Some Lunar Images from April 23

[Roy Foreman]

Not 100% happy with some of these as I have been battling with so-so seeing, but here they are for those that would like to see them.

16" F/4.5 Newt,  ZWO 183 MM, Red Filter, 15% pf 1000 frames at 25 fps

[Stu]

They look excellent to me Roy, you’ve done very well if the seeing wasn’t great. Lovely detail.

[teoria_del_big_bang]

Some great images  🙂

Steve

[vineyard]

I wouldn't grumble about any of those! Lots of lovely detail - you're doing that aperture justice.

[Roy Foreman]

2 minutes ago, vineyard said:

I wouldn't grumble about any of those! Lots of lovely detail - you're doing that aperture justice.

Thank you so much for your vote of confidence !  Maybe I'm being too harsh on myself,  but I always get the feeling that I could have done better.  And I will keep trying to do better if I can.  Seeing permitting of course !

Glad you like the images.

Roy

[vlaiv]

2 hours ago, Roy Foreman said:

at 25 fps

Why so low FPS?

What was your exposure length?

[Roy Foreman]

4 minutes ago, vlaiv said:

Why so low FPS?

What was your exposure length?

 

15 minutes ago, vlaiv said:

Why so low FPS?

What was your exposure length?

Hi Vlaiv

The ZWO 183 is a 20 megapixel camera with a maximum frame rate of 19 fps if you use the full imaging area, which I normally do.  For the images above I reduced the imaging area slightly which allowed 25 fps at around 14 ms exposure time. This was to try and reduce the effects of the not so good seeing.

The following night I experimented with imaging at prime focus and an imaging area of 1600 x 1200 px ( Native is around 5300 x 3500 ish ) and this allowed 63 fps at around 6 ms exposure time, making use of the tiny 2.4 micron pixels. I am processing these at the moment, but the results don't look as good as I usually get,  particularly with respect to sharpening.

I know higher frame rates are usually preferred, and I am starting to experiment with different setting to find an optimum.

I am, of course, open to suggestions if you have any.

[vlaiv]

27 minutes ago, Roy Foreman said:

 

Hi Vlaiv

The ZWO 183 is a 20 megapixel camera with a maximum frame rate of 19 fps if you use the full imaging area, which I normally do.  For the images above I reduced the imaging area slightly which allowed 25 fps at around 14 ms exposure time. This was to try and reduce the effects of the not so good seeing.

The following night I experimented with imaging at prime focus and an imaging area of 1600 x 1200 px ( Native is around 5300 x 3500 ish ) and this allowed 63 fps at around 6 ms exposure time, making use of the tiny 2.4 micron pixels. I am processing these at the moment, but the results don't look as good as I usually get,  particularly with respect to sharpening.

I know higher frame rates are usually preferred, and I am starting to experiment with different setting to find an optimum.

I am, of course, open to suggestions if you have any.

I would say tat 14ms is too high for most seeing conditions with aperture that large.

One of things we try to do in lucky imaging approach is to freeze the seeing. In order to do that we have to set our exposure at or below coherence time for given seeing. If we don't do that, changing seeing effects cause "motion" type blur on top of seeing distortion.

In most cases, 8" of aperture has coherence time of about 5-6ms. I'm guessing that larger aperture will have shorter time than that (but again - that depends on seeing on particular night).

ASI178 has 2.4µm pixel size and that translates into F/7.4 for red part of spectrum (at 650nm) for critical sampling. What F/ratio are you using for your scope? Going for longer FL than F/7.4 does not yield any additional detail.

Also - are you using coma corrector?

Fast newtonian is going to have very small diffraction limited field, depending on barlow used - you might not be able to fully exploit 9mm diagonal of ASI178. In that case - it is much better to use ROI and limit recording to diffraction limited part of FOV.

 

[CraigT82]

20 minutes ago, vlaiv said:

In most cases, 8" of aperture has coherence time of about 5-6ms. I'm guessing that larger aperture will have shorter time than that (but again - that depends on seeing on particular night).

Is there any kind of existing data table of coherence times? Maybe with aperture down one side and seeing level across the top?  Would be V.useful!

 

[Roy Foreman]

Vlaiv, you are absolutely right about exposure time and coherence. When I imaged the moon with my DSLR I came to the conclusion the 1/250 sec (i.e. 4 ms) was sufficient to freeze the seeing, and I am aware that this is what I should ideally be aiming for.

The images in this post were taken using a Baader FFC at 3x which incorporates a coma corrector, and results in F/13.5 - hence the longer exposure time.  This is why I have been experimenting with prime focus imaging at F/4.5 ( with coma corrector) and a smaller imaging area. Cannot draw any conclusions yet.

You are right when you say going longer than F/7.4 does not yield additional detail. I have found this also when comparing 3x images with prime focus ones. However the magnified images produce a finer tonal range, I think.

Thanks Vlaiv for you input.

[vlaiv]

43 minutes ago, CraigT82 said:

Is there any kind of existing data table of coherence times? Maybe with aperture down one side and seeing level across the top?  Would be V.useful!

I would also like something like that, however, now I can't find any articles on the topic that are not highly technical in nature.

What I've written above, I've read couple of years ago when I researched planetary imaging. I can't remember where.

It is mentioned all over the place - for example here:

https://ui.adsabs.harvard.edu/abs/2016ASSL..439....1B/abstract

Quote

Unlike adaptive optics, Lucky Imaging is a passive observing technique with individual integration times comparable to the atmospheric coherence time.

Here are some interesting starting points for research:

https://en.wikipedia.org/wiki/Greenwood_frequency

(defines minimum time in which atmosphere is frozen - important in adaptive optics systems)

https://en.wikipedia.org/wiki/Fried_parameter

Defines "size" of turbulent cells

In this wiki article:

https://en.wikipedia.org/wiki/Astronomical_seeing

There is this quote:

Quote

Another cheaper technique, lucky imaging, has had good results on smaller telescopes. This idea dates back to pre-war naked-eye observations of moments of good seeing, which were followed by observations of the planets on cine film after World War II. The technique relies on the fact that every so often the effects of the atmosphere will be negligible, and hence by recording large numbers of images in real-time, a 'lucky' excellent image can be picked out. This happens more often when the number of r0-size patches over the telescope pupil is not too large, and the technique consequently breaks down for very large telescopes. It can nonetheless outperform adaptive optics in some cases and is accessible to amateurs. It does require very much longer observation times than adaptive optics for imaging faint targets, and is limited in its maximum resolution

In the article I remember reading - there was relationship between t0 and r0 that depended on wind speed at altitude. It calculated time it takes for average seeing cell traveling at usual wind speeds to cross particular aperture - and then they related that to t0 - time that the seeing distortion is relatively constant.

[CraigT82]

Perhaps this page is the one you looked at...

http://slittlefair.staff.shef.ac.uk/teaching/phy217/lectures/telescopes/L10/index.html

It has a very neat explanation to why smaller apertures 'cut through' seeing: Essentially the smaller the aperture the smaller the section of the distorted wavefront it samples, and the smaller the section sampled the more planar it appears to the instrument. Brilliant! 

[vlaiv]

3 minutes ago, CraigT82 said:

Perhaps this page is the one you looked at...

http://slittlefair.staff.shef.ac.uk/teaching/phy217/lectures/telescopes/L10/index.html

It has a very neat explanation to why smaller apertures 'cut through' seeing: Essentially the smaller the aperture the smaller the section of the distorted wavefront it samples, and the smaller the section sampled the more planar it appears to the instrument. Brilliant! 

Explanation is the same, although I can't remember the page itself - it is good example of a brief explanation of phenomena involved.

Does not answer our question though - what are average exposure times that are needed to freeze the seeing - but it does point out dependence on used wavelength - and nicely explains why jet stream is responsible for such a poor seeing - due to wind speeds involved - it shortens coherence time considerably.

There is also dependence on scope size vs r0. As you say - when scope aperture is less than r0 - dominant wavefront aberration is tilt - image stretches and deforms rather than blurs.

It is also interesting to relate level of distortion in video like above to density of alignment points. Alignment points need to be dense enough to properly reshape back distortion created.

[jetstream]

@Roy Foreman nice images. I might try my 15" for lunar- do you use an EQ tracker?

[jetstream]

1 hour ago, vlaiv said:

Alignment points need to be dense enough to properly reshape back distortion created.

This is very interesting- is there some measure of this density? just use more-than?

[vlaiv]

1 minute ago, jetstream said:

This is very interesting- is there some measure of this density? just use more-than?

If I remember correctly - AutoStakkert has feature to automatically add alignment points. This is very handy for lunar shots as number of alignment points tends to be rather large.

Here is a screen shot from another thread here on SGL

There are rather big alignment points. Distance between alignment points is roughly the half of side of the square. If you look at above gif that I posted - whole gif could be covered with maybe 4 alignment points of above size - but in reality, we would need about 50-100 alignment points for that gif alone since size of tilt deformation is small.  You can select alignment point size in AutoStakkert - by selecting number of pixels.

Here is diagram roughly explaining how to select proper alignment point size:

If you have a feature in your movie that bends and stretches so that in one frame - looks like elongated blob in the top, in some other frame looks "normal" and in a third frame - it looks squeezed - you need to cover size of that feature with 2 alignment points rather than one (left part of diagram - not right).

In right part of diagram - large alignment point will stack complete feature as single thing and you'll get the blur of it shrinking and extending.

In left part of diagram - each alignment point will "figure out" that in top part - stretched blob - left and right sides are further than they should be - that in the middle part they are roughly where they should be (mean position over number of frames) and that in the bottom part - they are squeezed - and it will try to correct distortion.

You need roughly two alignment points per "wavelength" of deformation

If you look at above gif again - you'll see that larger crater is deformed but smaller is just bouncing around staying roughly the same shape:

In that image - I would try to make alignment points be roughly the size so that at least couple cover that larger deforming crater.

This is all related to that isoplanic angle from above linked lecture - alignment point needs to cover number of pixels that add up to isoplanic angle.

 

[Roy Foreman]

42 minutes ago, jetstream said:

@Roy Foreman nice images. I might try my 15" for lunar- do you use an EQ tracker?

Thank you - glad you like the images. Several people have asked about my 16" Newt. The OTA is from a Skywatcher Flextube Dob, and is mounted on a fork mount that I largely built myself - apart from worm wheel sets and electronics. Motors are connected directly to worm shaft, so no gearbox or belts. RA wheel is 14", Dec is 12". I have recently fitted a couple of large springs to the Dec worm housing to reduce backlash cause by thermal movement of the components. If it works well then I shall try and tackle the RA too.

[Roy Foreman]

2 hours ago, CraigT82 said:

Perhaps this page is the one you looked at...

http://slittlefair.staff.shef.ac.uk/teaching/phy217/lectures/telescopes/L10/index.html

It has a very neat explanation to why smaller apertures 'cut through' seeing: Essentially the smaller the aperture the smaller the section of the distorted wavefront it samples, and the smaller the section sampled the more planar it appears to the instrument. Brilliant! 

I have heard this many times before - large apertures are more affected by seeing than small ones. I have a Skywatcher 120 ED Equinox riding piggyback on the 16" so when I get the chance I will try and do a comparison. I could image with both at the same time but that would mean using different cameras, or do one immediately after the other. From the images I already have the 120 ED ones seem to be not far short of the 16", which bears out what was said in your post.