View source for Task answer notes from luisa

=Task 0=

I really hoped that I had successfully reached everyone with some content in June, or at least that people would have time to go back and go through it all again.  In June, my last words to you all were to go home, find a quiet few hours, and go through the wiki again with your notes.  I knew we'd gone through a lot, very fast, but I also knew that you all had been doing things right (=correctly, confidently, independently) while you were here, and that it was going to take some review with the materials on the wiki to get it all straight.  But it seems that I didn't successfully reach everyone, and apparently no one had time to go through it again. I will try to make sure next year's teams come later than June so that the visit is closer to getting tasks done and farther from vacations...  

If you are really feeling terribly underwater here, really, please, sit down with the [[Working with CG4+SA101]] page. I promise, there are TONS more words and instructions and step-through-it exercises there.  When I say I am proceeding as if you are familiar with the stuff from before, I really do mean it...  If absolutely the only thing you get out of this is successfully actually learning just one or two of the tasks from the summer and never getting into this, then that's fine, it's still more than you knew a year ago.

=Task 1=

''TASK: spot-check some of my photometry''

Still needs some spot checks. Have only heard from Russ that maybe one is wrong. Have not put in the time to resolve this one. 

NB: SPOT CHECKS does not mean DO EVERY SINGLE ONE. It means do enough to assure yourself that (a) you know how to do it, and (b) that I did it right in at least a few cases.  HOWEVER, having said that, in another few steps, you'll need to investigate each one separately in as many images as possible, so it is probably worth LOCATING the objects in a few different Spitzer frames, even if you don't do the PHOTOMETRY.  Are they all point sources?

<font color="red">Updates from Luisa, Fri Dec 17 15:28:53 PST 2010</font>. The way that I would go about this task is inexorably intertwined with investigating the images at each band, which was technically task 4.  I am pretty sure that my photometry is ok for most of the sources in the entire field, and most of the YSO candidates. For the source that got corrected midstream (the one previously-known YSO with the corrected coords) for some reason does not have a measured MIPS flux in my photometry. This is weird, and something that I ultimately want to track down.  There are also various kinks and wiggles in the SEDs of the YSO candidates that I want to check.

<font color="red">Updates from Luisa, Mon Dec 20 09:23:45 PST 2010</font>. In order to check my own photometry, I did things in two pieces. Before finding YSO candidates, I made a variety of CMD plots to make sure that the overall distribution looked like I expected it to. (See the wiki discussion of photometry for more on this.)  Then, I proceeded with the analysis below, selecting YSO candidates, looking at images, and making SEDs. If you look at my notes below, you'll see that I made notes where there was possible or likely photometry problems. I then did a more detailed check of the YSO candidate photometry. (This is an iterative process. It always is.) Here are more notes on that.  

WARNING -- in the process of doing this, i rediscovered the factor of ~18 that was lingering last June. I think Russ merged the wrong generation of mosaics into the big mosaic. This shouldn't matter for making pretty pictures, but will definitely matter for doing photometry! 

First, a raw check of the photometry, then notes on what I was looking for/expecting to find. 

Here is a table -- for each (surviving) object, I've tabulated my original photometry and a check I did. the check is redoing the photometry at the position we are reporting for the object in question. [[media:checkphotdec2010.xls|excel version of this information]]

  cat num           i1orig i1check diff    i2orig  i2check diff    i3orig  i3check  diff    i4orig i4check diff     m1orig m1check  diff
  073049.1-470209   18.15   18.15   -0.00   16.51   16.51   -0.00   14.54   14.53    0.01   12.57   12.57   -0.00    8.37    8.33    0.04
  073049.8-465806   14.26   14.28   -0.02   13.69   13.44    0.25   12.39   12.40   -0.00   11.05   11.06   -0.01    8.77    7.11    1.66
  073053.6-465742   14.18   14.18   -0.00   13.87   13.88   -0.01   13.53   13.53    0.00   13.04   13.06   -0.02    9.53    9.57   -0.03
  073057.5-465611   10.86   10.86   -0.00   10.43   10.44   -0.00   10.04   10.05   -0.01    9.26    9.27   -0.01    8.76    6.36    2.40
  073106.5-465454   11.28   11.28    0.00   10.63   10.63   -0.01   10.00   10.01   -0.01    9.09    9.10   -0.01    6.45    6.31    0.14
  073108.4-470130   13.21   13.22   -0.01   12.88   12.89   -0.01   12.50   12.51   -0.01   11.71   11.72   -0.01    8.68    8.89   -0.22
  073109.9-465750   14.35   14.35   -0.00   13.91   13.92   -0.01   13.52   13.54   -0.02   12.82   12.80    0.02   -9.00    9.70  -18.70
  073110.8-470032    8.75    8.64    0.11    8.37    8.22    0.14    7.88    7.89   -0.01    7.65    7.38    0.27    4.46    4.34    0.12
  073114.6-465842   10.07   10.07    0.00    9.63    9.64   -0.01    8.93    8.94   -0.01    7.88    7.87    0.01    4.24    4.11    0.13
  073114.9-470055   13.84   13.84   -0.00   13.54   13.54   -0.00   13.20   13.21   -0.01   12.55   12.57   -0.02    9.82    9.65    0.17
  073121.8-465745    9.26    9.26   -0.01    8.84    8.86   -0.02    8.73    8.36    0.36    8.18    8.16    0.01    5.97    5.75    0.22
  073136.6-470013    9.07    9.41   -0.34    8.81    8.87   -0.06    8.53    8.66   -0.13    8.19    8.66   -0.47    5.44    4.94    0.49
  073137.4-470021    8.50    8.17    0.33    8.21    8.13    0.08    7.98    7.74    0.24    7.19    7.19    0.00    4.23    4.12    0.11
  073143.8-465818   11.33   11.25    0.08   10.65   10.62    0.03    9.98    9.97    0.02    8.77    8.78   -0.01    5.55    5.46    0.10
  073144.1-470008   11.02   10.79    0.23   10.59   10.59   -0.01   10.21   10.23   -0.02    9.50    9.52   -0.02    6.86    6.65    0.21
  073145.6-465917   11.44   11.42    0.02   11.27   11.25    0.02   11.02   11.01    0.01   10.21   10.22   -0.01    7.47    7.00    0.46
  073243.5-464941   12.91   12.82    0.10   12.78   12.76    0.02   12.42   12.17    0.26   11.98   12.01   -0.03   -9.00   10.86  -19.86
  073326.8-464842    9.91    9.91   -0.00    9.70    9.70   -0.01    9.43    9.44   -0.01    8.56    8.58   -0.02    5.14    5.03    0.10
  073337.0-465455   14.04   14.04   -0.00   13.74   13.74   -0.01   13.39   13.40   -0.00   12.79   12.79   -0.00   10.64   10.06    0.58
  073337.6-464246   14.19   14.20   -0.01   14.14   14.15   -0.02   13.75   13.78   -0.03   13.44   13.55   -0.11    9.99    9.49    0.51
  073406.9-465805   13.27   13.27   -0.00   12.99   12.99   -0.01   12.67   12.69   -0.01   11.88   11.91   -0.03    8.63    8.76   -0.13
  073425.3-465409   11.73   11.75   -0.02   11.44   11.29    0.15   10.94   10.93    0.01    9.82    9.88   -0.06    3.54    3.39    0.15
  073439.9-465548   12.99   12.99   -0.01   12.33   12.34   -0.01   11.72   11.73   -0.01   10.90   10.92   -0.02 
  073542.2-470126   14.06   14.06   -0.00   13.09   13.09   -0.00   12.21   12.23   -0.01   11.24   11.26   -0.02 
  073548.5-470727   14.05   14.05   -0.00   13.28   13.29   -0.01   12.56   12.57   -0.01   11.69   11.71   -0.02 

Notes on objects:
*073049.1-470209 - 
*073049.8-465806 - noted in SED inspection as check i4/m1 - i4 well matched, i2 weirdly off, and m1 is in fact off.
*073053.6-465742 - noted in image inspection as faint m1 - m1 ok.
*073057.5-465611 - noted in SED inspection as check i4/m1 - m1 is in fact off.
*073106.5-465454
*073108.4-470130 - m1 a little off. original phot probably more accurate.
*073109.9-465750 - noted in SED inspection as where is m1 - now have an m1.
*073110.8-470032 - noted in SED inspection as check i4 - wildly different all bands except i3. what? 
*073114.6-465842
*073114.9-470055 - noted in image inspection as faint m1 -
*073121.8-465745 - because of multiple source issue, i'm expecting rotten photom here - actually well-matched except i3; m1 a little off. prob original better.
*073136.6-470013 - nothing should be unusual here, but irac photom wildly different all bands except i2, and m1 very different too. what?
*073137.4-470021 - i1i3 both very different.
*073143.8-465818
*073144.1-470008 - noted in image inspection as could be messy irac phot due to source proximity - got different i1 on redo. but this is faint obj next to bright obj. high error measurement anyway. m1 a little different, keep original.
*073145.6-465917 - m1 different. weird, keep original.
*073243.5-464941 - noted in image inspection as could be messy phot due to source proximity - actually well matched except for i3, but this is faint obj next to bright obj. high error measurement anyway. have an upper limit for m1 now (note that number given there is actually an UPPER LIMIT!).
*073326.8-464842
*073337.0-465455 - noted in image inspection as faint m1 - SED inspection implies check m1 - m1 is different, but original probably better because faint.
*073337.6-464246 - noted in image inspection as faint m1 - got i4 a little different - m1 is different, but original probably better because faint.
*073406.9-465805
*073425.3-465409 - got i2 a little different.
*073439.9-465548
*073542.2-470126
*073548.5-470727

Summary of photometry JUST changes/best fluxes to use; if not listed, keep orig fluxes... note that after all that, most of the fluxes are just fine!! but look at how much nicer the SEDs look in these few cases.  Sorry, but I only had time to make one example for you -- see examplenewsed.gif below..
*073049.8-465806 - use new m1 (10253 +/- 426 uJy,  7.11 +/- 0.05 mag)
*073057.5-465611 - use new m1 (20345 +/- 844 uJy,  6.36 +/- 0.05 mag)
*073109.9-465750 - use new m1 (942 +/- 66 uJy, 9.70 +/- 0.08 mag)
*073110.8-470032 - new IRAC fluxes look much better. use those.  [[media:examplenewsed.gif]]  (i1: 98177+/- 4918 uJy, 8.64+/-0.05 mag; i2: 92211 +/- 4620 uJy, 8.22 +/- 0.05 mag; i3 unchanged; i4: 71660 +/- 3591 uJy, 7.38 +/- 0.05 mag)
*073136.6-470013 - no change - original photom better
*073137.4-470021 - no change - original photom better
*073243.5-464941 - no change, except can add m1 LIMIT ( <324 uJy, >10.86 mag)


=Task 2=

''TASK: look at the short lists of YSO candidates''. 

''FAQ: wondering which files to use?'' *ysocand.tbl.

''which are duplicates?'' there are several that are rediscovered between the tiles. i'll leave it to you to identify the specific ones. from the entire catalogs, there are more than 1000 objects in common between the frames.

''are there any that should have been rediscovered but were not?'' no, that is -- if it was in both tiles, then it was identified as a YSO in both independent detections.


=Task 3=

''TASK: did we rediscover any of the [previously-identified YSOs]?'' 

yes, yes we did.  i get 5, or possibly 6; see next item.

''are there any in the maps that we did not rediscover?'' yes, possibly 1.

''why did we not rediscover them? (e.g., do they not have any IR excess? if not, why were they identified as young before? can you think of a reason for this discrepancy, if it exists?)''  I HAVE NOT COMPLETED THIS TASK but i can tell you that some strongly contributing reasons we did not recover one of them is because (a) it requires a larger than "usual" coordinate shift to match it to one of the catalog items (is that a true match? are there other sources in that region it could match to?); (b) for the closest match, we only have 2 irac bands in the catalog, which means that the gutermuth method could never have recovered it. (why do we have only 2 bands? is it on the edge? it it saturated? i don't know; i am leaving this to you.)

<font color="red">UPDATE 23 NOV</font> copy from email.

Mark is working on making SEDs, and had a question re: the source matching
with the previously identified objects:

mark says: I was able to match up 5 previous identified Reipurth & Pettersson
sources, but there's a final sixth that I'm having trouble matching
(see below).  Any ideas?

*CG-Halpha 4?	073121.8-465943	112.84104	-46.995456
*CG-Halpha 4?	073121.8-465744	112.84107	-46.96228
*CG-Halpha 4?	073121.9-465746	112.841607	-46.962904

hm. certainly when I did this before, there was one source where I wasn't
sure if I found the right match or not; it required a relatively large
shift to find the match.  ok, here is what i get for the matching.

*cg-halpha 1 -- 667" from anything. no match.
*         2 -- 1.45" from 073057.5-465611
*         3 -- 1.23" from 073110.8-470032
*         4 -- 1.66" from 073121.8-465943
*         5 -- 0.89" from 073136.6-470013
*         6 -- 0.82" from 073137.4-470021
*         7 -- 1.10" from 073326.8-464842
*         8 -- >10,000 arcsec from anything, no match
*PHalpha 92 -- >58,000 arcsec from anything, no match

Now, remember, this is a blind matching, which the computer is doing
without my intervention. 1.66" is pretty far, compared to the other
coordinates in the list (it's the largest one of all the matches). BUT,
the next closest thing is 6.2" away, much further than the rest.

So, let's go look at the images.  (NB: For one of the tasks below, i made a [http://www.youtube.com/watch?v=OVesekMydFk 5-min YouTube video] to demonstrate something awfully close to the following steps.) I asked IRSA's finder chart service
(http://irsa.ipac.caltech.edu/applications/FinderChart/ -- linked from the
fall work page also) to give me DSS and 2MASS images at this location:
112.84133 -46.995872
and in the DSS images as well as the 2MASS images, there's nothing really
very bright at this location, which already makes me suspicious -- the
prior optical surveys were very shallow, so they should have only found
the bright sources. The fact that there is no bright source here is
worriesome.

[[image:wrongcoord.jpg]]

Here is a screenshot of what I get when I look at the IRAC1234+MIPS1
image of this region, and the last pane is 2MASS J (I think) for
comparison. The green circle in the first panel is centered on the
position for CG4-Halpha 4; the other panes are aligned to cover the same
area, in the same orientation.  The objects at the location given for
CG4-Halpha 4 are faint even in 3.6 um!  And really nothing is there at all
in irac-3 or 4. It MIGHT be there in mips1, but wow that's a stretch. And
it's definitely not there in 2MASS.

SO, my conclusion is that we don't know exactly what object Reipurth and
Petersson were really looking at when they recorded CG4-Halpha 4. It could
have been that one of these faint guys really was whoppingly bright at the
time that they observed it -- after all, young stars do vary with time.
BUT whatever it was is either not really at this location (e.g., they
screwed up the coordinates), or perhaps there was an error in
transcription.

I went back to the Reipurth and Pettersson paper, and it reports in 1950
coordinates:
7:29:54.1, -46:51:19
when i convert this to J2000, I get
7:31:21.95, -46:57:45.1
note that this is -46:57:45, not -46:59:45. That's an easy error to make,
especially if transcribing something handwritten!

With that new position, the closest match is 0.71" away from
073121.8-465745. At this location in the images, there's something
whoppingly bright at the Spitzer bands, and it's seen at MIPS-24 too,
though for some reason this object is missing a flux at 24 um in our
catalog. The object apears to be saturated or somehow other compromised at
8um, so there's no flux there either. I can see it in the 2MASS images,
though it is missing a 2MASS flux in our catalog as well -- perhaps it is
saturated.

OK, so the implications of this are:
(1) we need to update (correct) the list of previously known sources in at
least 2 places on the wiki, and in all of your own notes and spreadsheets.
(2) someone needs to go manually measure the flux densities for him at
least at 24um.
(2a -- i need to track down why there is no mips-24 flux (and for that
matter, JHK and 8um) for this object in my catalog)
(3) the current best ysocandidate list DOES NOT ALREADY HAVE this guy included --
it's in the master catalog, but it didn't have a 24 um flux there either.
The previous ysocandidate list does have 073121.8-465943, which I had manually added on the basis of
the fact that it appeared (blindly) to match the position for the
previously known object. HOWEVER, now we know that this is erroneously
included due to this position error, and it should be dropped from our
list of YSOs.
(4) Now we know the answer to the question I posed in the Fall Work page
for task3 -- did we recover all of the previously known objects in our
fields? no. Are there any we did not recover? yes, one, because of missing fluxes.

=Task 4=

''TASK: for each of the YSO candidates, [investigate them at other bands]''

i guarantee you that at least 3 of the YSOs should be dropped because they are not really stars. not telling you which ones! you go figure this out ...

<font color="red">Update from Luisa Fri Dec 17 15:33:39 PST 2010</font>:
ITEM 1: I made a video that shows how to use the IRSA finder chart service plus ds9 to examine the images for each object. [http://www.youtube.com/watch?v=OVesekMydFk See the 5-min YouTube video] (this version has the watermark removed!).

ITEM 2: The YSO candidate list is starting with 30 objects.  I went through each of the Spitzer bands that we have, looking at each of the YSO candidates. Here are my notes on what I see in the Spitzer images. I did not look in all the 2MASS images. Did you find anything different? The spatial resolution at 2MASS is comparable to IRAC-1, so 2MASS should not reveal many galaxies that we did not suspect from the IRAC image, but it might reveal other useful information (such as "i can see the object here, so why do we not have a JHK flux for it in the catalog?").  No notes from me in the below means that there was nothing unusual or remarkable at i1234 or m1.

 SSTiau name	    	    
  073049.1-470209   
  073049.8-465806   
  073053.6-465742   very faint m1, but probably there.
  073057.5-465611   
  073106.5-465454   
  073108.4-470130   
  073109.9-465750   
  073110.8-470032   
  073114.6-465842   
  073114.9-470055   very faint m1, but probably there.
  073121.8-465744   multiple srcs on top of one, probably not unique detections DROP
  073121.8-465745   multiple srcs on top of one, probably not unique detections. this is best centered, probably best fluxes obtained here.
  073121.9-465746   multiple srcs on top of one, probably not unique detections DROP
  073136.6-470013   
  073137.4-470021   
  073143.8-465818   
  073144.1-470008   src is offset from core of bright star in i1,i2. real? in i3 can clearly see that it's really another source. computer wins!
  073145.6-465917   
  073243.5-464941   stuck in muck from bright star in i1,i2,i3. fluxes prob hi error. even in i4, hard to tell if real source. what does SED look like? i can't see it in m1.
  073326.8-464842   
  073337.0-465455   very faint m1, but probably there.
  073337.6-464246   very faint m1, but probably there.
  073406.9-465805   
  073425.3-465409   BOOMING bright m1.
  073439.9-465548   off edge m1 (no fluxes possible.)
  073501.1-465442   middle of gal, not true pt src much less YSO. offedge m1. DROP
  073504.7-465514   middle of gal, not true pt src much less YSO. offedge m1. DROP
  073508.1-465544   middle of gal, not true pt src much less YSO. offedge m1. DROP
  073542.2-470126   offedge m1.
  073548.5-470727   near diffraction spike i1, prob ok. offedge m1.


PROBLEM CHILDREN:

[[image:multiplesources.png]]
Look at how the red circles are nicely centered on the sources, even the fainter ones, in most cases in this zoom-in, except for that cluster of 3 sources on top of the bright object. (I've added some large red lines to draw your eye towards some of the little red circles centered on the YSO candidates.) Something hiccupped here in that one case -- the computer often has problems with the brightest sources because the PSF is a complicated shape when it gets really bright. The most likely best measurement of the real source is the one best centered on the source.  Drop the other two sources as false sources.

[[image:galaxynotyso.png]]
Here again, look at how the red circles are nicely centered on the sources -- I've added some large red lines to draw your eye towards the little red circles centered on the YSO candidates.  But there seems to be three "sources" within the galaxy. Here, the computer has been fooled by local brightness maxima in the galaxy.  These objects are not single young stars. They may indeed be large star-forming complexes in the other galaxy (they seem to have the right colors), but they are not single point sources, so they're not young stars in *our* galaxy. Drop them from the list.

I thought that 073144.1-470008 was going to turn out to be another problem child, but inspection at the longer bands (i3 and 4) reveals that the computer is correct, there really is a distinct source there.

SO now that we've dropped 5 things off the list, the running YSO candidate list is now down to 25 objects.

=Task 5=

''TASK: for the new YSO candidates[...]go get the optical''

OK, so this was WAY WAY WAY harder than I thought it would have been. there were TONS of duplicates within the optical catalog, as well as objects with no photometry at all.  I did this merging and have emailed around a merged catalog.

=Task 6=

''TASK: [...] make some color-mag and color-color diagrams''

I still think you can do this one. look at the IC2118 paper for specific examples and ideas.  I haven't seen any examples from anyone yet, so I don't know if you need help.

=Task 7=

''TASK: [...] make SEDs''

I sent around examples before, including ones with the optical included.  In order to incorporate the optical data, you need the wavelengths and zeropoints of the optical bands, which I sent in email.

<font color="red">UPDATE from Luisa Fri Dec 17 16:06:58 PST 2010</font>:
Mark Legassie managed to make some SEDs. The midwestern folks asked for me to make some notes on what I thought about each of the SEDs.

*073049.1-470209 very steep! could be very embedded source, could be xgal.  need spectrum to be sure. leave in our list as a mid-grade confidence object
*073049.8-465806 wow, that's a weird one. Either the m1 flux is wrong (too low by a lot)  or this might be an xgal. the high flux at 8um compared to the other bands  could suggest a high-mass young star or a complex of hi-mass YSOs in a  star-forming galaxy. i didn't note anything weird about the 24 um data in  my anaylsis above, so there is nothing screamingly obvious suggesting that the photometry is wildly wrong. need a spectrum. 
*073053.6-465742 nice. this is a keeper.
*073057.5-465611 also nice. possible that the 24 is a little low or the 8 is a little high.  we may wish to go back and check the photometry on those two points. i didn't note anything strange at 8 or 24 in the data, so there's nothing screamingly obvious suggesting that the photometry is wildly wrong.
*073106.5-465454 hm. broader than I would expect for a star. could be a real yso, could be xgal. need spectrum.
*073108.4-470130 nice. this is a keeper.
*073109.9-465750 pretty nice. where is the 24um flux? it is in the image, so i will go back and see if i can get a flux estimate to add in to the SED.
*073110.8-470032 nice. 8um is maybe a little low. would be useful to go back and check this flux.
*073114.6-465842 pretty nice. need spectrum to be sure.
*073114.9-470055 pretty nice. plain photosphere is steeper than this, so it has excesses.
*073121.8-465744 SHOULD BE DROPPED as dupe! the JHK and m1 fluxes accidently assigned to him should go to the next object.
*073121.8-465745 this is the legitimate object of this trio. the optical is appropriately assigned. not sure why no 8um flux -- probably take 5.8 and 8 um flux from prior obj, just by looking at shape of SED.
*073121.9-465746 SHOULD BE DROPPED as dupe! note that the optical from the prior source is identical here -- the optical got assigned to both srcs. note that the IRAC bands here are systematically lower than the above source -- it's not centered on the source, so it's not catching all the flux, so it's artificially low.
*073136.6-470013 nice. keeper.
*073137.4-470021 nice. keeper.
*073143.8-465818 nice. also a keeper. 
*073144.1-470008 nice. a keeper. (plain photosphere is steeper than this.)
*073145.6-465917 nice, keeper.
*073243.5-464941 this one has no m24, and really isn't in that image. this one is stuck in the muck of a bright star, and might have an excess only at 5.8 and 8 um (look, it seems to curve up). weak, marginal candidate. even getting a spectrum would be hard given proximity of bright star. 
*073326.8-464842 nice, keeper.
*073337.0-465455 small excess if any. very faint at 24um. weak candidate only because weak excess.
*073337.6-464246 nice, keeper
*073406.9-465805 nice, keeper.
*073425.3-465409 wow, this one is still just really really odd to my eye. it is clearly really and truly there and booming bright at m24. this may be the most interesting object in the set, because it's very embedded but clearly very bright. would like to get a spectrum, but may be too embedded to do anything from the ground!! list as mid-grade confidence jsut because we don't have any strong evidence that it's NOT a bg galaxy except for the 24um brightness.
*073439.9-465548 meh. no 24um possible. doesn't scream YSO to me, but maybe jsut because there are not a lot of points. we have stronger candidates. need spectrum.
*073501.1-465442 DROP
*073504.7-465514 DROP
*073508.1-465544 DROP
*073542.2-470126 meh. no 24um possible, weird that no 2mass or optical. weak candidate because of lack of data.
*073548.5-470727 meh. no 24um possible, weird that no 2mass or optical. weak candidate because of lack of data.

=Task 8=

''TASK: [...] SED slopes''

Let's wait and see if you get to task 7 before I just go ahead and do this one.

=Task 9=

''TASK: [...] 3-color images with sources indicated''

hint: use region files and ds9.