<html><body><div style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000"><div><br data-mce-bogus="1"></div><div> Hello Mohammad, <br></div><div><br></div><div> This is a pretty nice talk that you put together, I would have a few comments for your consideration<br></div><div><br></div><div><br></div><div> Slide 2 : reachable --> measurable<br></div><div> quark/anti-quark --> parton (it is not only qqbar but also gluons and you may prefer partons to be more strict)<br></div><div> cleanest .... ---> DVCS is considered the easiest way to access GPDs (DVMP is not "dirty")<br></div><div><br></div><div> Slide 3 : strictly speaking, GPDs are not a probabilities, you may call them overlap probabilities but this could be too technical</div><div> Last sentence --> it is just the other way around = BSA is sensitive to DVCS <br></div><div><br></div><div>Slide 4 : You should express C+ in the integral<br></div><div> Add somewhere in the text "Compton Form Factor"<br></div><div> <br></div><div>Slide 6 : PR-08-024 --> E08-024 within JLab nomenclature, you may also add spokesperson names <br></div><div><br></div><div>Slide 8 : yeild --> yield<br></div><div> What about contamination from incoherent channels that would mimic a coherent reaction ? In other words <br></div><div> can you put a probability number on the purity of your helium sample in the TPC ?<br></div><div><br></div><div>Slide 9 : theoretical calculations are done at a fixed kinematical point in x, Q2 and t; you are missing the Q2 information </div><div> for all the theoretical plots. Put a lot of caution in your wording since for a correct comparison to models you <br></div><div> need to integrate the models over the experimental acceptance, especially in the case of this low statistics experiment.<br></div><div> You are missing similar information for HERMES points<br></div><div> <br></div><div> Also, be aware for your presentation that comparing HERMES and CLAS in this t-dependent plot is </div><div> somehow misleading since HERMES and CLAS are not measuring the same quantity just because not exploring the </div><div> same xB Q2 phase space.<br></div><div><br></div><div> Slide 10 : H_im and H_re are squared in the alpha_3 term (this is the DVCS^2 amplitude)<br></div><div><br></div><div> alpha_1 is better of the order of 1, not 10 <br></div><div> "not sensitive to alpha_3" is something that you have to demonstrate and which is not showing up at this moment from your results<br></div><div> Note that the real part can be about 10, so squared is 100 and times alpha_3 put it at the same order of alpha_2</div><div><br></div><div> You may say that this analysis is not considering cos2phi dependence expected to be small magnitude but that you will investigate <br></div><div> later how accurate is this approximation<br></div><div><br></div><div> Slide 11 : globally same remarks for the comparison to HERMES and theoretical models + you are not showing any missing mass plot for the <br></div><div> incoherent channel<br></div><div><br></div><div> Slide 12 : you indeed seem to have a different trend but this may also come from basic re-interaction or final-state interaction effects like in</div><div> (e,e'p) reactions </div><div><br></div><div> Very nice job, </div><div><br></div><div> Eric<br></div><div><br></div><div> <br></div><div><br></div><div> <br></div><div><br></div><div><br></div><div> <br></div><div><br></div></div></body></html>