SoS for ELG still unresolved...

The SoS problem still pertains. The analysis to estimate the SNR2 from the z-band wavelengths did not work. The last two days I have been attacking it from a different perspective. The strategy is to make use of the standard stars in the plate. One could fit the std stars using the relation:

(S/N) = alpha * m^{-2/5}

which fixes the slope to -2/5 assuming poisson-dominated noise, where alpha is a fit parameter that is some proxy for the transparency. One could then apply those values to the relation

(S/N) = alpha^2 * m^{-4/5} * 1/(beta*sky)

which uses the same alpha, has a slope (-4/5) defined by the sky-dominated S/N relationship, has a term 1/beta to convert sky counts to the right units, and 1/sky to account for the sky background. beta is a semi-arbitrary variable - it could just be set such that we get SN2=22 at the minimum threshold.
I believe I have done the necessary edits to the idlspec2d program. Basically, it involved defining a new fitting routine for the ELG test plates (Should I name it fitsn_ELG or fitsn_vm (why not !!)?). I am going to run it tomorrow. Fingers crossed.

Contacted Hee-Jong and Demitri to know about the status of eboss22 tiling and eboss21 plate design. Both are working on respective problems. It is already getting late for eboss21. I am sure the drill shop is gonna scream at me .....