NIRC 3-5 Micron Observing
Using NIRC in the thermal infrared confronts the observer with a new set of problems. In particular, the higher background count rates and the increasing variability of the atmospheric emission both argue for fast timing patterns and small exposure times. Below we provide some information to help you plan and obtain 3-5 micron observations.

Exposure Times

Following is a list of estimated maximum exposure times for the broad NIRC passbands in the thermal IR. Note that these are estimated from background count rates taken under good to excellent conditions. Under skies with some cirrus or high water vapor content you may have higher background rates.

Timing Patterns

The standard timing patterns for 3-5 micron work are pat4xa and max12ur. Pat4xa is a slower timing pattern, but one which shows good detector characteristics: low read noise, and stability. Max12ur is a 12-bit timing pattern, with higher read noise It is also not nearly as stable as pat4xa. Note that to change between timing patterns requires some "settling time" for the new timing pattern to stabilize. This may be related to the thermal equilibrium of the detector, faster timing patterns heating the detector more. The most notable settling time is between max12ur and any other timing pattern. You may not see the baseline level of the images stabilize for 30-45 minutes. For this reason it is suggested that you try to schedule at least long blocks of time with a given timing pattern.

Note that max12ur does not work well with chopping! If you must use max12ur for your observations, do not use chopping. Do to the volatility of max12ur data, pair-subtracted images show residual vertical striping. Chopping accentuates thise striping.

Subarray Times

To reach low enough integration times you often will need to subarray the detector. Using a lower left subarray will allow a shorter integration time than using a centered subarray, although for larger subarrays the differences are not great. The trade-off is that NIRC and the telescope are geared to centering targets in the center of the full array. Centered subarrays are hence more natural to the system than lower left subarrays. Using a lower left subarray you may have to point at your target, then offset using the "mov" command. For example, let's say you are using a 24x24 lower left subarray. The center of this subarray is clearly (12,12). Hence, after pointing to the NIRC pointing origin you can type "mov 128 128 12 12" to move the target from the center of the full detector to the center of your subarray.

Below are plots of the minimum integration times with square subarrays using pat4xa and max12ur. Both centered and lower-left subarrays are shown.

In this plot the closed symbols represent pat4xa and the open symbols max12ur. As an example, consider the Lw filter. According to the table above, this will saturate in 4 msec. Looking at the plot above, we could use a centered 24x24 pixel pat4xa subarray, or a lower-left 32x32 subarray. We might even try a lower-left 40x40 subarray if conditions are good. If we required the maximum field of view we would use max12ur lower-left subarrays, and can get up to 88x88 or possibly 96x96.

As another example, the M filter requires 1.7 msec or less. This is just feasible with a centered 8x8 subarray in either pat4xa or max12ur, but to get maximum field we can use a lower-left 56x56 subarray and max12ur. For point sources, a 24x24 lower-left subarray with pat4xa might be a good compromise.

Below we recommend subarray sizes using the measured background rates in the 3-5 micron broadband filters:

Recommended maximum subarray sizes

Filter pat4xa max12ur
subc subll subc subll
Ls 144 152 any any
Lw 24 40 48 88
Lp 80 88 192 208
M 8 24 8 56


Troubleshooting

Frame Saturated?

  • Reduce tint to its minimum ("tint 0").

  • Change from patslow to pat4xa, or from pat4xa to max12ur.

  • Reduce your subarray size.

  • Change to a lower-left subarray.

Smearing of images?

  • Look at the compass roses to see if the smearing is in either the az. or el. direction.

  • Change the chop direction ("chpang") to see whether the smearing follows the chop direction.

  • Increase the "frmsettl" parameter to 2 ("modify -s nirc frmsettl=2").

  • Use a shorter chop throw.

Popular Links


Send questions or comments to: