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

Secondary Mirror Obstruction?

  • Please log in to reply
51 replies to this topic

#1 rannar

rannar

    Sputnik

  • -----
  • topic starter
  • Posts: 38
  • Joined: 19 Mar 2019
  • Loc: Texas

Posted 31 March 2019 - 09:35 PM

How much is too much in obstruction, and what are the effects of it? 



#2 slepage

slepage

    Ranger 4

  • *****
  • Posts: 313
  • Joined: 02 Jan 2010
  • Loc: San Diego

Posted 31 March 2019 - 09:52 PM

Depends on what you want to use your scope for.  For planetary the smaller the better, for imaging deep sky you can get away with a larger central obstruction.  As a general rule, as the central obstruction increases, resolution of a given aperture decreases. 

 

Steve


  • Earthbound1 likes this

#3 TOMDEY

TOMDEY

    Skylab

  • *****
  • Posts: 4427
  • Joined: 10 Feb 2014
  • Loc: Springwater, NY

Posted 31 March 2019 - 10:46 PM

Ironically, if you look at the MTF curves, high-frequency modulation actually increases with increasing central obstruction! And it is that anomalous high-freq modulation that determines limiting resolution.

 

RELATED:

 

Hubble planetary and star cluster imagery before the repair mission: Operations volitiously/intentionally defocused to the outer zone focus. In the presence of (the massive) spherical aberration, that gave anomalous bump in the MTF at high spatial freqs, and that allowed post-processed images of bright things (aka planets and star clusters) to give reasonably acceptable/presentable images, for that painful hiatus, during which the repair was being approved, designed and executed!

 

Image-processing where (nearly all of us) use various ~digital sharpening filters~ When used in excess, these filters present telltale artifacts like that white ring around the edges of planets and moon and ~plate of spaghetti~ H-alpha solar images... where only the selected freq presents. In that sense, most experts can pretty well tell what processing we have done.

 

[Most "scientific" images go entirely unprocessed. Only then, do the analysts and scientists start executing processing to extract the quantitative data... uncontaminated by prettiness. And those are the images the public rarely gets to see... because they generally aren't pretty!     Tom

Attached Thumbnails

  • 91 central obstruction MTF.jpg
  • 92 central obstruction MTF.jpg

Edited by TOMDEY, 31 March 2019 - 10:47 PM.

  • starzonesteve and Earthbound1 like this

#4 rgsalinger

rgsalinger

    Aurora

  • *****
  • Moderators
  • Posts: 4981
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 31 March 2019 - 11:46 PM

Looking at the graph, as I have read in the past, the resolution effect is small, the contrast effect is large. I thought that's to be expected since the diffraction limitation comes from the fact that the scope has a finite diameter and not from much of anything else. 

Rgrds-Ross



#5 ngc7319_20

ngc7319_20

    Apollo

  • *****
  • Posts: 1012
  • Joined: 25 Oct 2015
  • Loc: MD

Posted 01 April 2019 - 12:05 AM

How much is too much in obstruction, and what are the effects of it? 

An idealized planetary scope would have 15% or less obstruction (by diameter).    A scope meant for low power viewing, or imaging, could have 40% or even 50% obstruction.  

 

The main effect of obstruction is to move light out from the Airy disk, and into the Airy rings.  This reduces the contrast of fine details in the image.

 

Here is an illustration with simulated star images.  The images for 0% and 12% obstruction are pretty similar. By 25% obstruction there is noticeable amount of light in the first Airy ring.   At 50% obstruction there is much light in the first ring, and at 80% most of the light has been moved out of the Airy disk and into the rings.   (These images are scaled to have the peak at same display intensity, and gamma=0.5.)

 

ALL.JPG

 

As @TOMDEY says, the resolution increases (slightly) for large obstructions.  You can see that the central Airy disk is slightly smaller at 80% obstruction vs. the other cases.  But it would make dreadful images for a visual observer.

 

These images were made with the free software at http://aberrator.astronomy.net/.  (Has anyone made an online simulator for this?)


Edited by ngc7319_20, 01 April 2019 - 12:13 AM.

  • Galicapernistein, daquad, starzonesteve and 1 other like this

#6 Galicapernistein

Galicapernistein

    Viking 1

  • *****
  • Posts: 535
  • Joined: 24 Sep 2007

Posted 01 April 2019 - 01:43 PM

An idealized planetary scope would have 15% or less obstruction (by diameter).    A scope meant for low power viewing, or imaging, could have 40% or even 50% obstruction.  

 

The main effect of obstruction is to move light out from the Airy disk, and into the Airy rings.  This reduces the contrast of fine details in the image.

 

Here is an illustration with simulated star images.  The images for 0% and 12% obstruction are pretty similar. By 25% obstruction there is noticeable amount of light in the first Airy ring.   At 50% obstruction there is much light in the first ring, and at 80% most of the light has been moved out of the Airy disk and into the rings.   (These images are scaled to have the peak at same display intensity, and gamma=0.5.)

 

attachicon.gif ALL.JPG

 

As @TOMDEY says, the resolution increases (slightly) for large obstructions.  You can see that the central Airy disk is slightly smaller at 80% obstruction vs. the other cases.  But it would make dreadful images for a visual observer.

 

These images were made with the free software at http://aberrator.astronomy.net/.  (Has anyone made an online simulator for this?)

I’m surprised at how bright the ring is at 25%. That’s the point at which people say the secondary doesn’t matter. But it’s clear from your illustration that every percentage point matters. Great comparison btw.



#7 havasman

havasman

    Cosmos

  • ****-
  • Posts: 9708
  • Joined: 04 Aug 2013
  • Loc: Dallas, Texas

Posted 01 April 2019 - 05:04 PM

This website can teach a lot on that subject, starting with but not limited to the calculator linked here -  http://www.bbastrode...om/diagonal.htm

 

IMO, there is often too much worry spent on this subject outside an ATM situation. The Newtonian and other reflectors offered in the marketplace are most often pre-configured (mass market and midrange) and work great. The high end suppliers are extremely reliable advisors for such matters. Once an observer can define their observing predilections and habits, the preferred choices should be obvious. If the observer cannot do that then relying on market tested solutions is prudent.

 

Here is the most complete answer to your questions that I know of -  https://www.telescop...obstruction.htm


Edited by havasman, 01 April 2019 - 05:07 PM.

  • Jon Isaacs, Asbytec and ngc7319_20 like this

#8 Galicapernistein

Galicapernistein

    Viking 1

  • *****
  • Posts: 535
  • Joined: 24 Sep 2007

Posted 02 April 2019 - 06:40 AM

This website can teach a lot on that subject, starting with but not limited to the calculator linked here -  http://www.bbastrode...om/diagonal.htm

 

IMO, there is often too much worry spent on this subject outside an ATM situation. The Newtonian and other reflectors offered in the marketplace are most often pre-configured (mass market and midrange) and work great. The high end suppliers are extremely reliable advisors for such matters. Once an observer can define their observing predilections and habits, the preferred choices should be obvious. If the observer cannot do that then relying on market tested solutions is prudent.

 

Here is the most complete answer to your questions that I know of -  https://www.telescop...obstruction.htm

Secondary size is determined by F ratio and tube diameter anyway, so you can’t really separate the two. One of the many benefits of slow scopes is they allow you to use smaller secondaries, but switching to a slightly smaller secondary from one that provides a fully illuminated field would probably not be worth it to a general observer, although a dedicated planetary observer might benefit.



#9 Vla

Vla

    Messenger

  • *****
  • Posts: 489
  • Joined: 08 Mar 2014

Posted 02 April 2019 - 03:14 PM

How much is too much depends on more than one factor, including object of observation, other optical errors present, and observer's tolerance. The effect is transfer of energy from the central disk to the rings area, impairing image definition. But unlike aberrations, part of this transfer is followed by diminishing of the central maxima, which means that obstructed aperture can be compared with a larger aberrated one. If "c" denotes the relative size of a circular central obstruction in units of the aperture, drop in the central intensity of diffraction pattern normalized to 1 is given by (1-c^2)^2. But if we keep the flux unchanged, the drop is only 1-c^2, which means that the square of it represents the light loss due to obscuration. Still, the double square is a good approximation of the energy lost to the rings.

 

The effects of CO on image quality are routinely exaggerated. Pic below shows how much the central maxima shrinks for 33% and 50% obstruction (top, OSLO output). Maxima diameter at 33% CO is about 10% smaller, while the FWHM, which matters more for the resolution limit, is about 5% smaller. So, the obstructed aperture has at least 5% higher cutoff frequency, which does not show in the MTF normalized formalism (bottom left). If we factor it in, and use the approximate contrast cutoff for bright low-contrast details (Rutten/Venrooij), we see that the dreadful 33% CO effectively reduces aperture by less than 10% here, and the wrecking 50% CO by little over 25% - much less than what the common knowledge suggests. It means that the rest of the damage from the empirical accounts comes from (a number of) other factors.

Attached Thumbnails

  • co.png

  • Jeff Morgan and Kunama like this

#10 TOMDEY

TOMDEY

    Skylab

  • *****
  • Posts: 4427
  • Joined: 10 Feb 2014
  • Loc: Springwater, NY

Posted 02 April 2019 - 03:46 PM

Keep in mind that even a little more aperture makes up for a substantial central obstruction! That's because the entire airy disc shrinks and brightens in proportion to aperture. All of those curves above that I and others are presenting... are normalized to unity aperture and wavelength.

 

A six-inch scope with a 30% diameter obstruction resolves far better than an unobstructed five-incher. Just generate the non-normalized point-spreads and MTFs to see that in action!

 

PS: This is why a (good) modest-sized Dobsonian will always blow the socks off a good smaller refractor (any smaller refractor!) for both light-gathering and resolution!

 

But, gota admit... refractors make fine finder scopes on big Newtonian reflectors...    Tom

Attached Thumbnails

  • 12 Mars showing canali and clover leaf.jpg

  • Asbytec, gnowellsct, Stephen Kennedy and 3 others like this

#11 ngc7319_20

ngc7319_20

    Apollo

  • *****
  • Posts: 1012
  • Joined: 25 Oct 2015
  • Loc: MD

Posted 02 April 2019 - 04:41 PM

Well maybe I'm wrong, but it seems to me the problem with CO is in the middle spatial scales of the MTF graphs.  The contrast at 50% CO is cut in half here, vs. 0% CO.  On the PSF irradiance plots the trouble is in the wings at the edge of the plot.  The light in the first ring (just at edge of the plot) increases about 5 times going from 0% CO to 50% CO.  There is much area in the 2-D PSF out here... so the light being moved to the first Airy ring is big.

 

post-227720-0-64214600-1554236066_mod2.jpg


Edited by ngc7319_20, 02 April 2019 - 04:41 PM.


#12 ngc7319_20

ngc7319_20

    Apollo

  • *****
  • Posts: 1012
  • Joined: 25 Oct 2015
  • Loc: MD

Posted 02 April 2019 - 04:54 PM

 

A six-inch scope with a 30% diameter obstruction resolves far better than an unobstructed five-incher. Just generate the non-normalized point-spreads and MTFs to see that in action!

 

PS: This is why a (good) modest-sized Dobsonian will always blow the socks off a good smaller refractor (any smaller refractor!) for both light-gathering and resolution!

 

Well, this only works in a vacuum.  Under real conditions, seeing, etc.,  the 4" refractor will reach its resolution limit much more often than the 6" Dob.

 

Could you show the plots?  My understanding / experience was the equivalent obstructed scope was roughly (refractor aperture) = (obstructed aperture) - (obstruction) so the equivalent to the six-inch with 30% obstruction is a four-inch refractor.  And that assumes perfect seeing.  If your six-inch always shows perfect solid Airy patterns, then great.  But if the six inch is showing an undulating mess of star light, then the four-inch scope refractor is the one to use.


Edited by ngc7319_20, 02 April 2019 - 04:56 PM.


#13 Kunama

Kunama

    Aussie at large

  • *****
  • Posts: 4674
  • Joined: 22 Oct 2012
  • Loc: Australia

Posted 02 April 2019 - 05:19 PM

Well, this only works in a vacuum.  Under real conditions, seeing, etc.,  the 4" refractor will reach its resolution limit much more often than the 6" Dob.

 

Could you show the plots?  My understanding / experience was the equivalent obstructed scope was roughly (refractor aperture) = (obstructed aperture) - (obstruction) so the equivalent to the six-inch with 30% obstruction is a four-inch refractor.  And that assumes perfect seeing.  If your six-inch always shows perfect solid Airy patterns, then great.  But if the six inch is showing an undulating mess of star light, then the four-inch scope refractor is the one to use.

Much depends on the actual scopes, here is some interesting reading: http://www.astronomy...-reflector.html from Joe Cali. Joe is an experienced observer with over 40 years in the game and tells it like it is......



#14 TOMDEY

TOMDEY

    Skylab

  • *****
  • Posts: 4427
  • Joined: 10 Feb 2014
  • Loc: Springwater, NY

Posted 02 April 2019 - 08:01 PM

Well, this only works in a vacuum.  Under real conditions, seeing, etc.,  the 4" refractor will reach its resolution limit much more often than the 6" Dob.

 

Could you show the plots?  My understanding / experience was the equivalent obstructed scope was roughly (refractor aperture) = (obstructed aperture) - (obstruction) so the equivalent to the six-inch with 30% obstruction is a four-inch refractor.  And that assumes perfect seeing.  If your six-inch always shows perfect solid Airy patterns, then great.  But if the six inch is showing an undulating mess of star light, then the four-inch scope refractor is the one to use.

Yes, I do indeed agree that when the seeing is bad, a little refractor is worth hauling out; just leave the good scope inside for the good nights! And when the barometer drops to zero, haul out the monster scopes!

 

Actually, here are the curves for 25% central obstruction at 5, 6 and 8-inch apertures. I cranked up the aperture until the obstructed scope exceeded MTF of the 5-inch at all spatial freqs. Note that the 8-inch far exceeds the 5-inch whenever the seeing is 3 arc-sec or better. Where I'm located, 1 arc-sec seeing is not unusual and half-sec is often enough to want to take advantage of. And, only aperture can avail that. And, as a bonus... way more light. Big aperture is a win-win!   Tom

Attached Thumbnails

  • 19 scaled 5 6 8 MTF working.jpg

  • Eddgie and ngc7319_20 like this

#15 Vla

Vla

    Messenger

  • *****
  • Posts: 489
  • Joined: 08 Mar 2014

Posted 02 April 2019 - 10:02 PM

Well maybe I'm wrong, but it seems to me the problem with CO is in the middle spatial scales of the MTF graphs.  The contrast at 50% CO is cut in half here, vs. 0% CO.  On the PSF irradiance plots the trouble is in the wings at the edge of the plot.  The light in the first ring (just at edge of the plot) increases about 5 times going from 0% CO to 50% CO.  There is much area in the 2-D PSF out here... so the light being moved to the first Airy ring is big.

 

attachicon.gif post-227720-0-64214600-1554236066_mod2.jpg

Not wrong, but rather partial, by focusing on (and a bit exaggerating) the negative, and leaving out the positive. We need to take into account both for a balanced assessment. Relevant MTF is one on the right, and the contrast drop for 50% CO is little over 40%. But even that does not mean it is equivalent of that much smaller aperture. If we extend the lowest part of the MTF - the method by which is the mentioned rule actually obtained - we see that the corresponding aperture is 43% smaller. And if we project it through  the cutoff point, it corresponds to 36% smaller aperture, the two averaging at about 40% smaller. Likewise, 33% obstruction averages out at about 17% smaller - nearly half of what the rule proposes.

 

But this is still only a part of the worse part. Another relevant aspect within the worse part is contrast drop across the range of frequencies for dim low-contrast object. Here, the same approach based on the approximate cutoff for this type of details comes out at 24% smaller for 50% CO and 17% smaller for 33%. Taking the simple average with the previous one gives 32% smaller for 50% CO and 33% CO remains at 17% smaller effective aperture (by diameter).

 

But we still need to take into account the good part, i.e. better contrast of the obstructed aperture in the high-frequency range. Again using Rutten/Venrooij's approximations, we find that the 50% CO acts like 14%, and 33% CO like 7% larger aperture. Putting it all together, gives that 50% CO brings effective aperture down by 9%, and the 33% CO by 5%. Hmmm... This is, of course, simplistic approach, no pretense on exact figures, but it does throw quite a bit different light on the subject - again, based on the same methodology used to extract the D-d rule.

Attached Thumbnails

  • co2.png

  • howardcano likes this

#16 Vla

Vla

    Messenger

  • *****
  • Posts: 489
  • Joined: 08 Mar 2014

Posted 02 April 2019 - 10:13 PM

Well, this only works in a vacuum.  Under real conditions, seeing, etc.,  the 4" refractor will reach its resolution limit much more often than the 6" Dob.

Well, that's the trap. We want to know what the effect of CO is, and shouldn't mix in things like that. There are several things that favor smaller aperture, particularly quality refractor, but they do not fall under CO effect.


  • Asbytec likes this

#17 ngc7319_20

ngc7319_20

    Apollo

  • *****
  • Posts: 1012
  • Joined: 25 Oct 2015
  • Loc: MD

Posted 02 April 2019 - 10:32 PM

Yes, I do indeed agree that when the seeing is bad, a little refractor is worth hauling out; just leave the good scope inside for the good nights! And when the barometer drops to zero, haul out the monster scopes!

 

Actually, here are the curves for 25% central obstruction at 5, 6 and 8-inch apertures..... Note that the 8-inch far exceeds the 5-inch whenever the seeing is 3 arc-sec or better. 

Perfect!  Thanks for plots!  Yes that's consistent with what I expected.  An eight-inch with 25% obstruction would be roughly equivalent to a 6-inch refractor. So of course it will beat a five-inch refractor as you say.



#18 stargazer193857

stargazer193857

    Fly Me to the Moon

  • *****
  • Posts: 7440
  • Joined: 01 Dec 2013
  • Loc: Southern Idaho

Posted 02 April 2019 - 11:10 PM

How much is too much in obstruction, and what are the effects of it?


Too much is when the instruction gets heavy. This is an issue at f3 on big scopes.

#19 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 42351
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 03 April 2019 - 12:45 AM

There is a minimum size for a secondary to illuminate the center of the field maximally.
There is a maximum size and that is where the entire largest field chosen is maximally illuminated.
In between is a size where the brightness of the field edge falls but not significantly.

So there is not a lot of choice when choosing a secondary.
Want a smaller secondary? Get a longer f/ratio scope of the same aperture. Or get a larger scope with the same f/ratio. But don't just put a smaller mirror in your scope--it might end up reducing your aperture.

#20 TOMDEY

TOMDEY

    Skylab

  • *****
  • Posts: 4427
  • Joined: 10 Feb 2014
  • Loc: Springwater, NY

Posted 03 April 2019 - 06:36 AM

There is a minimum size for a secondary to illuminate the center of the field maximally.
There is a maximum size and that is where the entire largest field chosen is maximally illuminated.
In between is a size where the brightness of the field edge falls but not significantly.

So there is not a lot of choice when choosing a secondary.
Want a smaller secondary? Get a longer f/ratio scope of the same aperture. Or get a larger scope with the same f/ratio. But don't just put a smaller mirror in your scope--it might end up reducing your aperture.

Yes, and very concise! Once the PM size, F# and desired field are chosen... that constrains the central obstruction within a rather narrow selection-range!

 

Ultra-Fast systems also require field group, stress eyepieces and preclude the eye's best central pupil... such stuff.    Tom



#21 stargazer193857

stargazer193857

    Fly Me to the Moon

  • *****
  • Posts: 7440
  • Joined: 01 Dec 2013
  • Loc: Southern Idaho

Posted 03 April 2019 - 01:12 PM

Imagers want even illumination from edge to edge, even if they lose some light.
For visual, I take the on axis cross section of the light cone at the secondary and add 0.5".

#22 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 42351
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 03 April 2019 - 01:33 PM

Imagers want even illumination from edge to edge, even if they lose some light.
For visual, I take the on axis cross section of the light cone at the secondary and add 0.5".

That might work for some scopes, but not your 8" f/6 dob.

 

I assume you might use a 31mm Nagler in that scope, with a field stop of 1.65"

The size of the light cone at the secondary would be around 2.67" if narrowing from 8" to 1.65"

With your formula, you'd use a 3.1" secondary.

A 2.14" secondary works fine in an 8" f/6.  3.1" is too large.  Usable, sure, but unnecessarily large.

 

So perhaps you mean the diameter of the light cone for only the axial ray at the secondary, and assume a size of 0" at the focal plane.

But that doesn't work either, since then the diameter of the light cone at the secondary is 1.29".  Add a half inch and you have a secondary of 1.79" approximately, and that isn't large enough.

 

Perhaps your rule might work for some other scope, but it sure doesn't work for an 8" f/6


  • Jon Isaacs likes this

#23 oldtimer

oldtimer

    Vanguard

  • *****
  • Posts: 2132
  • Joined: 13 Nov 2008
  • Loc: Lake County Illinois

Posted 22 May 2019 - 08:20 PM

Question:  As a general rule of thumb, can visual only observers get away with a slightly smaller secondary that those doing imaging?.Shortly I will be acquiring a 8" F4 that seems to be optimized for imaging as the focal point is about 5 inches off the tube. It has a 70mm secondary. I am thinking of replacing it with either a 58mm or 62mm. I would be using it for low to medium power and not as a planetary scope.What do you think?



#24 Jon Isaacs

Jon Isaacs

    ISS

  • *****
  • Posts: 78358
  • Joined: 16 Jun 2004
  • Loc: San Diego and Boulevard, CA

Posted 22 May 2019 - 08:31 PM

I don't really see much reason to reduce the secondary size for low to medium power viewing.  

 

I calculate the light loss due to a 70mm secondary as about 12%, the light loss due to a 59mm secondary as 9%.  

 

Basically inconsequential. 

 

Jon


  • Pinbout likes this

#25 tommm

tommm

    Messenger

  • -----
  • Posts: 475
  • Joined: 16 Dec 2015

Posted 22 May 2019 - 10:39 PM

Yes, imaging systems typically use a larger secondary because they usually want a larger fully illuminated field of view that that typically used for visual use since the effects of a smaller one show up on photos but usually isn't noticed in visual use.  A 70mm minor axis secondary is about a 34% obstruction, 20 to 25% is more typical for visual use, but opinions vary with some using sizes above and below that range.  You can find lots of threads and opinions on CN regarding secondary obstruction size and its effects on contrast.




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