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

Long vs Short FL and Similar Image Scale: Images?

  • Please log in to reply
23 replies to this topic

#1 schmeah

schmeah

    Aurora

  • *****
  • topic starter
  • Posts: 4838
  • Joined: 26 Jul 2005
  • Loc: Morristown, NJ

Posted 24 March 2019 - 08:26 AM

Rather than rehash my thoughts from another thread, I copied my post below and thought I would start a new thread. The theory is great and is impressively detailed by Frank, John etc in other threads like the recent small vs big pixel thread. But much of the theory is over my head and I’m more interested in practical examples. I’m not referring to the dynamic range afforded by larger pixels/well depth, or faintest detail with greater aperture, but simply resolvable detail of small structure. Thanks for any examples!

 

Derek

 

 

“So a very good and practical point has been inferred here. If you can match the image scale of a long FL scope by using a chip with tiny tiny pixels, will you achieve similar detail as many have claimed? My images of small targets using my SCT are magnitudes more detailed than when I crop and re scale my refractor images of the same target to match the scale of the SCT. But my CCD has 7.4mm pixels so my scales vary from 0.65"/pixel to over 3”/pixel. I really would like to see the same imager under similar seeing and sky conditions compare a very long (>2000) FL image with the same target at very short FL with a similar image scale. Has anyone done this with cameras of very disparate pixel size? Because if this is indeed true, many of us would scrap these long FL scopes and premium mounts and just image and crop targets with tiny scopes and tiny pixeled cameras if the detail is truly the same. Someone pleeeease show me this.”



#2 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22700
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 24 March 2019 - 10:29 AM

You would need to maintain aperture to truly resolve the same detail. Just using a shorter scope and smaller pixels wouldn't be enough if the aperture was too small...as then, diffraction would increase and limit the amount of detail you could resolve. I have not had the opportunity to test the hypothesis myself, as I would need to have the very expensive scope and very expensive camera and very expensive mount, such as an 11" SCT + KA-16200 + AP1100 or something like that. From an image scale and diffraction standpoint, a 10" f/4 newt + IMX183 2.4 micron pixel camera should resolve the same detail as a 11" SCT + KAF-16200 6 micron pixel camera. But not just because the pixels are smaller...also because the apertures are roughly the same size. 


  • OldManSky likes this

#3 barrett_flansburg

barrett_flansburg

    Explorer 1

  • -----
  • Posts: 56
  • Joined: 30 Dec 2014

Posted 24 March 2019 - 01:54 PM

Let’s say you have an Explore Scientific 102mm f/7 refractor paired with a camera with the IMX183 2.4 micron pixels. Dawes limit for this scope is 1.14 arc seconds, so it should resolve anything that typical seeing would allow for long exposures. And the pixel scale is 0.69 arc sconds per pixel with this camera, so it would have almost 2 pixels across the Airy disk. Assuming your guiding error doesn’t blur things beyond the seeing, what is there to be gained by using a larger telescope than this? Maybe a 120mm refractor would be better, but probably nothing beyond that. 



#4 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8267
  • Joined: 12 Oct 2007

Posted 24 March 2019 - 04:40 PM

Well the theory for all this is pretty straightforward - but how things actually work in the real world is different.

 

There are two kinds of scaling in a thread like this - one where you keep the aperture fixed and just let pixel size and focal length get smaller.  And another where you scale the whole thing down linearly - including aperture.

 

In theory, for the first case you get just as much light in each pixel and the whole things scales down just fine.  And as long as the read noise stays the same - in terms of resolution there is no difference.  But this is a case where the well depth does impact dynamic range - and the smaller pixels will have smaller well depth - while they are receiving the same amount of light as the bigger version.  So your exposures would need to be shorter to avoid clipping.

 

In the second case you are getting impacted by diffraction as you scale down the aperture - and there is no stopping that effect as you get smaller.  You also have smaller well depth - but in this case it cancels with the smaller aperture and you can keep the exposure the same.  But the dynamic range still goes down.

 

But it's how things actually work in practice that matters to me.  As you scale down the focal length it is harder to have the optics perform as well in the faster system.  I think that even if I had tiny pixels I could not make hyperstar resolution equal what I can do at f/10.  And even f/7 loses a bit over f/10.

 

And if in addition you let aperture get smaller - diffraction will kill you if you are aiming for fwhm's below 2".  At that point you are talking laws of physics - and not just manufacturing quality.

 

At the same time I'm very impressed at how well smart phones can do these days, even with a tiny lens.  But that is not achieving sub 2" fwhm.

 

Another way to put it - if you have a large aperture scope but your fwhm's are fairly large, then you can scale the whole thing down and it should work about as well - except for well depth.  You would need to expose longer because each pixel is getting less light - since they are smaller but the f/ratio is the same.  So total exposure time would increase.

 

Frank



#5 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 24 March 2019 - 07:08 PM

Let’s say you have an Explore Scientific 102mm f/7 refractor paired with a camera with the IMX183 2.4 micron pixels. Dawes limit for this scope is 1.14 arc seconds, so it should resolve anything that typical seeing would allow for long exposures. And the pixel scale is 0.69 arc sconds per pixel with this camera, so it would have almost 2 pixels across the Airy disk. Assuming your guiding error doesn’t blur things beyond the seeing, what is there to be gained by using a larger telescope than this? Maybe a 120mm refractor would be better, but probably nothing beyond that. 

Actually the Dawes limit basically equals the FWHM, so 2 pix accross the FWHM.



#6 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 24 March 2019 - 07:24 PM

And if in addition you let aperture get smaller - diffraction will kill you if you are aiming for fwhm's below 2".  At that point you are talking laws of physics - and not just manufacturing quality.

Not exactly.

Spatial size of Airy Disk is a function of f-ratio.

It does not matter if you have a 8mm aperture on f/4 32mm camera lens or a giant 8m scope on a mountain top.

The number of pixels across the AD in both cases will be the same -- same sampling.

I measured roughly the same pixels across my FWHM for my f/4 180mm camera lens (45mm aperture) -- 2.2 pixels, as I did with my 200mm aperture f/4 Newt -- 2.7 pixels.

In both cases i am sampling min resolution about the same.

See info here:

https://www.cloudyni...ling/?p=9209669



#7 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22700
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 24 March 2019 - 07:41 PM

Not exactly.

Spatial size of Airy Disk is a function of f-ratio.

It does not matter if you have a 8mm aperture on f/4 32mm camera lens or a giant 8m scope on a mountain top.

The number of pixels across the AD in both cases will be the same -- same sampling.

I measured roughly the same pixels across my FWHM for my f/4 180mm camera lens (45mm aperture) -- 2.2 pixels, as I did with my 200mm aperture f/4 Newt -- 2.7 pixels.

In both cases i am sampling min resolution about the same.

See info here:

https://www.cloudyni...ling/?p=9209669

You are talking pixels. Frank is talking arcseconds. You didn't understand this before when you tried to make this argument. Having the same airy disc in PIXELS is NOT the same as the airy disc in arcseconds. If you have a 600mm focal length, with say 3 micron pixels. Then 2 pixels = 2 arcseconds. If you scale things up, say by a factor of two. Your focal length is now 1200mm, and the same 3 micron pixels are smaller in terms of angular size against the sky. Now 2 pixels = 1 arcsecond. If you scale your aperture up as well, so that diffraction in angular terms is also smaller, then your stars will NOT be the same size, even though they still span the same number of pixels.

 

Frank said it would be difficult to get stars below 2 ARCSECONDS with small apertures. And he is absolutely right, since diffraction alone could be 2", and seeing would be added blur on top of that.



#8 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 24 March 2019 - 07:57 PM

Frank seems to be implying that there is something special or important about 2" FWHM.

My point is you can still be well sampled (spatially) even if you operate at angular FWHM's much much larger than 2".


Edited by ks__observer, 25 March 2019 - 07:24 AM.


#9 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8267
  • Joined: 12 Oct 2007

Posted 24 March 2019 - 08:02 PM

Yes 2” is typical seeing limit for amateur deep sky work. You need a decent amount of aperture to achieve it but beyond that it doesn’t help much.

You can certainly do lower res wide field work - and that will be fine with a scaled down system except possibly for exposure time.

Frank

#10 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22700
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 24 March 2019 - 08:07 PM

Frank seems to be implying that there is something special or important about 2" FWHM.

My point is you can still be well sampled even if you operate at angular FWHM's much much larger than 2".

Sure...but that won't do what the OP is asking for. Which is to achieve the same resolution (angular size of resolved details) with small pixel cameras as with large pixel cameras.



#11 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 24 March 2019 - 08:24 PM

The OP is asking an interesting question that might have practicality given the ASI183 with its crazy small pixels and crazy high QE.

What is the effect of keeping image scale the same but loss of aperture?

You would not get as many stars.

But I would be curious to see some side by side comparisons as JH did in the other thread.


  • schmeah likes this

#12 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22700
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 24 March 2019 - 08:29 PM

The OP is asking an interesting question that might have practicality given the ASI183 with its crazy small pixels and crazy high QE.

What is the effect of keeping image scale the same but loss of aperture?

You would not get as many stars.

But I would be curious to see some side by side comparisons as JH did in the other thread.

To keep image scale the same, you adjust pixel size and focal length.

 

If you also reduce aperture, then you increase diffraction. So, despite image scale being the same, diffraction is larger, and you have limited how small the smallest details can be. As such, a small scope with a small aperture, even if you use small pixels, would not be able to resolve the same detail as a large scope with a large aperture and larger pixels, at the same image scale.

 

Seeing plays a role, but seeing is often over-estimated and is often measured (based on FWHM) to be smaller the larger the aperture. An 80mm scope might make you think you've got around 3" seeing. Switch to an 8" or 10" newt or SCT, and measure again, and you may find your seeing is closer to 2", or even below 2".


  • schmeah and Pompey Monkey like this

#13 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 24 March 2019 - 09:15 PM

I think it is about how much you want to zoom in.

I found these NGC 891 shots on AB -- one with a 127mm and one with a 400mm as aperture:

https://www.astrobin...391173/?nc=user

https://www.astrobin...8088/0/?nc=user

 

There is a tiny little galaxy on top of the main galaxy that is better resolved with the 16in.

But if you don't zoom i think they are about the same.

Even zooming it takes a little effort to see the difference -- and this is a 3x aperture difference.

I think stars look a better in 16in but it could be processing as well.



#14 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22700
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 24 March 2019 - 09:23 PM

I think it is about how much you want to zoom in.

I found these NGC 891 shots on AB -- one with a 127mm and one with a 400mm as aperture:

https://www.astrobin...391173/?nc=user

https://www.astrobin...8088/0/?nc=user

 

There is a tiny little galaxy on top of the main galaxy that is better resolved with the 16in.

But if you don't zoom i think they are about the same.

Even zooming it takes a little effort to see the difference -- and this is a 3x aperture difference.

I think stars look a better in 16in but it could be processing as well.

If you maintain aperture and image scale, then the details are the same. You could "zoom in" the same, even if the pixels of one of the cameras are much larger. And that is what the OP is asking about...is that really, truly possible.

 

From a physics standpoint, yes, as long as you DO maintain aperture and image scale, then a shorter focal length/smaller f-ratio with small pixels should indeed deliver the same details as a longer focal length/higher f-ratio with large pixels.

 

One of the points Frank makes, though, is that with longer focal lengths it is easier to correct for aberrations. I don't know at what focal length aberrations really become a problem, but if you try to compare something from say a Hyperstar at 600mm, with small enough pixels, to something from the same SCT at it's native focal length with large enough pixels to have the same image scale...then yes, the hyperstar version of the image, even if the image scale is the same, will definitely have more optical problems than the f/10 version of the image. Now, if you compare say a 10" f/4 1000mm IMX183 image, to an 11" f/10 2800mm KAF-16200 image...is the 1000mm image going to have a lot of optical issues? I don't think so. I think the biggest difference would be the presence of diffraction spikes....aside from that, as long as there isn't a flaring issue (which is often due to a poorly centered secondary or some other reflective thing in the light path) then newts can produce stars just as good as SCTs of similar aperture size.


  • schmeah likes this

#15 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 25 March 2019 - 03:40 AM

From a theory standpoint you are correct that if you maintain aperture and image scale you can crop zoom to get same pic.  Practice is different.  I know they say the Hyperstar is not diffraction limited.

 

But the OP actually asked about "tiny scopes and tiny pixeled cameras."

The two astrobin pics do not have the same image scale -- the 127 has both less optical resolution and less pixel resolution.

The 127 shot is by Jerry Macon, so probably the best 127 shot you will get.

Just looking at the extended object, and not zooming in too much, the images are pretty close -- in my opinion.

I think i like the 16in shot better overall -- stars smaller and the galaxy looks a little cleaner.

But for many who want to go with a smaller scope and smaller mount, then i think small pixels and crop zooming offers a lot of potential.

 

Actually if you pixel peap and zoom into the stars with the 16in the star shape is little off.  Also at f8 the stars might take up a larger spatial size then the f5.3 Televue.  Interestingly both scopes are about the same price.

We don't know the FWHM so we don't know how much each is spatially sampled.

 

It might seem higher resolution has an edge. 

But you can still get a lot out of both systems.


Edited by ks__observer, 25 March 2019 - 03:55 AM.


#16 schmeah

schmeah

    Aurora

  • *****
  • topic starter
  • Posts: 4838
  • Joined: 26 Jul 2005
  • Loc: Morristown, NJ

Posted 25 March 2019 - 05:32 AM

This is very helpful. Thanks. So at what point regarding aperture do diffraction and aberration become practically important? Can I ever match image scales on my FSQ85 or SVR90 to give comparable small detail as my Edge 9.25? Daniel here on CN has posted some impressive images at 650mm focal length using a 2.4 micron pixel ASI 183, which when zoomed in seem to approximate the detail of galaxies taken with my Edge at 2350 FL. 

https://www.flickr.c...06/47265273532/

Now of course that does not account for difference skill, sky conditions, etc. which is why I was hoping to find someone out there who can compare with his/her own equipment.

 

Derek



#17 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 25 March 2019 - 05:50 AM

Daniel's shot is amazing.

IMO there is often plenty of excess resolution in our AP shots. 

Looking at the "big picture" without zooming in much, I think almost everyone is oversampling -- we can reduce the pixel density in favor of SNR by a great amount.

Look at the crop zooms for the Leo galaxies here -- each crop zoom can stand on its own:

https://www.cloudyni...c/#entry9237792

 

With that said, taking a closer look at the NGC 891 shots on my good computer, I was on my Galaxy Note phone before -- I am left with the impression that aperture matters.

With due respect to w4sm (16in shot) -- the stars are mis-shaped and there is a lot of noise in the background -- Jerry's 127 shot is super high quality.
Yet notwithstanding w4sm's limitation, looking on the better computer screen I see a lot more detail in 16in aperture shot.  (Caveat: We don't know seeing conditions and other issues that might affect the results.)
Also, I think smaller stars really makes a big difference in image quality perception.

One big take away relevant to JR and Frank's style of trying to get "round" and "well sampled" stars -- it seems you guys want at least 3pix over the FWHM for star quality:

Based on w4sm's stars, which clearly need some work, is that even with (sorry to offend) crappy stars you can still get great extended object detail. 

If the star issue got cleaned up the extended object I am sure would look even better.
So I don't think you need great stars to get a great main target image -- especially at the loss of so much SNR.
To paraphrase the Godfather movie -- "leave the stars, take the SNR."



#18 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22700
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 25 March 2019 - 06:00 AM

This is very helpful. Thanks. So at what point regarding aperture do diffraction and aberration become practically important? Can I ever match image scales on my FSQ85 or SVR90 to give comparable small detail as my Edge 9.25? Daniel here on CN has posted some impressive images at 650mm focal length using a 2.4 micron pixel ASI 183, which when zoomed in seem to approximate the detail of galaxies taken with my Edge at 2350 FL. 

https://www.flickr.c...06/47265273532/

Now of course that does not account for difference skill, sky conditions, etc. which is why I was hoping to find someone out there who can compare with his/her own equipment.

 

Derek

Diffraction as the diameter of the airy disc from first minimum to first minimum in angular terms is:

 

A = 2.44 * γ / D * 206264.8

 

Where:

 

γ = wavelength of light in microns (i.e. 0.000555 for green light)

D = diameter of aperture in millimeters

 

The measure of the diameter of the airy disc is in effect from the edge of the halo to the edge of the halo, or close enough. This is wider measure of a star than FWHM. To convert from airy disc diameter to FWHM, multiply the angular size of the airy disc by ~0.45. So, for say a 250mm aperture and an 85mm aperture:

 

A250 = 2.44 * 0.000555 / 250 * 206264.8 * 0.45 = 0.5" FWHM

A85 = 2.44 * 0.000555 / 85 * 206264.8 * 0.45 = 1.5" FWHM

 

Now that you have diffraction in terms of FWHM, you could convolve that with your assumed seeing. If you have 1.5" seeing, your total spot sizes would be:

 

S250 = SQRT(0.5^2 + 1.5^2) = 1.6" FWHM

S85 = SQRT(1.5^2 + 1.5^2) = 2.1" FWHM

 

This is actually a meaningful difference. Having imaged with subs at 1.6" and 2.1" myself on several occasions, the difference is quite noticable in both star sizes, star peakness, and detail sharpness. Now if you actually have 3" seeing (rare, but it does occur in heavily populated areas with large heat plume, and under areas with high jetstream velocities and turbulence, such as right under the edge of the polar vortex):

 

S250 = SQRT(0.5^2 + 3^2) = 3.1" FWHM

S85 = SQRT(1.5^2 + 3^2) = 3.4" FWHM

 

A smaller difference, and also a less meaningful difference. In both cases, your stars are going to be more blurry, and that blur is dominated by seeing. If you have seeing this bad, a big telescope isn't going to do you much good.

 

So, there is the basic theory to determine your ballpark FWHMs for different scopes under known seeing conditions. Again, most people OVER-estimate their seeing, especially if they are measuring FWHM with small aperture scopes. Keep in mind, there are other sources of error and blur, so these formulas just get you in the ballpark. Blur terms convolve with each other much like noise terms...the larger terms will dominate the smaller, rendering smaller terms largely moot. However, if your airy FWHM is around 0.5", then it may not be a larger term...tracking error may actually be larger, depending on your mount and environment.

 

Anyway... I, too, would like to see someone put the theory to the test. ;) I would do it myself, and I guess in some sense I probably can. I have data with an IMX183 from a 150mm aperture scope, and I am now using a 106mm aperture scope. I could try to get useful comparison data from the new scope on an object I've imaged before. Of course, to actually get useful comparisons, it might take me until next winter, which wouldn't be all that useful to you now. And, a 150mm aperture is a good one for a refractor, but there are MUCH larger apertures to be had, and for a lot less money as well. You could easily have a 250, 300, 350mm aperture with an SCT, RC or Newt at much less cost. With small-pixel cameras, an 8-10" aperture newt should be plenty to reduce diffraction enough that even ~1.5" seeing would dominate.


  • schmeah likes this

#19 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 25 March 2019 - 06:12 AM

Now that you have diffraction in terms of FWHM, you could convolve that with your assumed seeing. If you have 1.5" seeing, your total spot sizes would be:

 

S250 = SQRT(0.5^2 + 1.5^2) = 1.6" FWHM

S85 = SQRT(1.5^2 + 1.5^2) = 2.1" FWHM

 

This is actually a meaningful difference. Having imaged with subs at 1.6" and 2.1" myself on several occasions, the difference is quite noticable in both star sizes, star peakness, and detail sharpness. Now if you actually have 3" seeing (rare, but it does occur in heavily populated areas with large heat plume, and under areas with high jetstream velocities and turbulence, such as right under the edge of the polar vortex):

 

S250 = SQRT(0.5^2 + 3^2) = 3.1" FWHM

S85 = SQRT(1.5^2 + 3^2) = 3.4" FWHM

 

Do you have a reference for this formula?

My preliminary analysis of my shots from 180mm FL to 1480mm FL seems to be that seeing smears the AD spatially for my results so that the final FWHM size is about 3x the ideal FWHM -- does not matter if it is the 180mm lens or the 1480mm lens.  I need to prepare the data a little more rigorously, but that is my impression at the moment.  Could be different seeing on different nights.



#20 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8267
  • Joined: 12 Oct 2007

Posted 25 March 2019 - 06:25 AM

One concrete benefit of having well sampled stars and the ability to zoom in at the level of the noise is that you can tell very faint and small galaxies from stars just by the profile over a few pixels.

 

If you look at this thumbnail:

 

get.jpg?insecure

 

and then click on it - and then zoom into full resolution - it gets noisy - but if you look close you can see differences in smudges at the level of pixels.  And I haven't done any sharpening or smoothing - which would create artifact and make it harder to distinguish stars from galaxies.

 

If the final presentation isn't intended to convey that level of detail - then there may be less benefit from long focal length and small pixels.  But most any deep image will allow views of small, faint galaxies in the background.  And being able to tell faint stars from faint galaxies is an indication the scene has captured and can convey the max info the optics can deliver.

 

Frank


  • schmeah and StevenBellavia like this

#21 ks__observer

ks__observer

    Viking 1

  • *****
  • Posts: 812
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 25 March 2019 - 06:40 AM

you can tell very faint and small galaxies from stars just by the profile over a few pixels.

This whole debate comes down to pixel density versus SNR.

Yes -- if your goal is to zoom in and look at galaxies over a few pixels, then set your system and processing to max resolution and "leave the SNR, take the stars."

But it you have a large extended object in your FOV and you want a pretty picture of the extended object, and you want to admire that FOV from a distance without much zooming in, then "leave the stars, take the SNR."

It is not one is better than the other.

Each tool in the tool box serves a different purpose.

What I care about is that people make an informed decision and not chase one method just because they think they have to.

 

Edit:

Great pic by the way.


Edited by ks__observer, 25 March 2019 - 06:44 AM.

  • StevenBellavia likes this

#22 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8267
  • Joined: 12 Oct 2007

Posted 25 March 2019 - 06:45 AM

 

It is not one is better than the other.

Each tool in the tool box serves a different purpose.

What I care about is that people make an informed decision and not chase one method just because they think they have to.

You are taking the words right out of my mouth.  The choice of sampling depends heavily on what the imager wants to convey.  Highly sampled and detail or wide field and deep are perfectly fine.  There is no 'optimum' one should aim for.  It is a continuous trade off of factors.

 

And it applies to this thread in terms of whether you need big optics or not.  Well - if you don't want detail then you don't need big optics and small pixels.

 

But I think the OP is asking if it is possible for small optics to match big optics in general.  And I would say no - if you are striving for max detail.

 

Frank


  • ks__observer likes this

#23 StevenBellavia

StevenBellavia

    Ranger 4

  • *****
  • Posts: 316
  • Joined: 11 Mar 2014
  • Loc: New York

Posted 25 March 2019 - 07:48 AM

One concrete benefit of having well sampled stars and the ability to zoom in at the level of the noise is that you can tell very faint and small galaxies from stars just by the profile over a few pixels.

 

If you look at this thumbnail:

 

get.jpg?insecure

 

and then click on it - and then zoom into full resolution - it gets noisy - but if you look close you can see differences in smudges at the level of pixels.  And I haven't done any sharpening or smoothing - which would create artifact and make it harder to distinguish stars from galaxies.

 

If the final presentation isn't intended to convey that level of detail - then there may be less benefit from long focal length and small pixels.  But most any deep image will allow views of small, faint galaxies in the background.  And being able to tell faint stars from faint galaxies is an indication the scene has captured and can convey the max info the optics can deliver.

 

Frank

Frank,

 

In your awesome image, you were sampling at 0.397 arc-sec/pixel.

Is there a way to tell what your actual achieved resolution is?

 

Do you have the FWHM values for the stars?

Or, if Saturn were in this view, (hypothetically speaking, of course), would you have resolved Cassini's division at 0.7 arcsec? (though "lines" might be different from point sources?)

Or are there any known double stars in there as a reference?

 

The Airy Disk for your 11" Edge, with 0.7 reducer should have been 0.92 arc-sec (for 510 nm light)

Resolution is about half of the Airy Disk diameter, so 0.46 arc-sec, which is close to where you were.

The 11-inch with reducer and ASI 1600 appears to be a nice setup.

 

Just curious.

 

Steve



#24 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8267
  • Joined: 12 Oct 2007

Posted 25 March 2019 - 08:18 AM

Frank,

 

In your awesome image, you were sampling at 0.397 arc-sec/pixel.

Is there a way to tell what your actual achieved resolution is?

 

Do you have the FWHM values for the stars?

Or, if Saturn were in this view, (hypothetically speaking, of course), would you have resolved Cassini's division at 0.7 arcsec? (though "lines" might be different from point sources?)

Or are there any known double stars in there as a reference?

 

The Airy Disk for your 11" Edge, with 0.7 reducer should have been 0.92 arc-sec (for 510 nm light)

Resolution is about half of the Airy Disk diameter, so 0.46 arc-sec, which is close to where you were.

The 11-inch with reducer and ASI 1600 appears to be a nice setup.

 

Just curious.

 

Steve

Thanks - I just checked and the fwhm in the stack is around 1.6".  The lowest I have gotten is around 1.2" but I am often below 2".

 

I have good seeing so I focus on high res.  And I'm also trying different guiding techniques for high res - so there is a common theme.

 

And I have a lot of light pollution at around 18.5 mag skies - so I can't do color well.  This is just with a C filter.

 

Yes - for me the ASI1600 and EdgeHD f/7 is a good mix.  And I also like it at f/10 with 0.28" per pix.

 

And to stay on topic - it would be hard to do this with a smaller refractor.

 

It's an important but subtle point.  If you look at the stars - the star spots have a clear trend with brightness - and the brighter ones look bigger.  But then you see small ones that are a bit fainter than they should be.  Well - that is a tiny spot containing maybe 100 billion stars - and you can tell the difference just by the subtle flatter profile of the galaxy even on the scale of a few pixels - or perhaps 2".  The background is dense with tiny, faint galaxies - and you need to look at the level of pixels to tell.

 

Oh - and the diffraction limited fwhm of the 11" at 550nm is 0.39" - so with fwhm's of 1.6" this is still seeing limited - but diffraction plays a non-negligible role.

 

A good image of saturn would be much sharper than this deep sky, long exposure image.

 

It is physically possible for a 5" or so refractor to make a comparable image, I think.  But it would be a challenge.

 

Frank


Edited by freestar8n, 25 March 2019 - 08:28 AM.

  • StevenBellavia likes this


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






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics