Poster:
|
acrv |
Date:
|
December 27, 2010 09:53:30pm |
Forum:
|
forums
|
Subject:
|
Re: epub and kindle formats |
Hi,
The older the books (and most of the IA books are venerable), the greater the number of errors. You can pick up at random, for example this:
http://www.archive.org/details/tableaudeparis01mergoogThe problem extends to full text format too.
I understand that is a technical issue due to the kind of characters of older prints, but why then to offer formats that can't be made (fully) readable?
Regards,
Cristi Andrei
Poster:
|
Jeff Kaplan |
Date:
|
December 27, 2010 10:30:06pm |
Forum:
|
forums
|
Subject:
|
Re: epub and kindle formats |
Thanks for your feedback.
As you say, the issue is OCR. It is inexact at best on some texts. The ePub and Kindle formats are created on demand using the text file. So, if the text file is inferior, the Kindle and ePub formats will also be. The eBook formats are offered system-wide so, as you have discovered, some are better than others. Another factor is that as scanning and deriving improves so will the OCR. Some of the newer texts are better than older ones for that reason. There has been some re-deriving to try to improve older scans.
This post was modified by Jeff Kaplan on 2010-12-28 05:30:06