This is good work addressing a practical problem for holographic cloud probes. The results are important for the atmospheric science community and I would expect that this paper will be published fairly easily after some minor revisions, mainly identified already by the other reviewer. Here are my minor comments:

(1)  The authors use the standard Huygens-Fresnel reconstruction method, which is indeed appropriate here. I do wonder though if they have considered the more accurate angular spectrum method, and if so, why they chose not to use it instead. Probably there is not much difference for such large particles. 

(2) Around line 55, the authors address the promising concepts of skipping reconstruction. Because the particles here are spheres, the problem is actually simplified quite a bit. I wonder if the authors are aware of the very fast, very simple method in Denis et al. to size spherical particles quite accurately without reconstruction? See: Denis, Loïc, Corinne Fournier, Thierry Fournel, Christophe Ducottet, and Dominique Jeulin. "Direct extraction of the mean particle size from a digital hologram." Applied Optics 45, no. 5 (2006): 944-952.

(3) Regarding clumping small glass beads around line 275, there is an effective method to deal with this effect. Simply use sonification as shown in Fig. 6 in Giri, Ramesh, and Matthew J. Berg. "The color of aerosol particles." Scientific Reports 13, no. 1 (2023): 1594.

The manuscript "In-Line Holographic Droplet Imaging: Accelerated Classification with Convolutional Neural Networks and Quantitative Experimental Validation" by Thiede et al shows a very interesting new technique to improve assessment of holographic instruments and improvement of data quality. I enjoyed reading this manuscript and suggest minor revisions. 

I believe the manuscript could be more concise in several areas when well-known topics are being discussed, even though they are presented nicely and thoroughy here. It might help to give focus on the actual new results. 

minor comments:

in table 2 it is hard to see which line the notes belong to, maybe separating them would make it clearer.

in line 322 you say 3 or 4 - when is it which? under which circumstances is it which?

same in line 385: is it 1 or 2? 

typos:

Line 148 typo - -

Line 256 typo an

Line 399 typo: filtering

Line 621: space missing