Struggling with Transforms in VPhot

Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 04/20/2016 - 00:58

I'm brand new to transformations, but couldn't put this off after getting hold of Gordon Myers' TG program. Using VPhot to generate the reports needed by TG makes things so much easier, but I'm struggling with applying a transform in VPhot. Here is a VPhot screenshot of my coeffs. as entered in VPhot, and a screenshot of the VPhot error (tested on a star with known magnitude in M67). The TG transform ini file is also attached. I'm obviously missing something and would appreciate some guidance.

Thanks

Michael

 

 

Affiliation
American Association of Variable Star Observers (AAVSO)
2 Filter Transformation

Michael:

To run a transformation in VPhot (or other software) you select an image taken in two separate filters, e.g., B and V, NOT a single filter image, e.g. V.  IOW, select a filter pair in your image list and click transform.

HTH, Ken

PS: There is a way to run single filter transformation BUT it requires that you know what the expected color of the target is. I think that is what you were trying to do?

Affiliation
American Association of Variable Star Observers (AAVSO)
2 Filter Transformation

Ah, thanks Ken (and Phil). I was using two images in each B and V, and got that to work (see attached). But, yes, I was looking for single image transformation and forgot about specifying the expected colour of the target.

Affiliation
American Association of Variable Star Observers (AAVSO)
Applying transforms in VPhot

Michael,

Did you select images in two filters (for example B and V, or V and I) before you clicked on "transform"?  VPhot needs images in two filters (for which you have transforms) to apply the transforms to your measurements.

Phil

(Oops!  It looks like Ken beat me to the reply.)

Affiliation
American Association of Variable Star Observers (AAVSO)
Why not use TA?

Using TA for the transformation process has a couple of advantages over VPhot:

- TA can do 4, 3, 2 and 1 filter transforms (the later if you provide the reference color)

- TA propagates the error of the transform into the result

- TA includes in the resulting AAVSO Extended Format record full documentation of the transform process in the comment/note section

- VPhot does not use the AAVSO recommended transform. For BV it would be Tb_bv and Tv_bv. The classic way is to use Tb_bv and Tbv. I believe VPhot is using Tbv and Tv_bv. There is not a big difference here, but its not the recommended method.

- If you've developed your coefficients with TG then you have in hand the INI file needed by TA.

- And its easier with TA. If you prepare a WebObs submission with a 100 stars it is just a few clicks and file saves to transform the whole set. With VPhot you will be faced with multiple mouse clicks for every pair of stars.

Give it a try and compare!  https://www.aavso.org/transformapplier

George

Affiliation
American Association of Variable Star Observers (AAVSO)
Why not use TA?

Oh cool!  I had already downloaded and installed TA (kinda got a feel for what it does), but then I got focused on doing everything, including transforms, through VPhot without knowing the differences and pitfalls. Thanks so much for pointing out the pros and cons.

I'm puzzled as to why VPhot does not use the recommended transform. Is it a work in progress? I didn't see any cautionary note in the VPhot UI or help doc.

I do have the ini file form TG. I'm wondering if it's correct. I posted my set-up and results here:

https://www.aavso.org/use-transformation-coefficients

I'm also wondering about supplying a reference colour. A variable can change its colour, and the AAVSO CCD Photometry Guide (p. 60) says:

"If for example, you were only able to create an image in one color, it is possible to substitute the published B-V values for the variable and comparison stars (if they exist) for the measured ones in the above equation. Please note that this method is prone to error since in many cases, the color of a variable star may change."

So, to get around this, one observes in two bandpasses and not worry about providing B and V (or any other two bandpass values), is that right?

Lots to learn, but I have great support!

Thanks.

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Using TA

The INI file you showed in your first post is exactly what TA will need. Just rename it back to .INI and store along with your observations.

And you are correct, we combine multi-filter observations in the transform process just so we don't have to find reference and target colors. If you do single filter photometry, the TA Help tab will describe how to introduce color information. If you have a case where you have 4 B observations and 3 V obs TA will do 3 BV transform pairs and then a single filter transform using the the last B-V color that it saw from the pair transforms.

George