Can what I’m seeing be real? - Forums

I just updated to the latest ASIAIR Plus firmware and this is what I’m seeing it report for my EQ6-R + Esprit 100 ED + Askar OAG + ASI174mm mini Bin1.

I’ve carefully checked the ASIAIR guide camera focal length (by setting it as my main camera and letting the ASIAIR plate solve it to 560mm).

it seems crazy to think I can get this kind of guiding with my equipment in my backyard at Bortle 7 Austin TX!?!?

Have others noticed this remarkable improvement in guiding with this new firmware?

Probably just a good nights seeing. I've had a few in that range with an almost identical setup (without ASIAir).

You're shooting at a medium focal length with a good mount and your guide camera has big pixels, I would be more shocked if it were worse.

Rob:
You're shooting at a medium focal length with a good mount and your guide camera has big pixels, I would be more shocked if it were worse.

Rob,

thanks for your reply but I’m confused.

The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

Wouldn’t tiny errors in calculating fractional pixel positions lead to larger reported errors as your guide camera’s pixel size increases and focal length decreases since each pixel would correspondingly cover a larger patch of sky?

Interesting. Obviously a good seeing night. What are your current guide settings?

John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

Rob:
John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

huh, didn't think about that before

good question

Not all stars fall nicely in the center of a pixel of the guide camera. Even with a grossly under sampled guide camera. You had a good night with your kit under the stars and hope you have many more just like this run!!!

CS Tim

Rob:
John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

It's very simple. Take a star lighting up many pixels in a radius. Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side. You can easily calculate how much the star has moved in that direction without necessarily having a higher resolution than the movement. Calculating the center of mass of a star will not necessarily lead you to the center of a pixel.

Miguel T.:
Rob:
John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

It's very simple. Take a star lighting up many pixels in a radius. Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side. You can easily calculate how much the star has moved in that direction without necessarily having a higher resolution than the movement. Calculating the center of mass of a star will not necessarily lead you to the center of a pixel.

Miguel has offered a much more verbose response than I did. I agree completely. I star that only illuminates a single pixel will never be used by a guiding software... Stars that are candidates for guides stars have will utilize many pixels...

CS Tim

Tim Ray:
Miguel T.:
Rob:
John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

It's very simple. Take a star lighting up many pixels in a radius. Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side. You can easily calculate how much the star has moved in that direction without necessarily having a higher resolution than the movement. Calculating the center of mass of a star will not necessarily lead you to the center of a pixel.

Miguel has offered a much more verbose response than I did. I agree completely. I star that only illuminates a single pixel will never be used by a guiding software... Stars that are candidates for guides stars have will utilize many pixels...

CS Tim

also multiple star guiding ?

TiffsAndAstro:
Tim Ray:
Miguel T.:
Rob:
John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

It's very simple. Take a star lighting up many pixels in a radius. Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side. You can easily calculate how much the star has moved in that direction without necessarily having a higher resolution than the movement. Calculating the center of mass of a star will not necessarily lead you to the center of a pixel.

Miguel has offered a much more verbose response than I did. I agree completely. I star that only illuminates a single pixel will never be used by a guiding software... Stars that are candidates for guides stars have will utilize many pixels...

CS Tim

also multiple star guiding ?

No matter how many stars are used for guiding, they must all meet the conditions to be considered reliable for movement detection. PHD2 will already make the right choice for you, even if you think some star look better. Multiple only mean the movement is reported by more than one star and the result can be averaged between them for more accuracy.

- Multiple pixel coverage
- No white clipping. Don't trust your human eyes. Over-saturated stars may appear nice to us but they're useless for guiding.

Miguel T.:
Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side.

I had not even considered a non-binary pixel reading for a star in guiding. 😅 In my mind it was lit or unlit. Thanks for your explanation, now I understand guiding more and can therefor stress less about it!

Rob:
Miguel T.:
Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side.

I had not even considered a non-binary pixel reading for a star in guiding. 😅 In my mind it was lit or unlit. Thanks for your explanation, now I understand guiding more and can therefor stress less about it!

The part you quoted was just to simplify the explanation. All it does in reality is calculate the center of mass, which naturally yields subpixel precision. Now you also understand why the ASI174MM is so popular. Big 5.86 um pixels for more sensitivity while arcseconds per pixel don't matter so much.

with multi-star option. this scenario that we have been talking about is just repeated with multiple stars in a single frame. then all the delta's are averaged together to calculate just how far the scope has moved vs just making this calculation based on a single star from the guiding sub….

this helps with limiting the effects of atmospheric turbulence so the guiding software isn't chasing the "seeing"…

I use this option on all of my PHD config's on every rig I use…

CS Tim

The real question is how long your guiding remains steady with those numbers. If guiding numbers like that are holding most of the night, then either ZWO vastly improved the ASIAIR Plus, you've got some really steady seeing, or both. But if those numbers are merely a snapshot of a particularly good moment in an evening's session, well that's pretty normal for the ASIAIR Plus if the seeing is adequate.

I've never had issues guiding with the Asiair, if seeing permits it then the guiding is consistently good. The past year using the AM5 values around 0.2-0.3 is pretty standard. Any more than that and I know the seeing is getting worse, thin clouds etc.

Miguel T.:
TiffsAndAstro:
Tim Ray:
Miguel T.:
Rob:
John Stone:
The ASIAIR reports those guiding numbers in terms of arc-sec which are absolute.

I might be confused, but the maximum resolution for that guide camera and telescope is 2.2 arc-sec/pixel, and the telescope itself has a limit of 1.16 arc-sec/pixel. How is it possible to detect deviations smaller than mechanically able to resolve?

It's very simple. Take a star lighting up many pixels in a radius. Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side. You can easily calculate how much the star has moved in that direction without necessarily having a higher resolution than the movement. Calculating the center of mass of a star will not necessarily lead you to the center of a pixel.

Miguel has offered a much more verbose response than I did. I agree completely. I star that only illuminates a single pixel will never be used by a guiding software... Stars that are candidates for guides stars have will utilize many pixels...

CS Tim

also multiple star guiding ?

No matter how many stars are used for guiding, they must all meet the conditions to be considered reliable for movement detection. PHD2 will already make the right choice for you, even if you think some star look better. Multiple only mean the movement is reported by more than one star and the result can be averaged between them for more accuracy.

- Multiple pixel coverage
- No white clipping. Don't trust your human eyes. Over-saturated stars may appear nice to us but they're useless for guiding.

I'm v v new and still f£#&king about with my guiding.
So I assumed multiple star guiding would be inherently better because of averaging 5 stars rather than guessing about one big bright fat over saturated blob.
Probably a mix of my own noob experiences and what I've read and watched.
My stars in phd2 tend to look like a dot surrounded by a comet coma but just changing exposure time (seemingly) at random if my guiding is bad to 1.5s or back to 1sec seems to influence it.
The zwo f4 120mm guidescope doesn't seem very good, but it's the only guidescope I've tried. Probably user error

Rob:
Miguel T.:
Now observe one or many of the pixels increasing in intensity by 5% on one side, and decreasing by 5% on the opposite side.

I had not even considered a non-binary pixel reading for a star in guiding. 😅 In my mind it was lit or unlit. Thanks for your explanation, now I understand guiding more and can therefor stress less about it!

dunno about asiair but in phd2 itself you can select target view and can see the calculated center of mass (?) of the target. It moves a fair bit around where my eyes think the center is.

TiffsAndAstro:
I'm v v new and still f£#&king about with my guiding.
So I assumed multiple star guiding would be inherently better because of averaging 5 stars rather than guessing about one big bright fat over saturated blob.
Probably a mix of my own noob experiences and what I've read and watched.
My stars in phd2 tend to look like a dot surrounded by a comet coma but just changing exposure time (seemingly) at random if my guiding is bad to 1.5s or back to 1sec seems to influence it.
The zwo f4 120mm guidescope doesn't seem very good, but it's the only guidescope I've tried. Probably user error

Multiple star guiding is/should be better.

Not only you're averaging the non-uniform atmospheric turbulence that could affect one star more than the others but also the sensor noise that could send the center of mass of a star out of wack. Imagine trying to calculate that accurately on just a single star when the noise add a "random" small value to each pixel. Longer exposure time can also help to average the bad seeing (2s/3s instead of 1s for example), as long as you don't over-saturate. I personally run at 2s at all time.

Now if the seeing is globally bad / the mount can't move with the precision we're asking it to / the guidescope is wobbling or terribly out of focus / PHD settings are non-optimal for your setup, no miracle can be made.

I use dew strips on my finder/guide-scopes. I also have seen the "comet comma" on the 120 and my finder/guide-scopes. Mostly I refocus the guide camera and then mess with the gain and the slider on the PHD bottom bar next to the Brain…. I hardly ever focus my finder/guide-scope. Maybe 10 sessions or an entire imaging season goes by in between focus runs… I have also disconnected the camera from PHD and reconnected the camera in PHD has solved this. .

The slider on the bottom is like the STF function in PI. I believe PHD calls it Display Gamma???

CS Tim

Tim Ray:
I use dew strips on my finder/guide-scopes. I also have seen the "comet comma" on the 120 and my finder/guide-scopes. Mostly I refocus the guide camera and then mess with the gain and the slider on the PHD bottom bar next to the Brain.... I hardly ever focus my finder/guide-scope. Maybe 10 sessions or an entire imaging season goes by in between focus runs... I have also disconnected the camera from PHD and reconnected the camera in PHD has solved this. [I haven't seen it in a while so all of this is from an old guys memory] .

The slider on the bottom is like the STF function in PI. I believe PHD calls it Display Gamma???

CS Tim

cheers for this. I refocused the guide scope in daylight yesterday but of course clouds arrived as soon as I 3ppa.
Didn't realise I could change gain via brain button. Mine is still on 48 zwo mm mini guide cam. What gain do you use if you have same camera?
And yeah that slider is just contrast/brightness for the display.
I use cheap dew heaters on both scopes. Not noticed any dew yet but in winter who knows? I'm new

I typically run the gain at Unity gain on my guide cameras which include several 120mmMinis, a 120mm-S, 174mm and a 290mm. The 120mini's are my workhorses for most guidescopes. I run a small SBONY 30mm GS on my Samyang rig. Most other rigs use a 50mm or 60mm Guide Scope in the 200mm to 260mm FL range.

Usually I run the exposure recommended by the various "assistant" utilities inside PHD. I do use an entire worm revolution as the time I run the guiding assistant utility vs the 2min minimum time to capture data for utility to chew on. Other than that default will get you down the road pretty far.

I use ASCOM with Green Swamp instead of EQMOD on my Skywatcher mountings and ASCOM and CPWI instead of EQMOD/GreenSwamp on the Celestron mountings. The SB MyT mount configuration is outside the scope of this thread…

I do use 30 sec of settling after a dither and I dither every 3rd frame.

CS Tim

I change the gain thru ASCOM not the Brain button… I was using the Brain as a Sign Post as not to confuse which slider I was referring to… Anyway… Getting all of this to work is really what I enjoy most about the hobby… its the Alchemy! All your questions where great and are common within the community. Always ask. Folks on this forum are terrific and are eager to help!!!

CS Tim

Not to knock your setup, but I wouldn't trust that those guide numbers are accurate. Your guide camera and scope combination yield 2.16"/pixel. Nothing wrong with that, and as others have explained the guiding software is perfectly capable of calculating centroids on stars with better accuracy than the resolution of the guide camera. However, getting numbers around 1/8th of the resolution of the camera is not likely yielding an accurate number.

That being said, I doubt it matters at all. As others have pointed out, you clearly had a night of good seeing with a setup that was working well. That's wonderful, and even with a 4" telescope that can yield really nice, subtle details in your images. Since you are using an OAG, you don't need to worry about differential flexure.

However, was the RMS error on the guide exposures really 0.27"? Or was it maybe 0.31" or perhaps it was 0.23"? I don't think I would "trust" numbers that are that far below the sampling resolution of the camera and scope if I were trying to draw conclusions like, "My guiding was working better than a previous night when I measured 0.35." Rather, I would simply conclude, "My guiding was working really well (or, more precisely, it wasn't having to work much at all). Be happy that it's performing, and move on to the next thing.

I'm not quite sure why we collectively seem to obsess quite so much about our guide numbers. Once I get a graph that indicates my guide camera and telescope are working well, I try to look at how my actual exposures are doing. For example, I often find that I get my best guide numbers with 2s exposures, but I get my best telescope performance (eccentricity and FWHM of stars in my actual image) with 1s exposures or 1.5s exposures, even though the guiding looks a touch worse. Basically, I am trading faster guide response for higher uncertainty in centroid measurement.

Since your system is working well, I would recommend you shift your attention now to what yields the best images–it might be you get sharper details with settings that generate marginally worse guide numbers.

- Jared