Affiliation
American Association of Variable Star Observers (AAVSO)
Sat, 01/28/2012 - 13:41
When I have an image in V and one in R, how do I get the Transform function to work in VPHOT when the scope already has all the coefficients entered?
First of all you need to register the transformation coefficients for your scope. This has already been done for many of the 'public' telescopes, but if you use your own scope, go to 'Admin' and then 'Telescope Setup'. Then select you telescope in the list.
There is a transformation coefficients section on the telescope page. Click on the 'Help' link for details on how to add the various coefficients. Note that you can also click on the 'public' scopes to view their detailed setup even if the link is grayed out. It can be helpful to look at some examples.
Now you are ready to apply the coefficients. First thing to do is to create a sequence for the field of interest. This is done "as usual", open the image, import stars from AAVSO or build it up manually, and save the sequence. Make sure the sequence have photometry for both of your filter bands. Then go to the image list, select two images of the same target, taken with different filters, and click the 'Transform' link above the list. You will be taken to the color photometry page. Select the sequence you have created in the drop down list, and click 'Show Report'.
The report page will show you the final estimates, as well as loads of detailed info - once again - have a look at the help page for detailed information. Click the 'AAVSO Report' link to report the transformed measurements.
Geir
kge wrote
"select two images of the same target, taken with different filters, and click the 'Transform' link above the list. You will be taken to the color photometry page. Select the sequence you have created in the drop down list, and click 'Show Report'."
Is there an optimum strategy for transforming lots of images, maybe comparable to the time series function? I haven't taken any lengthly time series lately, but I might. As a minimum, I take a set of three images for each target. It can add up over a night's worth of data and I would spend more time transforming them two by two than it takes to acquire them in the first place.
Related, I often use three filters: V, B and I. If I transform the VB pairs and the VI pairs using the same V image I get two measures of V for each measure of B and I in the AAVSO report. Must I manually edit out the extra V report or is there some way we could do three color transforms at one pass? (Sarty appears to have an algorithm for that.)
Hi Jim:
I see three issues here:
1. Why do you take three images? I assume it is to improve understanding of precision. If so, why not stack the three images and report one magnitude with improved SNR and precision.
2. I would not delete the second V magnitude from the report. It provides extra information about accuracy since you used two images and two filters to report two discrete standard magnitudes. This tells something about the accuracy.
3. Yes, one cannot currently run a time series of transformed magnitudes (quite sure about this?), even by running single filter transforms that you can do with single images, if you have a known/assumed color for the target. It may be a useful for Geir to put that idea in the wish list. Most people use time series for temporal analysis and do not worry about standard magnitude so much and/or they make a manual shift to adjust to a common magnitude from several observers. In the best of all worlds, I would think calculating transformed magnitudes of a time series within VPhot would be a good idea, even if it is just the single filter transform method.
I wonder what Arne and Geir think?
Ken
MZK wrote:
"Why do you take three images?"
It is Arne's recommendation, ergo, I do it (whenever it's convenient :)).
"one cannot currently run a time series of transformed magnitudes (quite sure about this?), even by running single filter transforms that you can do with single images"
Tom Krajci mentioned this the other day. How does it work and how would one implement it in VPHOT?
Yes Jim, my next sentence stated why you would want to take multiple images. It was meant to be retorical. Multiple images is a good thing to get at precision but you can stack them rather than reporting 3 values. ;-)
When you set up your sequence, open the target details and change the default color value (0.000) to its known or assumed color index. If you take images in two colors (e.g., B;V), you can calculate the instrumental color index (b-v). With your color index transform slopes (around 1.0) you can then calculate the standard color index (B-V). Not much difference!
Enter this known color index for the target and save sequence.
Then select the V (or other as needed) filter image for the target, use sequence and check the photometry. Look at the magnitude box and notice that it will be transformed.
MZK wrote:
"one cannot currently run a time series of transformed magnitudes "
I think I have figured out a way to do it with info supplied by VPHOT (and the comp star photometry). It would require a little set up at home (ie, create a text file with your target stars and the Bc, Ic and Vc (or whatever you are using) values of your comp star (would only work with a single comp star, not an ensemble (at least, not yet)). One would feed the AAVSO report generated by VPHOT to a VB script that would extract the instrumental values, run them through Sarty's algorithm using the stored comp star values and rewrite the AAVSO file with the transformed values.
The question is, before I run off and re-invent the wheel, has anyone else done this? Would anyone else be interested in using it?
Hi Ken,
You are right, you can not transform time series results in VPHOT.
It is already on the wish list, has been there for quite a while. So one of these days ....
Geir