One good corner when not using a reducer - How come?

[swag72]

I have come up with the two pictures below to look at the corners of my images.

The information regarding these images is as follows:

60s subs guided on an HEQ5. Skywatcher Evostar 120ED scope and a modded 1000D DSLR. For the purposes of this there has been NO REDUCER / FLATTENER used and in order to attain focus I have used a 50mm extension tube.

I am NOT using a reducer as I wanted to see the results on the corners of not using it. I have tried another camera and get exactly the same results, so I'm content that the camera is not causing this phenomena.

The first image shows a shot with the camera orthogonal in the imaging train. You can see the good stars at the bottom left and the 3 other bad corners.

The second image has been rotated by 180 degrees and the good corner has moved accordingly.

What I don't understand is how, with the reducer removed I have ANY good corners at all.

Can anyone suggest how I can get 3 poor corners and 1 good one when not using a reducer?

[Steve]

Have you checked your focuser's alignment? It is easily done with a laser and some paper or acetate over the objective with a clearly marked centre. Put the laser in the focuser then switch it on, the spot should exit the objective at it's centre. I am guessing it doesn't. Loosen the focuser's three fixing screws then manouver it til the spot is on the centre, then tighten. Easy

Probably a good idea to adjust your focuser's tension first to minimise the possibility of the camera's weight pulling the drawtube off-axis.

Like you say, you won't achieve abberation-free corners without a flattener when using a DSLR but you should at least be able to centre it so the corners are equal. You will probably get away without using a flattener with the 314L's smaller sensor.

HTH

[swag72]

Cheers for that idea Steve. Would an out of line focuser cause this problem then? Any way of testing it without a laser as I don't have one out here in Spain. Just a normal laser type of thing or a collimator?

[Merlin66]

Sara,

We discussed this previously....

A collimator type laser is much easier to use.

(you can make your own but it's a bit of fiddle fart...)

[swag72]

So looks like I will need to bit the bullet and get a collimator. Ken, in your experience, would a out of kilter focuser produce this effect?

[Merlin66]

Can't be 100% sure, but getting the focuser square and collimated to the objective will certainly make everything else easier to control.

( I use a two stage process involving a laser collimator and a Cheshire eyepiece....)

HTH

[swag72]

Is there any sort of test I can do without a laser to test it at this stage?

[ollypenrice]

I think Steve's idea is worth a go.

I use a collimatable laser from Astro Engineering. To collimate it you make a set of vee blocks in which it sits. Don't panic, the Vee blocks just consist of two pairs of 6 inch nails knocked into a block of wood. Each pair is knocked in at 45 degrees to the surface and they cross over in an X shape. The top of the X is the V which cradles the laser. Sit it on the vees and shine it at a well distant wall. Mark the point of contact. Rotate the laser and adjust the collimation screws till rotating the laser has no effect on the point of contact of the beam on the wall.

Steve may have lasers that don't need collimating, though?

I think it will be the focuser for two reasons. The first is that it has been changed and such operations involve the possibility of error.

My other reason for thinking Steve is right is this. For the sake of argument we have the camera horizontal/vertical during the shoot. Whatever part of the sky is captured on the top left of the chip is the good corner. Now we rotate the camera 180 deg and reshoot. Now the good corner is still whatever is captured on the top left. In other words the good corner is defined by the sky, not by the camera orientation. The good side changes when you change the camera angle. If that is the case, as I understand it to be, then the problem must lie at the optical interface of incident beam and camera. If the problem lay within the camera then the good side would follow the camera. But it doesn't.

Yes, I think it's the focuser off axis. If its adjustable, and it seems to be, you will soon be sorted.

Can you do it without a laser? Don't know, but what about this; do a defocused star test in corner one. How circular is the diffraction pattern? Move star to corner two. Ditto. Go round the 4 corners.

Go to the worst corner and adjust the focuser till the star test is better.

Repeat the 4 corner tests and retry, making iteratively smaller adjustments.

While this sounds theoretically OK to me it might just not work in practice, I really don't know.

Why is one corner OK? Problems are like that! I guess the correct optical centre of the beam is striking the chip somewhere between the centre and the good corner. The pixel resolution gives you a certain focal leeway and this leeway is exceeded on the three corners furthest from the axis of the incident beam.

Olly

[Merlin66]

Sara,

Don't have a spare laser, but I can lend you a Cheshire....PM me if it would help.

[swag72]

The good side changes when you change the camera angle. If that is the case, as I understand it to be, then the problem must lie at the optical interface of incident beam and camera. If the problem lay within the camera then the good side would follow the camera. But it doesn't.

But surely the good side DOES follow the camera in this case?

My good bottom right corner goes to the top right corner after the 180 degree rotation. The camera has been rotated by 180 degrees and so the corner moves.

I am sure that I am missing something very fundamental here, but if the focuser was not square for example and this was the issue, then when I rotated the camera then the good corner would remain in the same place.

[Merlin66]

Just to confirm what you're saying....

If the bottom right was good....and you rotate the camera 180 degrees, this corner would then be up in the top left...diagonally opposite....is this the case?

[swag72]

Yes, in the images posted above, this IS the case - the bottom left in the first image is good. That is with the camera sitting square in the OTA. So I rotate the camera about 180 degrees and in this case (image 2) the good corner has now moved to the top right. So the good corner has moved WITH the camera.

Surely that rules out the focuser and OTA?

[Merlin66]

Hmmmm

I need a cup of coffee and a clean sheet of paper....Let me give it some serious thought....

[steppenwolf]

The good corner has not moved with the camera. The good corner has stayed in exactly the same place on the focal plane but as the camera has turned and thus the focal plane has turned the 'good' light is now falling on a different set of pixels on the sensor. Thus, the camera is not at fault, it must be the focuser or the optics.

[swag72]

I will agree with you, but I don't understand why!! In my mind and with the help of a camera infront of me, turning it through 180 degrees ............. Grrr. I will just take your word for it!!

So has that gone some way to explaining why I have a good corner without a reducer, that incidently moves as the camera is rorated!!

Answers in thread, yes or no!!!!

Girls and technology spring to mind!! Certainly in my case anyway!!

[ollypenrice]

These things are murder! Here's next week's POW for SURE so the rest of you can FORGET IT, right??!! Ladies and gentlemen, the little known 1234 Nebula...

Large pic is nebula on sky. Small one is chip. Left hand side is pre-rotation, right hand side is with camera rotated.

It confirms Steve's verbal description, methinks. If the problem were in the part that rotated, the good part would rotate with it and 1 would remain sharp. I think Steve FLO is on the money.

Edit; CURSES, I have put top and bottom the wrong way round on the right hand image. The upside down top shoud be at the bottom but it will take too long to correct. I dare say some forty hour version of the Soap Bubble Nebula in a 24 inch RC might squeak past me now for POW. Glum scowl...

Olly

[swag72]

I must be the most STUPID and THICK person on the forum. Thanks for the drawing Olly, but what I am seeing is different to your drawing.

1 is sharp in the first image and 1 is STILL sharp in the rotated image. My sharp corner HAS NOT stayed in the same place as your drawing suggests. That drawing makes PERFECT sense to me, but is not a reflection of what I am seeing, surely. When I rotated the camera, the good corner has rotated with it. If I don't understand it from the next reply, I will give it up as a bad job and not try to understand the process any more and accept what you say as gospel!! :confused:

Grrr, my head hurts!!

Forget POW Olly, incidently I think this is an excellent candidate, but how about a DOW (Dunce of the week) award?!!

[steppenwolf]

How about this diagram which shows how I see it. Note how the four stars at the bottom left in the original image are nice and round and continue to be in the rotated version because the camera is not at fault (i.e. it doesn't have a tilted sensor) and the light falling on the position of the focal plane remains the same but the sensor has rotated 180 degrees:-

[steppenwolf]

Just to perhaps simplify this, this view shows the sensor actually rotated 180 degrees in the second image:-

[ollypenrice]

I'm lost again! I thought I had it!!

Olly

[barkis]

Surely the focus of the image stays put within the focuser, but as the focuser is misaligned, rotating the camera, and thereby the sensor which is not orthogonal, the image will only focus on the sensor corner in one place in it's 356 degree rotation.

The camera sensor intercepts the only true focus, because it is not rotating in a vertical plane due to the focuser misalignment.

Ron.

[ollypenrice]

Surely the focus of the image stays put within the focuser, but as the focuser is misaligned, rotating the camera, and thereby the sensor which is not orthogonal, the image will only focus on the sensor corner in one place in it's 356 degree rotation.

The camera sensor intercepts the only true focus, because it is not rotating in a vertical plane due to the focuser misalignment.

Ron.

...and if that's the case the true focus will fall on a different part of the sensor when the camera is rotated, but Sara is saying it doesn't.

The confusion couldn't be arising from Sara rotating the images for presetation? Sara?

Olly

[steppenwolf]

I am sure that Sara is not rotating her images for display, she is showing them exactly as they came off the camera both before and after rotation.

My take on this is as follows:-

Focused, 'clean of aberration light' is falling in one place only - the bottom left hand corner of the sensor when it is in its original position. The focused, 'clean of aberration light' falls in the same place behind Sara's telescope irrespective of the angle of the camera or indeed whether the camera is even there or not. Thus rotating the camera to any angle results in this 'good' light continuing to fall in the bottom left of the frame. If the camera is rotated 180 degrees then when Sara downloads her image, the software doesn't know the camera was rotated so the 'good' light now appears to us when the image is displayed to be in the top right of the frame.

[kookoo_gr]

It seems that it is either a slop problem or collimation

I would check for slop when placing the camera at the focuser and if it exists i would buy a hotech sca adaptor (actually i have bought one in order to make sure there is no slop when placing my camera/oag at my focuser)

Also i would check the collimation of the optics and especially the tilt of the secondary mirror, a good tool for checking its collimation is the concenter MkII collimator Spheretec

A good idea as mentioned as above is to check that the focuser's collimation

[Earl]

Can I make a suggestion.

Can you take 4 images of a star field each image been a clockwise 90 degree rotation of the camera.

if the flat area stays in the same corner on each image its a Light path issue, if it rotates with the camera rotation it is the camera, or the connection of the camera to the OTA.