AAVSO Extended format: Check and comp star magnitudes

Affiliation
American Association of Variable Star Observers (AAVSO)
Tue, 06/24/2014 - 15:22

Based on the T-C-K example (present and previous) I have been submitting observations with standardized instrumental magnitudes for the comp and check stars. However, I noticed in Deick Szabo's AAVSO report application that it is populating these fields with raw instrumental magnitudes (negative numbers). 

It occurred to me that submitting the raw instrumental magnitudes in these fields  allows AAVSO to extract "pseudo zero points" (Zpt -k'X) and also makes it possible to assess sky conditions for sets of data with more than a few observations more easily by detecting non-systematic variation in the comp or check raw magnitudes. With standardized magnitudes flux variations in the comp and check tend to offset each other and reduce the effect of variations in sky quality which makes assessing sky quality for a set of observattions more difficult. On the otherhand, if sky quality variations don't show up in the scatter of check star standardized magnitudes, it may not matter unless the separation between comp and check is significantly smaller than the separation between comp and target. 

Which values does AAVSO prefer in these two fields, standardized instrumental magnitudes or raw instrumental magnitudes?

Brad Walter, WBY

From the extended file format page:

2. A little more information:

#TYPE=EXTENDED
#OBSCODE=TST01
#SOFTWARE=GCX 2.0
#DELIM=,
#DATE=JD
#OBSTYPE=CCD
#NAME,DATE,MAG,MERR,FILT,TRANS,MTYPE,CNAME,CMAG,KNAME,KMAG,AMASS,GROUP,CHART,NOTES
SS CYG,2450702.1234,11.235,0.003,B,NO,STD,105,10.593,110,11.090,1.561,1,070613,na
SS CYG,2450702.1254,11.135,0.003,V,NO,STD,105,10.594,110,10.994,1.563,1,070613,na
SS CYG,2450702.1274,11.035,0.003,R,NO,STD,105,10.594,110,10.896,1.564,1,070613,na
SS CYG,2450702.1294,10.935,0.003,I,NO,STD,105,10.592,110,10.793,1.567,1,070613,na

Instrumental magnitudes

Hi Brad,

We would like to have the instrumental magnitudes for comp and check if possible.  That is what we request in the Extended File Format.  If you provide both the comparison star ID and the ChartID, then we would know the standardized magnitude for the comp by default, and the instrumental magnitudes would let us do a lot more with the data.

I have a blog post about this, available here, also linked from the CCD Resource and Reading List.

Matt

Affiliation
American Association of Variable Star Observers (AAVSO)
instrumental mag comparison stars

Good morning everyone. apropos of the instrumental magnitude of the star and the comparison star checkeo, if more than one comparison star photometry in the zero point would be a combination of all zeros of the stars used points is used, as would be reported in the extended format such information only when the instrumental magnitude of a comparison star calls?. to name one example, allows up to 9 MaxInmDL comparison stars.

Affiliation
American Association of Variable Star Observers (AAVSO)
Multiple Comp Stars (Ensemble Photometry)

Oliver,

You don't enter a comparison star name or magnitude if you are using multiple comparison stars (Ensemble Photometry). You enter ENSEMBLE for the comparison star name (CNAME) and na for the comparison star magnitude (CMAG). you should list which comparison stars you used in the notes column using either the chart label or the AUID number so that AAVSO knows what comp stars you used. I suppose that you could also enter the raw instrumental magnitudes of each comp in the notes section but the notes field is pretty short it is supposed to be 100 characters including spaces and punctuation, but I have found that it accepts far fewer than that.  If you entered LLL-MM,MMM "space" for each comp star, you use 11 characters per star and I think you would run out of space after 3 or 4 stars.  I don't think AAVSO stores any header information other than the defined fields. So I don't think adding extra comment lines to the header would save the information.  Can AAVSO staff comment if this is correct?

You can look at the example on the Extended File Format page 

 http://www.aavso.org/aavso-extended-file-format

which you can access from the link on the WebObs file upload page. 

I hope this answers your question. 

Brad Walter, WBY

Affiliation
American Association of Variable Star Observers (AAVSO)
instrumental magnitudes without color correction?

thank very much Brad. is correct what you say, I thought in a somewhat incoherent possibility to introduce the combined magnitude of the comparison stars used, but for the purposes of verification would be nothing practical.

Now, suppose to report to only use a star AAVSO comparison and another to checkeo, there are still a factor that I have not understood, the instrumental magnitude will introduce to amber, is literally Instrumetal magnitude (always negative in my software ) or this same magnitude put only corrected with the coefficients (to be used in the equation: mrawstar - mrawcomp + MC, in my case, DSLR and use their channels away from the standard system, so that the coefficients have to own magnitudes backing then apply mrawstar -. mrawcomp + MC, but performing it in reverse may obtain the same result (I have not done in reverse) that is, if the photometry is transformed instrumental magnitudes are sent literally backing or processed but not apply mrawstar - mrawcomp + MC and if the first case, there would be information of the coefficients used to obtain the final results..

Affiliation
American Association of Variable Star Observers (AAVSO)
Instrumental magnitudes what color correction

Oliver, 

I am ignorant of correct procedures with a DSLR. I know the color bands differ significantly from the standard bands, but  I have no experience using DSLRs for photometry and do not know correct methodology.  I probably should have qualified my post as relating to CCD photometry. Also, I may not have adequately understood the question. 

Raw instrumental magnitudes will always be negative numbers unless you add a constant  value to all measurements to make them all positive. There is no benefit to doing this since it shifts every raw instrumental magnitude by exactly the same amount so the difference between any star and the comp remains the same. The only reason I can see for making the shift is if someone prefers to work with positive numbers. 

The formula I gave, mstar = mrawstar - mrawcomp + MComp is a frequently used formula for calculating untransformed  AAVSO standardized magnitudes in differential photometry (within certain limitations such as FOV < approximately 1 degree and altitude above 30 degrees). So you are taking the difference between two negative numbers and add the sequence magnitude of the comp star. When you transform these magnitudes the transformations are not applied to mstar. Instead you transform in the following way:

Mstar = rawstar - rawcomp + T(CIstar -CIcomp) + Mcomp where the CIs are an appropriate transformed color index for the target star and the same sequence color index for the comp star.

I am not sure that I addressed your question. 

Brad Walter, WBY

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Instrumental magnitudes what color correction

Brad,

I did not make the question very clearly. You spoke of untransformed photometry. I spoke instead of a detail of the transformed photometry, but showing the same problem when reporting the instrumental magnitude in the extended format. my question is applicable to standard CCD with filters if taken the following example. small differen in band V and a little more in R and B in DSLR, this would be like saying that filters using a CCD with a standard low over the horizon, such as the transmission system modifies the light reaching the camera, this is far from the true response of the filter, the stars themselves tanpara comparison ckeckeo and then be corrected magnitude loss as a function of color on these before any equation to normalize the standard zero magnitude. I do like this:

comparison star: Vi = vicomp - e * (B-V) - z

I do not use zero value

comparison star: Vi = vicomp - e * (B-V) - 0

at this time the only star this magnitude corrected for loss of function in color (but still not standard size) and it is currently instrumental, then I establish the zero point:

zero point corrected color: 0c = 0 - ((mag Cat) - (Vi))

(the zero point is always thus with negative value)

and check star and target star only I correct the color, and I replace the zero point equation obtained:

V = vcheck - e * (B-V) - (-0c)

V = vobjeto - e * (B-V) - (-0c)

my question is, before taking the magnitudes transformed to standard magnitudes, I report the instrumental magnitude or the magnitude and color corrected but standardize the extended format?.

Affiliation
American Association of Variable Star Observers (AAVSO)
Instrumental magnitudes

Matthew, Thanks for the response. I certainly agree that the raw instrumental magnitudes for the comp and check provide more helpful information Your references did provide some clarification but I have a few suggestions that would really help eliminate confusion. 

1. Change the example on the Extended File Format page to reflect the values you want. The example appears to use

mrawstar - mrawcomp + MC for the comp and check stars, and for the comp that works out to be MC, the sequence magnitude of the comp star. If you want raw instrumental magnitudes entered, the comp and check star values should be negative numbers in the example.  

2. The term instrumental magnitude is imprecise. It seems people within AAVSO use it to mean raw instrumental magnitude, -2.5*LOG(GAIN*NETCOUNTS/EXPTIME), but in general, it means a magnitude that is not transformed to the standard system.  m_rawstar - m_rawcomp + Mc is an instrumental magnitude offset from the raw instrumental magnitude by the difference between the sequence magnitude  and the raw magnitude of the comp star.  

3. It would be more illustrative if the examples on the Extended FIle Format page were realistic. The values for the comp star in the standardized magnitude example might be possible but would be very unusual. The magnitude values for the comp star in the differential magnitude would never happen, e.g. the 105 comp star has a V magnitude of 10.021, very far from 10.5.However, this point is moot if you really want us to enter the raw instrumental magnitudes. Then the values should be negative numbers.

I'm not complaining. I'm trying to offer constructive suggestions to help AAVSO get better data. 

Brad Walter, WBY

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Maxim DL AAVSO file

I finally got a response from Diffraction limited with regard to what they report as instrumental magnitudes in the AAVSO report. They report the raw intrumrntal magnitudes of the comp and check with a constant offset of +25. They just don't tell you what they are doing anywhere in their manual, at least up through version 5.24.

It is interesting that when I compared the raw instrumental magnitudes I obtained using another program on the same images using the same measurement circle dimensions and annulus, I did not get a constant adder of 25. It was only approximately constant but varied by a few hundreths of a magnitude. Must be the difference in calculation methodologies. 

 

Brad Walter

Affiliation
American Association of Variable Star Observers (AAVSO)
MaximDL extended format

Hi Brad,

As long as CMAG and KMAG have the same numeric offset (25 in the case of Maxim, 19 in the case of IRAF, etc.), it is correct to report those instrumental magnitudes in the extended format.  Constants make no difference in our use of those values.

Magnitude differences between software packages are common.  As long as the target and comparison star have the same difference between packages, the reported magnitude for the target will be the same.  These differences can be caused by how each package determines sky, or handles partial pixels in a circular aperture, etc.  There is no gold standard for this, and so the resultant values will differ at some level.  They will usually be consistent within one software package, and the researcher can determine the offsets when performing his/her analysis.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
PEP extended format

 

The question arises as to what values should be reported in Extended Format for those of us doing two-color PEP.   For single-color reduction, the WEBPEP application reports the "chart" V magnitude for the comparson star, not the instrumental value.

Tom

 

Affiliation
American Association of Variable Star Observers (AAVSO)
PEP extended format

Hi Tom,

If you are writing the Extended Format file yourself, then use the instrumental magnitude for the comparison star, not its chart V magnitude.  WEBPEP should be doing the same; I'll mark that down for the next revision.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
PEP extended

 

Arne:

Thanks.  The reported WEBPEP check star magnitude is stranger, yet.  It calculates a differential instrumental magnitude for the check relative to the comparison, then adds the comparison standard magnitude...

Tom

Affiliation
American Association of Variable Star Observers (AAVSO)
PEP extended

Mira Pro does the same thing for its automatic AAVSO report. Actually, its not too strange. What is being submitted is what AAVSO calls in WebObs the standardized instrumental magnitude (to differentiate from differential magnitude, rather than connoting it is transformed to the standard system). It is the same calculation used for the target star except that for the comp it is a mathematical tautology and you end up with the sequence (standard) magnitude for the comp.  So in single comp photometry every magnitude is offset from its raw instrumental magnitude by Cseq - Craw. The problem is that this offset for the comp and check is not only unknown, that would be OK if the comp and check were offset by a constant value, but the offset varies from image to image. That makes it very difficult to transform or adjust zero points between data sets after the fact based any sort of rational. You simply have to "make them look right' in relation to some good photometry for which you have all of the info.

If you look at the WebObs instructions it says to submit the instrumental magnitudes for the comp and check. it would be clearer if the WebObs instructions specified the "raw instrumental magnitude," but I guess they don't because a raw instrumental magnitude with a fixed constant offset works just as well. It's just an unnecessary step

That is why I have stopped using the Mira Pro automatic AAVSO report and make my own. It takes a few minutes longer, but what the heck. 

Brad Walter

Affiliation
American Association of Variable Star Observers (AAVSO)
PEP extended

 

Brad:

PEP data reduction applies an adjustment for differential extinction between the comparison and the variable, so the check star calculation by WEBPEP, omitting the adjustment, is different from the target calculation.  

 

Affiliation
American Association of Variable Star Observers (AAVSO)
PEP extended

Tom,

That seems really strange. That means if they are not close together you will have a systematic error in your check values that makes you think something is varying when it isn't, and if you use the check for error estimation, it increases your variance. Hmmmmm!

Brad Walter

 

PEP Extended comment

Just for some background and clarification, PEPObs as it exists for the website was based on a much older tool that was used by the PEP committee in the early 1980s, predating the World Wide Web let alone WebObs.  We retained how the code originally output data during our 2004-2005 rewrite of that code (from BASIC into Perl) in order to make newer reports fully consistent with older ones.  The current PEPObs tool is based on my 2004 rewrite of the 1980s code, but converted into Python.  The language is different but the algorithm is essentially 30 years old.

I'm not surprised that the reporting format for PEPObs is different, since it predated the development of both WebObs and the Extended Format specification.  Reconciling those is certainly something that can and should be discussed and implemented.

Affiliation
American Association of Variable Star Observers (AAVSO)
Maxim DL Extended format

Yeah, I know that a constant offset doesn't matter for instrumental magnitudes, but I took a while for the diffraction guys to simply state what they were doing. Since I was not getting a constant difference between raw instrumental mag from another program and the Maxim DL instrumental mag (close but more scatter than I expected) I wanted them to confirm what they were doing. I don't use maxim. The program I use gives the sequence value for the comp and the standardized instrumental magnitude for the check in the AAVSO file output. Therefore the offset is not constant between images which is less useful. However, I can get the raw instrumental magnitude for all stars as well and will start using those for comp and check instead. The AAVSO file is just no longer generated automatically. It's not a big deal since I am not doing photometry on hundreds of data images each night. 

Brad Walter, WBY

Affiliation
American Association of Variable Star Observers (AAVSO)
magnitude instrument

if it is correct, because the loss of function of magnitude in the air mass is a linear relationship. if it was in the case of the lost function of magnitude in the zenith distance would be a curve.

thanks for informing me. that are instrumental magnitudes without color correction. that I needed to know to write the export of data reduction software on the end of write. I hope you guys can prove here.

Affiliation
American Association of Variable Star Observers (AAVSO)
Extinction for Transformations

The most general forms for extinction are 

v0 = v -k1v*X - k2v*(b-v)*X

(b-v)0 = (b-v) - k1bv* X- k2bv*(b-v)*X. Analogous expressions apply to any other color index.  
Please note that (b-v) is the instrumental color index not the standard color index. I think that is what you intend in your equations, but capital letters are often used to indicate standard magnitudes. 

k1 and k2 are the first and second order extinction coefficients. k2 is very small and is often not used except for extremely precise photometry or situations where there are large differences in the color of comps and target. k1 and k2 have different values for each filter and color index. 

X is the airmass. It is not shown in your equation. I assume it is rolled into "e" in your expressions but can't be sure. As long as you are above, say 30altitude airmass = secant(Z) = 1/cos(Z),  where Z is the zenith angle = 1 - altitude.  Below that you should use one of the polynomial expressions for airmass but you shouldn't be trying to do high precision photometry that low anyway. 

The zero subscript indicates an instrumental magnitude above the atmosphere. 

In generating transformation coefficients using star clusters such as M67 or NGC 7790, all the stars are so close to each other that X is essentially the same for all,particularly, if you image the cluster near the meridianso that X is close to 1.0. Therefore, in differential photometry the k1X terms of the comp and the other star your are measuring add to zero. In calculating the transformation coefficients using raw instrumental magnitudes it just moves the complete curve up or down. it does not affect the slope. So you can forget about this term at reasonably high altitudes. 

The second order extinction does have an effect because it is different for different color stars at the same airmass. In your expression I am assuming e = k2X, therefore it iwould be different for every airmass. Normally, you separate the extinction coefficient from the airmass because the second order extinction coefficients are fairly stable and small compared to first order and you can go long periods (say a year) between calculations or, if you know you have significant seasonal changes in air quality you may just determine second order coefficients for different seasons. You do not have to determine them every night. 

So if you are correcting for extinction when calculating your transformation coefficients form measurements in a cluster you get

v0 = v - k2v*(b-v)*X, and 

(b-v)0 = (b-v)- k2bv*(b-v)*X, and similar ones for the other color indexes.

Including the airmass as a separate factorallows you to use constant values of k2 at different airmasses. I have done this and have found that as long as I am near airmass 1.0 it has a very small affect on the transformation coefficients. If you are willing to go to the extra effort, then, as they say,"it couldn't hurt!"

You can determine k2 values easily by imaging the same standard cluster you are using to determine your transformation coefficients  over a wide range of airmasses, say, X2 - X1 >= 1.0 and plotting v vs. (b-v)*X and (b-v) vs. (b-v)*X (or the equivalent for any other color index) for the standard stars. The slopes are your k2 values. I would image at least three different airmasses each night you image the cluster for your transformation determination to check whether k2 values are being affected by changes in sky conditions in addition to airmass. This is probably  what you are doing, but I can't tell if "X" is rolled into the "e" in your equations. I assume it is, but sometimes assumptions get in the way of communication. 

I hope that I finally understood. Please excuse me if I simply re-stated things that are obvious to you. It is hard to know how much to include when responding in a forum. 

Brad Walter, WBY

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Airmass

Do not worry bred.

I see you understood the concept of what I did. I  replaced K´ by a variable slope in function of the airmass, this works perfect. there is one detail that is perhaps what makes it seem a lot of work, and I do not have to calculate, the software automatically calculates this for the central time of each image. and simply apply that slope and zero point. and ready. that in relation to air mass. as to what else I prefer maybe in a week or two and load the manual sortware a link with the only drawback that was Spanish. I can only write in my language, but for very large levels of detail, I am unable to make an accurate translation, but surely there will be someone who haga. translators are not as good in the syntactic analysis of the writing.

v0 = v -k1v*X - k2v*(b-v)*X

in mathematical terms, I replace two 1st and 2nd order terms by a single term. that would come to be: slope vs. air mass, let's call (A * X) and just apply it like this:

V = v - (A*X) * (B-V) - Z

I get b-v b-v vs B-V, which would come to be very rough on the first operation, but would not be accurate, then the B and V magnitudes obtained from these first operation, I get B-V again, and repeat the process until the final index does not change more. at this moment b-v is B-V and therefore the transformation is complete. my camera webcam celestron Neximage RGB in the blue channel requires 80 to 90 iterations, if I take into account in calculating 8 decimal, or 8-0 if you take into account 3 decimals.

in my case, I use the zero point in each operation, because it changes on each band, and praise red stars have different zero points in b , v, r, etc. so to get B-V is necessary to use the zeros to calculate the magnitude in each itercion standard.

 

Affiliation
American Association of Variable Star Observers (AAVSO)
different treatments

hello brad and tom

What he says Tom is completely accurate. Anyone who would do this, there appears to adjust the chek star and the target object with a setting that they apply different treatments. I no understand correctly if PEP (besides being a photomultiplier) is a software outside routune to report the data to AAVSO. but I have done on mine for zero point using a linear regression setting that determines the point zero function o comparison stars in function of the her standardized instrumental magnitude, and also as a function of her B-V, you can choose which, because as both color magnitude in each band is different for each, the zero point also it will be, I have found that the linear regression is more powerful than simply agarrr average values ​​compared because dontro average this this sistematoco error can correct, or the color or magnitude in each band (each measure).

this was precisely my question leveraging this issue, because if you wanted to rebuild the methodology used in our report, the ratios used to transform check star and the object under study would be different, would not be a mistake, but by using these methodologies .

though I have done is to eliminate the error due to loss of antiblooming on webcam or DSLR (which acts in very low level, even into the range of linearity) affecting you measuring star with brightness differences, but that is another subject.

 

Affiliation
American Association of Variable Star Observers (AAVSO)
What Scope is that?

Tom, 

In your picture, what scope are you standing next to? The ring around the perimeter makes me think I might be seeing  the back end of a Boller and Chivens. 

Brad Walter, WBY

Affiliation
American Association of Variable Star Observers (AAVSO)
B&C

This is off topic for this forum, but In May of 2012 I started a networking campaign to find a new home for the 1965 vintage 36" B&C housed at Princeton's Fitz Randolph Observatory. The observatory was officially closed in 2010 and was slated to be demolished to make way for new athletic facilities. I connected with Wayne Green and others. Wayne in turn connected with John Briggs of the Mittelman Foundation. They are heavily involved in scientific education and outreach and operate at least two observatories. They took the scope disassembled it, moved it to a storage facility and are now in the process of updating the controls and selecting a location for a new observatory. I have fond memories of stargazing at the Fitz Randolph observatory when I was an undergraduate and later at reunions. It was already lightly used when I was an undergrad in the 1970s. It had something of a revival during the OCETI program and then was used less and less Even the local astronomy club was no longer using it. 

It is now on its way to having a new home and up to date controls. I couldn't have hoped for a better outcome. The glass in this telescope is particularly good, even for a B&C. 

Affiliation
American Association of Variable Star Observers (AAVSO)
BAA / AAVSO observation formats

Greetings Dick,

I'm not sure how different these formats are.

And if the AEFF is going to be changed then we need to study this.

George
 

Affiliation
American Association of Variable Star Observers (AAVSO)
Must they be?

Dick:

An obvious question (to me anyway) is why they need to be the same? The BAA and AAVSO are two separate organizations.

BAA is kind enough to send their data to the AAVSO on a regular basis. We would just end up with duplicates that need to get removed later.

Ken