Docbook XML to Epub

"Regarding your first question, if you put a role attribute on a phrase (<phrase role=“red”>contents</phrase>), the value of that role will be carried to the HTML (e.g., <span class=“red”>contents</span>)
You can then apply css normally.

Regarding your second question, I’m not so sure what’s the best thing to do, but you might try the same approach, wrapping the inlinemediaobject in a phrase with a role, then key off whatever sequence of spans you get (probably something like span.XYZ>span.inlinemediaobject>img).

Dick Hamilton”

That is great! Thank you so much. The red span works great. I will see if I can do the same for the images.