Having multi-filter CCD observations transformed into the standard system is important: it makes our data cross-observer comparable and useful to the professional astronomers.
In 2014 we developed the tools to document and make this easier to do. Check out the documentation here.
And now M67, an excellent calibration target, is in the perfect position for evening observations.
What are you waiting for? Now is the time to move to the next level!
And there is help available: volunteers are listed below to help you learn each step of the process.
Goal of the campaign:
- Increase the number of observers who are submitting transformed data to WebObs.
24 out of 134 observers submitted transformed observations in Jan-Feb 2015
Let's see if we can improve this metric!
- Use the tools developed this last year for transformation
TG: Transform Generator
TA: Transform Applier version 2.30 or better
Process:
--- Get those M67 images! Its a convenient evening target in the month of March. Use best practice to get them BDF calibrated.
--- Extract the instrumental mags.
VPHOT is the most convenient way to do this. It automatically identifies the standard stars in the field and TG is prepared to work with its output directly.
Ken Menzies is available to help with questions:
If you use some other tool for extracting the mags then the issue will be using the proper labels for the stars. AUID's are always preferred and are available
with the photometry from VSP
--- Use TG to compute your transform coefficients
Installation and process details are available here.
Gordon Myers is available to help with questions.
You can get your coefficients from TG in a format compatible with TA (INI file) when you save your results.
--- Use TA to apply your transform your data
- You prepare you WebObs submission as you already do. The only adjustments to that process:
- Comp and Check stars should be identified by AUID. You may be able to use the VSP labels, but AUID's are preferred.
- The ChartID in your observation records needs to be a photometry page, not the picture chart.
Installation and process details are avaiable at here.
- Get the latest version of TA here.
George Silvis is available to help with questions.
Before you start submitting TA transformed data, you should review the results. TA has a feature built in to help you. If you check "Test TC" the transform process will be applied to your Check Star data. Review the results in the Report tab. If your observations can reliably match the transformed standard magnitude of the Check star, then you can submit your variable star measurements with confidence.
--- Tell us how you're doing by posting comments to this thread. What needs to be changed to this process to make it easier? We'll fix it!
Cheers,
George
SGEO
Here are my coeffients for M11 using the whole standard of comps (384) For BVI.
14" f/10 SCT and St-1001e Camera. Quite a load for VPHOT to process!
As stated, I'm pretty new to photometry (but not to imaging). I'm getting my feet wet by starting with generating my tranformations from images of NGC7790 from last summer, but the first number I'm getting for the color transform seems a bit high (don't really know if its reasonable): T_bv = 1.308. I've tried to follow the CCD manual as best as I can, and watched the CCD school video. Details:
Any feedback is appreciated, thanks. I have more nights to analyze but don't expect a big difference since the plot looks pretty tight.
Keith
George et al.,
I've been doing CCD photometry for about a year and a half now. My measurements have been either on pulsating HADS stars or eclipsing binary systems. I have submitted some of my measurements via WEBOBS, but have not been transforming my data. I have B, V, Rc, and Ic filters, and have taken measurements of the standard cluster, M67, with all four filters on four separate nights. On each night, I shot 10 images with each filter. These were uploaded to VPhot, and the photometry results were downloaded and processed in Transform Generator. I have the averaged coefficients, but am not sure what to do with them.
While attending a webinar I learned that there is a Transform Applier function in VPhot. I typically process my HADS, EB, or EW data offline using MPO Canopus, and do further analysis on the light curves with Period04, Peranso, and perhaps Binary Maker 3. If my results are worth sharing, I usually produce an AAVSO format file in the PhotoRed page of MPO Canopus, and upload this through WEBOBS.
I assume that using Transform Applier on the light curve data would require contemporanious measurements with multiple filters along with the transform coefficients. For example, I could not use B-filter data from one night along with V-band values from another. If I make measurements in repeated sequences of B, V, Rc, and Ic, will these be close enough in time? If so, I could transform my data before submitting it. Isn't this, however, a bit late in the process? Wouldn't it be better to transform the data before I do the period phasing, Fourier analysis, etc.? I suppose I just need help on the How To rather than on the Why To.
Allen
Allen:
Yes, the normal procedure is to image your target in a time series using repeated filter image sets (e.g., BVRIBVRI....). As a target pulsates or eclipses, its temperature/color will/may change over its period. Comparing different filter images from different phases on different days would certainly/likely not represent the same color/temperature! Transformation will correct your system's (filter/camera) magnitude for such color changes during the run. The image filter sets need to represent the same phase, so it makes sense to take them in a repeated series.
In VPhot you can run your time series with all filters in one step and generate the AAVSO Report. In the past, it was necessary to run each filter time series individually. This time series, multifilter report can then be imported in TA and run in one step. The transformed time series report is generated by TA and then you submit it to AAVSO. Note that TA requires one comp star, not an ensemble, to work properly. This is just a current reporting restriction of TA. LesvePhotometry can run a transformed, ensemble time series in one step.
I do not understand what you mean by "a bit late in the process'? The process is to collect your time series images, calculate your standard magnitudes by differential photometry, transform your magnitudes to correct for small differences between your system and the standard system, submit your data, and subsequently conduct any light curve analyses to calculate periods. In VPhot and other software, the differential photometry and transformation (i.e., standard magnitude color correction) may be done in one step or two steps. Light curve analysis can subsequently be conducted in period analysis software like VStar or Peranso or other.
Questions?
Ken
I submit this report for transformation
#TYPE=EXTENDED
#OBSCODE=VJEB
#SOFTWARE=VPhot 4.0.35
#DELIM=,
#DATE=JD
#OBSTYPE=CCD
#NAME,DATE,MAG,MERR,FILT,TRANS,MTYPE,CNAME,CMAG,KNAME,KMAG,AMASS,GROUP,CHART,NOTES
FO Aqr,2460209.62666,13.572,0.014,V,NO,STD,ENSEMBLE,na,122,12.153,1.0805,0,X29253AO,|CMAGINS=-6.383|CREFMAG=13.265|ENSTYPE=1|KMAGINS=-7.495|KMAGSTD=12.153|KREFERR=0.046|KREFMAG=12.184|VMAGINS=-6.075
with this set of transformation coefficients
[Setup]
description= TG - Version 6.9a, Telescope= CDK 400, Time created (UT) = 2023_09_22_10:07:37
[Coefficients]
Tbv= -1.910
Tb_bv= 1.432
Tbr= -1.409
Tb_br= 0.931
Tv_bv= -0.045
Tvr= -0.944
Tv_vr= -0.170
Tr_vr= -2.161
[Error]
Tbv= 0.162
Tb_bv= 0.054
Tbr= 0.045
Tb_br= 0.024
Tv_bv= 0.103
Tvr= 0.076
Tv_vr= 0.022
Tr_vr= 0.163
[R Squared Values]
Tbv= 0.983
Tb_bv= 0.997
Tbr= 0.999
Tb_br= 0.999
Tv_bv= 0.853
Tvr= 0.988
Tv_vr= 0.967
Tr_vr= 0.988
And I get this error message
#TYPE=EXTENDED
#OBSCODE=VJEB
#SOFTWARE=VPhot 4.0.35
#DELIM=,
#DATE=JD
#OBSTYPE=CCD
#NAME,DATE,MAG,MERR,FILT,TRANS,MTYPE,CNAME,CMAG,KNAME,KMAG,AMASS,GROUP,CHART,NOTES
FO Aqr,2460209.62666,13.572,0.014,V,NO,STD,ENSEMBLE,na,122,12.153,1.0805,0,X29253AO,|CMAGINS=-6.383|CREFMAG=13.265|ENSTYPE=1|KMAGINS=-7.495|KMAGSTD=12.153|KREFERR=0.046|KREFMAG=12.184|VMAGINS=-6.075
# Missing information; need Tv_bv and BVcolor.
Can anyone help me?
It looks like you have your transformation coefficients in the ini file, but have they
been loaded into your Telescope Profile? That's a necessary step for TA/VPhot to access them.
Because this is a single filter observation you will need to provide the estimated B-V color.
You're close!
George
Jean:
The normal transformation process requires images taken in 2 or more filters. Looks like you only used a single V filter? Do you only have one filter in your system?
With 2 or more filter images, the color of the target can be measured by TA. With only one filter image, it cannot be measured and you need to provide the color of the target in your aavso report. This single filter transformation is a more advanced procedure. If you look at the TA instructions, you should find what to do?
If you only have one filter, it is normal to submit untransformed data.
Ken
For most of my uploads I am managing to transform.
However VPhot seems to deny some of my obs the process.
It rejects some obs for targets that it as accepted before.
I have gone through the TA and TA reports and can see no difference between them bar the variable data results.
One set it will. One it won't.
Have wasted so much time trying to sort this that I have reluctantly reverted to Single Image Photometry with no transformation.
Hi all,
Just a basic question ... I'm doing DSLR photometry for many years, reporting in TG TB and TR.
Is it needed/useful that I begin such an apparently delicate process so as to transform my data ?
I'm ready to spend an evening on M67 ... but not sure of the result.
By the way, another basic question : should I do this for any different optic I use for phometry ?
(2 different refractor, 1 reflector, and some lenses...)
Thanks,
Christophe
It is up to you if you want to transform your tricolor images to photometric BVR. If you want your data to be scientifically useful you should transform. Lately, I have only been submitting TG when doing DSLR photometry. I feel that it is so close to V that the difference of transformed TG to V is minimal. Even transformed, I have not been happy with TB and TR.
If you are going to obtain transformation coefficients then they should be obtained for every setup. The coefficients are specific for that telescope(lens)/camera combo.
Barbara
Barbara,
I'm curious that you find your TG values are so close to V that you do not need to transform your DSLR data.
What is your Tv_bv? I presume it must be quite small. It's a few years since I last did DSLR photometry, but then my Tv_bv was quite substantial at -0.125. That means a target-comp B-V difference of 1.0 mag unit will theoretically result in a TG mag error of 0.125 (ignoring the sign of the B-V difference and error). As the relationship is linear, the error would be about 0.06 mag for a smaller target-comp B-V difference of 0.5, and so on.
Of course, the above generalization applies only to targets with radiation properties that approximate those of a black body.
Roy
Roy, indeed, I find that TG is very very near V, when comparing with other observers data.
Thanks Barbara. Good anwser.
I will try to enter in this somewhat complex process (from my point of view).
Fortunately, you don't always need to transform your DSLR data to be scientifically usefull, as in many alert here and there, researchers are directly accepting TG/TB/TR data. They have to calibrate the data by themself, I guess, or are using only timing (EB).
Christophe
Hello there,
Well well well.
I don't know what is the exact proportion of transformed observations in the full AAVSO database. I guess this is not a very large proportion. And that could be better, for sure. The reason of this Transformation Campaign, most probably.
Possibly, I know why so few transformed data :
transformation is fundamental, but transformation is by far too complicated to being applied by “ordinary” and humble observers.
The facts : I spend about one week on this only subject, since my precedent questionning.
I have read dozen of webpages and posts on this subject in the AAVSO forum and elsewhere, also read a few papers found on Arxiv or ADS … A friend helped me, and tried to explain this stuff by private mails, with examples and screenshots. And ?
And nothing.
I've installed TG/TA on my laptop. But AIP4WIN don't works properly. There are always some bugs at some level in this venerable but old soft (win11 64 bit), and the process stops unexpectedly. But I don't have access to Vphot, cause I'm not an AAVSO member (it's too expensive for me). Because there is no clear from A to Z available tutorial.
So what ? It's a pity.
My two cents...
I have spend nights on targets and send a few thousand lines of data in TG TR TB, and was being participating at some campaign for years. I suppose I have been of some utility here over this time …
We are now in 2024. Just one click and a previously unlocated image have been plate-solved. Three more clicks and the photometry have been extracted in 3 colors for a time serie of 500 individual images. What else ?
I'm fully convinced that transformation could be made in a fully automated manner. I mean you get those famous images of the standard fields, you say what filter was used, and the installed software on your laptop computes all that is needed to get the result in just one second, and it applies it to the observations.
No more maths. No more exhausting consultations of the UBVRI mags of dozen of stars. No more excel tables.
This is exactly what is done in ASTAP : one click for plate-solving, one more click to get V transformation, and the software ask SIMBAD, get about 600 stars Gaia photometry in a few secs from Vizier, and return the transformation coefficients for V in 30 seconds or so, from the beginning to the end of the process. Unfortunately, this option is only available for V. And only from Gaia magnitudes, not from standard mags.
If such a decision was taken at high level somewhere in AAVSO, the right and easy software will be soon available for all observers and contributors, and not only for the AAVSO members.
And in less than a few month, all imported data could be transformed, once for all, for a better science.
For now, I will stop spending time on this subject. Far too complicated to me, and totally disconnected from reality. Sorry sorry sorry.
Transformation in it's current form is far from easy. It's far from being simple and accessible.
It's difficult an unpleasant.
It's discouraging and exhausting.
No more, no less.
If AAVSO wants transformed data, simplification of the full process is needed.
I will continue sending untransformed observations in TG TR TB, and I hope so, also in BVR in the next future. Only untransformed BVR data, alas.
I finally will have enough money to buy a serious camera and some Baader filters, after years of saving money.
No more transformation process for me, unless it is severely simplified/fully automated, and freely open to active observers. One click, one result.
I Hope everybody here will understand my position - a position maybe shared by a large and silent number of observers ?
Just my two cents.
Thanks for reading.
Sorry,
Scientifically yours,
Cheers.
Christophe
(sorry, only bad english option available ;-) )
Christophe,
I sympathize with your frustration, because it does require study to understand TCs and how to determine them. In my view it is of great value to be able to do so.
If your photometry software allows you to select stars for measurement in an image and obtain and tabulate the instrumental magnitudes for those stars, and if you can use a spreadsheet, in my view the process is not difficult. OK, it takes a bit of time, but when it is done you can see and understand exactly what the process is.
The AAVSO uses standard fields (mostly clusters, I think), but Tycho II stars with low error estimates are fine, or some other source of standard stars.
Here is a link to an example spreadsheet, which includes simple instructions. The standard stars are E region standards from the southern hemisphere (my personal preference). You need to enter your own standards.
Edit: The spreadsheet did not display properly in the link I posted a few minutes ago. If anyone wants to see the spreadsheet, message me privately offline (click my observer initials and send an email), and I will send you the file.
Roy
.
Roy,
Thanx for taking some times to answer to my post. Indeed, I think this is a fundamental step toward good data samples. Otherwise, I will not spend time here ;-)
Yet, this is a very consuming, discouraging and painful task, specially if you considere that you have to do this for every color/filter and optical system you own and use to do photometry.
That means TG TR TB, plus B V R I, on two different cams, with 3 or 4 optics, if not more (lenses, refractors, telescope ...).
Last night, I couldn't sleep, and I finally arrived at the same conclusion than yours : what you said is exactly what I was thinking. I use AstroimageJ. I will get the ADU counts for each standard stars, use a spreadsheet to compute the instrumental magnitudes, and derive all the other stuff from this.
Nevertheless, I will have to pick up standard stars one by one, verify each coordinate in order to be sure, select and report star names and standard magnitudes manually, at the risk of introducing some error at some point in the process.
Pfffew !
All of this could - and should - be fully and definitely automated, for the benefit of all here, and for the science, of course.
Just one question, for which I did'nt found a clear answer :
When computing the transformation in V, should I use only yellow middle main sequence stars, or should I choose/pick up alltogether blue, yellow and red stars, as well, of all magnitudes with good enough SNR ?
Same question about B : should I stick only to blue stars, or could I select any star in the field, no matter it's color ?
Same for R an I.
Thanx in advance,
Christophe
(ps : I will contact you privately about the spreadsheet, so as to compare with mine ;-) )
Christophe,
For TCs, use as wide a range of colours as possible for every filter you use. Ideally, the stars should be distributed regularly across the range of colours (not just concentrated at the blue and red ends).
Roy
To determine transform coefficients to relate your measurements to the "standard" system you need to use reference stars that are known to have "standard" measurements. You should not just pick a bunch of Tycho II stars with low error measurements.
AAVSO provides charts and photometry for the standard fields: https://app.aavso.org/vsd/stdfields
There are workflows that can generate the transform coefficients without resorting to spread sheets. VPhot and Transform Generator is one. And I understand the the Tycho Tracker software also offers this feature. https://www.tycho-tracker.com/ .
George
I agree that it is ideal to use photometric standards for TCs but I dispute that valid TCs can be determined only from photometric standards. Years ago when I was still doing DSLR photometry and determined TCs many times mostly using E region standards, I occasionally experimented with other targets.
The attached screenshot of a spreadsheet (it seems the display corrupts if I try a link to the spreadsheet itself) shows plots for TCs using the B and V values from the planetarium programme Guide. My understanding (after contacting the creator of Guide) is that the values are derived from Tycho II photometry. Nowadays, I use more stars for each plot, but in my opinion the plots in the attached are OK for calculating TCs. If anyone disagrees, I would be interested to know why you think the data is no good.
One of the reasons I suggested the use of a spreadsheet for tabulating and plotting the data for TCs is that it is a very good learning tool for the understanding of TCs.
Here is the link to the screenshot: DSLR TCs ARX 2016.png
Roy
This is a useful tool for getting a first look at transformations. The Tycho-2 stars, corrected to the standard system, might be OK in a pinch, but you really do want to do this every genuinely clear night with bona-fide standards. The Tycho-2 errors and your own observational errors need to be added in quadrature for this to be legit. I will also say that in addition to the spreadsheet plot of the fit, you also want to look at the residuals from the fit star-by-star (the R^2 value is not really relevant) to look for systematics --- and of course with something like 20 stars, as you suggest.
The paper from 2006 by Taylor & Joner in regard to transformations gives a lot of useful examples in this regard:
https://ui.adsabs.harvard.edu/abs/2006PASP..118.1716T/abstract
See for example Figures 1 and 2, and related text sections 8.3 and 8.4 about why the two sorts of plots are not redundant. This also shows that proper transformations need not degrade data quality. The examples and exercises in the paper apply of course to many sorts of transformations.
\Brian
Thanks Brian,
The initial reason for the post was to try to help Christophe. I hope I have not made it more difficult for him. If he did want to try transforming his data and if he was, for example, having any difficulty finding standards, I thought the Tycho stars option was worth mentioning to get him started.
I should also just have accepted that George's comment was pertinent and not responded. However, it did stimulate your (as always) informative contribution.
Re R squared - the peviously attached screenshot was from 8 years ago, and I have long since stopped displaying that value because of its lack of relevance.
Roy
Hello there !
Thanx to all for answering to my previous questions.
Unfortunately, I have a few more ...
So, I've done a spreadsheet that should be good enough to find my color transforms and other related stuff. (I hope so ...)
I also had some holes in the cloud cover last week, so I could get some decent images of Melotte 111 with various exposure times.
I suppose SNR, here, is a very important point. What is acceptable about this ? Is an SNR 3, 5, 10, 20 or more, needed on measured standard stars to achieve a good determination of transform coeff ? What is the lowest limit ? The less the SNR, the more standard stars I could measure, on the other hand.
Can I add data from different imaging nights, and from different exposure times ? The brightest stars were saturated at 120 secs exp., so I had to go down to 60 secs.
Thanks in advance,
Christophe
There is perhaps some confusion here. A signal-to-noise of 3 represents an image only just reliably detectable over the background. For good precision, one would want S/N of at least 50 (giving ~0.02 mag errors at minimum), and you might want to shoot for S/N >200 (i.e. 0.005 mag nominal photon-statistical errors) for standard star calibration. And get multiple images (say five) in each filter at those levels, not just single ones, in order to average the results.
\Brian
Thanks Brian.
Yes, I know that SNR 3 is very low and is used to characterize SNR of very faint nebula barely visible on deep sky images. And SNR 20 is sometimes asked as the lowest limit in some alert here.
The fact is in Melotte 111, most of the standard stars are in the range 100-110, or below. That's a rather wide range comparing to the brightest stars which are in the range 60/80 in the same field. So, I have to choose between having a good SNR on the lowest standard stars, with saturated bright stars or, conversely, bright stars correctly exposed with fainters stars with poor SNR.
So, the reason of my question : what is an acceptable SNR in this particular case ?
If SNR must be at least over 50, I will have to image the cluster again, with a longer exposure time, because most of the standard stars are actually in the range SNR 30-50 on my images. The faintest of them are in the range 8-20. And yes, they are different in blue/green/red.
I currently get 10 images in a serie.
I forget to say : I'm imaging with a DSLR, so the QE is not so good. My exposure time will be around 5 minutes, on a 50 mm refractor. This is a very small optic, but very useful to observe some very bright target like b Per, ou 31 Cyg.
Clear sky,
Christophe
Christophe,
I'll give you my approach to this, essentially the same as Brian's, just explained a bit differently. First let me say if you are going to use the entire FOV of your system be sure you have a good flat.
This cluster has a lot of good white and bluish stars, but very few red stars. Ideally you want a wide range of colors (B-V's) when calculating transforms. There might be some other standard clusters that would give you a wider range. For now let's stick with Melotte 111.
First I'd identify reddest stars that will appear in your images. There are two standard stars in that cluster with B-V's of about 1.3. (If you load the VSP chart, then click on "Load Photometry Table for This Chart", you will see all the information you need for each standard star.) Make sure the brightest of the red stars (comp 78) will appear in your image. You may need to adjust your pointing to be sure.
Take some test images to find the longest safe exposure you can use. This is one that gives a good SNR (better than 100) in star 78 but is still safely below your linearity limit.
If your tested linearity limit is 60K ADU's I'd use an exposure that gives maximum pixel values in 78 of about 40K or 45K ADU's. This should give you a very good SNR. Five minute exposures would be a good place to start.
Next identify the bluest stars in the image. Do test images (using that red star exposure time) to see how many of these exposures you would need to stack to give a SNR in several the bluest stars of around 50. (SNR increases approximately with the square root of the number of exposures you stack). You may need to do some experimenting depending on how faint these blues stars are, and you may not be able to reach a SNR's of 50. Just do the best you can. With CCD sensors there is eventually a diminishing return on how many images you can stack to advantage. This would probably not be a problem with a CMOS camera.
Reminder: You'll need a good flat, and you should know (by you own testing) your linearity limit.
Phil
Phil's and Roy's responses are much better than mine, based on actual use of these devices(!). Given the increasing use of these small telescopes + cameras, I am wondering if there is reason to identify some groups of fairly bright stars in wide fields suitable as standards. I don't have a good feeling for the field sizes that would be good, or the magnitude ranges. Would 1-deg or 2-deg fields be too large, and maybe mag 5 to 8, as Roy suggests? I am thinking of star-groups or asterisms such as Upgren 1 (up in the evening now, but not a wide color range), Brocchi's cluster, or Collinder 21, an asterism in the autumn/winter sky. There are more such in the northern sky. I think the southern hemisphere is served pretty well by the traditional E-region standards, which are typically mag 6-9 and cover 1 to 2-deg fields. In general V magnitudes and B-V colors should already be available for most such stars, so assembling the star-lists ought to be straightforward.
\Brian
Brian
First let me say I have not used DSLR's for photometry myself, just old fashioned CCD cameras with SCT's and BSM systems, so FOV's of 30' to 90'. I think a collection of pre-selected comps would be welcomed by DSLR's observers of bright stars.
Most AAVSO observers are not used to dealing with extinction since most comps and targets, until recently, have shared the same image with a relatively small FOV, less than ~2 degrees. Might this become an issue for bright comps several degrees away?
Phil
Since you are…
Christophe,
Since you are using a DSLR camera, I recommend that that you select the brighter stars for your transforms. In Melotte 111 there are 18 of them from 6th to 8th mag in V, for example, with a good range of colours. You may not be able to use all of them because adjacent stars may be too close, and hence inappropriately placed in your aperture or annulus.
Roy
I'm trying to learn how to transform my time-series measurements. I have not been able to get TA to work for me, and have been experimenting with the two-color transform in VPHOT. The target star for this test is V2808 Aql, and I used an ensemble of twelve comp stars and one check star from the APASS database. Interestingly, I found that the difference between the untransformed and transformed values was less than the uncertainties in the measurements. The AAVSO report files for the transformed results indicate that the values are transformed. My transformation coefficients are based on averages of several sessions on M67 and NGC 7790. I'd appreciate any help I can get.
Allen
Allen,
For starters, I tried to make a finder chart for V2808 Aql, but VSP could not find it. It also is not in VSX. Perhaps this was a typo. Did you make a VSP chart for this star? If so, what is the number of the chart?
Phil
duplicate post