Your simple mirrors are stranger than you think

[Ben the Ignorant]

 

[Louis D]

I get more confused trying to figure out why mirror diagonals flip things vertically for refractors and CATs, but Newtonian diagonals don't do the same thing for Newts.  It seems like the same principle, doesn't it?

[andrew s]

Excellent explanation of the mirror paradox.

Regards Andrew 

[Ben the Ignorant]

On ‎16‎/‎09‎/‎2017 at 15:41, Louis D said:

I get more confused trying to figure out why mirror diagonals flip things vertically for refractors and CATs, but Newtonian diagonals don't do the same thing for Newts.  It seems like the same principle, doesn't it?

Interesting remark, Louis. I am going to rest a newtonian tube on a table, aim it at a landscape, and try various orientations of the focuser, plus various orientations of my head to sort this out.

[harrym]

I think it's to do with even/odd numbers of mirrors. In the Newtonian there are two mirrors and each mirror flips the image once, so the image you see isn't flipped (though it is rotated 180°). However, a refractor with a diagonal has just one mirror (in the diagonal) and a catadioptric with a diagonal has three, so the image is flipped.

[Louis D]

1 hour ago, harrym said:

I think it's to do with even/odd numbers of mirrors. In the Newtonian there are two mirrors and each mirror flips the image once, so the image you see isn't flipped (though it is rotated 180°). However, a refractor with a diagonal has just one mirror (in the diagonal) and a catadioptric with a diagonal has three, so the image is flipped.

Rotated 180 or flipped, it still looks wrong terrestrially.  How many more mirrors would it take to make it look right side up and not flipped left for right?  Would they have to be oriented in a specific way like porro prisms (at right angles to each other) in a monocular?

[andrew s]

Don't forget a conventional astronomical refractor inverts the image about the optic axis !

Regards Andrew

[Ben the Ignorant]

I put a newtonian tube on a table, facing an electrical post, a pine tree to the left, and a leaf tree to the right. With the focuser held vertically, I turned my back on the landscape, and saw everything "correct" in the eyepiece. Post is upright, pine tree to the left, leaf tree to the right.

Later I rotated the tube around its horizontal axis so the focuser went from vertical to horizontal, and sat at the standard place for looking through a newtonian: on the side. The post's bottom was first in the left part of the image, and when rotating the tube, the post's bottom turned from 9 o'clock to 12 o'clock.

Will have to try that with my little SCT. Funny that a newtonian does work as terrestrial scope, but only if you turn your back on your target!

[Ben the Ignorant]

3 hours ago, harrym said:

I think it's to do with even/odd numbers of mirrors. In the Newtonian there are two mirrors and each mirror flips the image once, so the image you see isn't flipped (though it is rotated 180°).

But the main mirror reflects light along its optical axis, whereas the secondary does so 90° away from the main mirror's axis. This is tougher to represent in the mind's eye than I thought!

[harrym]

2 hours ago, Louis D said:

Rotated 180 or flipped, it still looks wrong terrestrially.  How many more mirrors would it take to make it look right side up and not flipped left for right?  Would they have to be oriented in a specific way like porro prisms (at right angles to each other) in a monocular?

Has to be an even number for the same reason I mentioned before. You could do the Porro prism effect with 4 mirrors I think? (Edit: but this might cause distortions...)

Does anyone know how those 45 degree image erecting diagonals work?

[Louis D]

25 minutes ago, harrym said:

Has to be an even number for the same reason I mentioned before. You could do the Porro prism effect with 4 mirrors I think? (Edit: but this might cause distortions...)

Does anyone know how those 45 degree image erecting diagonals work?

They use Amici prisms that invert and revert as well as bend the light either 45 or 90 degrees.  I think inverting and reverting is the same as rotating 180, correct me if I'm wrong.  I have never grasped how Amici prisms work their magic, though.  The 2D diagrams fail miserably to explain what is happening in three dimensions.

[Louis D]

57 minutes ago, Ben the Ignorant said:

I put a newtonian tube on a table, facing an electrical post, a pine tree to the left, and a leaf tree to the right. With the focuser held vertically, I turned my back on the landscape, and saw everything "correct" in the eyepiece. Post is upright, pine tree to the left, leaf tree to the right.

Later I rotated the tube around its horizontal axis so the focuser went from vertical to horizontal, and sat at the standard place for looking through a newtonian: on the side. The post's bottom was first in the left part of the image, and when rotating the tube, the post's bottom turned from 9 o'clock to 12 o'clock.

Will have to try that with my little SCT. Funny that a newtonian does work as terrestrial scope, but only if you turn your back on your target!

I think the old Zuka spotting scope used a single extra diagonal type mirror to send the image to your eye while holding the scope on your shoulder, bazooka style.  I assume the final image was upright and not reversed given it's use as a terrestrial spotting scope.

[Ben the Ignorant]

I had never heard of that Zuka scope, Bing finds no picture of it, and Google finds only three, from the same Astromart page! Must be pretty rare; I'm not sure what kind of diagonal was added in order to use it parallel to the natural line of sight. I think we all need to work on our visualization skills!

[harrym]

28 minutes ago, Louis D said:

They use Amici prisms that invert and revert as well as bend the light either 45 or 90 degrees.  I think inverting and reverting is the same as rotating 180, correct me if I'm wrong.  I have never grasped how Amici prisms work their magic, though.  The 2D diagrams fail miserably to explain what is happening in three dimensions.

Inverting and reverting always gives an image that isn't reflected, but the angle of rotation depends on the positions of the mirrors/prisms relative to the direction you're viewing from. For example, a vertical Porro prism appears to rotate the image 180° (and reverses the direction of the light), but a horizontal Porro prism doesn't appear to rotate the image at all. This is highly counterintuitive but it seems to be correct. See the image below, and always imagine you're viewing the light coming towards you.