Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

Polar Alignment | Astrometry style


Gonzo

Recommended Posts

  • Replies 35
  • Created
  • Last Reply
Hi Gonzo,

Looks like an interesting thread but it's completely above my head.

What software are they talking about?

Can you give us a brief description of the technique and what makes it stand out?

same here. looks good, but have no idea how it was done.

Link to comment
Share on other sites

Hi Gonzo,

Looks like an interesting thread but it's completely above my head.

What software are they talking about?

Can you give us a brief description of the technique and what makes it stand out?

In a nutshell, you would need to have the following on your computer:

astrometry.net-0.38.tar.gz (Astrometry.net)

astrometry indices (you must send them an email to give you access to the download area

imagemagick

How it works:

align your mount to Polaris as normal.

take a picture and parse it through the script (http://web1036.sabine.webhoster.ag/Bilder/np/polar.zip)

It will produce an image showing where the NCP is.

If you are miss-aligned, adjust your mount and take another picture.

parse that new picture again through the script, repeat as needed.

Link to comment
Share on other sites

I don’t really get it L how this software can know how my mount is aligned from single picture? I am sure there must be other conditions met.

If you read the comment from the author of the script in the link I've posted, it states:

"

For polar alignment with this methode you have to point your telescope roughly to the NCP. That means finding polaris and pointing to it. Then take a picture, solve it, check the crosses, move the telescope with Alt/Az screws, take another picture, check the crosses and so on. After that you should still do an 1/2/3 point star alignment.

"

:p

Link to comment
Share on other sites

I am still not convinced. In this method you are aligning your scope with NCP, not mount. In order to align your mount, you would have to have scope perfectly aligned with RA axis. Its not always possible to be 100% accurate (I think its called cone error).

Link to comment
Share on other sites

you also need to read this thread http://stargazerslounge.com/diy-astronomer/108643-help-me-test-my-polar-alignment-procedure.html and combine the two methods

---

To start off, you need to roughly align the RA axis and set your camera pointing to declination +90 or as close as you can get it (so that Polaris and Lambda will be in the picture). Lock the declination at this stage and don't change it until PA is finished. The two images will be taken at different RA angles, the first with the short side of the sensor horizontal and the second with the long side of the sensor horizontal. Review the two images and record the (x,y) pixel coordinates of the two stars.

---

Link to comment
Share on other sites

you also need to read this thread http://stargazerslounge.com/diy-astronomer/108643-help-me-test-my-polar-alignment-procedure.html and combine the two methods

---

To start off, you need to roughly align the RA axis and set your camera pointing to declination +90 or as close as you can get it (so that Polaris and Lambda will be in the picture). Lock the declination at this stage and don't change it until PA is finished. The two images will be taken at different RA angles, the first with the short side of the sensor horizontal and the second with the long side of the sensor horizontal. Review the two images and record the (x,y) pixel coordinates of the two stars.

---

I see...I think I will stay with PHD Drift Alingment :p

Link to comment
Share on other sites

The above method is more for EQ1/2 and mounts without a polarscope.

Here's the sequence of pictures I took this evening and put through the script.

image_service.php?photo_id=1409&type=midimage_service.php?photo_id=1411&type=mid

image_service.php?photo_id=1408&type=midimage_service.php?photo_id=1406&type=mid

image_service.php?photo_id=1405&type=midimage_service.php?photo_id=1407&type=mid

Here's a random 10s exposure picture I took after my polar alignment of the EQ1 (and using the motor)

image_service.php?photo_id=1412&type=mid

I'm not seeing any star trails, OK this is not your top notch alignment or greatest picture you'll see but it's more of a proof of concept thing :p.

Now using my mini photo tripod, I took another 10" exposure picture to find out if there was going to be star trails at that exposure :icon_salut:

Click and see by yourself.

image_service.php?photo_id=1413&type=mid

I'm getting an HEQ5 early this year, but this method is great for my EQ1.

Will be writing an improved script for it soon.

Link to comment
Share on other sites

The method in post #4 will not work unless you somehow know that the centre of your image is dead centered on your RA axis. That is difficult to achieve with an EQ1 or similar. That's why my spreadsheet method relies on TWO images, taken at different RA positions. By seeing how the stars have been rotated in the two images I am able to compute the position of the RA axis. And by "plate solving" the images I know where the NCP is.

Once I get astrometry.net installed on my Windows laptop, I will see if I can make a "blind" version of my spreadsheet that doesn't require manually recording 4 star positions.

Link to comment
Share on other sites

The proof that it works really well would be a 3-4 minute exposure at zero declination without trails.

I'm not sure at what declination you shot those 2 test images, so I plate solved them. The first one is centered at declination 87.2 (in Ursa Minor) and the second one at declination 18.9 (between Gemini Orion and Taurus). Do you think it's a fair comparison?

Link to comment
Share on other sites

The proof that it works really well would be a 3-4 minute exposure at zero declination without trails.

I'm not sure at what declination you shot those 2 test images, so I plate solved them. The first one is centered at declination 87.2 (in Ursa Minor) and the second one at declination 18.9 (between Gemini Orion and Taurus). Do you think it's a fair comparison?

two different pictures, one on the mount, the other one on the tripod as said earlier.

wow, I didn't think that sharing an interesting link would be received that way.... :)

Link to comment
Share on other sites

two different pictures, one on the mount, the other one on the tripod as said earlier.

Ok, but the difference in declination means that in those same 10 seconds, stars appear to move very different angular distances. At those two declinations, the rates are in the ratio of about 1:3.

wow, I didn't think that sharing an interesting link would be received that way....

It is interesting but I think it's wrong and I tried to explain why. Apologies if I caused offence.

Link to comment
Share on other sites

As we have clear sky tonight, I'll try again that method and do a fair comparison.

one with the motor on and one without, 30sec exposure of the same stars :)

Will post results later on.

And if it doesn't work, my up-coming HEQ5 pro will do ;)

Ok, but the difference in declination means that in those same 10 seconds, stars appear to move very different angular distances. At those two declinations, the rates are in the ratio of about 1:3.
Link to comment
Share on other sites

If you visit

Polar Alignment Error Calculator

and plug in the values you used for your tracked photo you will see that the polar alignment error could be quite large and still show very little star trailing. Your image scale was 17 arcseconds per pixel so let's say that we wouldn't be able to detect a half-pixel trailing, say 8 arcseconds for the first field (called "Drift"). Then we put 0.16666 for the second field called Time and 87 degrees for the third field called Declination (as that was the declination of the centre of the image). The calculator comes up with a polar alignment error that is anything up to 3500 arcminutes, or about 60 degrees. That means that your polar alignment could be out by about 60 degrees and you'd still be getting an almost undetectable star trailing on that part of the sky.

For the static shot, I estimate the star trailing at about 10 pixels which makes 170 arcseconds. If you put that value into that website's calculator and also put in a declination of 19 degrees you see that the polar alignment error is 4120 arcminutes so a static shot is roughly what you'd get if you misaligned a tracked platform by about 70 degrees.

Link to comment
Share on other sites

yes, try and shoot for one minute and pick a target in Orion, that's pretty much on the celestial equator where the stars drift the fastest. Even with one minute you will only be testing your polar alignment to about 1 degree accuracy.

Link to comment
Share on other sites

Here's my test.

Polar aligned using the above method ( I am no the author, just a user)

image_service.php?photo_id=1417

I then moved the camera towards Orion as you asked.

(sorry, couldn't be be asked in rotating the pictures)

Must say that I have very very slightly nudged the mount before taking the following pictures.

image_service.php?photo_id=1414&type=mid

30sec iso 800 using RA motor

image_service.php?photo_id=1415&type=mid

50sec iso 400 using RA motor

image_service.php?photo_id=1416&type=mid

50sec iso 400 NOT using RA motor

Link to comment
Share on other sites

I looked at the 3 belt stars with the Windows magnifier and you can just about make out a flattening of the stars of about one pixel. That would give you a polar alignment error of less than 1 and 1/4 degree.

Now, your first image, of Polaris, with the red crosses, shows the two crosses intersecting to about one pixel accuracy and the image scale is 100 arcseconds per pixel so you would think that the Polar Alignment error would be better than 100 arcseconds or 1.67 arcminute or 0.028 of a degree.

Do you see that in order to prove to yourself that you have achieved such a tiny polar alignment error you would need to take a much longer exposure, something like 40 minutes?

That is why I am not convinced that the closeness of the two red crosses tells you anything about the error in polar alignment. There is nothing in the procedure that was described in that website that makes me think that the centre of the image coincides with the axis of rotation.

Link to comment
Share on other sites

Here's how I see it, I did not write the script, I was just sharing it.

Last week, I was unable to take pictures with an exposure more that 4 or 5 seconds. Today, I was able to take a 50 seconds picture with minimum star trails.

I have an HEQ5 pro lined up, and probably a CCD camera (not sure which one yet, more research is needed).

So when that day happen, I'll start worrying about the ultimate polar alignment.

I know you're trying to help and you're The expert, but throwing numbers like this without explaining how you get them is not helping the newbies like me.

I'll try your method once I know how you get all those numbers.

PEACE :)

Link to comment
Share on other sites

throwing numbers like this without explaining how you get them is not helping the newbies like me.

You are quite right. Which number was giving you grief? I got the image scale from the astrometry.net solution and the polar alignment error numbers from that web calculator I pointed at Polar Alignment Error Calculator

Today, I was able to take a 50 seconds picture with minimum star trails.

That is great but it doesn't prove that the method you came across works. You can get that result with any method that puts you within a degree (that's two full moon diameters) of the celestial pole. OK, how did i get that number? I just plugged in 17,1,0 in the fields of that calculator and hit Calculate. The resulting alignment error is about a degree. Why did I put 17? because at your image scale that's the limit of the perceptible drift.

So when that day happen, I'll start worrying about the ultimate polar alignment.

I was trying to help you build up some intuition and understanding. Maybe someone else reading this is able to get more out of it.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.