AUTOGUIDING ENCODER MOUNTS USING ASIAIR DISCUSSION: Page 2

AUTOGUIDING ENCODER MOUNTS USING ASIAIR DISCUSSION

SFn8CieR

w7ay
Man oh man you really know your stuff!!
Fantastic explanation of encoder mounts and multistar guiding with an asiair using SNR, and the reason for increased gain in the camera to filter out the bigger bright stars that have more influence (due to SNR) which reduces the overall average and benefit of multistar guiding.

In my case with a ASI174 MM mini guide camera, ZWO recommends gain of 189 for lower read noise. You say another 10db (=100 units on asiair), to set the gain to 289, to rid the bright stars in the field of view. Is that gain figure right or just use any gain figure to begin with and then adjust accordingly if bright stars are being used for guiding?

To me, asiair is the lazy mans way to take astrophotos, which suits me down to the gound atm. I bought a mini computer recently with Nina, PHD2 etc on it and maybe I will try that one day. I have taken some great images using the asiair without really knowing much at all about guiding, with a fast scope, which is much easier to guide but now with my 130-PHQ it is much more critical to understand this stuff. Also one day I will buy a larger scope, maybe a 10" RC and guiding that will be a bigger challenge. PS. I thrive on challenges as I need to keep myself mentally stimulated and this is doing it for me..

Nick

w7ay

SFn8CieR PS. I thrive on challenges as I need to keep myself mentally stimulated and this is doing it for me..

Yep, the only reason this hobby is keeping me interested after 60 years. It is definitely not taking photos, but trying to better understand the fundamentals.

Have fun, Nick!

Chen

SFn8CieR

Hi Chen,
It has been cloudy here so many nights I have not had a chance to try out the guiding yet. However I contacted Ioptron again to ask about this timelapse thing they previously mentioned. Here is their reply..

Hi Nick,
Since your mount has encoder, it might have some benefit with long cadency guiding, that is, guiding corrections are issued on intervals of, say, 5 to 10 seconds to guide out the low-frequency errors, such as polar alignment error, atmospheric refraction, thermal fluctuation, flexure, etc, and no need to use rapid guiding in order to allow the encoder working diligently between pulses and keeping the mount on track. You may set exposure time long, 5-10 sec (instead of 1-5 sec with non EC model), which will dampen the effect of seeing variations right away or set guiding cadence with two parameters: the guide camera exposure time and the time lapse (time between guiding pulses). 2-3 sec exposure plus 5-10 sec timelapse should be a good start point.

My train of thought has always been that an encoder mount needs less corrections from guiding software due to very low PE. I have tried this and failed each time probably because my Calibration Step and RA&DEC Durations have not been set up properly. My RA&DEC durations were too high so guiding pulses were too long. Also Calibration Step was no good because I was getting 20 or so steps to calibrate. So I am going to try this again but with the revised numbers.
I will also try short exposures as I think that this will work with the information you have given me.

Not sure about using timelapse though as asiair does not give the option of setting the time between exposures.
What is your thoughts on all of this?
thanks Nick

w7ay

SFn8CieR I will also try short exposures as I think that this will work with the information you have given me.

Long exposures can have problems with mounts that moves a lot while exposing a guide frame, so I won't recommend really long exposures. I would rather use shorter exposures, but issue correction pulses less often.

What many people don't realize is that when you are taking a guide exposure to estimate the new cetroid of a star, that exposure does not give you a pinpoint star. The star is moving while you are taking the exposure! So you end up with a elongated star, and that is harder to estimate xcentroid. At the end of the exposure, the star is at one of the two ends of the elongation... but the software has no idea which end of the elongation was the start, and which is at the end! The result is that you get poor guiding.

With strainwave gear mount, the star can move half an arc second (angle) if you use a one second (time) exposure. Because of this, strain wave mounts need very fact exposures, so as not too fall into this elongation trap.

The star does not move from point to point. In fact, the movement is like a sawtooth. You can see a desciption of that in the discussionn here (post #6, #74, #84 in a very long thread):

https://bbs.zwoastro.com/d/15989-getting-the-best-performance-from-my-am5/6
https://bbs.zwoastro.com/d/15989-getting-the-best-performance-from-my-am5/74
https://bbs.zwoastro.com/d/15989-getting-the-best-performance-from-my-am5/84

Once you understand the sawtooth graph, everything will become clear and autoguiding will no longer be mysterious. Please try to understand the sawtooth. I promise it is worth it for the rest of your autoguiding life :-).

The problem is that people have always assume that when you take a guide exposure you actually get the location of a star. You don't. You get a line that is traced by the sawtooth.

Now, you have a good mount, so your sawtooth is very small. The ZWO mount is so poor that I have given up helping people. The ZWO mount is a lost cause.

Chen

SFn8CieR

w7ay
Boy you were'nt kidding when you said a lot to read...whew!
Made a lot of sense and very interesting to read.

You mentioned for smooth mounts, the slope of the PE is small, and the approximation of using the centroid is good enough. So my question is, the Asiair uses SNR for autoguiding and not centroids, so is that the same thing? I could be wrong here but Star Centroids need to be estimated during guiding whilst SNR is more accurate especially if multi star guiding?

When you say to reduce or slow down the guide rate, do you mean the siderial guide rate say from 0.5 to 0.25?) to apply a gentler slope to the leading edge of the sawtooth, or, do you mean increasing the correction pulse? I am a little confused here.

Still have to let all this sink in and maybe read it over a few more times.
thanks Chen,
Nick

w7ay

SFn8CieR Asiair uses SNR for autoguiding and not centroids

No, no. ASIAIR does use the average of the centroids of multiple stars. But each centroid is firtst weighted by its SNR.

This means that if you have a very bright star (or two), the result will be mostly contributed by that bright star. I.e., you are back to 1-star guiding.

If you start with 1 second guide exposures, a 2 second exposure should reduce the variance of the error by a factor of two (RMS by a factor of 1.414). If you double again, and use 4 second exposure, the variance of the error is now 4x smaller than 1 second exposure (RMS error is a factor of 2).

When you double the number of stars, you are doing the same thing (assuming ergordic statistics). I.e., the variance of the error caused by atmospheric turbulence is reduced by a factor of 2 each time you double the number of stars -- but only if they all have the same SNR in ASIAIR. That is why you should be using better software than ASIAIR. ASIAIR has too many problems once you look deep enough -- ZWO simply don't understand the mathematics underlying astronomy. Time keeping is another thing that ASIAIR fails misreably.

When you say to reduce or slow down the guide rate, do you mean the siderial guide rate say from 0.5 to 0.25?) to apply a gentler slope to the leading edge of the sawtooth, or, do you mean increasing the correction pulse? I am a little confused here.

As I said, you need to keep thinking and try to understand that sawtooth. If I anser your question, you will have one more, and I cannot keep answering questions. If you understand the sawtooth, you will be able to figure it pout yourself.

Chen

SFn8CieR

Does a guidescope need to be accurately centered on the same stars on the main scope to get accurate guiding?
Can it be a little offset or is this not really relevant?
cheers Nick

w7ay

SFn8CieR Does a guidescope need to be accurately centered on the same stars on the main scope to get accurate guiding?

Of course not. The guide scope has nothing to do with your main scope. It is there to tell your mount to ponit to a constant coordinate in the sky. I.e., guide scope has to do with the mount, not the main scope (and that is why differential flexure has to be taken care of, unless you use an OAG -- which has too many other problems of its own).

You can be imaging M81, but point your guide scope at the celestial equator.

The only times where it matters are 1) there is differential flexure; but that exists even with both scopes pointed at the same place, and flexure should be fixed for its own sake anyway. 2) the target is much closer to the horizon than the guide scope (or vice versa). Atmospheric refraction near the horizon can be pretty bad.

https://en.wikipedia.org/wiki/Atmospheric_refraction

As time passes, the change of a star position near the zenith is very small, while the change is very rapid as the other scope rises from the horizon.

Remember too that most autoguiders are guiding by the pixel in the plate, not by the arc seconds of a star in the sky. Because of that, if you are imaging something close to the pole , the RA axis will show very small pixel error while the real angular RA difference can be quite large. This is why when testing and reporting guiding errors, you should do it neat the celestial equator, or the numbers will look too optimistic.

Chen

« Previous Page