Drift alignment

Astrophotography has never been easier

Moderators: Sam, Tech@ZWO, ASIAIR@ZWO

Post Reply
Suttonrog
Posts: 24
Joined: Sun Oct 27, 2019 10:35 am

Drift alignment

Post by Suttonrog » Wed Jan 22, 2020 7:04 pm

My ancient Skywatcher EQ5 Pro bearings finally seized, so I have upgraded to an EQ6 R pro. Unplugged the 5 and plugged in the 6- no problems.
Like many people, I do not have a clear view north to do a proper PA, so out came my laptop to do a PHD drift align and quickly had it tracking well.
Do ASI have any plans to include drift alignment in future issues so that I can ditch my laptop?
Rog

User avatar
ASIAIR@ZWO
Posts: 487
Joined: Tue May 29, 2018 3:29 am

Re: Drift alignment

Post by ASIAIR@ZWO » Thu Feb 06, 2020 11:02 am

Suttonrog wrote:
Wed Jan 22, 2020 7:04 pm
My ancient Skywatcher EQ5 Pro bearings finally seized, so I have upgraded to an EQ6 R pro. Unplugged the 5 and plugged in the 6- no problems.
Like many people, I do not have a clear view north to do a proper PA, so out came my laptop to do a PHD drift align and quickly had it tracking well.
Do ASI have any plans to include drift alignment in future issues so that I can ditch my laptop?
Rog
Yes, it's on the plan already.
ZWO | ASIAIR Support
Location:lon=120.6 lat=31.3
Suzhou China

MSellars
Posts: 7
Joined: Mon Sep 30, 2019 8:38 am

Re: Drift alignment

Post by MSellars » Mon May 18, 2020 4:15 pm

I have the same problem in the Southern hemisphere - I cannot see the SCP (Octans) to do a PA.

Please, please, please . . . . ZWO I cannot wait for drift align support in ASIair

vladi
Posts: 3
Joined: Tue May 12, 2020 9:09 am

Re: Drift alignment

Post by vladi » Thu May 28, 2020 12:39 pm

+1 A very welcomed feature! Thanks for putting it on the plans!

W7AY
Posts: 224
Joined: Fri Jan 25, 2019 3:38 am
Location: Portland, Oregon, USA

Re: Drift alignment

Post by W7AY » Thu May 28, 2020 6:19 pm

Instead of resorting to full drift alignment (it takes a fair amount of time for the object to drift enough), you may still be able to use ASIAIR Polar Alignment even if you cannot see the pole.

Short answer: if you can see a clear spot in the sky between DEC +/- 70º and DEC +/- 90º (+ for northern hemisphere, - for southern hemisphere), and in addition, 60º in RA west of that clear spot is another clear spot in the sky, you can use ASIAIR's polar alignment process. If you do not mind reduced accuracy, you can extend the area in the sky to DEC +/- 60º.

Long answer: I have tested ASIAIR's PA on my mount simulator, and have found that the hour angle of the mount at the start and finish of the ASIAIR PA 60º slew does not affect polar alignment accuracy.

For example, place your mount's declination at somewhere between 89 and 90 degrees on the prime meridian side of the hemisphere (i.e., tilt your mount's DEC just a tiny bit towards the zenith).

Many people start at zero hour angle (i.e., OTA pointed almost at NCP), but that is not necessary.

In my tests, I have found that the starting hour angle does not matter at all. You can for example start with the RA that is 2 hours east of the meridian, and let ASIAIR slew it to 2 hours west of the meridian to perform the PA adjustments.

(In fact, the +2 HA to -2 HA configuration is most convenient for my RainbowAstro RST-135 mount because if I start at zero hour angle, it is difficult to lock one of the Altitude locking bolts when the mount has finished slewing.)

The ASIAIR only needs to plate solve before the slew, and again after the slew has finished (and while you are adjusting the altitude and azimuth bolts). So, as long as your OTA has a clear view of the sky at the start of the slew, and at the end of the slew, you are all set.

ASIAIR does not need images of the sky in between the starting and ending position of the 60 degree slew. And more importantly, ASIAIR does not need to see the pole itself.

From my simulation, I have not seen a polar alignment error of more than a few seconds of arc by starting at different hour angles.

For polar alignment, ASIAIR will also accept DEC angles that is as much as 30 degrees from the pole. It will tell you that you are more than 30 degrees off if you deviate too much. So now you have a huge piece of the sky to perform polar alignment -- 30º worth of DEC and 120º worth of hour angle.

However, try to stay above 80 degree declination if possible (+80 for northern hemisphere, -80º for southern hemisphere), because the accuracy of ASIAIR polar alignment appears to suffer when you get very far away from the pole. If memory serves, the error is large when you are 30 degrees in declination away from the pole (but even that might be acceptable, depending on your plate scale).

That being said, you do not need to have very precise polar alignment to capture good images. This is a simple calculator for field rotation (i.e., what happens if your RA axis does not point exactly at where the earth's pole points to):

http://celestialwonders.com/tools/rotat ... rCalc.html

The result is in micrometers (microns). You can check your camera specs to find your camera's pixel size (typically between about 2.5 microns to 5 microns). If your error is less than the pixel size (or even the Airy disk of your optics), you do not need more polar alignment accuracy.

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

The polar alignment accuracy will also depend on atmospheric refraction. This is not a problem if the polar region is more than 60º in altitude (i.e., your latitude is over 60 degrees). However for lower latitudes, atmospheric refraction compensation is less accurate, and you also need to factor in temperature (Bennett's formula uses 10ºC to estimate the refraction before temperature correction, for example) and pressure. I think the ASIAIR just uses "typical" temperature and pressure (I have never seen their code, so I don't know exactly what they use to compensate for atmospheric refraction.)

See here to learn about atmospheric refraction:

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

You can see that it is completely moronic and a waste of time to try to adjust polar alignment for errors of less than 10 arcsecond, because the estimation error for the pole's location is greater than that, and with typical plate scales, field rotation will not be large enough to be noticed. Just because ASIAIR tells you the polar alignment error is zero, does not mean that your RA axis is actually pointed at the pole.

Good luck,
Chen

vladi
Posts: 3
Joined: Tue May 12, 2020 9:09 am

Re: Drift alignment

Post by vladi » Wed Jun 10, 2020 9:12 am

Wow! Thanks a lot, Chen, for explanations and insights! It is really helpful! Especially the fact that there is no need for more or less exact alignment on to polar star! This will save a lot of time while setting up.

Still, IMHO, the need to see the northern side is a serious issue for many potential "astro backyards" ;) I cannot speak for the whole world, but I believe most houses and buildings are rotated the way that yards, balconies and terraces - basically all good locations to put a telescope, are facing south or south-west. This ensures that there is the longest time for sunlight to fall on those places. It does make a sense for any non-astro use ;) In this situation the view on the meridian is not obscured in many cases. Plus, there is a view onto the east or west. For some larger places, it would be possible to walk away from the walls of a house to still be able to see the Polaris. Still, people would like to place they setups possibly close to the walls of their houses, since there is either a cover from rain or a shorter distance to WiFi, power, beer, etc. This also requites some flexibility on the FOV onto meridian and onto east and south to escape possible abstractions in the surroundings. This is why I still believe that even a larger flexibility in the classical polar alignment will not eliminate a strong in interest in a software-supported drift alignment or other techniques not requiring a view to the North.

I tried the drift alignment in PHD2 a few times. This is very lengthy process of similar steps repeating all over. You just need to understand what is going on and make sure to not make small mistake, since it throughs your back for 10-15 min. Maybe it is about me, but I never completed it in faster than 20 min. Still, the potentials are large. The move of stars on the meridian with a 200mm focal length is very fast, you can just watch that on the screen w/o additional measures with an aperture of 2sec and below. The sequence of steps is actually very simple IMHO (https://www.cloudynights.com/articles/c ... vice-r2760). Unfortunately, PHD2 does not makes it so simple for some reasons. I guess for effort for an implementation of a good user guidance (who still needs to rotate the adjustment bolts) would be a fraction of the effort of the current polar alignment.

PS. KStars and Ekos also provide a support for drift alignment with a view on south and either on east or west, but the workflow there, IMHO, is the same lengthy and hard to understand for a beginner as in PHD2.

Just my 2 cents...

--
Vladimir

W7AY
Posts: 224
Joined: Fri Jan 25, 2019 3:38 am
Location: Portland, Oregon, USA

Re: Drift alignment

Post by W7AY » Wed Jun 10, 2020 4:16 pm

Hi Vladimir,

I think that we are in complete agreement that some kind of tool is needed when you have no clear view at all of the polar region. I was simply pointing out the misconception that the ASIAIR needs to see the pole itself to do Polar Alignment. Even SharpCap allows you to be up to 5º away from the pole.

I also agree with your experiences with PHD2 . The mosquitos usually got the better of me in the days I was drift aligning with PHD2.

You can already use the high magnification settings (reverse pinch the ASIAIR app's preview screen) to perform the D.A.R.V. procedure. However, a more precise crosshair that has double lines could be useful too, since D.A.R.V. is really a visual tool.

However, ZWO appears to be hell-bent to make ASIAIR a button-pushing experience for beginners (notice that they won't even add an "advance" window to expose some of PHD2's functions, for example). And they may consider the D.A.R.V. user interface to be beyond beginners' reach.

I have recently been investigating using plate solves to perform the steps in D.A.R.V., instead of visually waiting and watching for the "V" shaped drift to form. (instrumentation is one of my main interest in astrophotography -- the actual images are just a side interest :-) If I come up with anything, I will definitely suggest it to ZWO. ZWO appears to have very capable software and algorithm folks on the ASIAIR project, but as far as I know, except for Mr. Wen (Sam) himself, the developers are not amateur astronomers themselves. So they might be less knowledgable of the history of astronomical instrumentation, or even the need for some of the tools.

Another thing that I am looking into is whether the DONUTS algorithm can be used to estimate polar alignment corrections. DONUTS is a multi-star auto-guiding method that uses correlations of the FFT of successive autoguiding frames to detect deviations in tracking, It is very resistant to "seeing" problems. DONUTS was developed by professional astronomers not that long ago, but it is available in INDIGO; multi-star autoguiding is also available in the Lacerta MGEN-3. Granted, plate solving (if there are enough stars) also averages multiple stars and is also less affected by "seeing."

One of the disadvantages of many drift methods (including D.A.R.V.) is that you require patches of clear skies at two locations, and the need to be very close to the Meridian for one of the locations. Frank Barrett has a series of white papers on the subject of Polar Alignment a few years back,

http://celestialwonders.com/articles/polaralignment/

In the last one (http://celestialwonders.com/articles/po ... oning.html) he states,

[quote]"However, when doing a simultaneous correction of both axes, it will be necessary to choose the reference star carefully. In this case, it will be necessary to choose a reference star on the meridian. Here’s why. Only on the meridian will an adjustment in the altitude axis result in a star’s apparent motion in an altitude-only direction. Once you deviate from the meridian, the star’s motion will appear to move in both altitude and azimuth.[\quote]

I.e., if the star is not smack on the Meridian, then the drift has both altitude and azimuth components, making it more difficult to manipulate the alt and az bolts.

Interestingly, IF we can use plate solving instead of using visual drift align, the plate center *can* be placed smack on the Meridian, and that may be able to provide a good hint for people who are searching for better polar alignment techniques.

If you have not yet come across the descriptions of the "Collins' method," the following is also worth reading:

http://www.shadycrypt.com/pages/Collins ... vation.htm

The diagrams show the kinds of drift one gets near the east and west horizons for people at different latitudes.

The need to be close to the east or west horizon can also affect accuracy of determining the pole. Stars near the horizon will suffer from atmospheric refraction, and correcting for refraction is not perfect.

Polar Alignment is definitely an intriguing subject :-). The holy grail :-).

Chen

Post Reply