As you guessed in the last discussion, the ppm to pdf (or eps) 
conversion causes some loss in the Fig. 10. I added both ppm 
and pdf versions on the repository. I am also trying to find 
a cleaner way of doing this conversion...