The problems only exist with images not text.
I have created a tiny RTF file with an image. The tests I have run show that conversion from RTF to Epub in Calibre does corrupt the image.
Using MS Word to save as webfiltered HTML then import to Calibre and convert the resultant zip file to epub preserves the image. Therefore Word is obviously creating an HTML file that Calibre can work with properly without any changes occurring to images.
It appears that Calibre's own conversion to HTML produces the error since I have tried the route of RTF to HTMLZ (and zip) to epub within calibre and the fault reproduces exactly everytime.
I will post the bug report as you suggest and wait in hope.
Thanks for the interest and response.
|