second, my experience with all-in-one devices has been that the software bundles are frequently sub-par, and may not work phenomenally well. This is especially true with some companies’ Mac support (I’m looking at you, HP), with Mac bundles often lacking features found in the PC software. Sometimes this can be made up with additional software (either OS X-included software or third-party), but the potential hassle was one thing to consider. The ScanSnap series has been noted for its excellent Mac support and software.
- Third, I already have two printers (an very capable Canon i900D photo printer and a small Samsung laser printer). The thought of purchasing a third printer in the form of a multi-function device when I didn’t want to get rid of either existing printer was not appealing.
- Fourth, while I was sure that the Fujitsu products did exactly what I wanted to do (set up a paperless office), few reviews that I could find of multi-function devices discussed in any detail the capabilities of the ADF on the multi-function devices. Ultimately, I decided to buy something I knew would do what wanted rather than take a risk buying something I thought might work.
The conclusion I reached for my circumstances was that the extra expense was justified by what I was getting in return.
Physical impressions
The scanner itself is marketed as a portable solution for people who need document-scanning services while on the road. It is definitely small enough to include in larger laptop bags when closed. The weight is given at Fujitsu’s site as 3.08 lb. (1.4 kg).
The scanner can be powered in two ways: via an external power adapter with brick or via USB. USB power requires a special cable (included) that does not also serve as the USB communication cable. When powered by USB, you will need two available USB ports; one for the power cable and one for the communication cable. Fujitsu also notes in their documentation that scanning speed while powered over USB may be significantly slower than normal. If using a USB hub, the hub must itself be powered in order to provide power to the scanner. This may be a consideration for people who have a severe limitation of available USB ports (e.g. Macbook Air users).
When opened, the document feeder is sufficient to hold approximately ten sheets of paper of normal thickness. There are adjustable guides that center paper in the scanner; the maximum width is the American Letter standard. There are guide marks for Letter, A4, B5, and A5 paper sizes, but the guides slide smoothly to accommodate any paper size of less than Letter width.
The only button on the scanner is the Scan button, which initiates scanning. There is no power button; the scanner will automatically turn on when the cover is opened, and will automatically turn off with the cover is closed. There is one latch that when depressed will open the scanner so that document jams may be cleared. Otherwise, there are no physical doodads to adjust (or break).
Scanning performance
In my experience, the document feeder worked extremely well. After scanning nearly 2000 pages, I experienced only three paper jams which were easily rectified by reloading the paper. In one instance, I had to load the pages into the feeder in a different order to get the scanner to feed the documents correctly, but this appears to have been a one time quirk of that particular document set.
Although the paper feeder only holds ten sheets at a time, the scanner will prompt you when it finishes scanning whether to continue with the current set (allowing you to load additional sheets) or whether to finish the current set. This enables easy scanning of larger sets of documents despite the relatively small feed tray.
The scanner supports duplexing, meaning that it can scan two sides of the page in one pass of the document through the scanner. Duplexing can be turned off in the scanner software’s settings. I found that it was convenient to do so when I was scanning certain statements from my bank, which included the same disclosure information on the back of every sheet. The extremely small and dense print of the disclosure statements resulted in very slow optical character recognition (OCR; more later), and I didn’t view the information as critical for my uses. Otherwise, the duplexing worked very well.
Scanning speed was very good. Fujitsu’s technical specifications state the scanner can process approximately eight pages per minute in “Normal Mode”. At MacWorld’s recommendation, I was scanning in “Better Mode”, which is rated at six pages per minute. This seems a conservative estimate to me, although the limiting factor appears to be how quickly your computer can process the images. Over longer scanning operations (10+ double-sided sheets) the scanner seemed more likely to pause every few sheets as my three year-old PowerBook G4 1.67 GHz processed sheets that had already been scanned. Operations with only a few sheets seemed to finish about one sheet every five seconds (estimated).
Software
Note that the ScanSnap S300M is a Mac-specific product and ships with Mac-compatible software. PC users should consider the ScanSnap S300, which ships with Fujitsu’s PC-compatible software. Other than the case color and the included software, I believe the two products are virtually identical.
The scanner software includes several auto-detection algorithms that attempt to correctly orient (including correcting for skew) the pages being scanned, whether or not the page is blank, and whether the document is color or black and white. In practice, theses algorithms worked most of the time, with a relatively few number of pages outputting with the incorrect orientation (usually upside down) or with a blank page being scanned in. Zero documents in my tests exhibited skew, even when they fed slightly skewed through the scanner. All of these various options can be adjusted in the settings of the scanner software.
One very pleasant surprise was how well the scanner automatically detected the paper size and provided accurate sizes within the PDFs it created. Although the maximum width is Letter, the length of documents can be larger than Letter length. There is also support for scanning, in one set, sheets of different sizes (commonly for me pages where the payment stub had been removed from one sheet). The ScanSnap software correctly adjusted in each case to sheets of greater and lesser length, and the output PDF correctly displayed those pages as a different size.
Once a set has been completed, the software presents the user with four options: Scan to folder, Scan to e-mail, Scan to print, or iPhoto. Each options is relatively straight forward and includes a brief explanation for the user. In most cases for those interested in the paperless office, the first option will be the one selected.
Orientation problems were easily corrected in either Apple’s Preview.app and Adobe Acrobat 7.0. I discovered that documents with heavy creases often resulted in blank pages being scanned. While the included software has a welcome preview window at the document save stage in the Scan to folder screen which lets you view each page that has been scanned, viewing is its only capability. It would be nice if basic 90° rotation and page delete options were included on this screen.
The scan to folder screen allows you to save to a local folder, a network folder, or .Mac (now MobileMe). Although the software gives a default name to the document this can be easily changed from this window. A nice touch at this screen is the Name history feature, which allows you to select from any of the prior ten names chosen. This comes in handy if you are scanning a number of document sets that will have similar names. I highly recommend ordering your documents before you start scanning, so that you can make full use of this feature. Not only will your saved documents have a consistent naming scheme, but you’ll save yourself on typing. (An example are my credit card statements, which I have saved all as yyyymm_bankname_statements.)
Documents are scanned in as PDFs in which each page of the PDF is an image of the page scanned. Although these are PDFs, because the pages are all images and not actually text they are not searchable, text cannot be selected, and the contents of the document cannot be indexed by desktop search applications. This is where OCR software comes in. (Not OCD, although I imagine a touch of OCD is extremely useful for this type of project.) OCR is software that looks at the image of the page and attempts to recognize what letters the image is expressing. OCR has been around for quite a while, and is fairly mature technology. Its ability to accurately recognize characters is quite good, especially for standard fonts with some variation among different manufacturers.
At the settings recommended by MacWorld, scanned documents worked out to between 250k-500k per page, depending on color/b&w, paper size, an amount of text on each page. After scanning about 2000 pages, I used about 600 MB of hard drive space.
Included with the scanner is I.R.I.S. Software’s CardIris software for scanning and extracting contact information from business cards. While CardIris does perform OCR on the scanned cards, it is a very specialized application and cannot be used for character recognition on most documents.
As noted in the MacWorld review from last year, one deficiency with this software package is the lack of included OCR software for general documents. Fortunately, Fujitsu has been running a rebate for most of this year where I.R.I.S. Software’s well-reviewed ReadIris is sent to you on return of the rebate certificate and proof of purchase. Unfortunately, I did not receive my copy before writing this review. However, I was able to perform OCR on my documents using Adobe Acrobat Professional 7.0 (later versions should work, too) along with the AppleScript provided by MacWorld. The AppleScript is a folder action that runs when a document is saved to a specified folder, launches (or switches to) Adobe Acrobat (there is a version for ReadIris, too), and then automagically begins OCR on all pages. Instructions are included with the AppleScript.
After scanning
Managing your documents is key in order to be able to find your data on demand. While some prefer to manually manage their documents via a folder scheme, others prefer to use various search programs to find information. Each has its drawbacks and benefits.
Folders
One advantage to folders is that the analogy is well-understood by just about all computer users. Folders structure navigation is ubiquitous among virtually all computing platforms. The disadvantage is that upkeep and management overhead is typically high. Unless you regularly and consistently keep up with saving your documents according to your scheme, things will quickly get out of control.
Help is, however, available. Hazel is an application that will monitor whatever folder you tell it to and will perform actions on those folders’ contents based on rules you specify. The rules editor is very similar to both Entourage’s and Apple Mail’s e-mail rules, so most users should be able to figure out the process fairly easily. For instance, one could easily set up a rule that said to move any documents in the “OCR This” folder with “Bank1″ in the file name to the “Bank1″ folder in documents.
I use Hazel for other purposes, but haven’t yet set it up to aid me with this project. Ultimately, Hazel is only as good as the rules you set up, so it still would require a bit of planning.
Search
At the other extreme is the search scheme, whereby documents are stored in only a few (or one) folder, and then located by means of a search. This can be done in a desktop search application like Google Desktop Search or Spotlight, which both will index file names, metadata, and file contents for supported file types; or it can be done using various other applications, such as Leap, Yep, or DevonThink.
Leap and Yep are both from Ironic Software. In one sense Leap is a superset application of Yep in that they both do a similar job indexing file contents and tagging files with search terms. They both provide useful browsers that can focus in on one tag or a selection of tags. The main difference between them is that Yep is focused specifically on PDF documents, while Leap is more generally for any file type. Assuming that you’re saving your documents in PDF format, Yep may be sufficient for you paperless office needs.
DevonThink creates a fully searchable database of any folder structure you provide to it. The Pro Office version also integrates an import and OCR function with full support for Fujitsu products.
The advantage to this scheme is that it requires very little planning on your part. Virtually every document can be stored in a few locations (or a single location). The disadvantages are twofold: First, to find a document again requires that you remember some specific, relatively unique data about that document. If you only remember something general about the document, your search might return an absurd number of results. Second, if you can’t remember something relatively unique, searching through your save location(s) which might include hundreds or thousands of documents can be very time consuming.
The middle path
Ultimately, I’ve decided to take a middle path that I think will enable me to take advantage of both approaches. I’ve selected DevonThink Personal to provide the document index and search functions as I already have an OCR program and an acceptably quick workflow with the AppleScripts from MacWorld.
However, I?m also making sure to name my files consistently and with key information that will make browsing for them easier. I also have a few top-level folders in Documents that will break down my scanned documents into broad categories (Banking, Insurance, Medical, etc.). In this way I hope if I have to go digging, at least I?ll have a head start.
Conclusion
The Fujitsu ScanSnap S300M is one of my better computing purchases in the last several years. The software “just works” in the best Macintosh tradition, and the hardware is simple enough to be idiot-proof, but well designed and very competent at what it does.
As stated above, flatbed scanners are very good at certain tasks; if your needs run to those tasks (photo scanning, scanning oddly-shaped objects, scanning documents wider than Letter width, etc.), and not to document scanning, a flatbed scanner might be better for you. But for those of us seeking to get a handle on the various paper statements we receive, it?s difficult to imagine a product that would handle the move to the paperless office as smoothly as the S300M.
There was one casualty in my crusade against paper. My longtime friend, a Fellowes shredder, started spitting out ball bearings near the beginning of this project (hence, the large stack of paper in the first picture). This was despite the fact that I?m very careful about not feeding too many sheets of paper at once (and almost feed less than the rated maximum). Now I have to buy a new shredder. So long, Mr. Fellowes!
Originally posted at my blog, http://davidvoegtle.net/blog.
Technorati Tags: acrobat, adobe, devonthink, fujitsu, hardware, ocr, readiris, reviews, s300m, scanners, scanning, scansnap, technology
