Creation of Complex Script E-Book Using Images for Every Word?

[sungkhum]

Hello,

I am looking for a way to create an e-book with the Khmer language (which is UTF-8). No reader currently supports Khmer that I am aware of, so I wanted to see if there is an "easy" way to convert each Khmer word a document (a plain text document is fine) to an image and use the images as the content of an e-book. I've seen images used for formulas and Greek and Hebrew, but never as the whole content of a book. Is it possible? How should I go about doing it? Is there currently any way to automatically convert every word of text to images to be used in an e-book?

Thanks,
Nathan

[wallcraft]

What you need is a format that allows embedded fonts, and a publically available font that supports Khmer. One good option is ePub, but it is not available on the Kindle.

Unfortunately, Kindle's AZW (MOBI) format does not support embedded fonts. So if you want something that works on a Kindle the best option is a PDF with embedded fonts that has a page size optimized for a small screen. Most PDFs don't work well on the Kindle because thay assume a letter/A4 sized page. Since you are creating your own PDF from a word processor (say), you can optimize the PDF either by changing the page size or by using a big font size. All that matters is how many words fit on a line (you want fewer words on a small screen).

[ldolse]

Agree with Wallcraft on ePub.

For the Kindle you could use the Topaz format to accomplish this, but I don't know if Amazon provides the tools for small/independent publisher to create a Topaz book - there's no public domain/open source way to do it.

[sungkhum]

Thanks for your replies. I am interested more in doing an "image" based book (an image for each word would be best because then screen size would be less of an issue as the lines could break at each word) because Khmer Unicode really isn't supported by many platforms, so embedding a font wouldn't solve the issue.

Are there any other options other than Topaz that will convert each word into an image for use with an e-book reader?

Thanks,
Nathan

[ldolse]

Topaz is the only option that does what you want exactly the way you describe.

PDF is your next best option - if you use Acrobat professional it can add 'hints' that will tell a reader where it can re-flow the text, but not all readers support that. Also make sure to create the pdf with the most common screen size in mind - 800x600 pixels.

[sungkhum]

Ok, thanks. I guess I will start looking around for ways to create the images from the words automatically and then revisit the ebook part.

Thanks again,
Nathan

[Starson17]

Originally Posted by ldolse

Topaz is the only option that does what you want exactly the way you describe.

He might want to consider djvu as a format. It's useful for scanned books. I think of it as a cross between PDF and Topaz. It has multiple layers and divides the book into images and text. It compresses the images, like PDf, but it also handles scanned images of individual letters, sort of like Topaz. It compresses the first individual letter image it finds, then locates all the similar images of that letter on the page and links to the single scanned image of that letter. That way it doesn't have to store an image of the entire page, nor multiple images of the same letter.

The multiple layers allows a low res image (100dpi IIRC) to be displayed as you scroll through the pages quickly, and then the high res image gets overlaid as you stop. It's open source. Wikipedia has a summary article on it I believe.

[ldolse]

Originally Posted by Starson17

He might want to consider djvu as a format. It's useful for scanned books. I think of it as a cross between PDF and Topaz. It has multiple layers and divides the book into images and text. It compresses the images, like PDf, but it also handles scanned images of individual letters, sort of like Topaz. It compresses the first individual letter image it finds, then locates all the similar images of that letter on the page and links to the single scanned image of that letter. That way it doesn't have to store an image of the entire page, nor multiple images of the same letter.

The multiple layers allows a low res image (100dpi IIRC) to be displayed as you scroll through the pages quickly, and then the high res image gets overlaid as you stop. It's open source. Wikipedia has a summary article on it I believe.

Is there a reader which supports djvu? Agree the format is pretty appropriate to his requirements, but I wasn't aware of any readers (at least the usual suspects) supporting it... I don't believe djvu supports reflow, IMHO Topaz's ability to reflow is what makes it a pretty awesome format for this sort of application, shame Amazon doesn't provide tools to allow small publishers to leverage it.

[Starson17]

Originally Posted by ldolse

Is there a reader which supports djvu?

EBookDroid on Android is what I use.

I don't believe djvu supports reflow, IMHO Topaz's ability to reflow is what makes it a pretty awesome format

AFAIK, you are correct - no reflow. It seems more oriented to reducing the size of the file as compared to pdfs for scanned pages that typically have lots of images - like textbooks and technical literature. They hoped to get it used for making scanned pages more accessible over bandwidth limited links.