Following up from the fully-corrected pi0-results (in the link below), below I present the contributions to the systematic uncertainty of the fully corrected direct-photon spectrum and the result itself.

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-09282018-run-9-pp-unfolding-results-different-pttrk-cut-9-11-gev-pi0-triggers-r-0

Like in the link above, I used tracks with 'pTtrk = 0.2 - 15' GeV/c in the jet-finding.  The values of 'B' (the level of background in the gamma-rich trigger) that I used in the gamma-direct subtraction are tabulated below.

To assess the uncertainty due to the estimation of 'B', for each systematic variation I did the gamma-direct subtration 3 times: once using B, once using 'B - dB', and once using 'B + dB' (where 'dB' indicates the uncertainty, systematic + statistical, on 'B').  Then the systematic uncertainty on the fully corrected spectrum is estimated in the same manner as in this post:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-09222018-run-9-pp-unfolding-systematics-9-11-gev-pi0-r-02

The sources of systematic uncertainty included below are the same as are in the uncertainty on the result in the 1st link.  By far and away, the uncertainty on 'B' is the biggest source of uncertainty on the result...

Pythia Uncertainty:

Efficiency Uncertainty:

Regularization Uncertainty:

Prior Uncertainty:

The plots below compile the uncertainties incorporating the variation of 'B'.

Pythia and Regularization Uncertainty:

Efficiency Uncertainty:

I didn't put the uncertainty due to the choice of prior since it's identically zero everywhere.  Lastly, the two plots below show the fully corrected result with the full error band vs. Pythia8 (direct-photon trigger).

(The only difference between the two is the number of bins plotted.)