Well known RVb DF Cyg has two periods - short about 50d and long about 780 d.
However TESS data shows clear low amplitude third oscillations with period 0.25d.
Is it real pulsation mode or something other?..
Hello Degen,
What sort of original data did you used to make those graphs ?
Christophe
Peranso downloaded just one TESS set: sector 15, 2 min, #17280, 2019-08-15…2019-09-11.
It’s hard to beleive to so fast process in lax post-AGB star, so I suspect instrumental effect.
But there is also crasy idea about central white dwarf 
Because of the big TESS pixels (21"), contamination from a nearby star is a possibility in crowded fields. I note a WISE eclipsing variable shown in VSX about 50" SE that has period 0.500 days, about 4 mag fainter than DF Cyg. Depending on the extraction method used for the TESS data (there are several such), you might be getting a greatly subdued trace of that star in the DF Cyg lightcurve.
\Brian
I fact, I mean : did you download an ready made LC, or did you perform the photometry by yourself ?
In the later case, as Brian said, it may be a contamination by a nearby star, in your target aperture, or in the comp aperture.
It could be also that your comp is the source of the pulsation.
Maybe you could check Aladin/Gaia in search for nearby stars responsible of the contamination ? And try another comp.
Last but not least : try to do the same photometry in another sector. The instrumentals should be different from one sector to another.
Yes, yes, for WISE J194856.2+430131 TESS gives exactly the same (ready) light curve, so it is obvious instrumental effect.
A lot of thanks for the explanations!
(upload://dGf7MMbUl1OdpB7FFODMYVfR6dL.png)
Hello Degen,
Your wise LC is very interesting.
Such a total similarity in both LCs is not supposed to occur for two stars separated by one or two arcmin. If two stars are located in the exact same pixel (separation a few arcsec), their respective light is blended, and the resultant LC is contamined by each other.
This is not the case here.
So the issue is to be found somewhere else : Peranso ? Possibly the link to the downloaded LC ?
Anyway, what you have to do now, is to download the original TESScuts and perform your own photometric analysis.
Never rely upon other work, allways question/verify data and result.
C
The task exceeds the capabilities of the human brain, but Chat GPT fortunately helped to get DF Cyg light curve directly on MAST website, without Peranso. In flux it has the same shape with fast oscillations.
WISE has not been resolved by name, but coordinate search gave the same LC, so nothing changed. I do not have ideas on the reason of artefact.
Wow, Kepler gives EXCELLENT 4-years light curve, classic RVb without any hints on fast oscillations!
Just a few words about your last LC.
This is absolutely not the same temporal definition, nor the same magnitude scale, in the Kepler LC than in the TESS LC.
Possibly the Kepler LC could have been smoothed at a level where it should be difficult to see some 0.02 mag amplitude lasting a few hours only.
Step of points is about 30 min, so time resolution is enough. But indeed curve is suspiciously smooth…
…
I can also confirm the oscillation is due to contamination of WISE J194856.2+430131 / TIC 272951492 .
Pixel level data, in TESS sector 15 for DF Cyg / TIC 272951532. Pixels outlined in white are the aperture chosen by the SPOC pipeline:
If one excludes the pixels near the contaminant, the oscillation largely disappeared.
Other ready-made TESS lightcurves from QLP pipeline have similar contamination issue. In sector 74, QLP uses a circular aperture of 3 pixels radius [*], so the contamination is even more severe.
One can inspect it interactively in the webapp:
https://tess-tpf-fqhnyorhza-uw.a.run.app/tpf?tic=272951532§or=15&magnitude_limit=
[*] QLP aperture can be found in QLP data’s header BESTAP. In sector 74, its value is 3.0:4.0:3.0, indicating a (circular) aperture of 3-pixel radius, and background annuli with a 4-pixel inner radius and 3-pixel width.
One way to inspect the header is to use lcviz on MAST website. Look for metadata in the plugin sidebar.