Odd diffraction spikes

[Swillis]

Hi,

I got out for three nights in a row with with my (new to me) Samyang135 (+600d on SA2i). All went well and I got just over 2hours of 60sec subs. Stacked and processed in Siril.  After processing I noticed a couple of the brightest stars (near the edge) have odd diffraction spikes. 

I was using a 49mm stepdown ring so was not expecting any spikes at all. The lens itself was set at F2

The issue appears that it may be mainly near the edges but hard to tell as the brightest starts are at the edge. 

Whole image for inspection, but needs some better processing (something to do while it's so cloudy I guess)

Thanks

Simon

[pipnina]

If it's only happening at the edges, then it seems likely to me that edge-of-field illumination is partially going past the edge of the iris aperture blades, which would make the aperture it's looking through no longer well-rounded and induce spikes in the image.

I'm not familiar with this lens though, what do star shapes look like when stopped down, say to f5.5 or f8?

[ollypenrice]

135/49=2.7. It sounds as if this will be close to your true F ratio. I'd say the lens was doing pretty darned well at that and a bit of star-rounding in Photoshop will sort out most of the problem - if it qualifies as a problem.

Olly

[vlaiv]

It is interesting that it only shows on one side of image.

Stars on left side, although bright enough - don't show this effect.

This usually points to cause being closer (to focal plane) than actual aperture iris.

When iris is the cause - stars are affected all over the FOV - much like spikes from newtonian. When whole field is affected - that means cause is in collimated beam. When only part of image is affected - cause is in converging beam.

Something similar happens when OAG prism protrudes on one side of the FOV - then only stars on that side of FOV get single spike due to this.

My guess is that it is actually a hair or something similar - on the back side of the lens - on one side of it.

[Swillis]

Thanks for all your replies. 

2 hours ago, vlaiv said:

interesting that it only shows on one side of image.

The image was cropped more on the bottom and left, so might explain this. I will have to go back a check the uncropped image. If only my original framing of the target was better...

But I will check the back of the lens just in case. 

3 hours ago, pipnina said:

edge-of-field illumination is partially going past the edge of the iris aperture blades

This was one thing I considered, but I believe others are using similar stepdown rings, but could be using smaller sensors. I will see how it goes next time. 

2 hours ago, ollypenrice said:

I'd say the lens was doing pretty darned well at that

It was doing very well in deed, I am very pleased with it. 

Thanks

Simon

[StuartJPP]

On that specific star it looks like the same issue as this to me.

 

 

 

However the rest of the field looks pretty free from the same artefact (though the other dimmer star not far away and close to the edge appears to be starting to have the same issue).

It is the brightest star and is at the fringes so maybe the mirror box or something else is causing an obstruction?

 

[Swillis]

9 hours ago, StuartJPP said:

On that specific star it looks like the same issue as this to me

Thanks, could it be the step down ring that's causing it then?

I will have to see if it happens again. I have checked the lens and camera, and can't see a stray hairs sticking anywhere.

I reprocessed the data and only cropped the very edge out this time

There are a few stars at the bottom which have a strong vertical spike, and two fainter spikes in a cross.

Thanks again

Simon 

[Swillis]

I took a look through the lens with the stepdown ring on and I can see the effect described above

On 28/09/2022 at 11:17, StuartJPP said:

On that specific star it looks like the same issue as this to me.

 

 

 

However the rest of the field looks pretty free from the same artefact (though the other dimmer star not far away and close to the edge appears to be starting to have the same issue).

It is the brightest star and is at the fringes so maybe the mirror box or something else is causing an obstruction?

 

So convinced this was the cause, I did a bit of a test the other night, single image with and without the stepdown ring. 

The image shows one star from the top left (upper images) and one star from the bottom right (lower images)

The effect appears different in each corner, but the stepdown ring doesn't appear to be the cause. I should have tested the lens at f2.8, so the aperture was similar but that would have meant removing everything as my mounting bracket obscures the aperture ring. (Maybe one for the next test)

I guess Vlaiv might be on the right track with something closer to the sensor. If the effect was caused by a hair would this restrict it to one side of the image rather than all the edges?  Could it be something with the construction of the camera?

Thanks again

Simon 

[vlaiv]

7 minutes ago, Swillis said:

If the effect was caused by a hair would this restrict it to one side of the image rather than all the edges?  Could it be something with the construction of the camera?

Yes - that's why I thought it could be the problem.

Here is a little diagram showing what is happening:

If obstruction is in region where light beam is collimated (where aperture sits) - then it will affect whole field as it will block some of light coming in at any angle. That is represented by upper blue line in diagram - it intersects beams at both angles.

Once beam starts to converge after the lens, then blockage that is causing diffraction can impact only some part of the field - depending where object that is blocking the light is. In above image - blue line is only intersecting light rays converging into right part of FOV - so only stars in right part of the FOV will be affected - but stars in left part of the FOV won't experience this as light is not "touched" by object.

[Swillis]

Thanks @vlaiv for the description. That makes sense.