Previous Article Next Article Table of Contents
IR Cameras: Decisions, Decisions (1Sep93) (from KPNO, NOAO Newsletter No. 35, 1 September 1993) The alert reader, counting on fingers, will have noticed that there are three IR cameras available for spring semester 1994. The following table compares some important instrument parameters. |-------------------------------------------------------------------------| | Instrument | COB | IRIM | SQIID | |------------------|--------------|--------------------|------------------| | Detector type | 256 x 256 | 256 x 256 | 256 x 256 | | | InSb | HgCdTe | PtSi | |------------------|--------------|--------------------|------------------| | QE | 90% @ JHK | 30,40,45%@J,H,K | 7,5,3%@J,H,K | | | < =80% 3+ um | --- | 1% L' | |------------------|--------------|--------------------|------------------| | Well depth, e- | -200,000 | -250,000 | -1,000,000 | |------------------|--------------|--------------------|------------------| | Cosmetics* | good | fair | superb | |------------------|--------------|--------------------|------------------| | Filters | wide variety | J H K K'; 2.12, | J H K L' all | | | (see above) | 2.16, 2.22 narrow | simultaneous | |------------------|--------------|--------------------|------------------| | Scale @ 1.3-m | 0.93 "/pix | 1.96 "/pix | 1.36 "/pix | | 2.1-m | 0.5 | 1.09 | --- | | 4-m | 0.3 | 0.60 | 0.40 | |-------------------------------------------------------------------------| * "cosmetics" is a judgement taking into account response uniformity, bad pixels, recovery from saturation, and electrical stability APART from QE. All three instruments will be converted to the operating system called "Wildfire" during fall 1993, so the system efficiencies external to the dewars will be the same. So, how does one choose? SQIID is recommended for 1.3-m programs requiring modest sensitivity and spatially extensive coverage, and/or for which simultaneity of data is desirable. Our experience is that its very high response uniformity, stability, fast recovery from saturation, large well depth, and low number of defects are very advantageous for programs expecting large dynamic range variations (e.g. mapping star forming regions) or smooth, spatially extended low level variations (e.g. colors of galaxy envelopes) if the integration times are reasonable. These factors, plus the simultaneity of data, "buy back" a lot of the apparent QE disadvantage. With the appearance of IRIM and now COB, SQIID's ecological niche on the 4-m has narrowed to programs which are confusion (resolution) limited rather than sensitivity limited and for which the simultaneity of data is a strong advantage. IRIM and COB show comparable performance in their region of overlap, broadband imaging. Dark current and read noise in each are quite low. Both have the sensitivity advantage of a high QE detector compared to SQIID. COB has a much larger filter complement and capability beyond 2.5 microns which IRIM lacks, so it is the default instrument when either feature is fundamental to the program. For broadband imaging, consider the pixel size vs. the scientific goals. IRIM's pixel size on the 4-m is a better match to present levels of 2 micron seeing, probably giving it the edge for sensitivity limited programs. If the problem is resolution limited, COB has smaller pixels on the 2.1-m than IRIM on the 4-m. For sheer field of view, IRIM on the 1.3-m wins hands down. The bottom line? If you have been using either SQIID or IRIM and find it satisfactory for your program, keep right on using it. If you need the spectral resolution or long wavelength coverage afforded by COB, it is the only choice. For broadband imaging in need of more sensitivity than SQIID can provide, look at the tradeoffs of pixel size vs. field of view, sensitivity, and appropriate telescope with IRIM and COB. If either COB or IRIM are suitable for your program, please say so; this gives greater flexibility in scheduling, which is always helpful and occasionally decisive. Ron Probst
Previous Article Next Article Table of Contents