The automated processing worked fine on January CTIO .9-m run.

Here are some results on the analysis, from satisfying to less satisfying :
1 - The automated processing of the images is fine. A comparison of Ronchi 200 and Ronchi 400 is attached.
The Ronchi 400 enable a better resolution on the bluer lines. But it doe not do so well in the red part. Maybe because of the enhancement of optical distorsions (ronchis.pdf).

2 - Comparison of various attempt to calibrate the telescope throughput :

• Manufacturer spreadsheet
• CBP run 5
• Monochromatic flats run 1
• Template = star SED * model Tatmo

I would say that the background light (high) in the monochromatic attempt and (very high) in the CBP attempt is an issue.