Originally Posted by Araucaria
A friend gave me a copy of a long and well-known book in mobi format. The file size was then about 2500KB. I have the hardback original which is 350 pages or a bit more, plus a couple of illustrations. To my annoyance, I found that a page had been repeated in digitising it, starting half-way through a sentence. In fact it had been repeated a second time, too. Not having a mobi file editor, I used Calibre to convert it into an epub file, then opened it in Sigil and deleted the repeated passages (just two pages). I then used Calibre to convert it back to a mobi file.
To my surprise, the resulting mobi file is less than half the size of the original, just over 1000KB. But apparently it's just the same as the original (but two pages smaller). For the rest of the book, I couldn't spot any difference at all in what was on my Kindle.
I'm sure there's a simple explanation, but I can't imagine what it is. Can anyone explain?
Many conversions leave kruft from the original source or previous (program x) conversion.
Calibre has its own process and may clean up a lot of inline code (style sheet stuff...)
Sometimes... Just sometimes. A odd original coding may result in a larger file.