|
Those tools don't actually impact the input ( just decrypt ) so unless you are using a really old out of date version, or the book is Topaz and not azw3, I really doubt they are the issue.
Try unpacking your azw3 using KindleUnpack ( there is a standalone version - see the Mobi forum here and a calibre plugin version - you can load it as an official plugin ) and check the box to save the raw format to see exactly what is inside your azw3 file. It will also generate an epub for you from the azw3 contents that can be unzipped and checked as well.
I have found some poorly made official azw3 files where the fragment table start locations are wrong because the input was not in utf-8 and many smart-quotes were used all throughout the text messing up the byte counts.
You could also try calibre in debug mode to see if it generates any warnings about incorrect offsets when trying to convert your azw3.
Hope something here helps.
KevinH
|