ABBYY FineReader Sale

[anamardoll]

The ABBYY FineReader OCR software came up in a thread I posted about converting paper-books to e-books. I happened to noticed when I was checking out their site that there's a dale going on until July 19th -- the FineReader Express 9 is $50 (50% off) and the FineReader Professional 10 is $170 (regularly $400).

Has anyone used this program before? I'm wondering what level of accuracy the program provides for a "regular" book (i.e., text from top to bottom, with no fancy layouts). I think $170 would be a steal if we were talking a 99% conversion accuracy from image to text, but anything less than that would entail a LOT of proofing. Since e-books from publishers come with OCR mistakes, I'm assuming that probably this software doesn't boast 99% accuracy or comparable........?

Just wondering if anyone has used this.

[elcreative]

Finereader generally recognised as one of the best OCR programs around... still needs some proofing but should get 99.9%... trouble is the 0.1%

[Elfwreck]

For regular, clean-font books, it's over 99% accurate; often 2-4 pages between errors if the scans are good, and those tend to be things like names that aren't in its internal dictionaries. And that's FR 7; 10 should be more accurate because OCR tech has improved in the last six years.

Note: 99% accuracy is one character wrong every couple of sentences. It's a lot better than that. 99.9 is still 1 character wrong per page.

I'm attaching a sample from a public-domain book that I'm in the process of converting. (I'm done with the main text; am trying to decide how/whether to deal with the index.)

This was scanned at 600dpi, so the scans are good, but the font is older than current books & has a tighter line-spacing than a lot of modern books. 10-pg PDF extract, and Word output from FineReader 7 with no corrections & keeping all the auto-detected formatting.

Normally, I wouldn't keep the line breaks. I'd also try to remove the headers & page numbers before OCR; I've got various ways to do that but FR 10 might have better ones. And I'd run through FineReader's internal correction process, which is easier to deal with than comparing the Word export to the PDF and making changes that way.

[mr ploppy]

I use version 8, it came free with a scanner that died years ago. Works very well but sometimes struggles with italics. But it does keep them as italics in the end document, which is a bonus. You can have a split view with the page image on one side and the resulting text on the other, and it highlights words it's not too sure about.

[anamardoll]

I went ahead and tried their trial version last night and was hooked. Purchased the full version and started ripping up an old Asimov backlist book. Now I just need a better paper cutter....

[Elfwreck]

Originally Posted by anamardoll

I went ahead and tried their trial version last night and was hooked. Purchased the full version and started ripping up an old Asimov backlist book. Now I just need a better paper cutter....

Most copy shops have the (big scary) industrial-sized paper cutters that go through the whole book at once; they charge a dollar or so per book to use it. (Probably require a staff member to operate it.) If you have a friend in a shop, you might be able to get them to cut books for free when they're not busy.

If you get access to one of those, pick up a few throwaway books from thrift stores to practice cutting the spines off without cropping into the words.

[Greg Anos]

I use Finereader 9 full version, with an Optiscan 3600. Error rate for me is around 1 in 5 pages for straight text. Since I am one of those "zero-defect tolerance" people, it doesn't save me any proofing time, but fewer changes.

Finereader 10 has "cloud" features, so I am not interested. I use Finereader on my "sterile" machine...

[Elfwreck]

Originally Posted by Ralph Sir Edward

I use Finereader 9 full version. Error rate for me is around 1 in 5 pages for straight text. Since I am one of those "zero-defect tolerance" people, it doesn't save me any proofing time, but fewer changes.

Finereader 10 has "cloud" features, so I am not interested. I use Finereader on my "sterile" machine...

Cloud? Really? For the standard version? Eep. But the info-page seems to say the cloud features are for the Corporate & Site licenses, not Pro.

Also: Do you know what resolutions 9 allow syou to save PDF images at? 7 lets me save out as a range of sizes from 72dpi to 600 dpi; 8 just has low-medium-high, and I think high is 300 dpi. (Maybe if I'm lucky, "high" is "match original." That's what happens when I save 400dpi scans at 600 on FR7.)

I prefer to work with 400 dpi scans, and save out as 300 dpi when I'm keeping images, but sometimes I want to keep 600dpi scans as original.

[mr ploppy]

Originally Posted by Elfwreck

Most copy shops have the (big scary) industrial-sized paper cutters that go through the whole book at once; they charge a dollar or so per book to use it. (Probably require a staff member to operate it.) If you have a friend in a shop, you might be able to get them to cut books for free when they're not busy.

If you get access to one of those, pick up a few throwaway books from thrift stores to practice cutting the spines off without cropping into the words.

I've never tried it because I use a camera so that I don't need to destroy the books, but couldn't you just heat the glue up with an iron and then pull the pages off?

[anamardoll]

Originally Posted by mr ploppy

I've never tried it because I use a camera so that I don't need to destroy the books, but couldn't you just heat the glue up with an iron and then pull the pages off?

Yeah, I've tried that with a hair dryer, but with mixed results on older books. Cutting has seemed more precise, thus far, but I may be DOIN IT RONG.

[Greg Anos]

Originally Posted by Elfwreck

Cloud? Really? For the standard version? Eep. But the info-page seems to say the cloud features are for the Corporate & Site licenses, not Pro.

Also: Do you know what resolutions 9 allow syou to save PDF images at? 7 lets me save out as a range of sizes from 72dpi to 600 dpi; 8 just has low-medium-high, and I think high is 300 dpi. (Maybe if I'm lucky, "high" is "match original." That's what happens when I save 400dpi scans at 600 on FR7.)

I prefer to work with 400 dpi scans, and save out as 300 dpi when I'm keeping images, but sometimes I want to keep 600dpi scans as original.

I just saw their hype about the cloud, I didn't pay too much attention about the details.

I scan straight into the Finereader OCR program. i don't save intermediate files.

[Elfwreck]

Originally Posted by Ralph Sir Edward

I just saw their hype about the cloud, I didn't pay too much attention about the details.

I scan straight into the Finereader OCR program. i don't save intermediate files.

Work has high-speed scanners that I can sometimes use, so I can scan at work & then import the pages into FR. (You'd think that someone would make a nice simple portable TWAIN scanning program that scans to tiff or, sigh, PDF. Haven't found one yet.)

[graycyn]

Thanks for posting the sale! I was thinking about buying at the sale price above, but then today had the notion to check Amazon.com before I pulled the trigger "just in case". Glad I did!

Amazon.com has ABBYY Finereader 10 at $130.99 w/free shipping. Only 7 left in stock now tho.

I caved.

[Ravensknight]

I've gotten an old version kicking around. Used it to convert the Corum books. A new version sure would be nice though...

[harriska2]

Originally Posted by anamardoll

Now I just need a better paper cutter....

I got a slightly used come 9770 19" guillotine stack paper cutter. $399 on ebay. I used the new blade it came with, had it sharpened 2 times, and purchased another blade for $140. We have just figured out that you can self sharpen the blade with a whetstone. Takes a couple minutes to take off the blade and you sharpen as best as you can (about 4 minutes). We sharpen one side then the other side we break the edge slightly. This method saves a ton of money and time (shipping or driving to a specialty place and $18 per sharpening). Plus when they sharpen they take a *bunch* of blade off. I've cut about 400 books and am pretty much done with my library. This cut so well I was able to reglue several of the books back together.