Optical train issue? Odd Shadows - can you figure it out?

[astrorabbit]

Hi Guys,

I've been scratching my head with a fellow star gazer as to where my issue may lie, we just can't figure out where I'm getting these odd shadows in my images from. I'm calling them shadows as they seem to be formed only by bright stars and record as dark spots in my images.

My optical train is pretty simple: SW 200PDS -> LP filter -> Coma Corrector -> Nikon D7100 (stock, not modded in any way)

I have this issue with and without any filters, I've cleaned the optics and sensor in the past with no difference. The issue only shows with really bright stars as they seem to be strong enough for the shadow to record. 

Here's a image processed in such a way to try and clearly show the issue, its a three minute sub at ISO 800, around the brightest stars I get six shadows in a hexagonal shape with a central dark spot too.

Please help!

Thanks in advance for your thoughts

[Merlin66]

Just when you think you've see everything along comes something like this..........

No ideas at this stage, but I'd be interested in the solution.

[astrorabbit]

Just when you think you've see everything along comes something like this..........

No ideas at this stage, but I'd be interested in the solution.

I'm really confused by it, I'm thinking it must be a shadow as it shows as dark spots on the image, the shape of the spots being elliptical make me believe it may be something to do with the secondary mirror however I can't understand why there is a clear hexagonal pattern with the central dark spot.

Hopefully someone has had this issue and cured it before!

[Merlin66]

I think you must have the experts beaten on this one.....no one seems to be able to comment.

[astrorabbit]

I think you must have the experts beaten on this one.....no one seems to be able to comment.

Nope, I posted the pic on astro group I'm in on facebook and had the members stumped there too.

Think I'm going to go step by step taking all the links out of the chain one by one untill I find the issue, I'll try a different camera, coma corrector, filter, clean the optics with brillo pads until something works!

[Merlin66]

Yeah,

A hundred views and no comments.....

[andrew s]

Try a flat field and see what that shows? It may help provide some insight. Regards Andrew

[andrew s]

I would also try a star shot without the coma corrector. You need more info to try to pin it down. Regards Andrew

[denmonkey]

Do the only thing you can, start taking things away from your set up and trying replacements If you can and at least narrow it down to the specific piece of gear causing the issue and go from there. Since it shows in a few places my money is on a reflection from something maybe the filter. Does it happen without it ?

[denmonkey]

Sorry spotted you said no.

[Mukv]

Greetings

not too sure about the cause but looking at the pattern in the bright stars top right the focus points may line up.

not sure of the orientation , I'm on my laptop and don't have a graphics package to try to overlay one over the other. 

Andy

[D4N]

There isn't any auto noise reduction in the camera that's going haywire on the bright stars is there?

It's hard to imagine what on the scope could cause that with it being dark spots and not light spots.

Sent from my iPad using Tapatalk

[ollypenrice]

I can't do any more than pass on a hunch based on a PM I received recently regarding Nikons (about which I know nothing at all.) The cameras do some in-cam processing and you might find a lot of useful info from a group called Nikonhacker. I've informed the person who contacted me about your thread and he might be able to put you onto the trail. It seems that the dark spots are radially placed around a bright source so, personally, I doubt that flats will help, but who knows?

Olly

[Alien 13]

Greetings

not too sure about the cause but looking at the pattern in the bright stars top right the focus points may line up.

nic grid.PNG

not sure of the orientation , I'm on my laptop and don't have a graphics package to try to overlay one over the other. 

Andy

This does look like the pattern in the OPs image i wonder if its light getting through the viewfinder?

Alan

[Merlin66]

Ahhhh

Stray light in the viewfinder .....sounds plausible....

[sharkmelley]

In Photoshop I've overlaid each bright star accurately with using their diffraction spikes and the "shadows" line up perfectly as well - so the pattern is consistent around each star.  The hexagonal pattern is actually an elongated hexagon.

The pattern on the Multi Cam 3500 DX autofocus module doesn't have any obvious hexagonal component to it.  In addition, stray light coming in through the viewfinder wouldn't neatly arrange itself around the bright points (i.e. stars) seen through the lens.  So I can't see how that would be the explanation.

I've never seen anything like this before.

Mark

[Merlin66]

Hmmmm'

(Clutching at straws....)

Running out of ideas.......

[ollypenrice]

So, optical or electronic? I think it could be either. If you rotate the camera what happens to the dark marks? Do they follow the rotation or remain where they were relative to the tube? If they rotate with the camera they are either coming from the camera or from the optics close to it. If they don't, it must be the main optics. Or so I'd have thought.

Olly

[IanL]

One question, think you have said this but will ask anyway.  have you performed any processing on the sample image other than stretching it?  I.e. is it a single image with no darks, bias, flats or stacking applied?  That needs to be your starting point as any number of things could happen with more processing.

If it was a shadow cast by something roughly the size of a dust grain, then it is easy enough to work out the distance using this:

http://www.wilmslowastro.com/software/formulae.htm#Dust

If your sample image above hasn't been re-scaled (i.e. there is a 1 to 1 relationship between pixels in the sample and the original image) the shadows are about 10 pixels wide on average, and assuming your 200PDS is running at f/5, the D7100 has 3.9micron pixels then a small dust-like object casting the shadows would have to be roughly 0.2mm in front of the sensor surface.  Clearly that makes no sense at all, as you have a regular pattern of shadows corresponding with the bright stars, and something that small and close would only cast a shadow in one place on the sensor.

I guess you might have some bigger object in the optical train a bit further away that is conspiring to produce a dark shadow pattern like the one in the sample, but again I'd expect it to be a single shadow not multiple copies corresponding to stars.  Once you start getting out in to the realms of the secondary mirror there is nothing that could produce a defined shadow in an image - you can wave your hand around in front of the OTA or obscure part of the secondary and all you're going to see is an overall darkening of the image as any obstruction that far out will out of focus.

So what are you left with?

1. Diffraction could produce a symmetrical pattern of some sort.  It doesn't have to be the classic spike pattern from a set of secondary vanes, it could be something different depending on the shape of the obstruction.  But of course you'd expect it to appear as a bright pattern not a dark one.

2. Reflection artefacts - could be some thing bouncing around in the optical train, especially if the viewfinder and/or AF points get involved as suggested above. In terms of the pattern produced this does seem credible.  If each bright star was producing a reflection from something you'd expect it to line up around the star itself like you have.  But again if it is reflection you'd expect bright artefacts not dark ones.

3. On camera processing of some sort?  Well I can't think why processing would produce a pattern of that sort in and of itself, but it could be interacting with 1. or 2. above to excessively darken bright spots relative to other bright areas (the stars).  By way of analogy you can get all sorts of odd things going on with badly made or badly applied flat frames and create bright artefacts of dark ones instead of eliminating them, so I wonder if there is some processing that is going wrong on camera (or off camera if it turns out that you've done more to the image than just stretching it).

I concur that you need to do a few more tests.  Certainly try rotating the camera relative to the scope and target if you can and see if the pattern changes.  At the very least you'd then have a way of fixing it - take a bunch of frames at one orientation and then a bunch of others at a few slightly different orientations and align/stack them together.  Provided the dark spots don't overlap, they'll get rejected if you have enough frames with different orientations.  You'll have to crop it a bit (or a lot) depending on how much you rotate as there will be outer regions with insufficient overlapping frames, but it would give you something to use.

You might also try taking a few images of a bigger/bright target against a dark background (e.g. a lightbulb) and see if you get any similar patterns on a larger scale.

[sharkmelley]

Is there anything in the optical train with a hexagonal layout e.g. mirror clips on the primary or secondary, any other support with bolts in hexagonal orientation.  Same question with the focal reducer.  Or anything inside the camera or the camera attachments?

Mark

[happy-kat]

Use a different camera entirely but with all else identical to rule in or out the Nikon.

[tekkydave]

Use a different camera entirely but with all else identical to rule in or out the Nikon.

Definitely the first thing to try. If still there then remove anything else in the optical path you can but one at a time.

[tekkydave]

Could it be some kind of defect within the sensor chip itself?

[Uranium235]

Use a different camera entirely but with all else identical to rule in or out the Nikon.

Its the only way to be sure.

[happylimpet]

Its a processing artifact - either in camera or in software. Nothing else can make dark spots.  Is your software accidentally using light frames as flats?  Or lights as darks?

Either that or something in the camera.  Turn off all the clever assed noise reduction etc.  Personnally, my moneys on the camera , as you referred to the image as a sub so I assume unprocessed.