ASAS Tutorial

Here are the original messages and instructions from the chart team beginning in 2003 regarding the use of ASAS-3.

It would be nice to re-write all this into one cohesive document, but I haven't had the time to do it yet. If someone else wants to help distill this down, I would be glad to help edit and check it. Until then, this is better than nothing.

Original ASAS-3 tutorial from Mike 9 August 2003
(revised links November 30, 2011)

I have mentioned ASAS-3 data in several posts. I will try to de-mystify this
subject for you and hopefully get some of the new members active in mining
this data for stars in the southern hemisphere from -90 to +10 or +15 dec.

The home page describes the project in some detail so here is the link.

You can plot light curves and search for variables and any number of pretty
neat things here just for fun if you want to.

To gain access to the photometric information start here:

Scroll down on the left menu to the "search photometric catalogs".

Click on I-Band, the V-band link doesn't always work. I don't know why and I
don't care. As you'll see it doesn't matter.

A new window pops up. Take the top right split screen and scroll down to
where you can enter coordinates, near the bottom of the page.

Enter the coordinates of the star, or the center of the field of interest.
Be sure to use the format requested.
Specify the area of search. 900 seconds is 15x15minutes, an e chart. Don't
go much bigger than this or you will get a mountain of data.
Click on "V-band" for the photometry. (See I told you it didn't matter)

Hit submit.

A list will pop up of all the stars measured in the field, the number of
measurements and the V mag and errors.
I usually copy and paste this into a text editor and print it so I have a
hard copy and a file in the PC.

The stars are conveniently ordered by increasing declination to ease finding
them in your planetarium software or whatever you are using.

I have done quite a bit of comparison to fields with other reliable
photometry (ala Henden) and compared it to Tycho 2 V Bessell and it stands
up well, to around 12.5. After that it gets a bit shaky.

You'll see errors are usually less than 0.03 for stars brighter than 12.5V.
This is pretty damn good. I would have no problem in adopting these values
on AAVSO charts.

So, why is this useful?

Well, there are a great number of variables that will never be calibrated by
Arne for safety reasons. As I understand it, the 1m telescope rides on a
liquid bearing of mercury, highly toxic. If he tries to point the telescope
too far south mercury will spill out all over the place and people will die.
Astronomy is indeed a serious business!

So, this data covers the southern sky where Arne will never go with the USNO
1m scope. There are many brighter variables that could have new and improved
charts made using this data and Tycho 2 only for comp stars. Southern
observers have a distinct advantage over the northerners in this respect.
TASS is not nearly as organized or accessible, even though it's intent is
much the same (more on this later too).

Even in fields that Arne can calibrate, the data is usually saturated for
stars brighter than 12 or 12.5V. So this ASAS-3 data fills a void between
Tycho 2 (down to 10.5) and Henden (from 12.5-19+)

So, if some of our new southern observer/members would like to research the
existing charts for southern variables and can document how screwed up they
are, and what might be done to fix them, you will get my attention. If there
are no existing charts for variables you feel are in some way important or
deserving of new charts, write up a brief proposal and be prepared to do
some of the work.

There is much to be done that CAN be done now, if enough of you are able to
do the work. Mati is a gold mine of excellent information for southern
variables, and he and Bruce (a transplanted Aussie) are experts in the art
of creating sequences. If you have questions, ask. We will help.


Determining magnitudes from ASAS-3 data (from Bruce Sumner)

      This tutorial assumes that you have read and used Mike's excellent instructions (ASAS-3 tutorial, 9 August 2003) and can obtain a list of ASAS-3 magnitudes for a particular field. You should review that article before proceeding.

      I assume that you are in ASAS, have searched for a field by variable name or by coordinates, and have the results on your screen. Firstly it will be necessary to drag the middle frame divider towards the top so that the bottom two frames take up most of the screen.

      The bottom left panel shows all the stars in declination order. If you do not find your star where you expect it to be, you may have to go down the list to where it appears to repeat. This happens because there are often two results for each star and they are separated into two lists, on following the other. So take care to find the proper entry for your star.

      In the bottom left panel you should now click on the star for which to want to find a magnitude. A graph of the individual magnitudes of your chosen star will appear in the bottom right frame. If you can read off a magnitude to a tenth of a magnitude i.e. you can obtain a visual magnitude, from the plot, then this star is suitable as a comparison star. If the star is obviously variable, skip it, and find another star. It cannot be stressed too much that you MUST, in every case, look carefully at the plot and determine a visual magnitude from the star plot.

      The next step is to obtain the actual magnitude of the star to V=0.01 magnitude.

      Click the 'data' button at the bottom of the plot. This will bring up a long list of the individual magnitudes, headed with a summary of this data. It is in the summary area that our magnitude will be found. First note that the long listing is normally split into several sub-lists, each headed by a summary and followed by the list of individual measures.

      So which summary to use? Scroll through the summaries and find the one with the largest number of observations in the 'number of observations' field. This number will generally be well over 100. Note that summaries based on fewer that 100 observations cannot be used if possible. Here is an example for comp star #1 from the sequence for WZ


#dataset= 2 F0728-24_212
#desig= 071844-2705.6
#cra=    7.312311  07:18:44.3
#cdec= -27.092959 -27:05:34.7
#class= 0
#cmag_0= 11.821
#cmer_0=  0.048
#nskip_0= 1
#cmag_1= 11.809
#cmer_1=  0.036
#nskip_1= 1
#cmag_2= 11.816
#cmer_2=  0.038
#nskip_2= 1
#cmag_3= 11.831
#cmer_3=  0.046
#nskip_3= 1
#cmag_4= 11.856
#cmer_4=  0.056
#nskip_4= 1
#ra=    7.312309  07:18:44.3
#dec= -27.092850 -27:05:34.3
#ndata= 176

#   HJD    MAG_1  MAG_0  MAG_2  MAG_3  MAG_4  MER_1 MER_0 MER_2 MER_3 MER_4 GRADE   1868.78085 11.848 11.834 11.874 11.897 11.939 0.020 0.040 0.018 0.021  .026 A 142

      Check that the number of observations (ndata) is over 100. In this case the largest number is 176. This is the third summary for this star. The first summary has ndata=5 and the second summary has ndata=17.

      Now you will notice that there are five magnitudes and associated errors in each summary. These magnitudes are obtained by using different apertures in the processing of the magnitudes. These magnitudes generally vary slightly from aperture to aperture. Here are the apertures used by the ASAS team:

MAG_0 = 2 pixels = 30 arcsec   V fainter than 12.0

MAG_1 = 3 pixels = 45 arcsec   V = 11.0 – 12.0

MAG_2 = 4 pixels = 60 arcsec   V = 10.0 – 11.0

MAG_3 = 5 pixels = 75 arcsec   V =  9.0 – 10.0

MAG_4 = 6 pixels = 90 arcsec   V = brighter than 9.0

      So which aperture should be used for our magnitude? Luckily the ASAS team has already determined this for you. Notice the headings for the individual measures following the summary. For our example you will see that the first column of magnitudes is headed with 'MAG_1'. Each star you investigate will have a different order for the individual measures. The first column is always the ‘correct aperture’ for each star, and follows the above table. In the above example the first column is MAG_1, and in the summary cmag_1 gives a magnitude of V=11.809+/-0.036 This is the magnitude for this star. This aperture is confirmed as the correct one for this star in the above table where MAG_1 = cmag_1 is used for stars between magnitudes 11.0 to 12.0.

      Always round the values to 0.01 magnitude, both for the magnitudes and the errors. Quoting higher accuracy is statistically insupportable. For our example the magnitude should be quoted as V=11.81+/-0.04.

      Now getting back to the list of magnitudes in the right panel, why can't these be used as they appear? Because the magnitudes shown here are always from Aperture 0, and aperture 0 only gives accurate magnitudes for stars of magnitude 12 and fainter.

      However the plots in the right frame are for the correct aperture for that star, as they are the individual magnitudes that appear in the first column of individual measures. Which is why the plot can be used to find a visual magnitude.

      It should be noted that sometimes you will see that the magnitude error is smaller for an aperture different from the ‘correct aperture’ determined from the aperture table. Don’t be tempted to use this magnitude, ALWAYS use the magnitude associated with the ‘correct aperture’ found in the table. 

      So there you have it. Here are a few additional hints.

1. ASAS-3 can only be used to determine visual magnitudes. Under no circumstances can ASAS magnitudes be used as comparison stars in photometric work. This is basically because ASAS magnitudes are untransformed, the magnitudes are derived from only filter only. 

2. ASAS-3 should only be used between magnitude 8.0 and 12.9. Never go fainter than magnitude 13. At 13.5 errors are often in the order of 0.2 or 0.3 magnitude increasing to 0.4 or more at magnitude 14. Magnitudes for bright stars up to magnitude 6.0 can often be obtained if care is taken, but generally there is more reliable photometry available for bright stars from other sources.

3. ASAS-3 can only be used in uncrowded areas, i.e. away from the Milky Way. Only the smallest aperture (MAG_0) will normally be uncontaminated with neighbouring stars. Crowding in Sgr and Sco, for example, often cause errors in ASAS-3 magnitudes of around 0.2 or 0.3 magnitude, even for bright stars. In these types of fields if you need magnitudes you have to choose stars that are completely isolated, but still expect larger than normal errors.

4. There is no colour information available for ASAS-3 stars. ASAS-3 can only provide untransformed V magnitudes. 

5. ASAS pixel size is quite large at one pixel = 15 arcsec. 

Noting all the above caveats, ASAS-3 is a superb source of homogeneous CCD(V) magnitudes, with absolute errors of only a few hundredths of a magnitude. They are generally better that Tycho-2 magnitudes, and for non-Milky Way fields in the Southern Hemisphere they are currently unbeatable between magnitude 7 to 13.

Bruce Sumner

Additional Notes from Sebastian Otero
Sun, 22 Aug 2004 16:49:13 -0300

> > I remember seeing a list from you some time ago about which apertures
> (0-4)
> > you use for ASAS-3 magnitudes. Can you pass it please, I want to revise my
> > instructions. Thank you.
> > Bruce
> MAG_0 = 2 pixels V= <12.0
> MAG_1 = 3 pixels V= 11.0 - 12.0
> MAG_2 = 4 pixels V= 10.0 - 11.0
> MAG_3 = 5 pixels V= 9.0 - 10.0
> MAG_4 = 6 pixels V= down to 9.0
> Hi, Bruce,
> This is what I use and is actually what the ASAS team is using for their
> first column "right aperture" photometry online.
> I took a look at your tutorial and it looks like it's ok but a "hole"
> between mag. 9.0 and 10.0 made the larger apertures be off by 1:
> ""As a guide, cmag 4 is used for bright stars, about magnitude 7, cmag 3 for
> magnitudes 8 and 9, cmag 2 for magnitudes 10 and 11, cmag 1 for magnitudes
> between 11 and 12 and cmag 0 for fainter than magnitude 12.""
> When I find a nearby star, I use a smaller aperture, the largest possible
> when the difference between consecutive apertures gets small. For instance:
> #cmag_0= 8.666
> #cmer_0= 0.048
> #nskip_0= 1
> #cmag_1= 8.659
> #cmer_1= 0.036
> #nskip_1= 1
> #cmag_2= 8.645
> #cmer_2= 0.038
> #nskip_2= 1
> #cmag_3= 8.531
> #cmer_3= 0.046
> #nskip_3= 1
> #cmag_4= 8.361
> #cmer_4= 0.056
> #nskip_4= 1
> I use 8.65 since the decreasing "normal" fading trend related to smaller
> apertures is recovered at aperture 2 meaning no more contamination from the
> companion.
> Anyway, these stasr should only be used if no other photometry is available
> since we are prone to errors using them.
> Remember that now the ASAS-3 bright limit is also brighter due to a change
> in exposure times (three 1-second images instead of a 3-seconds one). It can
> be used up to 6.0 or even brighter (but some isolated saturated datapoints
> may start appearing specially brighter than 5.7) Anyway, there's no need
> to have V data for these bright stars at least for sequence purposes, since
> there are plenty of photometry.
> Cheers,
> Sebastian.

More recently-

> From: "Sebastian Otero" <>
> To: <>
> Sent: Tuesday, July 29, 2008 12:11 AM
> Subject: Re: [Aavso-sequence] off topic: choosing the appropriate ASASphotometry
>> Hi, Aaron,
>>> Hi, all. This does not have to do with sequence making per se. However,
>>> I need some ASAS advice from people who actually use it so you all
>>> came to mind first. Plus the answer to my question will probably be
>>> valuable to new sequence makers who haven't used ASAS yet.
>>> The question is about choosing ASAS photometry. I remember how to
>>> do it in the old days by matching the order of the APERTURE fields with
>>> the magnitude of the star. However, I haven't used ASAS in years and
>>> don't know how to choose the correct field in their current system.  Right
>>> now do I just take the photometry with the highest "grade", ignoring
>>> everything with "D" or below? Is it as simple as that or is there another
>>> trick to the new system?
>> The only thing that changed with the new reduction is that the quality of
>> the results is worse...
>> And there is an strange thing that I noted: the mean magnitudes in the
>> left panel of the ASAS new form are the same as the mean magnitudes
>> in the old catalogue (and also the number of observations quoted, even
>> though now there are a lot more of data).
>> That figure should be ignored, since the magnitude quoted comes from
>> mean magnitudes from the sometimes corrupted all catalogue causing
>> wrong numbers.
>> As Marc said you need to go to the data itself and see the mean magnitude
>> figure from the right aperture, MAG_0 for stars <12.0; MAG_1 for stars
>> between 11 and 12, MAG_2 for mag. 10-11; MAG_3 for mag. 9-10 and
>> MAG_4 for everything brigther than 9.0.
>> Now, since the new catalogue
>> ( has those zero point
>> problems, the best mean magnitudes come from the old catalogue at:
>> A look at a light curve to make a comparison between the old reduction
>> and the new one speaks for itself regarding the quality of the results:
>> New reduction:
>> Old reduction:
>> This is HD 59462, a constant star of V= 8.405 (GCPD).
>> Lots of "false eclipses" in the new reduction. Even images that gave
>> "normal" data before now give wrong results. The overall scatter increased
>> a lot too.
>> The mean magnitude from the entry with the largest number of observations
>> in the left panel is 8.439.
>> However, clicking on the GetData button we get V= 8.412 from the MAG_4
>> mean magnitude of the largest dataset in the NEW CATALOGUE and V= 8.419
>> from the old catalogue.
>> Other example, HD 109892.
>> the new catalogue has two large datasets, one with V= 8.473 and the other
>> with V= 8.488.
>> It is V= 8.450 in the old catalogue.
>> V= 8.44 in the GCPD but with measurements between 8.42 and 8.46.
>> But sometimes different datasets have very diferent mean magnitudes...
>> This happens a lot in the Musca area.
>> That was HD 109993 in Musca. Of the two main datasets, one goes V= 8.024
>> and the other V= 8.207 !!!!
>> Which one is right?
>> Let's go to the old catalogue. Ignoring the early saturated data, we see a
>> very different light curve, with both datasets matching:
>> The mean V mag are V= 8.169 and 8.164
>> GCPD says V= 8.158 and Hipparcos V= 8.16.
>> And the slightly fainter value only happens because some saturated
>> datapoints at the early times are used to derive the mean mag., if not the
>> zero point is completely okay. So if you want V mag. for bright stars you
>> have GCPD and HIP but in case you need ASAS, just take a look at the
>> hundred of individual observations after the saturation problems were
>> corrected (perfect data between HJD 2452950 and 2453250 for bright
>> stars) and you can get the V mag that way (in the old catalogue)
>> In short, sometimes you can get good mean values from the new catalogue,
>> but sometimes not, or you have to choose between datasets so you need
>> external confirmation. The old catalogue does not have that kind of problem
>> and has a very good zero point comparable to standard V in all fields.
>> When there are no data in the old catalogue (some fields are corrupted) you
>> can use the new one but it is safer if you use a couple of GCPD/Hipparcos
>> stars in the field to apply a zero point correction if needed.
>> Finally, the new reduction has actually improved things a little at the
>> faint end. Data at V= 13 or 14 makes more "sense" then before... But
>> everything bright, 7, 8, 9, 10th mag has become really really worse.
>> Cheers,
>> Sebastian.

Another case in point example

From: "Sebastian Otero" <>
Cc: "AAVSO-DIS" <>
Sent: Wednesday, September 03, 2008 8:18 PM
Subject: Re: [AAVSO-DIS] VSP and EU Del / U Del comp stars

>> It is interesting, that Hipparcos indicates it was constant. Especially
>> since the ASAS-3 data show some sign of variability between 6.2-6.8:
>> but maybe the error of these data are bigger than the real amplitude -
>> if there is any real variability.
> Those are saturated observations since the star is very bright, which
> translate in fainter results.
> A clue to realize when a measurement is saturated or not:
> Look at the magnitudes in the first too columns, MAG_4, reduced with the
> largest aperture, the right one for such a bright star, and MAG_0, that
> corresponding to the smallest aperture. When there is a difference of more
> than 0.03 or so mag. between them, something is wrong.
> An example of a saturated magnitude, followed by a good one:
>   2947.54545  6.541  6.720  6.695  6.587  6.555    0.037 0.039 0.032 0.028
> 0.033  A 78969
>   2950.53033  6.244  6.265  6.240  6.241  6.244    0.050 0.042 0.035 0.035
> 0.043  A 79404
> The mean mag of the star seems to be V= 6.25 or so in ASAS, in agreement
> with the V= 6.24 in Hipparcos.
>>> Finally, NSV 13285 is just a suspected variable, with no published
>>> observations.  The Tycho/Hipparcos epoch photometry can be obtained
>>> through VizieR and shows some hint of a small variation in the 0.05mag
>>> range or less.  Personally, I'd consider NSV 13285 as one of
>>> those "guilty until proven innocent" stars; show me some real
>>> photometry before making them suspect!
>> So, you mean we can probably use this comp star in the future too?
> ASAS-3 also suggests variability at the 0.05 mag. level (7.07-7.12),
> typically seen in K giants.
> It doesn't seem to be a problem for visual use.
> Cheers,
> Sebastian.