[Last Change: 08 Nov 2013 (rev 3) — Page History]


The Chandra Optics were not calibrated in the IR or UV, but certainly do reflect and focus images from IR and UV sources. The detector assemblies include filters designed to reduce or eliminate the possibility of such photons' accumulation by the detector. Several observations on both ACIS-I and ACIS-S were made to quantify any IR and UV light leakage, with Betlegeuse observed for the former and Vega for the latter.


The complete list of ACIS RAW mode data for Betelgeuse (as of 2014 Dec) is:

OBSID Target CCD Dither Yoff Xoff SIMODE Date-Obs
02593 Betelgeuse I0 Nominal 4.70 -5.68 TE_00528 2002-01-14 13:04:32
03366 Betelgeuse I1 Nominal -4.00 -3.68 TE_00524 2002-01-14 14:46:20
03367 Betelgeuse I2 Nominal 4.70 2.50 TE_00524 2002-01-14 16:32:40
01856 Betelgeuse I3 Off 0.00 0.00 TE_0013C 2000-09-22 11:08:25
03678 Betelgeuse I3 Nominal 0.00 0.00 TN_005EE 2003-02-06 07:57:42
05048 Betelgeuse I3 Nominal 0.00 0.00 TN_005EE 2004-02-02 08:48:16
05963 Betelgeuse I3 Off -0.33 0.00 TN_006E8 2005-02-02 13:46:12
06447 Betelgeuse I3 Nominal -0.33 0.00 TN_00736 2005-11-23 10:17:58
17374 Betelgeuse I3 Off -0.5 -0.1 TN_00C5A 2014-11-20 04:42:26
01855 Betelgeuse S3 Off 0.00 0.00 TE_0013E 2000-09-22 12:52:55
03679 Betelgeuse S3 Nominal 0.00 0.00 TN_005F0 2003-02-06 06:57:42
05049 Betelgeuse S3 Nominal -0.33 0.00 TN_005F0 2004-02-02 07:51:10
05964 Betelgeuse S3 Off -0.33 0.00 TN_006EA 2005-02-02 14:49:43
06448 Betelgeuse S3 Nominal 0.17 0.00 TN_00738 2005-11-23 11:22:57
17376 Betelgeuse S3 Off -0.6 -0.15 TN_006EA 2014-11-16 08:42:19

The complete list of ACIS RAW mode data for Vega (as of 2014 Dec) is:

OBSID Target CCD Dither Yoff Xoff SIMODE Date-Obs
02586 Vega I3 Nominal 0.00 0.00 TE_0013C 2002-02-19 13:03:35
00036 Vega I3 Off 0.00 0.00 TE_0013C 1999-11-05 09:44:17
03685 Vega I3 Nominal 0.00 0.00 TE_0013C 2003-01-27 07:50:40
05046 Vega I3 Nominal 0.00 0.00 TE_0013C 2004-02-05 06:10:26
05961 Vega I3 Off -0.33 0.00 TN_006E8 2005-02-20 21:55:31
06445 Vega I3 Off -0.33 0.00 TN_006E8 2006-01-14 18:30:04
02587 Vega S3 Nominal 0.00 0.00 TE_0013E 2002-02-24 00:38:27
03686 Vega S3 Nominal 0.00 0.00 TE_0013E 2003-01-27 03:23:16
00037 Vega S3 Off 0.00 0.00 TE_0013E 2000-06-07 04:40:58
05047 Vega S3 Nominal -0.33 0.00 TE_0013E 2004-02-05 05:10:57
05962 Vega S3 Off -0.33 0.00 TN_006EA 2005-02-20 22:48:22
06446 Vega S3 Off 0.17 0.00 TN_006EA 2006-01-14 19:25:55

This list includes data where the dither was not turned off. It may be possible for some adventurous soul to correct the dithered data.

For completeness, four observations were carried out with SI modes corresponding to 6-chip TE mode:

OBSID Target CCD Dither Yoff Xoff SIMODE Date-Obs
00984 Vega I0123S67 Nominal 0.00 0.00 TE_002A4 2001-02-01 07:24:10
00985 Vega I23S5678 Nominal 0.00 0.00 TE_002A2 2001-02-01 08:40:25
02594 Betelgeuse I23S1234 Nominal -1.00 0.00 TE_002A2 2001-12-07 12:33:18
03365 Betelgeuse I0123S23 Nominal 4.50 -4.50 TE_002A4 2001-12-16 00:50:10

Observations without dither

This is a list of those data taken with dither turned off. Click on the image to get a FITS file. See below for how the images were created.

OBSID Target Yoff Xoff nmed Image
36 Vega 0.00 0.00 11 <noautolink>00036.png</noautolink>
1856 Betelgeuse 0.00 0.00 9 <noautolink>01856.png</noautolink>
5961 Vega -0.33 0.00 11 <noautolink>05961.png</noautolink>
5963 Betelgeuse -0.33 0.00 11 <noautolink>05963.png</noautolink>
6445 Vega -0.33 0.00 13 <noautolink>06445.png</noautolink>
OBSID Target Yoff Xoff nmed Image
37 Vega 0.00 0.00 5 <noautolink>00037.png</noautolink>
1855 Betelgeuse 0.00 0.00 5 <noautolink>01855.png</noautolink>
5962 Vega -0.33 0.00 5 <noautolink>05962.png</noautolink>
5964 Betelgeuse -0.33 0.00 5 <noautolink>05964.png</noautolink>
6446 Vega 0.17 0.00 5 <noautolink>06446.png</noautolink>

Data Reduction

ACIS RAW images are images as read out from the detector (before event processing is performed) with additional bias columns added (eight per node).

Observations were processed with rawmedadd more , a Perl script using PDL which

  1. bias subtracts each frame using the bias columns
  2. groups the frames
  3. generates a median image from each group
  4. sums the median images.

The median process is used to help remove cosmic ray blooms. The number of frames in each group was adjusted by eye to remove the majority of cosmic ray blooms. There is significant residual (non-source related) structure in the resultant image.

Obsid 00037 included some time at the beginning during target acquisition. The raw images for these 89 frames were excluded from the directory prior to running rawmedadd.

The data were processed with the nmed parameter as specified in the above table.


Richard J.Edgar, Oct 25, 2013

We had a detailed look at the data which are presented here:

Chandra observed both Betelgeuse and Vega in the raw mode, looking for optical transmission of the optical blocking filter for ACIS. Each star was observed on S3 and on I3. There are 10 observations which were not dithered, greatly simplifying the analysis. One round of four observations (two stars by two chips) was done in 1999/2000, another in 2005, and two observations of Vega were done in 2006. A few dithered observations were done between 2000 and 2005.

A summary of the results is as follows.

[1] Betelgeuse is about 5 to 6 times as bright as Vega, as measured by ACIS (either S3 or I3). Since Vega is by definition magnitude 0 in all bands, this ratio leads us to believe that the light leak is in the infrared, between I and J bands, for star-like spectra.

[2] Either star is detected at about 50 times the signal on S3 compared to I3. Said another way, the sensitivity of I3 to optical/IR light is about 2% that of S3. We assume this is also true of the other front-illuminated (FI) chips on the I array. There is some anecdotal evidence for spatial variations in the OBF's optical transmission.

[3] The fluxes of both stars decreased a reasonably consistent factor of about two between 1999/2000 and 2005/2006. It is unknown if that decrease continued. For the analysis here, we assume it has not, that is, we assume the optical sensitivity of ACIS is as it was in 2005/2006, which errs on the conservative side.

[4] About 25% of the optical signal in the central 3x3 pixel region of the optical point spread function ends up in the central pixel.

[5] The absolute brightness of Vega in 2005 was 123.7 ADU per second in a 3x3 pixel window around the maximum (measured in ten-second raw frame exposures).

There are two threshold magnitudes to consider. First, if stars are sufficiently faint, they will produce less than one ADU in a typical ACIS frametime of 3.2 seconds. Optical emission from stars at this magnitude or fainter may be ignored for planning purposes.

A second, brighter, threshold magnitude consists of an optical signal equal to the split threshold used in detecting and grading x-ray events in the ACIS on-board analysis. If pixels exceed this value, nearby x-ray events will be misgraded (and possibly discarded onboard, if all 8 adjacent pixels exceed the split threshold) and their energies will be overestimated by the amount of non-x-ray signal collected in the pixels considered part of the event.

We now consider the brightness thresholds for these two effects.

From [4] above, Vega (with I = 0) deposits 123.7 ADU per second in a 3x3 pixel island. Multiplying 123.7 by a frametime of 3.2 sec gives 385.8 ADU/3.2 sec frame. The central pixel can have up to 25% of this signal, which gives us 99.0 ADU in each frame.

In order to cut this signal to 1 ADU, we need stars 1/99 as bright as Vega in the band where the light leak is (between I and J). log(99)/log(2.5) = 5 magnitudes. Thus any star with both I and J band magnitudes fainter than 5 is observable on S3, without any significant effect on pulse heights.

For FI chips, whose sensitivity is 2% that of S3 (fact [2] above), log(99*0.02)/log(2.5) = 0.75 magnitudes. Thus any star fainter than 0.75 magnitudes in I and J can be observed without worry.

If one must observe brighter stars than this, consider the effect of the split threshold, which is 13 in TE mode for ACIS. The brightest optical pixel is probably also the brightest x-ray pixel, if the source is close to on-axis. Adjacent pixels will average one-eighth of 75%, or 0.09 of the 3x3 pixel island optical signal, and we wish to force this to be less than 13 ADU. Thus the threshold is log(0.9*99/13)/log(2.5) = -0.4 magnitudes. Any star fainter than -0.4 magnitudes in I and J will get less than the x-ray split threshold worth of optical signal on S3, and essentially all of the events will be telemetered. See POG 6.19 for advice on observing modes, and contact the CXC help desk with difficult or confusing cases.

Similarly, for FI chips with 2% of the optical/IR sensitivity, the threshold is correspondingly brighter: log(0.09*99*.02/13)/log(2.5) = -4.7 magnitudes. Thus any star fainter than magnitude -4.7 in I and J bands can be observed without losing x-ray events. Again, see section 6.19 in the POG for other strategies.

Observations of Jupiter give similar results. The signal fades by a factor of about two over the life of the mission (in this case, from 1999/2000 to 2011), and the total signal is consistent with what's expected given that Jupiter is resolved (and covers many pixels) and the stars are not. See Elsner et al. (2005:JGR 110, A01207) for some details on observing strategies and data reduction and analysis.

Shortening the frametime by using a subarray or minimizing the number of CCD chips used lowers the optical signal in direct proportion. Split thresholds can be adjusted if need be. If the bandwidth allows, observing in the VF mode provides corner pixels for each event which can be used to estimate the bias level on an event-by-event basis. For very bright sources, even the event threshold can be adjusted if needed. Such observations (and their analysis) are challenging, but can be done.

If any bright optical source is in the field, it is important to take the bias frame off-target, so that the optical image of the source is not present in the bias frame.

Colors of Betelgeuse were taken from Cousins, A. W. J. (1980), South African Astronomical Observatory Circular, 1, 234

This page is maintained by the CXC Optics Group
CXC logo

The Chandra X-Ray Center (CXC) is operated for NASA by the Smithsonian Astrophysical Observatory. 60 Garden Street, Cambridge, MA 02138 USA.   Email: cxcweb@head.cfa.harvard.edu Smithsonian Institution, Copyright © 1998-2013. All rights reserved.