[phenixbb] observed reflection numbers of refinement greater than the experiment

Hi all, I have a problem with reflection numbers of phenix.refine. I did the phenix.refine then i found that the number of observed reflections given for the refinement 75747 (48-2.85A) is greater than the number of observed reflections for the experiment 40434 (50-2.85A). It shoud be around 40000. . This is a peak of MAD data about 2.85A. I used the TLS parameters to refine the structure. I really don't know why because this is my first time use phenix to do refinement. Any advice and suggestion are welcome! Thanks. Carl --------------------------------- Power up the Internet with Yahoo! Toolbar.