Fancy formatting on Kindle -- how's it done?

[ldolse]

Originally Posted by cybmole

yes - in Kindle for PC, viewing the sample, I see that the opening word - "Introduction", which is in a handwriting style font on a white rectangular background, does not re-scale with the main chapter font; it must be a standalone image

so should there be an option to preserve the scans/images in mobi to epub conversion ?

Amazon != mobi, as was stated it's a Topaz file. Calibre has nothing to do with the creation of the textual/image output that shows up in your Calibre library when you process a Topaz file - the same third party plugins that strip the encryption are creating the HTML that Calibre sees. Note that it's HTML once Calibre gets it - not topaz or mobi. As Dwanthny noted, this is often completely unedited OCR text generated by Amazon or the publisher. The image files are extracted from SVG by the plugin. Whether there are any possibilities to improve it would be the responsibility of the maintainers of those plugins.

The simplest thing to do with Topaz files is just to read them on a Kindle or with Kindle software...

[cybmole]

OK - back to OP's original question, which was how to to this in reverse i.e. get FROM his HTML creation TO a fancy formatted ebook.

1. for publishing in epub format, all the things mentioned so far can be done in HTML / EPUB i.e. multiple fonts ( via embedded fonts) , drop caps, shaded backgrounds. much discussion & guidance re fonts in the sigil forum

2. for publishing via amazon i.e. in their mobi format, the tricks which mobi does not support would have to be done by creating images and then using img tags in html to reference them. calibre should convert that from html into MOBI

[cybmole]

Originally Posted by ldolse

Amazon != mobi, as was stated it's a Topaz file.

the sample downloads as B004AHM8T0_EBSP.azw into kindle for PC. ( with the usual .APNX and .MBP extras.

I thought that a .azw extension indicated that it was actually mobi, and that topaz had a different extension - AZW1 or TPZ ? ( that is what mobilereadwiki is telling me )

[bhoyt]

Thanks to all for your help. I now understand how the topaz files do their tricks. Very tricky indeed.

Yes, I know that epub allows much more than mobi. I'm just trying to make as nice looking a book as possible for the Kindle so that limits me to mobi. I still haven't discovered all that can be done in that format. A list of all possible tags and attributes that the Kindle 3 can handle would be great. Anyone have such a list?

[DoctorOhh]

Originally Posted by cybmole

I thought that a .azw extension indicated that it was actually mobi, and that topaz had a different extension - AZW1 or TPZ ? ( that is what mobilereadwiki is telling me )

The drm file attached in post 4 is azw1, but more than that if you view it in a text editor like Notepad++ you'll notice the first 4 characters is TPZ0 indicating it is a Topaz file. IIRC If it was a MOBI drm file somewhere in the first line you could see MOBIBOOK and there would be no TPZ0.

[cybmole]

that suggests that amazon samples don't always have the same format as corresponding full books

i try the sample in notepad++ & do not see same characters as you.
I see
Histories_and_Fallacies then lots of NUL then bookmobi.....

NB his link to book title was at amazon.com, I have no kindle account there so I took my sample by finding same book sample in Kindle store at amazon.co.uk... maybe they use different formats in different regions ???

[ldolse]

Topaz is literally scans of the original pages of the original paper book. The scans then go through proprietary Amazon publishing software to convert the individual characters/images to a library of custom glyphs. The cool thing about Topaz is it's the only pdf-like format that supports reflow really well. But unless you're Amazon there is no way to create one.

Hence there is no way to go from a proper ebook format to a Topaz book...

Regarding different variants of the book, it's possible that it might be Topaz in one marketplace and mobi in another - a Topaz version is sometimes released first to get a book to market quickly, and then followed up with a proper mobi version. I imagine it's also possible in those cases they wouldn't update the sample...

[DoctorOhh]

Originally Posted by cybmole

that suggests that amazon samples don't always have the same format as corresponding full books

In my experience the sample and purchased book were always the same format.

Originally Posted by cybmole

i try the sample in notepad++ & do not see same characters as you.
I see
Histories_and_Fallacies then lots of NUL then bookmobi.....

That definitely is a mobi book. But it is not the same book as the one he attached. The book he attached was World of Difference, A: Putting Christian Truth-Claims to the Worldview Test.

Originally Posted by cybmole

NB his link to book title was at amazon.com, I have no kindle account there so I took my sample by finding same book sample in Kindle store at amazon.co.uk... maybe they use different formats in different regions ???

The sample for the book he attached is topaz so I think it is safe to say the sample and the full purchased book are still the same format.

[bhoyt]

Sorry, I confused everybody by referencing a different book than the one I attached. The attachment was "World of Difference, A: Putting Christian Truth-Claims to the Worldview Test" which was the free sample I was looking at on my Kindle 3. It had the beautiful fonts and line wrapping that puzzled me. The reference to "Histories and Fallacies" came from a wrong link I clicked on. I hope that hasn't caused to much confusion for anyone.

I now understand a lot that I didn't from your discussion. Thanks, again.

[cybmole]

Originally Posted by bhoyt

I now understand a lot that I didn't from your discussion. Thanks, again.

so do i :-)