Can you really do photometry with the little ZWO Seestar S50? Well, yes. But probably not to the level that some would have you believe.

I read several posts on Starry Night claiming a photometric accuracy of 0.01 mag using 10 x 10s stacks. But I'm skeptical of that claim.

The graph shown here is for a stack of 6 x 10s, and the photometry was performed on the green channel signal using PixInsight against the PPMX Catalog in Strasbourg, France. I asked for stars down to mag 15, but it only gave me down to about mag 12.5. Still, that's pretty neat to have. What it gives is the (proper) V channel magnitudes, while I'm using my Bayer green channel as a cheap substitute. I know that some folks have transforms for going between my G and the proper V, but I didn't bother with that effort here. That transform won't affect the scatter we see here.

The image here shows the field of T CrB that I grabbed the other night. Trying to find the field in the first place can be tricky since the little Seestar S50 doesn't allow you to enter RA and Dec. But with star hopping and patience, I finally nabbed it. I stacked all 6 images and did a drizzle integration to 1x scale, then tore out the G channel after removing the background gradient. I didn't bother with any noise removal nor deconvolution. That won't matter for aperture photometry.

The Aperture Photometry produces a report for all stars in the field that it was able to match against the catalog, containing the RA & Dec of the star, its V mag, my Signal to Noise ratio, and my measured flux in the aperture. I used an 8-pixel circular aperture here. It does background estimation and subtraction before presenting the flux.

There is an unknown zero point offset between V and my measured magnitudes (mag = -2.5 Log10 (Flux)). But we can estimate it on the assumption that the majority of stars will not be hugely variable. Using an iterated fitting for that zero point, using graded weighting, initially based on SNR, as described in the writings of Astronomer Peter Stetson, we can arrive at a pretty good estimate for that zero point constant, then add it to our measured magnitudes to get our estimate for V.

Now take the scatter in that zero point as the inherent scatter of our measurements, and you find a band of uncertainty around the zero point. In this case it is +/-0.1 mag. And that tells me that you aren't going to reach 0.01 mag uncertainty in just 4 more 10s frames. So that's why I"m skeptical of the 0.01 mag claims...

[ Uncertainty in mag measure follows the same inverse Square Root law as SNR. If you need to improve your SNR by a factor of 10, you need 100x measurements. Same for improving magnitude uncertainty by a factor of 10. In my case, I got to +/-0.1 mag in 6x10s. To reach 0.01 mag I'd have to stack 600x10s, or about 1.7 hours of integration. For myself, life is too short to spend that much time on just one star field. I'm delighted that I can reach 0.1 mag uncertainty in just 1 minute of integration on a super simple-to-use telescope. YMMV ]

But just as neat, if not more so, see those stars showing 3 or more sigma deviation? That means our image captured several stars that are believably brighter and dimmer than the nominal catalog V magnitudes. I haven't yet tracked down just which stars in the image field those are, but they aren't all faint stars, which makes the result all the more believable.

Science can be a lot of fun...

[PS: After converting reported decimal RA & Dec to HMS and DMS format, I can see that the bright star in the batch, the one showing almost 0.5 mag brighter than the catalog V -- that star is T CrB. ]