> When I set download depth e.g. 2, it download pictures 
only 
> from starting page. It didn't download pictures from 
lower 
> level pages (html files on lower levels are downloaded 
> well).

2 == 2 clicks from the starting index.html page, that is, 
the top URLs and the images & html just below.
(html might be linked from the top page)

You'd better use filters (scan rules), by the way, which 
are generally more powerful ; such as:
-* +www.foo.com/bar/* +*.gif +*.jpg +*.png +*.css +*.js