Overheads

The following table lists overheads as determined during the LBC-Blue and -Red commissioning runs in 2006-2007 and 2007-2009 respectively (*) as well as from image timestamps.

Overheads for a single-filter OB amount to about 25% of the total open-shutter time.

Operation	Overhead
Telescope preset*	1-3 minutes
Collimation (dofpia or dohybrid)*	15 minutes at beginning of night 3-5 minutes after slews to new targets and every 30-40 minutes for longer programs
Pointing/Co-Pointing correction	5 minutes
Sending OB to telescope*	10-15 seconds
Mirror preset (focus)*	4-30 seconds (longer at beginning of OB, shorter between exposures)
Camera preset*	~20 seconds
Exposure preset and post-exposure operations*	10-12 seconds
Image readout*	LBCB 20-22 seconds, LBCR 17-21 seconds (see below)
Filter change	7-24 seconds
Time between dithered exposures	~40-50 seconds

 

Detector overheads

The overheads for image readout in the table above come from the Commissioning Reports released in 2009 and are for a full-frame readout. For whatever reason, the detector overheads measured recently are slightly higher, but still higher for the Blue than for the Red channels. Several series of biases taken over the last year 2018-2019 lead to the following results for the full frame detector overheads.

When observing binocularly with LBCB and LBCR, and especially with exposures of unequal durations (for example, taking a long U-Bessel exposure and 2 shorter i-SLOAN exposures), taking into account the blue and red detector overheads can reduce the amount of ‘dead’ time.

 

nrows	t_overhead(s)       LBCB	t_overhead(s) LBCR
4608	34	28

Sub-windowing

Sub-windowing is often used for high cadence observations, e.g. for exoplanet transits. For the LBCs, only the number of rows can be restricted; the sub-window must cover all 2098 columns. See Setting up sub-windowing below for instructions on how to set up sub-windows in your LBC scripts.

Sub-windowing overheads
Detector overheads for several different sub-windows are listed below. These were measured from a series of ~25 biases taken on 20190203, and the values in the table are the average values +/- the standard deviations. The values below are for all 4 chips being read out. But since each chip is read out independently, the overheads when reading out all 4 chips and only chip 2 are similar.

It appears that there is little advantage of setting up subwindows that cover fewer than ~1000-1500 rows, as at that point, irreducible overheads dominate.

nrows	Y_extent (arcmin)	YMIN	YMAX	t_overhead(s) LBCB	t_overhead(s) LBCR
4608	17.3	1	4608	34.2 +/- 1.5	26.8 +/- 1.8
3000	11.25	1416	4415	26.7 +/- 0.9	19.0 +/- 0.2
2000	7.5	1916	3915	22.2 +/- 0.5	15.5 +/- 1.6
1500	5.6	2166	3665	20.3 +/- 1.3	15.4 +/- 1.9
1000	3.75	2416	3415	18.9 +/- 1.3	15.5 +/- 1.3
500	1.875	2666	3165	18.9 +/- 2.0	15.5 +/- 1.4

Setting up sub-windowing:

How to center the sub-window on the optical axis

The sub-window should ideally be centered on the rotator center, which is also the optical axis. The LBC scripts have always required that the same subwindow be used for both Blue and Red, so only the average value of the Y coordinate of the rotator centers on Blue and Red is used. This is 2916.

Rotator Center LBCB       (x,y) [pixels on Chip 2]	Rotator Center LBCR        (x,y) [pixels on Chip 2]
(1033,2919)	(1079 2912.5)