[Clascomment] OPT-IN: Measurement of Unpolarized Cross Sections and Polarized Cross Section Differences for Deeply Virtual Compton Scattering

[Kijun Park]

This is a nice piece of paper for the follow up PRL 115 212003, 2015.
I have few comments and concerns for this paper.
L#; line number in the paper

It is a bit weird that specific kinematic bin range in the “title” and this makes very long title.

L44: TMD -> TMDs

Fig.1 l(k), l'(k') should be used consistently with other notation. E.g. e(k), e’(k’) ?

L196: impressive accuracy of luminosity calculation.

L215: (e-e')^2 should be (k-k')^2

L225: remove "as well"

L257: energy -> momentum (this is a p_e cut in Fig.5)

L270: remove "in table form"

L302-304: suggests: "Additionally, several fiducial cuts are applied on the detector edges where acceptance is poorly understood."

L347-349: suggests: not using "itemize", just one complete sentence.

Eq(9): I don't know how author calculated both \theta_e, \theta_p. Eventually, \theta_p (not calculated one) is used in Eq(10).
>From my understanding, we should start with couple of assumptions for such kinematic corrections.
One of them is that large angle of proton is some level precise. Second, the most of momentum distortion is due to lack of understanding of torus field.  Lastly, such effect could be larger at forward angle region due to detector mis-alignment.
Then we can calculate E0 by Eq(9) with \theta_e_(calc) from elastic event Eq(10).

Eq(14); sin^2(\theta_2/2)

L454: Author claims there are significant amount of BG, but any systematic BG subtraction study is mentioned later.

L496: \phi_{\gamma} is same as \phi before ? 
A notation should be consistently used.

Fig.9-12: there are vertical/horizontal lines are shown but no explain what they are. 1\sigma, 2\sigma cuts ?
"PRELIMINARY" can trigger any confuse to reader. It seems your plots are preliminary although I realized that you didn't mean.
why don't you use "INITIAL" to avoid confuse ?


Fig.13: bottom-right plot, x-axis is overlapped. 

L537: efficiency -> occupancy ?

Eq(21): dQ^2 -> \Delta Q^2

Fig.15-16:  (degrees) -> (deg) or consistently used to other place
“A” in the fit function is confused with “A” acceptance. Perhaps, “a” (up to you)

SECTION X.

Authors itemized four major systematic uncertainties. There are couple of problems I can see.

1/ DVCS-PRL2015 paper claimed the acceptance is one of major source of systematics and 5.3%
I wonder why this is not major source in this analysis.

2/ PID, \pi0 subtraction systematics are never mentioned although I expected these are relatively small.

3/ Eq(22): I don't understand why author claims 5% systematic uncertainty from Eq(22) which shows 4%.
What do you mean "conservative" ? Is this due to sector dependent ?

4/ L755, Table 1: Total estimation of systematic error is 10.3%. I have no idea how this number comes from.
Moreover, Global normalization error should not be in the total systematics.
I assumed all systematics (considering PRL 115 (2015) as well):
ExclusiveCut:5.5%, Fid.Cut:4.2%, Rad.Corr:3%,Acc.Corr:5.3%,PID:1.6%,\pi0-BG:1%
Then I can have 9.4% in total point-to-point systematic uncertainty.



Fig.17: -<t-> binning is quite impressive. Especially this analysis explores the larger –t bin compare to PRL115 (2015).  I expect a binning effect should be discussed.

Fig.18: axis labels can barely be read.

Fig.19: vertical space is needed between Eq. and Fig.19


Fig.20: right plot: \alpha should be \epsilon (this should be same as \epsilon in page 13.)
Gev^2 -> GeV^2 for both x, y axis


Fig.21-26: systematic error bands are invisible in the print-version (on paper).


Table II, -t (GeV^2), \phi(deg)

SECTION XII, Fig.27.:
The plot makes me confuse and present less strength of the paper.
Although author took into account different beam energy between two data, the cross-section ratio shows ~20% systematic difference.
Instead of showing Fig.27, perhaps author can add one sentence in SECTION XII with averaged deviation between two data sets (e.g. 10-20% or whatever number you have).



ACKNOWLEDGEMENTS
need to add agencies: NSF, INFN, CNRS, NRF, SURA, etc...