I found way to vastly improve displaying of text! - Page 2

igorsk · 10-03-2006, 06:51 PM

"Different OCR options"? What, are there any besides FineReader?

scotty1024 · 10-03-2006, 06:56 PM

Here's a shot taken with a Sony T9 in "Magnifier Mode" with Fluorescent light from above. The PDF document is using the iLiad to render the Tiresias font.

bingle · 10-03-2006, 06:59 PM

Quote:

Originally Posted by The GreatGonzo

Well, I was hoping for input from more than one user - for starters, the HP Scanner/Printer Paul is using is way out of my league, so if somebody has found a more economic solution, I wouldn't mind hearing about it.

Also, I'm pretty clear on how to scan and how to have a pdf file as the end result - but I would like to know more about the different OCR options.

I was wondering about this also. After reading that other thread, I went looking and found some pagefeed scanners that can handle 50 pages at a time for only around $300. Does anyone have any experience with "low-end" pagefeed scanners like that?

It seems once you go beyond 50 pages, you have to move up to professional prices.

scotty1024 · 10-04-2006, 12:36 AM

I've used the HP units and they aren't too bad. 50 pages at a time is 100 sides at at time if it is a auto-duplexor. Your average paper back is under 200 sides so two groups of less than 50 and you're done.

Besides, anything more than that and you're a freakin' pirate right?

This past weekend I got myself a new set of magazine shears. Home Depot had some Titanium Nitride Fiskars on sale. Today I used them. They cut the spine off Infoworld like a hot knife through butter. You don't even have to pull out the staples!

bingle · 10-04-2006, 02:36 AM

Quote:

Originally Posted by scotty1024

I've used the HP units and they aren't too bad. 50 pages at a time is 100 sides at at time if it is a auto-duplexor. Your average paper back is under 200 sides so two groups of less than 50 and you're done.

Besides, anything more than that and you're a freakin' pirate right?

I own a lot of books that aren't available as ebooks. And I might be willing to destroy them to turn them into ebooks! (for personal consumption only, of course...) But I'm not willing to spend the time to do it on a flatbed. 50 sheets at a time would be great, and as you point out, that's really 100 pages at a time.

I was thinking a lot about this as I browsed a recent library sale - 1.00 or so for an ebook, even with the cost and effort of conversion, is pretty nice.

This is the scanner I was looking at. Thoughts?

masc · 10-04-2006, 02:57 AM

As of converting pdf's to images we could use the imagemagick library. It handels it quite well but I don't know wich font hinting they using.

Malder1 · 10-04-2006, 03:42 AM

scotty1024,
I found that HTML files iLiad displays very well with excellent sharp edges. Problem only in PDF.

Malder1 · 10-04-2006, 12:13 PM

I wrote to iRex today about this problem with PDF and they said me that my letter will be forwarded to their developers.

scotty1024 · 10-04-2006, 02:49 PM

Quote:

Originally Posted by bingle

This is the scanner I was looking at. Thoughts?

I personally haven't used any Fujitsu scanners. Based on the specifications though what I like is the direct to searchable PDF feature, handles a wide range of paper sizes/weights, two flips and its open (they try to make the scanner jam free but you still get jams). Scan speed of 24 sides per minute at 400 DPI B&W would be fast enough for me.

The only warning flag I could see was the lack of TWAIN support. But straight to PDF is plenty good 'nuff these days.

The other item you may wish to spend money on is a manual guillotine paper cutter (http://www.abcoffice.com/mancut1.htm). Pricey, unless you can find one used, but when it comes to making that nice clean, 90 degree (in 3D) cut to get the spine off, accept no substitute: except taking it to Fedex Kinko's.

The key is the clamping, well, that and a blade you can put several hundred foot pounds of cutting force into without breaking the blade or the handle.

If you have patience you can use Paul's method of debinding. But if you can get one used, there's no going back. I had one, but it "disappeared" one day... so I'm back to Fedex Kinko's cutter at $1.49 a whack.

Paul Moews · 10-04-2006, 02:51 PM

I too have been very impressed by the Iliad.

I want to preserve the look and feel of old books and when the Iliad
came out I was excited at the thought of displaying books sized below
4 x 6 inches just as they were when freshly printed. Some of them
have very nice fonts.

Unfortunately they don't look so good on the Iliad - jagged edges from
downscaling either by Adobe Acrobat or by the Iliad software. Malder1's
ideas and this thread have enabled me to make things a bit better.

I took a page that fits within the display field of the Iliad - ca.
4 x 5.8 inches and scanned it in in color at 600 dpi - then reduced it
to 158 dpi-bicubic reduction- in adobe photoworkshop and converted to grayscale.
Much better when displayed on the Iliad but a bit washed out. Going back to
adobe photoworkshop I increased the input levels to make the output darker.
Again much better. I could collect the png's with adobe acrobat to make a
PDB which was much closer to the original.

This would only be practical if I could write a batch command for adobe
photoworkshop - might be ok if the books pages are uniform so that all pages
needed the same input level increase.

Unfortunately I have scanned in over 300 books at 600 bpi as bitmaps. They
look fine in adobe reader on a good display and I don't fell much like
scanning them all in again.

My scanner is a Fujitsu scan snap - with delicate pages I often feed them in
one at a time - it doesn't take much more time than feeding in stacks and
there are fewer sheets stuck together. It was only $350.

Any comments - do I really need to go to exactly 158 dpi? Can I do anything
to fix up the books I have already scanned?

scotty1024 · 10-04-2006, 03:06 PM

Quote:

Originally Posted by Malder1

scotty1024,
I found that HTML files iLiad displays very well with excellent sharp edges. Problem only in PDF.

So far Tiresias in xpdf is just looking anti-aliased to me. I'll go shoot some pictures of minimo in action. And shoot some pictures of some of the core Adobe 35 type faces in xpdf as well.

scotty1024 · 10-04-2006, 03:27 PM

Quote:

Originally Posted by Paul Moews

Any comments - do I really need to go to exactly 158 dpi? Can I do anything
to fix up the books I have already scanned?

Paul, search your feelings, you already know the answer...

I doubt, with all the hell everyone (including myself) is giving iRex about power, that you'll get them to turn on bi-cubic in the iLiad. Although it would be nice if they added a configuration option so you could turn it on...

Meanwhile, back in the real world...

It sounds like you have a stack of bitmaps. You just need a good batch TIFF(?) processor that can color match them to the iLiad display, bi-cubic scale them and then collate them into PDF documents.

Do you have a Mac?

Malder1 · 10-04-2006, 04:22 PM

I suggest to write to iRext Technologies concerning this problem with PDF. We have excellent device and especially screen. And temporal problem with incorrect text antialiasing in PDFs.

p.s. HTML really looks wonderful on iLiad, but large HTML files load very long time.

Moonraker · 10-04-2006, 04:57 PM

Quote:

p.s. HTML really looks wonderful on iLiad, but large HTML files load very long time.

I wish my HTML looked wonderful - am I missing something?

HTML on my iLiad does not flow correctly The top and bottom lines are frequently messed up.

Malder1 · 10-04-2006, 05:40 PM

HTML has problem with formating of text. But I talked that text in HTML files looks very well on iLiad and text in PDF looks bad on iLiad.

10-03-2006, 06:56 PM	#17
scotty1024 Banned Posts: 1,300 Karma: 1479 Join Date: Jul 2006 Location: Peoples Republic of Washington Device: Reader / iPhone / Librie / Kindle	Here's a shot taken with a Sony T9 in "Magnifier Mode" with Fluorescent light from above. The PDF document is using the iLiad to render the Tiresias font. Attached Thumbnails Last edited by scotty1024; 10-03-2006 at 07:01 PM.

10-04-2006, 02:51 PM	#25
Paul Moews Enthusiast Posts: 32 Karma: 12 Join Date: Jul 2006	some improvement in scanned images of text I too have been very impressed by the Iliad. I want to preserve the look and feel of old books and when the Iliad came out I was excited at the thought of displaying books sized below 4 x 6 inches just as they were when freshly printed. Some of them have very nice fonts. Unfortunately they don't look so good on the Iliad - jagged edges from downscaling either by Adobe Acrobat or by the Iliad software. Malder1's ideas and this thread have enabled me to make things a bit better. I took a page that fits within the display field of the Iliad - ca. 4 x 5.8 inches and scanned it in in color at 600 dpi - then reduced it to 158 dpi-bicubic reduction- in adobe photoworkshop and converted to grayscale. Much better when displayed on the Iliad but a bit washed out. Going back to adobe photoworkshop I increased the input levels to make the output darker. Again much better. I could collect the png's with adobe acrobat to make a PDB which was much closer to the original. This would only be practical if I could write a batch command for adobe photoworkshop - might be ok if the books pages are uniform so that all pages needed the same input level increase. Unfortunately I have scanned in over 300 books at 600 bpi as bitmaps. They look fine in adobe reader on a good display and I don't fell much like scanning them all in again. My scanner is a Fujitsu scan snap - with delicate pages I often feed them in one at a time - it doesn't take much more time than feeding in stacks and there are fewer sheets stuck together. It was only $350. Any comments - do I really need to go to exactly 158 dpi? Can I do anything to fix up the books I have already scanned?

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Actually found "Text to Speech" Useful!	Gideon	Amazon Kindle	16	03-17-2009 03:02 PM
Is there a way to improve the look up function ?	enzo.sandrolini	Bookeen	8	07-26-2008 12:10 PM
to iRex: how to improve the ui	vvoi666	iRex	25	07-01-2008 09:29 PM
Cybook not found in linux, found in win XP	fjf	Bookeen	15	01-18-2008 06:57 PM
How would you improve the V3?	try_again	HanLin eBook	26	01-11-2008 06:11 AM

10-03-2006, 06:51 PM	#16
igorsk Wizard Posts: 3,442 Karma: 300001 Join Date: Sep 2006 Location: Belgium Device: PRS-500/505/700, Kindle, Cybook Gen3, Words Gear	"Different OCR options"? What, are there any besides FineReader?

10-04-2006, 12:36 AM	#19
scotty1024 Banned Posts: 1,300 Karma: 1479 Join Date: Jul 2006 Location: Peoples Republic of Washington Device: Reader / iPhone / Librie / Kindle	I've used the HP units and they aren't too bad. 50 pages at a time is 100 sides at at time if it is a auto-duplexor. Your average paper back is under 200 sides so two groups of less than 50 and you're done. Besides, anything more than that and you're a freakin' pirate right? This past weekend I got myself a new set of magazine shears. Home Depot had some Titanium Nitride Fiskars on sale. Today I used them. They cut the spine off Infoworld like a hot knife through butter. You don't even have to pull out the staples!

10-04-2006, 02:57 AM	#21
masc Enthusiast Posts: 47 Karma: 10 Join Date: May 2006 Location: Austria Device: iLiad	As of converting pdf's to images we could use the imagemagick library. It handels it quite well but I don't know wich font hinting they using.

10-04-2006, 03:42 AM	#22
Malder1 Connoisseur Posts: 52 Karma: 10 Join Date: Aug 2006	scotty1024, I found that HTML files iLiad displays very well with excellent sharp edges. Problem only in PDF.

10-04-2006, 12:13 PM	#23
Malder1 Connoisseur Posts: 52 Karma: 10 Join Date: Aug 2006	I wrote to iRex today about this problem with PDF and they said me that my letter will be forwarded to their developers.

10-04-2006, 04:22 PM	#28
Malder1 Connoisseur Posts: 52 Karma: 10 Join Date: Aug 2006	I suggest to write to iRext Technologies concerning this problem with PDF. We have excellent device and especially screen. And temporal problem with incorrect text antialiasing in PDFs. p.s. HTML really looks wonderful on iLiad, but large HTML files load very long time.

10-04-2006, 05:40 PM	#30
Malder1 Connoisseur Posts: 52 Karma: 10 Join Date: Aug 2006	HTML has problem with formating of text. But I talked that text in HTML files looks very well on iLiad and text in PDF looks bad on iLiad.

Advert

Advert