Typos in ebooks - Page 11

bizzybody · 02-02-2011, 03:47 AM

Yet another common OCR bugaboo is reading the pair cl as a lowercase d. If OCR software just had a list of such common goofs and lists of words that often have recognition problems, then present them to the user with the surrounding words for correction, that would help a lot.

Another one I've seen a lot of is turning italic sans-serif uppercase I and lowercase l into forward slashes.

Much of it depends on the quality of the paper, which affects how much the ink spreads, but fer cripes sake, there shouldn't be a mixup between rn and m if the software simply compared the width of rn VS m. I don't see how it should be possible for it to see a lowercase m as rn, especially not with 100% of the lowercase m's in a book as I've seen a few times, and in the same ones every instance of rn was rendered as an m.

As for any device or OS that doesn't support unicode, it's not "defective", it just doesn't support unicode. It's also highly unlikely any such will ever get unicode support. Therefore e-book creation software should have the OPTION to create non-unicode output when making an e-book for reader software such as Mobipocket or TealDoc which has a version for those platforms.

HarryT · 02-02-2011, 03:55 AM

Quote:

Originally Posted by bizzybody

As for any device or OS that doesn't support unicode, it's not "defective", it just doesn't support unicode. It's also highly unlikely any such will ever get unicode support.

Why do you say that? A fair number of devices in the last few years have switched from Mobi to ePub support, via a firmware upgrade, and hence have acquired Unicode support in the process. One example are the reading devices made by Bookeen, where the user can choose to install either a Mobipocket or an ePub firmware. It's really not unlikely at all!

ATimson · 02-02-2011, 09:12 AM

Quote:

Originally Posted by bizzybody

As for any device or OS that doesn't support unicode, it's not "defective", it just doesn't support unicode. It's also highly unlikely any such will ever get unicode support. Therefore e-book creation software should have the OPTION to create non-unicode output when making an e-book for reader software such as Mobipocket or TealDoc which has a version for those platforms.

And then, what, retailers should start selling "Unicode" and "non-Unicode" versions of each book? That'll work real well.

Jellby · 02-02-2011, 10:38 AM

Fortunately, Ernest Vincent Wright wrote Gadsby in 1939 for readers that don't support the letter "e"

JSWolf · 02-06-2011, 06:07 PM

Quote:

Originally Posted by ATimson

And then, what, retailers should start selling "Unicode" and "non-Unicode" versions of each book? That'll work real well.

Just wait until I post my list of errors in Seize the Fire here and at the Trek BBS. It'll be a pretty good list.

As for Unicode, there is a solution. If ADE does not support the characters needed, just embed a font that works. Simple solution to the problem. Why publisher don't get how to do things like that, I have no idea.

rePete · 04-08-2011, 05:43 PM

Quote:

Originally Posted by raac

This doesn't happen with paper books

Perhaps it's because I'm a slow, plodding reader (I don't skim!!) and chew the dialog - I'm not in a hurry. Consequently, I find typos in MANY pb/hb that I annotate in the front of the book. When finished, I email the publisher with my list of typos so they can fix them next time around.

Got the next revision to one of O'Reilly's technical books by doing that

Tome Keeper · 04-15-2011, 10:10 AM

I'm not sure all issues are to do with OCR. At the moment I am very annoyed that the last 14 books I have read from Amazon have had typos! Mainly the being te or just e and several words in a row without spaces. I can cope with te, especially as the books in hard copy were rare and by my favourite author, but so many missing spaces is impossible to ignore!

GeoffC · 04-15-2011, 10:20 AM

intheendofdays

Welcome to Mobileread ....

Those examples you have listed can be OCR errors, as well as formatting problems. Did you tell Amazon ?

Giggleton · 04-15-2011, 10:36 AM

Quote:

Originally Posted by rePete

Perhaps it's because I'm a slow, plodding reader (I don't skim!!) and chew the dialog - I'm not in a hurry. Consequently, I find typos in MANY pb/hb that I annotate in the front of the book. When finished, I email the publisher with my list of typos so they can fix them next time around.

Got the next revision to one of O'Reilly's technical books by doing that

There's quite a few people that do this, I sent an email to Amazon asking about automating this process, with prerelease books sent to kindles as a reward for finding typos. Sounds like a fun game to me. One day we'll get there, not sure if Amazon will be taking us though.

Tome Keeper · 04-15-2011, 01:37 PM

Quote:

Originally Posted by GeoffC

intheendofdays

Welcome to Mobileread ....

Those examples you have listed can be OCR errors, as well as formatting problems. Did you tell Amazon ?

Heard too many stories about people getting banned from Amazon for returns and complaints to let them know. As I said only a few effect my ability to read them.

sourcejedi · 04-15-2011, 04:28 PM

Quote:

Originally Posted by intheendofdays

Heard too many stories about people getting banned from Amazon for returns and complaints to let them know. As I said only a few effect my ability to read them.

The cases I heard were for multiple returns of physical goods, where Amazon was taking a loss. I'm not saying they were all treated fairly, but a _complaint_ about an ebook shouldn't entail any risk. And they have a relatively good reputation for ebook customer service. [Kobo took something like 5 round-trips to realize their free sample downloads on the website were the crap automated OCR from the Internet Archive]. Amazon are supposed to be relatively good at passing on a simple "this ebook is not up to scratch" to publishers, as opposed to others who expect readers to provide full lists of typos.

Personally I'd report anything I found in a paid book; publishers need to know there's a problem. In the few cases where you've found it an annoyance, it seems a pity just to sit on that information

.

sourcejedi · 04-15-2011, 04:46 PM

Quote:

Originally Posted by bizzybody

Since comers is such a rarely used and archaic word in English, almost always with the word all before it, any time English OCR software thinks it sees "comers" it should be flagged, tagged and bagged as corners

<grin>

I found one instance of exactly the same error in #1 of Diane Duanes 'Young Wizard' series. That particular passed my eye completely. I found it after running a scanno-finding pass.

The guys at Project Gutenberg Distributed Proofing have accumulated a database of the commonly corrections they have to apply, and automated the process of scanning for them with a tool called GuiGuts. It was beautiful watching it find things like that.

It doesn't substitute for proofreading -- and it's still a manual process -- but it'd work great as a backstop, or a verification pass. (If it picks up a lot of errors, it's probably also missed a similar number of errors, but it serves as a good suggestion that the book hasn't been proofed as well as the Distributed Proofers would have managed).

Hellmark · 04-15-2011, 05:11 PM

I remember reading one book, where they apparently forgot to edit the last page, and so the final page of the book was so littered with OCR errors that it was unintelligible. Luckily my girlfriend had a copy of it on the book shelf, so was able to finish the story.

SeaBookGuy · 04-15-2011, 05:49 PM

In a recent library book, "t" often became "l" so that the word "to" would appear as "lo", etc.

1611mac · 04-15-2011, 06:03 PM

I mostly read classic works from the late 1700's to 1800's so my "books" are mostly ocr scans from places like google books, etc. These are free works so they have not been cleaned up.

Add to that that some are in "old english" which tends to drive OCR software nuts.

But thru it all... I am grateful to be able to read the works and most are surprisingly pretty easy to read.

.

04-15-2011, 06:03 PM	#165
1611mac eWanderer Posts: 523 Karma: 1441998 Join Date: Jul 2010 Location: NC, USA Device: iMac,iPad3,iPhone5-Kindle Fire,Touch,PaperWhite	I mostly read classic works from the late 1700's to 1800's so my "books" are mostly ocr scans from places like google books, etc. These are free works so they have not been cleaned up. Add to that that some are in "old english" which tends to drive OCR software nuts. But thru it all... I am grateful to be able to read the works and most are surprisingly pretty easy to read. . Last edited by 1611mac; 04-15-2011 at 06:26 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Looking for examples of typos in eBooks	Tonycole	General Discussions	1	05-05-2010 04:23 AM
typos or mistakes in ebooks	delcimai	Sony Reader	15	02-14-2010 11:53 AM
Typos during conversion	ddavtian	Calibre	11	10-20-2008 12:57 AM
eBooks and Typos	seldan	Reading and Management	9	10-08-2007 12:35 PM
ebook typos	sugarbear2403	Sony Reader	6	10-09-2006 11:47 PM

02-02-2011, 03:47 AM	#151
bizzybody Addict Posts: 302 Karma: 8317682 Join Date: Apr 2007 Location: Idaho, USA Device: Various PalmOS PDAs, Android Phones, Sharper Image Literati	Yet another common OCR bugaboo is reading the pair cl as a lowercase d. If OCR software just had a list of such common goofs and lists of words that often have recognition problems, then present them to the user with the surrounding words for correction, that would help a lot. Another one I've seen a lot of is turning italic sans-serif uppercase I and lowercase l into forward slashes. Much of it depends on the quality of the paper, which affects how much the ink spreads, but fer cripes sake, there shouldn't be a mixup between rn and m if the software simply compared the width of rn VS m. I don't see how it should be possible for it to see a lowercase m as rn, especially not with 100% of the lowercase m's in a book as I've seen a few times, and in the same ones every instance of rn was rendered as an m. As for any device or OS that doesn't support unicode, it's not "defective", it just doesn't support unicode. It's also highly unlikely any such will ever get unicode support. Therefore e-book creation software should have the OPTION to create non-unicode output when making an e-book for reader software such as Mobipocket or TealDoc which has a version for those platforms.

02-02-2011, 10:38 AM	#154
Jellby frumious Bandersnatch Posts: 7,570 Karma: 20150435 Join Date: Jan 2008 Location: Spaniard in Sweden Device: Cybook Orizon, Kobo Aura	Fortunately, Ernest Vincent Wright wrote Gadsby in 1939 for readers that don't support the letter "e"

04-15-2011, 10:10 AM	#157
Tome Keeper Addict Posts: 355 Karma: 1001201 Join Date: Apr 2011 Location: UK Device: Kindle 3 Keyboard, Surface Tablet, Android Smartphone, Laptop, Netbook	I'm not sure all issues are to do with OCR. At the moment I am very annoyed that the last 14 books I have read from Amazon have had typos! Mainly the being te or just e and several words in a row without spaces. I can cope with te, especially as the books in hard copy were rare and by my favourite author, but so many missing spaces is impossible to ignore!

04-15-2011, 10:20 AM	#158
GeoffC Chocolate Grasshopper ... Posts: 27,599 Karma: 20821184 Join Date: Mar 2008 Location: Scotland Device: Muse HD , Cybook Gen3 , Pocketbook 302 (Black) , Nexus 10: wife has PW	intheendofdays Welcome to Mobileread .... Those examples you have listed can be OCR errors, as well as formatting problems. Did you tell Amazon ?

04-15-2011, 05:11 PM	#163
Hellmark Wizard Posts: 2,592 Karma: 4290425 Join Date: Jun 2009 Location: Foristell, Missouri, USA Device: Nokia N800, PRS-505, Nook STR Glowlight, Kindle 3, Kobo Libra 2	I remember reading one book, where they apparently forgot to edit the last page, and so the final page of the book was so littered with OCR errors that it was unintelligible. Luckily my girlfriend had a copy of it on the book shelf, so was able to finish the story.

04-15-2011, 05:49 PM	#164
SeaBookGuy Can one read too much? Posts: 2,029 Karma: 2487799 Join Date: Aug 2010 Location: Naples, FL Device: Kindle PW 3, Sony 350 and 650	In a recent library book, "t" often became "l" so that the word "to" would appear as "lo", etc.