"If the image the image is solved then the orientation will be reported by a north-east indicator. The reported angle is between the north arrow and the Y axis counterclockwise (rotate east from north) except when the image is flipped. For flipped images it is clockwise (rotate east of north). So for both flipped and unflipped images it is the angle reported between Y-axis and north arrow in the direction north to east."
I agree with the first part but I don"t understand the last part (in bold). Why the change (for unflipped image) from an angle between north and the Y axis to an angle between the Y axis and north. It is not the same thing!
Take this image for instance (it's a fit I saved with Astap to hopefully get rid of all the row-order issues). The important thing is that it opens in the correct unflipped orientation: https://1drv.ms/u/s!AgPwp5gOmTQUg7df42CwqNy4rIfX8g?e=hbLcFp
Once open in Astap and solved, Astap gives an angle of 67°.
I believe this is wrong and that the angle should be around 290° instead and should read like this: in this image, UP is 290° counterclockwise from North.
I have added a schema that try to demonstrate that and some results from Nina and telescopius that show a 290° angle for the same field of view with these other tools.
The angle reported in ASTAP is taken from the old style solution. This defines the orientation CROTA and flip status CDELT. So the fits keywords (CROTA1, CROTA2, CDELT1, CDELT2 for the two axis
To define the orientation an image relative to the sky you have to specify two things:
1) Orientation
2) Flipped status.
If you want to place the image like a puzzle piece correctly in a sky map you will need these two pieces of information. The unflipped image you could place under angle of 67 degrees in the sky map.
If you try to place the flipped image under angle of 293 in the sky map it will never fit. You have to unflip it and then place under an angle of 67 degrees.
For this reason ASTAP places an letter F at the north east indicator as soon an image is flipped. See attached. In ASTAP you can flip an image under tools, rotate an image.
This is written in Wikipedia:
In astronomy, position angle (usually abbreviated PA) is the convention for measuring angles on the sky. The International Astronomical Union defines it as the angle measured relative to the north celestial pole (NCP), turning positive into the direction of the right ascension. In the standard (non-flipped) images, this is a counterclockwise measure relative to the axis into the direction of positive declination.
So if you have a flipped view you have to rotate in the opposite direction
This is complicated topic. Happy to discuss this longer.
Han
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thank you Han for your quick response. This is indeed a complicated topic!
And it's made even more complicated by fits format considerations which, I think, are not necessarily relevant because for instance, thanks to Astap's blind solve feature, I can reproduce the exact same thing with a jpg file (that I'm joining here).
First of all, do we agree that an image is flipped if east is to the right of north? If east is to the left of north, then the image is not inverted and is therefore an accurate representation of reality.
I believe this is always true and this is what we see in your Untitled.png file.
I disagree with your file M81_M82_Annotated1.jpg . As Wikipedia says: " The International Astronomical Union defines it as the angle measured relative to the north celestial pole (NCP), turning positive into the direction of the right ascension" (emphasis are mine)
In my view you are turning in the right direction but from the wrong arrow. Angle must be calculated from the North arrow towards the Up arrow and therefore the angle is not 70° but 290° (I rounded the angles for clarity). -70° would also be correct
I agree with you that the Astap method is perfectly fine and would be enough to place the image in the sky but I believe it does not follow the convention established by the IAU. This is why the angle reported is different from other tools.
Here are a few other examples:
Pixinsight gives an angle of -70° for this image (because it always normalizes angle in the [-180°,+180°] range
Carte du ciel also gives an angle of 290° for this field of view
Angle must be calculated from the North arrow towards the Up arrow and therefore the angle is not 70° but 290° (I rounded the angles for clarity). -70° would also be correct
I have plotted your image in HNSKY planetarium.
1) You are turning the the image 293 degrees counter clockwise to get the image north up as in the sky.
2) ASTAP is indicating the rotation of the sky 67 degrees counter clockwise to get the image orientation.
Both are correct and the only difference in the starting point either sky or image . The only thing what really helps to understand the orientation is a north-east indicator
The reported angle is the rotation from sky to image orientation counter clockwise. The camera for taking the image was rotated with the same angle clockwise.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
1) You are turning the the image 293 degrees counter clockwise to get the image north up as in the sky.
2) ASTAP is indicating the rotation of the sky 67 degrees counter clockwise to get the image orientation.
Everything is relative, you can rotate any frame of reference and indeed both solutions are equally correct. It is just a matter of convention. I think however it would be nice if the different astronomy tools would agree on this kind of thing.
The North-East indicator is really useful but it doesn't quite fit this scenario:
1. I see an image online which I want to shoot
2. I use Astap platesolving functionnality to get its coordinates and position angle
3. I use this information for framing my target in my acquisition software
Nico
Last edit: Nicolas 2024-08-20
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I would expect you set the frame orientation and position in a planetarium program and send the position and angle to Nina or set the angle using the Nina tools.
Using ASTAP angle you would have to calculate 360-angle for counter clockwise setting.
ASTAP is just reporting the image rotation (keyword CROTA2) while you want the camera angle. They are mirrored by definition.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Camera rotation in Nina is determined on the sky with a platesolve (same thing for an Asiair for instance). It would be nice if all tool using platesolve would agree on this angle.
But I won't bother you any further on the subject, there are indeed many ways to get around this problem and Astap nevertheless remains a superb astronomical tool.
Nico
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The solution in the ASTAP header is fully compatible with Astrometry.net or any other solver. It is defined by the four CD keywords CD1_1, CD1_2, CD2_1, CD2_2 and the ref celestial position CRVAL1, at reference pixel CRVAL2 and CRPIX1, CRPIX2. This is al standarized.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I have some questions about the angle reported by Astap.
I have read this page where the reported angle is explained
http://www.hnsky.org/astap.htm#viewer_angle
And I don't understand this:
"If the image the image is solved then the orientation will be reported by a north-east indicator. The reported angle is between the north arrow and the Y axis counterclockwise (rotate east from north) except when the image is flipped. For flipped images it is clockwise (rotate east of north). So for both flipped and unflipped images it is the angle reported between Y-axis and north arrow in the direction north to east."
I agree with the first part but I don"t understand the last part (in bold). Why the change (for unflipped image) from an angle between north and the Y axis to an angle between the Y axis and north. It is not the same thing!
Take this image for instance (it's a fit I saved with Astap to hopefully get rid of all the row-order issues). The important thing is that it opens in the correct unflipped orientation:
https://1drv.ms/u/s!AgPwp5gOmTQUg7df42CwqNy4rIfX8g?e=hbLcFp
Once open in Astap and solved, Astap gives an angle of 67°.
I believe this is wrong and that the angle should be around 290° instead and should read like this: in this image, UP is 290° counterclockwise from North.
I have added a schema that try to demonstrate that and some results from Nina and telescopius that show a 290° angle for the same field of view with these other tools.
Last edit: Nicolas 2024-08-20
Hi Nicolas,
The angle reported in ASTAP is taken from the old style solution. This defines the orientation CROTA and flip status CDELT. So the fits keywords (CROTA1, CROTA2, CDELT1, CDELT2 for the two axis
To define the orientation an image relative to the sky you have to specify two things:
1) Orientation
2) Flipped status.
If you want to place the image like a puzzle piece correctly in a sky map you will need these two pieces of information. The unflipped image you could place under angle of 67 degrees in the sky map.
If you try to place the flipped image under angle of 293 in the sky map it will never fit. You have to unflip it and then place under an angle of 67 degrees.
For this reason ASTAP places an letter F at the north east indicator as soon an image is flipped. See attached. In ASTAP you can flip an image under tools, rotate an image.
This is written in Wikipedia:
So if you have a flipped view you have to rotate in the opposite direction
This is complicated topic. Happy to discuss this longer.
Han
Attachements:
Thank you Han for your quick response. This is indeed a complicated topic!
And it's made even more complicated by fits format considerations which, I think, are not necessarily relevant because for instance, thanks to Astap's blind solve feature, I can reproduce the exact same thing with a jpg file (that I'm joining here).
First of all, do we agree that an image is flipped if east is to the right of north? If east is to the left of north, then the image is not inverted and is therefore an accurate representation of reality.
I believe this is always true and this is what we see in your Untitled.png file.
I disagree with your file M81_M82_Annotated1.jpg . As Wikipedia says: " The International Astronomical Union defines it as the angle measured relative to the north celestial pole (NCP), turning positive into the direction of the right ascension" (emphasis are mine)
In my view you are turning in the right direction but from the wrong arrow. Angle must be calculated from the North arrow towards the Up arrow and therefore the angle is not 70° but 290° (I rounded the angles for clarity). -70° would also be correct
I agree with you that the Astap method is perfectly fine and would be enough to place the image in the sky but I believe it does not follow the convention established by the IAU. This is why the angle reported is different from other tools.
Here are a few other examples:
Pixinsight gives an angle of -70° for this image (because it always normalizes angle in the [-180°,+180°] range
Carte du ciel also gives an angle of 290° for this field of view
Nico
Last edit: Nicolas 2024-08-20
I have plotted your image in HNSKY planetarium.
1) You are turning the the image 293 degrees counter clockwise to get the image north up as in the sky.
2) ASTAP is indicating the rotation of the sky 67 degrees counter clockwise to get the image orientation.
So
1) image --> sky, counter clockwise
2) Sky --> image. counter clockwise.
Both are correct and the only difference in the starting point either sky or image . The only thing what really helps to understand the orientation is a north-east indicator
Han
An other interesting aspect is to get you image orientation, you have to turn the camera 67 degrees CLOCKWISE or 293 degrees counter clockwise.
I will add to the documentation that the ASTAP reported angle is from "sky to image orientation"
I have added the following to the documentation :
The reported angle is the rotation from sky to image orientation counter clockwise. The camera for taking the image was rotated with the same angle clockwise.
Everything is relative, you can rotate any frame of reference and indeed both solutions are equally correct. It is just a matter of convention. I think however it would be nice if the different astronomy tools would agree on this kind of thing.
Nina changed this with its version 3.0:
https://f002.backblazeb2.com/file/ninasetup/Releases/3.0.0.9001/RELEASE_NOTES.html
Siril will change it in its next version:
https://gitlab.com/free-astro/siril/-/commit/f6793bc71466dbd0ead9d942d1126e9e0cd28b98
The North-East indicator is really useful but it doesn't quite fit this scenario:
1. I see an image online which I want to shoot
2. I use Astap platesolving functionnality to get its coordinates and position angle
3. I use this information for framing my target in my acquisition software
Nico
Last edit: Nicolas 2024-08-20
I would expect you set the frame orientation and position in a planetarium program and send the position and angle to Nina or set the angle using the Nina tools.
Using ASTAP angle you would have to calculate 360-angle for counter clockwise setting.
ASTAP is just reporting the image rotation (keyword CROTA2) while you want the camera angle. They are mirrored by definition.
Camera rotation in Nina is determined on the sky with a platesolve (same thing for an Asiair for instance). It would be nice if all tool using platesolve would agree on this angle.
But I won't bother you any further on the subject, there are indeed many ways to get around this problem and Astap nevertheless remains a superb astronomical tool.
Nico
The solution in the ASTAP header is fully compatible with Astrometry.net or any other solver. It is defined by the four CD keywords CD1_1, CD1_2, CD2_1, CD2_2 and the ref celestial position CRVAL1, at reference pixel CRVAL2 and CRPIX1, CRPIX2. This is al standarized.