expose.param:

        20.0            ! magnitude limit
        10.0            ! S/N
        5.0             ! Photometry aperture (radius in arcsec)
        0.61            ! Effective Telescope Diameter (meters)
        15              ! Readnoise (e- rms)
        0.57            ! plate scale (arcsec/pix)
        0.7             ! Wavelength independent throughput
        qe.TEK24        ! CCD quantum efficiency file (list: UBVRI JHK griz)
        sky.SC          ! Sky brightness file (list: UBVRI JHK griz)

   expose.output:

        #                 S/N =   10.00
        # photometry aperture =   78.54    (arcsec^2)
        #      telescope area =    0.29         (m^2)
        #          read noise =   15.00      (e-/pix)
        #         plate scale =    0.57  (acrsec/pix)
        #        photons/
        # Band   m^2-sec    sky      e    m(lim)  t(rn)[sec] t(snr)[sec]
        #         (m=0)
            U   0.410E+10  20.80   0.070  20.00   0.168E+04   0.591E+04
            B   0.142E+11  21.50   0.140  20.00   0.453E+03   0.620E+03
            V   0.879E+10  20.30   0.280  20.00   0.125E+03   0.952E+03
            R   0.107E+11  19.50   0.350  20.00   0.396E+02   0.119E+04
            I   0.732E+10  18.40   0.175  20.00   0.423E+02   0.923E+04

NOTE that expose.f uses two input lists: qe.TEK24 and sky.SC. These
currently contain VERY rough guesses for the detector QE and sky
brightness. The column headed 't(rn)[sec]' gives the time to reach
background or source limit on a given exposure. In general, your
exposures should be several times this value, all else being equal
(e.g. if guiding/tracking is not an issue). The column headed
't(snr)[sec]' gives the total time to reach the specified S/N.