[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: more details of signal-to-noise calculations for Mark IV
I thank Michael for his nice computation. The computations that I did for
the design of the Mark IV had somewhat more pessimistic numbers for the
photons from a mag 0 star. This led to the assumption that we would want to
make exposures in the same range as the Mark III. 6 minutes or so. I
checked everything - i.e. the lens focal length - on scaling from the Mark
III. The focal length determines the star spacing at any given magnitude,
and thus when we get into trouble (always) separating stars. If a one
minute exposure is long enough, then the electronic read out time is a
little long - 50 seconds. It is possible that with better lenses we will be
able to gain over the Mark III because the smaller spot will increase the S/N,
It sure would be nice to run with the shorter exposure. This would allow
covering a lot more sky and we could take more frequent pictures of the
whole sky. Note that it is a SQRT. Taking 1/4 the exposure costs only a
factor of 2 in the S/N.
All this just to point out that nature is perverse. I have designed the
system so that it will be relatively easy to upgrade to a faster read out.
But we really want the fast read out at the beginning of the survey where we
would like to cover the sky as quickly as possible, and "skim off" the easy
data. Once we have done this we will likely want to take longer exposures
to go deeper, and then the slower ADC will not be much of a penalty.
Michael may remember my first naive post on this subject about 4 years ago.
"Would an all the sky all the time survey to mag 15 be useful?" Michael
answered "Yes!", and I have been trying to do it ever since. The goal is in
At 03:12 PM 10/28/98 -0500, you wrote:
> Tom Droege asked for specifics on the method I used to calculate the
>relative contributions of sky noise and readout noise in Mark IV images.
>Okay, because _you_ asked for it :-), I include the gory details below.