Sony Partners With Google To Bring More Than 500,000 Books To The Reader

[MV64]

I just recently started using Google heavily for scans of old books. This is amazing. Lots of obscure things that history lovers like me will enjoy. Probably a lot cleaner than Gutenberg texts, too. Not sure how well ePub does on a Sony, though.

[Patricia]

Originally Posted by MV64

I just recently started using Google heavily for scans of old books. This is amazing. Lots of obscure things that history lovers like me will enjoy. Probably a lot cleaner than Gutenberg texts, too. Not sure how well ePub does on a Sony, though.

No, the EPubs are not cleaner than Gutenberg texts. Gutenberg stuff has been proofread. Admittedly, the earlier Gutenberg texts have a fair number of errors, but what you get with the Sony/Google ePubs is a straightforward mechanical conversion of their ocr'd texts. Sometimes the result is rather good. Other times it is appalling.
I have converted a number of the Google texts and find that even the good ones take the best part of a day to restore the italics, emdashes and remaining errors.
However, that being said, this is a free resource. Let's not complain. We have more books freely available, and that's a very good thing.

[jimad]

The free Google PD books via Sony are apparently converted monkey-free via a very good OCR, but still have a lot of errors compared to Gutenberg.

Gutenberg in turn is now offering EPUB format on their books, some including illustrations, but currently the Gutenberg file format translations to EPUB have a lot of errors.

I've tried taking the Gutenberg EPUB books, open them in Stanza, and Export them to .mobi to make them compatible with my Kindle. It works, but seems to lose any illustrations, and adds in some more errors of its own.

Still, for simple books it works well enough.

If you are a Kindle person you are still much better off first trying to find your free book prepackaged as a .mobi file already, such as at manybooks or freekindlebooks.

I hope Amazon decides to add non-DRM EPUB support native to Kindle, and that Gutenberg makes progress cleaning up their conversions to EPUB format!

How many different ways can the myriad players find to fight amongst themselves???

[frabjous]

All right, I know these are free, and no one has any right to complain, but you'd think someone would at least scroll through these (even without reading them to correct OCR errors) to catch glaring errors.

These screenshots is from their ePub version of Bertrand Russell's Our Knowledge of the External World (1914). One shows the edge of one page, where random marginalia from the pbook is displayed as an image. The other shows fingertips of whoever scanned the book.





Crazy.

[igorsk]

Originally Posted by jimad

I've tried taking the Gutenberg EPUB books, open them in Stanza, and Export them to .mobi to make them compatible with my Kindle. It works, but seems to lose any illustrations, and adds in some more errors of its own.

Stanza converter is very very basic and doesn't keep any formatting. Why not use Mobi Creator or at least Calibre?

[jgray]

Originally Posted by frabjous

The other shows fingertips of whoever scanned the book.

Quick, see if you can match those fingerprints and maybe we can nail the culprit

Yes, that is a pretty good example of a very bad scanning job.

[vivaldirules]

In the past, I've downloaded several few books from Google and the Internet Archive in PDF format and have used PDFLRF on them to read them on my Sony. That is far preferable to trying to decipher poorly OCRed material. But I still maintain that this (Google/Sony) is a step in the right direction.

[Ervserver]

I think Google hired a bunch of school kids to do the scanning at very low pay

Originally Posted by jgray

Quick, see if you can match those fingerprints and maybe we can nail the culprit

Yes, that is a pretty good example of a very bad scanning job.

[BobLenx]

In spite of OCR scanning errors, fingers on the scanner (love that page) and even complete gibberish in some sections of some of the books, this is a MASSIVE project that will grow rapidly. Google is doing an incredible job of scanning in books that are out of copyright. There are half a million now for our Sony eReaders. Within two years we may well be looking at several million and still growing. And of course Goggle's goal is to eventually have every book ever written available on the Internet. Yes they all need to be proofread, and eventually they will be. But with this effort and others such as Gutenberg, clearly the digital reading revolution is well on its way.

Many of these books will be virtually one of a kind - never before available to the public. The concept of what is being handed to us is staggering.

I for one am thrilled by this news.

[fugazied]

Google is digitising a LOT of books, thousands of books have copyright expiring too, so the list is growing quickly.
Very cool!

[Ervserver]

Google has a form to fill out for issues with the scanned books, maybe they would rescan problem books

http://books.google.com/support/bin/...rt_books_issue

[pilotbob]

Originally Posted by BobLenx

And of course Goggle's goal is to eventually have every book ever written available on the Internet.

Wait, I thought that was Amazon's goal. And I think they have patented it.

BOb

[amgoforth]

Originally Posted by HarryT

It still gives one a starting point from which to create a nice "cleaned-up" text; that's the important thing.

Harry how does one go about cleaning it up? There are a few I might like to do that to.

[peitsao]

Ok, OCR errors happen too frequently, even when I have students copy work I want to put online for a class, but epub and OCR are essential for college classes where papers require proper references and class discussions require a page number everyone can refer to. I don't care what students bring to class, be it pda, iphone, ebook or laptop with the required book on it as long as they can find the correct page number for highlighted text. If libraries get on the epub, OCR track, AMAZON doesn't stand a chance with their crappy page numbers that cannot be used for a term paper, and that changes every time you change the font size.

[Studio717]

I've been enjoying the PDFs of old books from Google Books for quite a while now, so this is good news. I read in the Wired piece that Google Books through Sony won't work with the 500. Does anyone know how to get it to work?

[Edited to add: I read on another thread about the 500 not being able to read epub and to (possibly) use Calibre to convert epub to what it can read.]

Since I also have a 505, I've tried it. Hmmm...



Here's my problem: I'm trying to locate and download a certain volume of Carlyle's. The problem is that there are no volume numbers on the listings AND the Google preview button seems to NOT be showing the right volume. I'll click it and it looks like the right one, but then when I download it, it's a different volume altogether. Frustrating to say the least!

A related problem is that the volumes all have the same name and the Sony downloader seems to be writing over each subsequent download, even though the volumes are different.

This was on a Vista machine. I'll dig out my netbook (XP) and see if that works any better.

Anyone have any suggestions? Am I doing something wrong?

(Btw, I've run into fingers (and whole hands) in some of the PDFs. It's funny, but I would enjoy the laugh more if they then went back and rescanned the pages they screwed up. I think they may have once or twice, but that could have been Internet Archive books instead.

It's heartbreaking when the page they completely mess up is the one page you need - in a book that is old and impossible to find anywhere else. (Yes, I relate this from experience.)

Another fun one is when they have volume one of a two volume series, and volume one ends right before what you need - and there's no volume two available.)