58 replies to this topic

[yuzameh]

I've seen many a 'green' magnitude off by a magnitude or more from contemporaneous visual and purported V magnitudes in databases.

 

Consequently my cynicism knows no bounds when people tell me something unfiltered (especially if unclear NIR and UV ends potential leakages have to be considered) is the same as using a monochrome device with proper Johnson V filter AND Johnson B filter so that proper transformations can be done on stars that are not white.

 

Best bet is to suck it and see.

 

Pick some reference stars from somewhere, constant stars, measure a star in whatever way you wish whether letting the machine do it all, via imaging two stars and setting one as the reference value against which the other will be measured, or using some sort of ensemble approach, then check your result and see if it agrees with the catalogued Johnson V magnitude for the star, THEN tell yourself if you are just getting green channel magnitudes or sufficiently proximate Johnson V ones.

 

Too much theory and not enough primary testing with personalised practical calibration against valid reference objects in amateur times series photometry.

 

Mathematics isn't a science, you can give proof of anything using algorithms irrespective of any reality.

[robin_astro]

TG to V is a reasonable thing to attempt but take care when transforming from TB and TR. It  is not only the filter passband shape but the out of band leaks with the R and B channels in RGB sensors.(in the visible, even with a UV/IR block). Things might look ok but you can then get poor results for example when measuring cool stars in B and hot stars in R. (This has even been an issue with some photometric filters)

 

Robin

Edited by robin_astro, 04 May 2024 - 06:12 AM.

[robin_astro]

Point to raise in Brian Kloppenborg's upcoming AAVSO webinar perhaps ?

 

https://www.aavso.org/webinars

 

 

Robin

Edited by robin_astro, 04 May 2024 - 06:35 AM.

[Ed Wiley]

Robin,

 

That is very good advice. I will be interested to hear about transforming TB amd TR. If you look at some untransformed reports of TG they seem to follow untransformed results for V (examples, V Boo Z Uma, R Cyg; the CV needs zero point correction). Since most observers do not transform, that is comforting. The few R Aqr untransformed TG reports show less scatter than some of the V reports. I am hard put to even find any TR or TB. Can anyone on the forum point to some TB and TR untransformed observations or those that are transformed? I have only been looking at LPVs, perhaps that is not the most productive class to search for TB and TR data.

 

Ed

Edited by Ed Wiley, 04 May 2024 - 11:33 AM.

[pbealo]

Looking at V1117 HER, from over 4000 observations there are just qty 2 TB, TR and TG data points.

[pbealo]

BTW: having difficulty using TG to determine TCs for my SeeStar. I easily got M67 images into VPhot, which separated the color planes, but once I make the AIP compatible reports, TG reports an error "no valid reference data to compute transforms"...

 

Peter

 

 

Edited by pbealo, 04 May 2024 - 04:24 PM.

[Ed Wiley]

Peter B.,

Those TG and TB data points, sparse as they are, look good relative to V and B. I wonder about red stars. Any examples out there? Perhaps this will be discussed at the AAVSo Webinar.

Ed

[apearce]

Hi All

 

I agree that transforming TB to B and TR to R is something to watch out for.  Of course, it's the same process as transforming from TG to V, however I agree with Robin that these are more problematic.  I generally don't report transformed TR to R magnitude estimates from my Seestar.  My experience so far is that the transformed V magnitudes from the Seestar compare quite favourably with values derived from monochrome cameras with a V photometric filter so can't see a reason to not report these.

 

Andrew

[yuzameh]

Hi All

 

I agree that transforming TB to B and TR to R is something to watch out for.  Of course, it's the same process as transforming from TG to V, however I agree with Robin that these are more problematic.  I generally don't report transformed TR to R magnitude estimates from my Seestar.  My experience so far is that the transformed V magnitudes from the Seestar compare quite favourably with values derived from monochrome cameras with a V photometric filter so can't see a reason to not report these.

 

Andrew

You've got to remember that when you are mostly, as you have for a long time, monitored CVs and, as in more recent times, discovered novae, a lot of the things you are looking at are white, or more or less white (especially when in outburst), and even when reddened aren't very red, just extincted a bit (lost some blue but not actually inherently red).

 

That can make V vs green channels look good.

 

On the other hand, I've seen green from other scopes miss visual by a large mark.  On the third appendage we are specifically talking about seestar here, so maybe fair does.  Maybe not.

Edited by yuzameh, 05 May 2024 - 03:01 PM.

[jgraham]

Just a quick update, I have been using my Seestars (now 3 of them) to take photometric images. I bought my first Seestar specifically for this purpose and liked it so well I bought another one. Then, I bought a third when they went on sale. With 3 Seetars going at once I can cover a lot of stars pretty quickly. I was getting a bit overwhelmed, so for now I have decided to concentrate on following T CrB. I have been using my cameras to take photometric images for nearly 20 years and my approach is probably a bit dated, but it seems to give good results. One of these days I'll update my techniques, but for now I'll keep do'n what I do. For the Seestar I take a 3 minute set of 10 second subs (18 subs), save the source images, stack them in DSS, extract the green channel in Nebulosity 4, measure a reference star, a calibration star, and the variable star in Autostar Image Processing (ASIP), reduce using spreadsheet that I developed many years ago (based on differential photometry? I can't remember. I seems to work.) Record the result as Tri-G. (I used to do Johnson V with a monochrome camera.) The graph below summarizes my observations for T CrB

 

 

In addition to T CrB I have been following a dozen or so other stars and the results compare very well with V and Tri-G data being reported on the AAVSO website.

 

As soon as my T CrB campaign raps up I'll be turning my attention back to my main observing list.

 

Fun stuff!

 

 

[Xilman]

One thing worth keeping in mind: if all one is interested in is temporal behavior the filter used is generally of very minor significance. Unfiltered is perfectly fine for determining the orbital periods of binary stars, mid-transit times of exoplanet transits, and rotational light curves of asteroids.

 

Exoplanetary work does have a minor dependency on color, but that affects the ingress and egress times more than anything; mid-transit times are essentially unaffected.

 

Paul

[jgraham]

Just a quick update...

 

My Seestars (now 3 of them) have proven to be photometry machines! They are so easy to use that it's usually not a problem to set one outside and grab a quick 3-minute observation of T CrB through reasonably clear gaps in the clouds. When running all 3 it's easy to get overwhelmed with data, but that's a good thing most of the time when our weather gets back to its more normal maybe 1 clear night per week. During those time I take data like crazy and then use the cloudy days/nights in between to reduce the data. Once we get past the T CrB campaign I want to spend some time taking a closer look at the bright/faint limits of the Seestar. I hoping for a short-exposure firmware update before then. I told ZWO if I had 1 wish for the Seestar it would be a 1-second exposure as an option. Having control over the exposure and gain is wish #2, but one wish at a time...

 

Just for yucks... my latest update on T Crb taken with the Seestar.

 

 

These are all 3-minute total integration times using the default 10-second exposure, source images saved as FITS, stacked in DSS, color separated in Nebulosity 4, the green channel reduced in Autostar Suite Image Processing using AAVSO reference charts, stars, and photometric data.

 

Enjoy!

 

[jgraham]

We always take a Seestar with us when we travel so that I can grab my observations of T CrB if it is clear, and we have set it up in some pretty interesting places. This one was probably the most fun. We are currently on the South Carolina coast and I have been setting up the Seestar out on the beach. Tonight I had to lift the Seestar up a bit to catch T CrB before it set behind a row of hotels just off of the beach, so I set it on top of a row of beach chairs tied up just above the high tide line...

 

 

 

The sky conditions were pretty challenging, but the Seestar performed brilliantly, and the data looks great!

 

A summary of my Seestar observations through this evening...

 

 

Neat stuff.

 

[gw_dra]

Wow, looks great! Unfortunately since I started this thread back in April I've gotten too busy to try it myself yet. But it's great to see what the little Seestar is capable of!

[glabella]

+1

The only limitation that I have had with the Seestar is on the high end. It look like magnitude 6ish to 7ish is about as bright as I would like to go, though I need to take a closer look to see where the actual limit is. Fortunately, I don't have a lot of stars in that range, though there are several that I am interested in. The only one that I'm currently watching that is that bright is R CrB 'cause its an old favorite and it's in the neighborhood with T CrB. After talking with the folks at ZWO at NEAF it sounds like they will be offering greater low level control of the camera (a Pro-Mode) in a future firmware update. I told them at the tippy top of my wish list was _shorter_ exposures so that I could go after brighter stars.

BTW... the Seestar does a great job on asteroids as well. I love going after tiny worlds way out in the asteroid belt.

Never a dull moment.

wouldnt just lowing the gain on the seestar accomplish being able to measure brighter stars so they dont oversaturate at 10 sec.  I know people are using higher gain settings in the 110 and 120 range and perhaps higher however i suspect they dont really know how it is affecting the noise level since most have poor sky conditions from night to night... 

[jgraham]

Good question. I’ll need to dig a bit deeper. I have been so busy measuring T CrB I haven’t had time to pursue this. However, it would be nice to do this in preparation for measuring T CrB when it erupts.

[Xilman]

Since getting the Seestar I have measured precisely one VS from a single stack of FITS images and have a two questions concerning measurement noise.

 

The electrons per ADU influences the sensor's noise level, hence the SNR, which is (almost) inversely related to the estimate error in the magnitude estimate. For the Seestar, what is the electrons per ADU value as a function of gain, and any other variable for that matter?

 

Further the dark current varies with the temperature of the chip but the CCD-TEMP header reports that quantity. What is the dark current sensor as a function of temperature? The application of auto-darks will remove the mean dark current at each pixel but will add the noise inherent in the dark at that pixel.

 

So far I have only used cooled CCD cameras and the answers to each of the questions above are generally given in the technical documentation. My searches for them in the IMX462MC documentation have been unsuccessful to date.

 

 

Thanks,

Paul

Edited by Xilman, 29 October 2024 - 07:25 AM.

[mrm6656]

Xilman:

 

The accessible Sony data sheet doesn't include much detailed info. But some of the info that you seek is available from ZWO who produce planetary cameras with the IMX462. Check out the page at

https://www.zwoastro...oduct/asi462mm/

Zwo have no separate data sheet that I could find for the color version of their 462 cameras, but the electronic data that you want are chip properties and should be independent of any color matrix on top of it.

 

      --- Mike

[Xilman]

Since then I have found this information https://www.firstlig...ing-camera.html

 

The e- per ADU value given there is around 1.0 for a gain of 100 and almost flat for higher gains, which is plenty good enough for my purposes as the SNR is only weakly dependent on the e- per ADU value.

Edited by Xilman, 05 November 2024 - 01:53 PM.

[robin_astro]

Since then I have found this information https://www.firstlig...ing-camera.html

 

The e- per ADU value given there is around 1.0 for a gain of 100 and almost flat for higher gains, which is plenty good enough for my purposes as the SNR is only weakly dependent on the e- per ADU value.

Note though that with that camera at gain 100, because the A to D converter  is only 12 bit this limits the maximum electron count to 4k (though the software might rescale this to trick you into thinking you are getting 16k). This is less  than the 11.2k full well depth of the sensor so you overflow the AtoD converter well before you reach full well depth.  (How this might relate to what happens inside the Seestar might be totally different though)

 

Cheers

Robin

[jgraham]

Just a quick note…

 

I have tried not to over-think using the Seestar, I just take what it gives me and run with it. The data seems spot-on from at least magnitude 15 to about 8. Above 8 you start to push saturation using the standard 10sec subs. (With my usual gear I try to stay below 80% saturation and my cameras tend to be linear up to at least 90% the last time I checked.) I haven’t pushed the low end yet as none of my current stars are below magnitude 15.5, but I’d like to reach at least magnitude 16. (I can reach my typical photographic limit of magnitude 18 with my usual gear.)

 

When using the Seestar for photometry you have the camera save all of the source images as FITS. These are calibrated by the Seestar, but no other processing is done to them. I typically take 20 subs (3m 20s total integration), stack in DSS, separate the R, G, B data, then reduce the G data.

 

The strength of the Seestar is that so far it is producing excellent data, it is remarkably quick and easy to use, and we take one with us whenever we travel and I can often keep track of my variable stars everywhere we go. I have set my Seestar up in some pretty interesting places and it has been an absolute joy to use. I ended up with 3 of them, which makes it possible to really crank through my active star list pretty quickly. I’m on the pre-order list for the new S30. I’m not sure I’ll be using it for photometry, but I’m at least going to check it out.

 

Should be fun!

[josjavpol]

I congratulate you on your work. I find it incredible that with a simple and modest team like Seestar, you are able to achieve these goals! 

 

 

Besines: Detecting exoplanets with simple equipment: "I used a tele-photo lens (the Pentacon 135 mm f 2.8), a CMOS camera (ZWO ASI 120 MM) and an equatorial mount (Skywatcher EQ3-2)"

 

https://www.cloudyni...pest-equipment/

 

Do you think it would be possible with Seestar?

Edited by josjavpol, 27 November 2024 - 08:58 AM.

[Xilman]

Note though that with that camera at gain 100, because the A to D converter  is only 12 bit this limits the maximum electron count to 4k (though the software might rescale this to trick you into thinking you are getting 16k). This is less  than the 11.2k full well depth of the sensor so you overflow the AtoD converter well before you reach full well depth.  (How this might relate to what happens inside the Seestar might be totally different though)

 

Cheers

Robin

The Seestar uses gain=80 by default. I am completely unfamiliar with CMOS detectors so have little idea what that means in terms I do understand: full well depth, e-/ADU and ADC resolution. Do I divide or multiply them by 0.8 and if so, which?

[robin_astro]

The Seestar uses gain=80 by default. I am completely unfamiliar with CMOS detectors so have little idea what that means in terms I do understand: full well depth, e-/ADU and ADC resolution. Do I divide or multiply them by 0.8 and if so, which?

I've just used the plots that come with the camera specs (like those in the link you posted) to translate camera gain to e-/ADU  but I suspect it will vary between cameras (and suppliers) and  I don't know if  there is a similar graph specifically for the Seestar

 

Cheers

Robin

[robin_astro]

 I don't know if  there is a similar graph specifically for the Seestar

 

 

They could be measured along with the noise characteristics from pairs of bias darks and flats though. 

 

Cheers

Robin