attending: Eric, Mansi, Gina, Dan R, Brad, Suvi, Jason, Eran, Rich

Wide-ranging discussion of filter and exchanger requirements.

Dan (ZTF optical engineer): quantification of deviations from strict
top-hat.  Band edges are not cost drivers; transmission, out-of-band
rejection, and transitions between are cost drivers.

Group consensus: maintain SDSS g' and Mould R bandpasses as a
baseline.  Eric to check if there are significant gains to be made in
sky background reduction by any bandpass tuning.  For R-band, sky will
help determine trade between maintaining PTF->ZTF baseline (valuable
for variable stars) vs potentially better calibration to SDSS/PS etc.

No strong opinions expresses about requirements on transmission,
ripple, out of band rejection, etc--Eric to follow up with Ia folks
not on the call.  ZTF not budgeted to compete with precision cosmology
experiments, but we'll try to maximize the value.

Filter exchanger: Consecutive g/r exposures of a field would be better
served by a fixed filter design (cheaper/easier/faster).  Instead, we
ask how we would maximally use an exchanger.  General consensus that
exchanges every 30 minutes (=> 20 per night) is probably a good max
value.  Implies exchange times of <= 90 seconds to avoid >5% losses in
observing (although with smart scheduling we may be able to hide the
exchange in a slew).

Next meeting: Dec. 4: follow-up of outstanding filter/exchanger items;
discussion of straw man cadences and how to evaluate them