Fri, 07/09/2021 - 19:32
I've been doing variable star photometry with my canon camera and telescope for a while now but I was interested in looking at exoplanets and asteroids as well.
With the variable star stuff I'm transforming the magnitudes but I wondered if that was necessary with exoplanets and asteroids. With exoplanets you are just looking for a slight change of magnitude during the transition so I figured it wouldn't really matter which wavelengths you were looking at. With asteroids they are illuminated by the sun and if you want to see their rotation once again it didn't seem to matter which wavelengths you look at.
The reason for the question is that the exoplanet guidelines seem to suggest that a DSLR would not be suitable.
Cheers
Steve
Hi Steve,
for exoplanets you only need differential photometry.
I've caught some exoplanets with my 8 inch Newtonian 200/900 mm and my Canon 200 D, filter TG. Some BAV colleagues do get results at even Residuals S = 3 - 5 mmag (mag-fit) accuracy with a Celestron C11 and a Canon 600 D.
See e.g:
http://var2.astro.cz/ETD/
http://var2.astro.cz/ETD/etd.php?STARNAME=XO-6&PLANET=b
http://var2.astro.cz/EN/tresca/transit-detail.php?id=1606144632
..for which equipment can reach which star, with which DSLR or CMOS or CCD camera.
wbea
PS: If you like asteroids, take a look at the book: Brian D. Warner : A Practical Guide to Lightcurve Photometry and Analysis
PPS: Man! somebody has caught an exoplanet with only an 4 inch refractor (100/500 mm)! this is great!
http://var2.astro.cz/EN/tresca/transit-detail.php?id=1586499909
I've got…
Thanks Bernhard,
I've got the Brian D. Warner book. I bought it for variable stars but it gave me the idea to try asteroids as well I've got an 8inch skywatcher Newtonian so I have high hopes to catch something. I just need a clear night or two.
Cheers
Steve
Cool!
you can take a look at:
NASA (They use their own reduction software)
https://exoplanets.nasa.gov/exoplanet-watch/
ESA (..has also an own reduction software, but ESO allows also data from your own software! (-:)
https://www.exoclock.space/
Exoclock is giving you ephemerides for exact your telescope focal length and aperture, and also for selected sky regions (if one is looking only in one direction...).
At ETD, one can look for targets by himself :
http://var2.astro.cz/tresca/transits-prediction.php?lang=en
for a given date, and D (min) , V (mag).
and of course take a look at the AAVSO Exoplanet Section (they have also their own reduction software (-;), and the excellent AAVSO youtube webinars!!.
At first try a bright star <= V=10 , with a deep transit depth ( 0.02 ....0.08 ), to see what you are able to obtain with your DSLR.
wbea.
Are you…
Hi Bernhard,
Are you averaging the two green channels? Would you get greater sensitivity if you averaged all the channels including TB and TR? That would be like 2x2 binning I suppose.
Steve
Hi Steve,
The Bayer Mask has Tri Color Red: TR, ...TG, TG; TB
I use Muniwin
https://sourceforge.net/projects/c-munipack/
which is averaging: (TG1 + TG2) / 2
I would not mix TB or TR! This will be leading to false results.. There is no binning at DSLRs, as far as I know...
PS: yesterday i cought this one (-:
http://var2.astro.cz/tresca/transit-detail.php?id=1626002690&lang=en
wbea
I understand…
Hi Bernhard,
I understand when you are measuring variable stars that you have to measure the wave bands separately but if you are only looking for a dip in light intensity during a transition of an exoplanet I don't see why it should lead to a false result to add the counts R+G1+G2+B and I thought it might give a slightly better sensitivity.
Steve
well.... you can do an experiment: (-;
(just kidding)...or you can take a look at any variable star at the lightcurve generator. There you will see that R;B;G od TR, TB, TG are not in the same region of light (standard mag)
wbea
PS: what you can do is, doing all the work for TR and all the work for TB, then you'll get the different dips (depths) of TR, and TB as seperated LCs.. (But not mixing them)
I believe you but I'm not sure why I'm wrong. I think the time period of the transit must be the same whichever wavelength band you look at. Also the time between transits must be the same. I found an equation which said that the ratio of the change of flux (height of the dip) and the flux from the star is equal to the ratio of the planets radius squared and the star radius squared. That's basically the ratio of the areas. I think that's how they work out the radius of the planet. I think that ratio must be the same whichever wavelength band you look at. So even if the height of the dip is different the ratio will be the same.
So I think you could just look at the total counts of all wavelengths and I think that would give you a better sensitivity because you are looking at finding a ratio of bigger numbers. I should say I'm not at all sure of myself about this.
A very interesting topic for me.
Steve
Hi Steve,
in short: the starlight generates different amplitudes through different UBVRI filters. Therefore, you cannnot add TG+TR+TB into one lightcurve, you have 3 LCs. Although, the TOM (time of minimum) will be the same.
Your are welcome to take a look into the DSLR + EXOPLANET Manual, it is a huge and great ressource. But take your time:
https://www.aavso.org/observing-manuals
See also:
https://en.wikipedia.org/wiki/Photometric_system
https://www.wikiwand.com/en/UBV_photometric_system
http://astro.vaporia.com/start/passband.html
The Bayer mask from DSLR are not made initially for doing photometry but it's working.
If one like to have the resolution of the whole sensor array -> the way is leading to monochrome CMOS or CCD, with the UBVRI Filters. But there you have to make Flats + Darks for every filter... (The pretty picture astronomers are using RGB + so on ....filters)
wbea
"... the starlight generates different amplitudes through different UBVRI filters. Therefore, you cannnot add TG+TR+TB into one lightcurve, you have 3 LCs. Although, the TOM (time of minimum) will be the same."
I wrote a paper on this using a variable star as the test (Axelsen, R. JAAVSO 45, 92, 2017). For my DSLR, the precision was greater with the "white light" (R+G+B) differential photometry that for either standard (non-transformed) V or transformed V. Of course, this data can be used only for timing light curve events, not for determining true magnitudes. In my opinion it could be used in this manner for asteroid light curves.
The stimulus for doing the above experiment was thinking about the fact that if only green channel data is used, the light from one half of the pixels is ignored.
Roy
Thanks Roy, that was my thinking too. I think Bernhard is right but I need to think it over some more.
Steve
luckily for men I…
Hi Steve,
luckily for men I did not wrote: "don't do it", merely... just try it. (-;
One of my favoured sayings comes from a math book: "Everyone was saying, "that" is not possible! Until someone came, who did not now that, and just did it!
wbea
Hi Roy,
thank you for your input! I'll read it carefully:
https://app.aavso.org/media/jaavso/3244_lDtwW61.pdf
@ Steve: Steve you are lucky, Roy already did your experiment and published it (-;
OK, I had in mind to upload stardata to the AAVSO database, therefore filters should not be mixed, to let the data be comparable with others.
But Roy you are right, for EXOPLANETS, only Differential-Photometry (V-C) is important. Not so much to obtain the correct standard magnitude of a star...
The side effects of mixing TR,TB, TG ( in the end, if there are some, or not at all...) are maybe having less impact to the end-result, than to get lower QE or SNR with only the green TG filter!
I'll re examine my Exoplanet data with all RGB, and check the results. (-:
Thanks Bernhard
Another idea I had was this. Bernhard's chart was a plot of magnitude against time but would it be better to plot intensity against time instead? Intensity is the total count of the pixels covering the star image. The equation I found related the ratio of fluxes and the ratio of star/planet radii. But Flux is proportional to intensity not magnitude since magnitude is a logarithmic value. By plotting intensity the dip during the transit is magnified so would be easier to see and for the purpose of calculating the radius ratio would be a more useful value anyway.
I did some calculations based on observation of a magnitude 12.3 star. The intensity came out at about 6000 using maximdl. Berhard's star had a dip in magnitude of about 0.02 magnitudes. For a 12.3 magnitude star that's about 0.16% dip. If you work out the equivalent in intensity it's more like a 2% dip - ten times as much.
Steve
Hi Steve,
The Exoclock-Project is accepting flux or magnitude. ??Muniwin is only giving magnitudes??
Other SW is giving also the flux?
wbea
I use MaximDl and you can get it to output the 'intensity' of the star. This is the raw ADU value of the centroid. Not really flux but I guess it must be proportional to flux.
It was just an idea I had. I haven't really followed it up. I just thought that magnitude being logarithmic might have the effect of attenuating any dip.
Bernhard, I haven't used exoplanet filters.
I don't know of any side effects.
Just one other comment. The images for the data in the paper were taken with no autoguiding, and the polar alignment was good bur not perfect. Therefore the field drifted slowly during the night. In my experience, autoguiding will improve sensitivity.
Just realized that the above is not correct. The last phrase should be "... autoguiding will improve precision."
Roy
I've read your…
Hi Roy,
I've read your paper, you did a lot of work.
Why did you choose the star V703 Scoorpi?
Bernhard
I've studied delta Scuti stars for several years, of which V703 Sco was one. I thought I had found frequencies of pulsation that differed from those reported in the literature (two periods are reported, 0.1152d and 0.1499d), but this turned out not to be sustainable. Thus I had a lot of unpublished time series photometry on V703 Sco, and when I thought about DSLR photometry using ADUs from all three channels combined for differential photometry, there was already enough data for this star to test the idea.
Roy