Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Refractor or Reflector for most optical resolution

astrophotography equipment imaging reflector refractor
  • Please log in to reply
38 replies to this topic

#26 Astro_Francis

Astro_Francis

    Lift Off

  • -----
  • topic starter
  • Posts: 4
  • Joined: 14 Dec 2017

Posted 05 September 2018 - 08:44 PM

Thank you for the responses. I completely agree that the full moon is not the most interesting phase of the moon. However, the camera owner is interested in capturing images of the full moon. It's a prototype camera and they want to show the camera's high resolution/pixel count. The goal is to get the most "badass" pic of the full moon with the camera, with emphasis on being able to resolve the high res image when zooming in. I see from the responses that may not be possible, regardless of telescope. We are planning to take the shots outside of Las Vegas to have the best seeing conditions. My first assumption was that my f/5 8" reflector would not have enough resolving power to utilize their camera. Thank you for the clarification on pixel size (in microns) x 5 = optimal focal ratio. Is the idea that a powerful telescope would resolve a distant star or moon feature so sharply that the photosite would not pick it up? It seems like more more aperature at a given focal length would yield a sharper image, even when undersampling. Is that false? How does binning come into play?

#27 rgsalinger

rgsalinger

    Fly Me to the Moon

  • *****
  • Moderators
  • Posts: 5108
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 05 September 2018 - 08:47 PM

Most of these comments are really interesting but I'm afraid that many of you are forgetting the seeing when making your recommendations. It's fine to be able to capture all of the "information" by using tiny pixels. Unfortunately, that is never ever going to be your limiting factor.

 

Last Saturday night I'm using two telescopes. One is 6" and one is 25" in diameter. The smaller one is refractor, the larger one a RC. The difference in resolution, measured by fwhm, was around .3 arc seconds.Last Saturday it was 2.2 to 1.9. The two cameras, although very very different, have very close to the same sampling rate - around .8 arc seconds per pixel. What's going on here is that the seeing is around 1.8 seconds, at best, where I image. So, I can get more aperture and smaller pixels until doomsday but there's never going to be much difference in "resolution" because of the seeing. 

 

The OP has already chosen his camera. The best thing for him to do is to find a scope that comes close to the optimal sampling as predicted by the Nyquist limit. That's conventionally equal to 1/3 of the local seeing. Anything smaller than that will yield no additional detail. Before buy a scope read the imaging book, please.

 

 

Rgrds-Ross 



#28 RedLionNJ

RedLionNJ

    Soyuz

  • *****
  • Posts: 3693
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted 05 September 2018 - 09:30 PM

Thank you for the responses. I completely agree that the full moon is not the most interesting phase of the moon. However, the camera owner is interested in capturing images of the full moon. It's a prototype camera and they want to show the camera's high resolution/pixel count. The goal is to get the most "badass" pic of the full moon with the camera, with emphasis on being able to resolve the high res image when zooming in. I see from the responses that may not be possible, regardless of telescope. We are planning to take the shots outside of Las Vegas to have the best seeing conditions. My first assumption was that my f/5 8" reflector would not have enough resolving power to utilize their camera. Thank you for the clarification on pixel size (in microns) x 5 = optimal focal ratio. Is the idea that a powerful telescope would resolve a distant star or moon feature so sharply that the photosite would not pick it up? It seems like more more aperature at a given focal length would yield a sharper image, even when undersampling. Is that false? How does binning come into play?

The basic limitations are two-fold.  One is the "seeing", or atmospheric steadiness (this actually includes the air inside the scope, as well as the mirror or lens components and surfaces). Very, very gradual temperature or humidity changes along the optical path lead to better seeing than abrupt ones. Abrupt changes tend to lead to 'roiling' effects, which blur the inbound light.

 

The other basic limitation is the wave nature of light. These waves may seem exceptionally small, but they're still appreciable in size - and equate to roughly the smallest piece of information you can capture. At visual light wavelengths (say, 550nm), your resolution will be HALF that which could be achieved at a wavelength of 275nm.  But to filter out longer wavelengths can get very expensive and problematic AND there is a very high likelihood any given sensor is only sensitive over some range from maybe 400nm to 900nm.

 

So you can't simply "magnify indefinitely" (for lack of a better term) and expect to get more and more detail out. For a 12-inch aperture, for example, I image at an optimum f/14.5 for my 2.9-micron pixels and I can never get any resolution finer than about 0.13 arcsec.  And I'm fine with that, as I get it maybe once a year (thanks to seeing).

 

If we were to assume the full moon is 30 arcminutes across, then you could get nearly 14,000 pixels of resolution out of it, with ideal equipment (at 0.13 arcsec per pixel).  But due to limitations mentioned previously, this wouldn't be possible in practice.



#29 Tom Glenn

Tom Glenn

    Surveyor 1

  • -----
  • Posts: 1939
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 05 September 2018 - 10:06 PM

Most of these comments are really interesting but I'm afraid that many of you are forgetting the seeing when making your recommendations. It's fine to be able to capture all of the "information" by using tiny pixels. Unfortunately, that is never ever going to be your limiting factor.

 

Last Saturday night I'm using two telescopes. One is 6" and one is 25" in diameter. The smaller one is refractor, the larger one a RC. The difference in resolution, measured by fwhm, was around .3 arc seconds.Last Saturday it was 2.2 to 1.9. The two cameras, although very very different, have very close to the same sampling rate - around .8 arc seconds per pixel. What's going on here is that the seeing is around 1.8 seconds, at best, where I image. So, I can get more aperture and smaller pixels until doomsday but there's never going to be much difference in "resolution" because of the seeing. 

What type of imaging were you doing, deep sky or planetary/lunar?  I see you are in Carlsbad, which is not that far from me, but at my house, I have a couple of dozen days per year in which a 25 inch scope would produce lunar orbiter quality images and blow a 6 inch scope out of the water.  And with my 9.25 inch scope, I have many examples on the moon in which I was not limited by seeing, but rather am at the limit of my scope's resolving power.  So seeing is usually the limiting factor, but not always, depending on your location.  The seeing is also heavily influenced by local factors, such as houses, concrete, etc, as well as altitude above the horizon.  So what might be good at one location will be bad just several hundred yards away.  I see that the OP is planning to image outside of Vegas, which unfortunately, as a desert climate, will typically not have very good seeing.  



#30 moonwatching ferret

moonwatching ferret

    Mercury-Atlas

  • *****
  • Posts: 2856
  • Joined: 26 Jun 2016
  • Loc: sebastian fl

Posted 05 September 2018 - 10:11 PM

imaging a full moon you can do with a decet planetary imaging cam with a large sensor. seems to me th best full moon images here were done with large instuments doing mosaics which is a talent in its own right trying to get all your images without changes to the shadows on the moon then processing everything to match. I would just stick to a decemt planetary cam and a decent size reflector I would post some of my images but at the scale i image the moon at it would take a hundred avi to do a full moon unless i just use my asi 224 and the 16 incher at f7



#31 moonwatching ferret

moonwatching ferret

    Mercury-Atlas

  • *****
  • Posts: 2856
  • Joined: 26 Jun 2016
  • Loc: sebastian fl

Posted 05 September 2018 - 10:14 PM

Thank you for the responses. I completely agree that the full moon is not the most interesting phase of the moon. However, the camera owner is interested in capturing images of the full moon. It's a prototype camera and they want to show the camera's high resolution/pixel count. The goal is to get the most "badass" pic of the full moon with the camera, with emphasis on being able to resolve the high res image when zooming in. I see from the responses that may not be possible, regardless of telescope. We are planning to take the shots outside of Las Vegas to have the best seeing conditions. My first assumption was that my f/5 8" reflector would not have enough resolving power to utilize their camera. Thank you for the clarification on pixel size (in microns) x 5 = optimal focal ratio. Is the idea that a powerful telescope would resolve a distant star or moon feature so sharply that the photosite would not pick it up? It seems like more more aperature at a given focal length would yield a sharper image, even when undersampling. Is that false? How does binning come into play?

I would say a half moon or gibbous phase moon would be much better to get deets in the terminator im sure you coud use the cam through a reflector. I knoew refractors are nice but I would just leave them for what there best for and thats dso work . heres a tycho with my 16 at f21

2018-08-22-0315_6-L_AS_f5000_g6_ap104_conv.jpg thats high resolution


Edited by moonwatching ferret, 05 September 2018 - 10:17 PM.


#32 rgsalinger

rgsalinger

    Fly Me to the Moon

  • *****
  • Moderators
  • Posts: 5108
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 06 September 2018 - 01:47 AM

Of course if the plan is to use the camera to take very fast images of the moon, then you can get under the seeing some of the time. My fault here for being focused on DSO's in this forum which is about solar system imaging. In that case, then, the bigger the better is correct. I've not tried to do any planetary to date with the 25" but we do have a suitable camera so I'll have to try it sometime in the future.

Rgrds-Ross



#33 Astro_Francis

Astro_Francis

    Lift Off

  • -----
  • topic starter
  • Posts: 4
  • Joined: 14 Dec 2017

Posted 06 September 2018 - 01:34 PM

So it would seem that the consensus is that a reflector would be better for this project than a refractor. I am still a little confused as to why a telescope should not be faster than 5x the pixel size in microns. The camera has a pixel pitch of about 2 microns. I currently have an 8" F/6 Orion XT8. Would upgrading to a Skywatcher Quattro 12" F/4 increase the resolution of the images? Or is the idea that the camera would not be able to detect the increase in optical resolution? Thanks all



#34 Tom Glenn

Tom Glenn

    Surveyor 1

  • -----
  • Posts: 1939
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 06 September 2018 - 02:33 PM

Think about it this way.  Light passes through the telescope aperture and an image is brought to focus at the image plane, which is coincident with the camera sensor.  The image here is a "real" image.  You could place a pice of paper, or your hand, in the light path at the image plane and see the image of the moon projected onto it.  Instead of that, of course, we place a camera there to capture the image.  The amount of information present at the image plane is dependent on the telescope aperture.  But how much of this information gets transferred to the sensor depends on how many pixels it is spread onto.  To use an extreme example, imagine a camera sensor that has only 288 pixels (this is a random number but allows me to do the following simulation).  I took one of my moon images that was posted above and used a Photoshop filter to simulate what this image would look like if extremely pixelated.  The following result is what the image would look like if it was captured on a 16x18 pixel grid.  This was then resized so that you can actually see such an image.  The entire moon was present at the image plane of the telescope, in very high detail, but here it was only captured on 288 pixels, so the result is that almost no information is conveyed.  

 

moon_pixelated.jpg

 

Every telescope has a resolving limit that is only dependent on the aperture.  The relevant values are the Dawe's limit and the Rayleigh limit.  In order to capture this level of resolution on a camera sensor, you need to be in the general vicinity of spreading out the resolving limit of the scope over about 3x3 pixels.  You can change the amount of pixels the information is spread over by adjusting the focal length of the scope with barlow lenses.  You can use a variety of calculators to determine the "sampling rate" of your optical system, but it also turns out that the 5x pixel rule yields approximately the appropriate F ratio that you want for maximum resolution. 

 

All that  said, it sounds like you and the camera owner may have unrealistic expectations.  It sounds like he or she wants you to use your telescope to showcase the ability of the camera with a high resolution image of the moon, but neither they nor you have any experience with lunar imaging.  The camera is actually the least important aspect of high resolution imaging, and in fact, spectacular high res images can be obtained with very cheap cameras.  Also, you wouldn't actually be taking a single image, but rather would have to collect a large volume of video data to process into the final image.  While it's very easy to take a simple photo of the moon, taking a large scale, high resolution image is a much more advanced topic.  You will find examples on this forum, including several I have taken in the past year, but constructing large high res lunar mosaics beceomes quite labor intensive and requires substantial experience at all phases, from image acquisition to data handling (100s of gigabytes) to processing and post-processing. 



#35 Garyth64

Garyth64

    Soyuz

  • *****
  • Posts: 3897
  • Joined: 07 May 2015
  • Loc: SE Michigan

Posted 06 September 2018 - 04:05 PM

"So it would seem that the consensus is that a reflector would be better for this project than a refractor."

 

No.  The consensus would be to go with the larger scope, which in this case is the reflector you mentioned.

 

But, for DSOs, a faster scope would be better.  For the moon and planets, a slower scope would be better.

 

Between an 8" f/6 and a 12" f/4, both will give the same image size of the moon, eg.  (.009 x 48" = .43").  But  the image in the 12" will have more detail.  The camera will be able to tell.  More detail spread over the camera's pixels.

 

(did I say that all correct?)



#36 Tom Glenn

Tom Glenn

    Surveyor 1

  • -----
  • Posts: 1939
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 06 September 2018 - 05:07 PM

 

Between an 8" f/6 and a 12" f/4, both will give the same image size of the moon, eg.  (.009 x 48" = .43").  But  the image in the 12" will have more detail.  The camera will be able to tell.  More detail spread over the camera's pixels.

 

(did I say that all correct?)

This is actually not correct.  Assuming we are imaging at prime focus with no barlow, then both scopes would be operating with a focal length of 1200mm (a 305mm aperture at F/4 and a 203mm aperture at F/6 have the same focal length...1200mm).  Because the focal length is identical in both cases, the sampling rate in both cases will be identical at 0.34 arcsec/pixel (assuming a 2um pixel size as stated by the OP).  This means the image resolution will be identical with both scopes, and in both cases would be under-sampled, meaning not operating at the full potential of the scope.  To reach critical sampling with this camera, the scopes would need to operate at about F/10, which for the 8 inch scope would mean a focal length of 2030mm, and for the 12 inch scope would be 3050mm.  In this case, the image scale of the two would be different, with a sampling rate of 0.2 arcsec/pixel for the 8 inch scope, and 0.14 arcsec/pixel for the 12 inch.  In other words, the resolving power of the 12 inch will be greater, but in order for this to be captured in the image, the focal lengths will also be different.  If imaged at the same focal length, there will be no discernible difference in the two images.  The faster scope would do better for DSO imaging with shorter sub-lengths, but this will be irrelevant for lunar imaging.  And in fact, a faster scope has more off-axis aberrations to worry about, and is more difficult to collimate.  



#37 Garyth64

Garyth64

    Soyuz

  • *****
  • Posts: 3897
  • Joined: 07 May 2015
  • Loc: SE Michigan

Posted 06 September 2018 - 06:15 PM

"Because the focal length is identical in both cases, the sampling rate in both cases will be identical at 0.34 arcsec/pixel (assuming a 2um pixel size as stated by the OP)."

 

So the image resolution is the same in a 12" f/4, an 8" f/6, a 6"f/8, and a 3" f/15, because the all have the same focal length. . . because of the camera.  Ok.



#38 Tom Glenn

Tom Glenn

    Surveyor 1

  • -----
  • Posts: 1939
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 06 September 2018 - 06:21 PM

 

So the image resolution is the same in a 12" f/4, an 8" f/6, a 6"f/8, and a 3" f/15, because the all have the same focal length. . . because of the camera.  Ok.

A 3 inch scope at F/15 would have roughly the same image scale, but would be oversampled beyond the limit of the scope.  The Dawes limit of such a scope is only 1.5 arcseconds, so nothing below that could be resolved.  Imaging at 0.34 arcsec/pixel in this scope would not give any resolution below 1.5 arcseconds, unlike with the larger scopes in which it would be sufficient to resolve about 1 arcsec (roughly 3x the sampling rate is a good rule of thumb, provided this value is greater than the Dawes limit).  And this example also speaks to the 5x pixel size rule, because we know that optimal sampling occurs at F/10 with this hypothetical 2um pixel camera, and so F/15 is obviously significantly over this value.  


Edited by Tom Glenn, 06 September 2018 - 06:23 PM.


#39 Garyth64

Garyth64

    Soyuz

  • *****
  • Posts: 3897
  • Joined: 07 May 2015
  • Loc: SE Michigan

Posted 06 September 2018 - 07:09 PM

Thanks.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: astrophotography, equipment, imaging, reflector, refractor



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics