Star Test Symmetry Revisited (long)
Derek Wong
which may be relevant to the discussion of star test symmetry last month
(AP Mak thread). He simulated a 6" f/10 Newtonian with 20% central
obstruction. See the results at:
http://www.home.earthlink.net/~dawong/
(apologies for the tacky Earthlink Template page).
The patterns are all divided to make it easy to compare patterns inside
of focus and outside of focus.
The first pattern shows the perfect symmetry of a perfect paraboloid.
The second ("Peak") shows a 1/10 wave peak to valley with better than
1/42 wave rms. The error consists of both 3rd order and 5th order
spherical aberration. The third ("Flat") shows a scope with the exact
same 3rd and 5th order error, but with a much more symmetric looking
star test. Nils Olof says "PEAK shows that there is a clear secondary
'breakout' on one side but not on the other", FLAT shows otherwise.
The results appear to corroborate what Roland said--you can have a 1/10
wave scope with very good rms values which has a very asymmetric star
test, and that hand correcting a scope to improve the symmetry may not
improve performance at all. I don't know if you would see such a
Newtonian mirror in practice, but a complex Maksutov system with
aspheric curves is a different story.
Please note that errors are possible, and comments are welcome. Can
someone simulate this in another program like Zemax to confirm? Nils
Olof knows a lot more about optics and the DIFFRACT program than I do,
so I put his e-mail at the bottom of the web page. The "ugly" math
discussion is below for those of you not sleeping :-)
Derek
*****************
Nils Olof said:
Say you can design away the 3rd order spherical abberation and have some
remaining 5th order, with a Zernike polynomial of the form 20 r^6 -30
r^4 + 12 r^2 -1, where 0<=r<=1.
Suppose you want to fool the starry-eyed crowd and deliberately add some
3rd order: the Zernike polynomial. 12 r^4 - 12 r^2 + 2!!! This cancels
the x^2 term and makes the center very flat:
Let Z1(r) = 20 r^6 -18 r^4 +1,
try that for breakout with some 20% obstruction!
(This is FLAT)
BUT if you subtract it:
Let Z2(r) = 20 r^6 -42 r^4 + 24 r^2 -3,
this is maximally steep near center but flat at the edge
(This is PEAK)
I predict the difference will be obvious!
BTW both functions have 4.16 (exactly) units P-V, so you may use a
factor 1/41.6 to get 1/10 wave P-V. The polynomials are orthogonal, and
... adding OR subtracting them does not change best focus.
Derek said:
(skipped a lot of math proving that the rms of both functions are
equivalent and discussion of orthogonal functions)
If f(r) represents wavefront deviation in a circle of radius 1,
rms(f) = k * sqrt (integral r = 0 to 1 of f(r)*f(r)*2*pi*r dr)
where k is 1/sqrt(pi)
The rms of Z1 AND Z2 is about 1/42 wave rms (scaled with 1/41.6)
Nils Olof:
I guess Zernike polynomials are a powerful theoretical tool, but they
wouldn't help directly in generating the patterns - - - while trying to
take a nap today, the thought struck me that Jim B's DIFFRACT would do
the job!
http://www.atmsite.org/ProjectsArticles/Carlin/couder/index.html
I made a small DOS Pascal routine to generate a deviated spherical
mirror 152 mm f/10, that I could run and get the diffraction pattern
(corresponding to that of a paraboloid one with source at infinity, and
with 20% obstruction (by diameter)). So I did - enclosed are the
results. Left half is defocused -6mm, right half +6mm, (and I have an
occulting bar to make the peak brightness not set by the center).
PEAK.gif is with the "peaked center", 1/10 wave P-V aberration I sent
you, FLAT.gif is with the flat-centered one (lacking x^2 term), but
strongly tilted edge instead - the one for which I predicted the
breakout would be fairly symmetric. NONE.gif is for an undeviated
sphere, otherwise similar - and the 2 image halves are similar as they
ought to be.
Bryan Greer
> The results appear to corroborate what Roland said--you can have
> a 1/10 wave scope with very good rms values which has a very
> asymmetric star test, and that hand correcting a scope to improve the
> symmetry may not improve performance at all. I don't know if you
> would see such a Newtonian mirror in practice, but a complex Maksutov
> system with aspheric curves is a different story.
Hello Mr. Wong,
As usual, Nils has done a great job analyzing these aberrations. Let's not
forget what the core issue was from last month's star test thread, though.
There's never been a question that some refractive/cat designs show
asymmetrical extrafocal patterns. The real question was if you "tweak"
those designs (i.e. aspherical surfaces) so that their star test patterns
become more symmetrical, does it further improve the PSF (or Strehl)? In
other words, does a better star test (as defined by symmetrical extrafocal
patterns) _always_ mean better optical performance?
I've played with Gregory Mak and fast APO designs (thank you Thomas and
Bratislav) in OSLO. On the Mak, for instance, every time I change a surface
(R1) that improves the star test symmetry, the PSF value also slightly
improves. This supports my original contention that a more symmetrical star
test does mean better performance. However, I have no way of knowing if I'm
tweaking the correct surface (or even the right type of correction), since
this really isn't my specialty. I hardly consider my limited results as
"proof".
I forget who it was, but someone in that thread made the contention that you
could even modify an all-spherical design (which shows considerable
asymmetry in the star test) to produce a more symmetrical star test, and the
Strehl would actually suffer! I can't disprove this, though I'd love to see
an example of it. I believe it was Roland who pointed out that from a
practical standpoint, radical aspherizing would introduce more roughness and
counter any small gains. This _does_ seem plausible, and of course he would
know better than me.
But keep in mind, these exceptions to the rule (if they exist) won't come
into play for the vast majority of people who are trying to use the star
test. The star test is straightforward for the majority of telescope
designs typical amateurs come across.
Finally, in the example you stated (the Newt with 3rd and 5th SA), removing
the SA will certainly improve the PSF value and produce a more symmetrical
star test.
Sincerely,
Bryan Greer
Columbus, OH
(please remove *takeout* from my address for private replies)
ar...@selena.kherson.ua
Derek Wong <daw...@earthlink.net> wrote:
> Nils Olof Carlin has generated some interesting diffraction patterns
> which may be relevant to the discussion of star test symmetry last month
> (AP Mak thread). He simulated a 6" f/10 Newtonian with 20% central
> obstruction. See the results at:
>
> http://www.home.earthlink.net/~dawong/
>
> (apologies for the tacky Earthlink Template page).
>
> The patterns are all divided to make it easy to compare patterns inside
> of focus and outside of focus.
>
> The first pattern shows the perfect symmetry of a perfect paraboloid.
> The second ("Peak") shows a 1/10 wave peak to valley with better than
> 1/42 wave rms. The error consists of both 3rd order and 5th order
> spherical aberration. The third ("Flat") shows a scope with the exact
> same 3rd and 5th order error, but with a much more symmetric looking
> star test. Nils Olof says "PEAK shows that there is a clear secondary
> 'breakout' on one side but not on the other", FLAT shows otherwise.
>
> wave scope with very good rms values which has a very asymmetric star
> test, and that hand correcting a scope to improve the symmetry may not
> improve performance at all. I don't know if you would see such a
> Newtonian mirror in practice, but a complex Maksutov system with
> aspheric curves is a different story.
>
> someone simulate this in another program like Zemax to confirm? Nils
> Olof knows a lot more about optics and the DIFFRACT program than I do,
> so I put his e-mail at the bottom of the web page. The "ugly" math
> discussion is below for those of you not sleeping :-)
>
> Derek
>
Derek,
This is simple misleading, that perfect from practical point of view
system, MUST has a symmetrical diffraction patterns. Simple myth.
At this moment we are quite busy manufacturing 10" F/14 Mak-cass
systems for AP . I bet, you will be not able to judge which system is
better looking on their interferorgams. All of them leave nothing to
wish better, but they are different in off-focus images. You will be
not able to differ them looking on planets or double stars.
I can add for a whole picture, that their 3-rd order spherical correction
typically better than 1/20 wave.
Valery Deryuzhin
ARIES INSTRUMENTS.
Sent via Deja.com http://www.deja.com/
Before you buy.
Derek Wong
> Hello Mr. Wong,
Bryan, kind of formal for someone you met at RTMC a couple years ago and
who bought one of your fine diagonals :-)
> The real question was if you "tweak"
> those designs (i.e. aspherical surfaces) so that their star test patterns
> become more symmetrical, does it further improve the PSF (or Strehl)? In
> other words, does a better star test (as defined by symmetrical extrafocal
> patterns) _always_ mean better optical performance?
Ah, but this is exactly what was shown here--the answer to your last
question is no. Both "Peak" and "Flat" have exactly the same size
aberrations, 3rd and 5th order and rms values. "Flat" has a much better
star test than "Peak", so a better star test does NOT always mean better
optical performance. This is a simple Newtonian, although again I don't
know if these mirrors would ever be produced in the real world.
> I've played with Gregory Mak and fast APO designs (thank you Thomas and
> Bratislav) in OSLO. On the Mak, for instance, every time I change a surface
> (R1) that improves the star test symmetry, the PSF value also slightly
> improves.
I'm NO optical designer, in fact many people here including you know
tons more about this subject than I. I can believe your statement above
for the vast majority of scopes.
However, I can imagine that there would be scopes which would have a
star test like "Peak". You could correct them to "hide" the
aberrations, either by refiguring them to "Flat" or by increasing the
surface roughness very slightly.
> But keep in mind, these exceptions to the rule (if they exist) won't come
> into play for the vast majority of people who are trying to use the star
> test. The star test is straightforward for the majority of telescope
> designs typical amateurs come across.
Yes, and from what I've heard you are pretty darn good at it! Jeff
Medkeff said you called all of the aberrations on one refractor which
matched the interferogram.
> Finally, in the example you stated (the Newt with 3rd and 5th SA), removing
> the SA will certainly improve the PSF value and produce a more symmetrical
> star test.
Of course, although I wouldn't want to tweak a 1/42 wave rms scope!
This is what I have taken from the patterns:
1. It is POSSIBLE to have an excellent scope (1/10 wave p to v
wavefront with 1/42 wave rms) with a very asymmetric star test. Yuri
sent me a real Mak design with 1/8 p to v and 1/32 rms with aberrations
similar to peak.
2. Judging complex aspherical designs with small central obstructions
with symmetry/breakout may be difficult--Suiter even mentions this in
his book. This means that if I ever end up with a Mak with a small
obstruction I'll be prepared to be roasted by people (not you!) who
pronounce the scope bad due to asymmetry.
I'm going to talk to Mils Olof about Roland's other statement--that a
Mak with a big central obstruction can "hide" significant aberrations
and have a symmetric star test. This is not to make a statement about
quality, but rather to increase our learning.
By the way Nils Olof mentioned a few people he would love to have input
from, and you were one of them.
Thanks,
Derek
Dan Chaffee
On Tue, 14 Mar 2000 03:00:05 GMT, ar...@selena.kherson.ua wrote:
>This is simple misleading, that perfect from practical point of view
>system, MUST has a symmetrical diffraction patterns. Simple myth.
>At this moment we are quite busy manufacturing 10" F/14 Mak-cass
>systems for AP . I bet, you will be not able to judge which system is
>better looking on their interferorgams. All of them leave nothing to
>wish better, but they are different in off-focus images. You will be
>not able to differ them looking on planets or double stars.
>I can add for a whole picture, that their 3-rd order spherical correction
>typically better than 1/20 wave.
Dear Valery,
You have given a specific example of the well corrected Mak Cass's
asymmetry in the star test, but before this lapses into another
possible hotbed of confusion, (if not debunking of "the book
from the devil's library":-), let me ask you this question:
Have you ever seen a well collimated, equalibriated newtoinan
with near perfect star test symmetry(including smooth surface and
no sign of astigmatism) be _out performed_ by another newtonian
of equal aperture/obstruction/thermally stable/collimated that shows
stronger differences in the star test asymmetry---on planets?(assuming
equally good seeing, eyepieces,etc of course) I'm speaking of pure
newtonian, not mak-newt, or a newtonian with optical window.
What's missing in Nil's Olof's simulation is a real life comparison
of the actual damage done to a low contrast, extended object. So it's
not clear from this that the less symmetrical star test could actually
be the better performer(or as good) --even with what is theoretically
the same rms error.
Dan Chaffee
Kansas City
Derek Wong
>
> This is simple misleading, that perfect from practical point of view
> system, MUST has a symmetrical diffraction patterns.
Valery, I hope you read enough of the post that you see that we arein
agreement with you on this for complex designs, and that the diffraction
patterns demonstrate this quite nicely.
> At this moment we are quite busy manufacturing 10" F/14 Mak-cass
> systems for AP . I bet, you will be not able to judge which system is
> better looking on their interferorgams.
I'm sure you are right with this point, but...
> All of them leave nothing to
> wish better,
Wrong. I want a 12" Mak with a 20-23% obstruction which weighs close to
50 pounds :-)
And yes, the seeing here justifies it.
Derek
Dan Chaffee
Hi Derek,
>> The real question was if you "tweak"
>> those designs (i.e. aspherical surfaces) so that their star test patterns
>> become more symmetrical, does it further improve the PSF (or Strehl)? In
>> other words, does a better star test (as defined by symmetrical extrafocal
>> patterns) _always_ mean better optical performance?
>Ah, but this is exactly what was shown here--the answer to your last
>question is no. Both "Peak" and "Flat" have exactly the same size
>aberrations, 3rd and 5th order and rms values. "Flat" has a much better
>star test than "Peak", so a better star test does NOT always mean better
>optical performance.
At best , this example of Zernike polynomial expansion only indicates
that a less symmetrical star test _may_ be an equal performer to one
close to symmetry, not that a scope testing closer to symmetry would
produce worse images than one less symmetrical. Yet, a paraboloid's
star test lacking in symmetry may very well (and always has been
in my experience) be a poorer performer to one more similar either
side of focus.
This is all the more reason to work toward symmetry for those of us
who figure our own newtonains.
Dan Chaffee
lude...@my-deja.com
> Derek,
>
> This is simple misleading, that perfect from practical point of view
> At this moment we are quite busy manufacturing 10" F/14 Mak-cass
> systems for AP . I bet, you will be not able to judge which system is
> better looking on their interferorgams.
Valery,
a question: Lets estimate you are able as you say to produce same size,
same focallenght ,same central obstruction MCT with same RMS and same
P.t.v. but one with symmetrical image and one with asymmetrical image,
would you see in an testreport made by an indepentent company show for
both systems the same RMS, same P.t.v., same energyconcentrationand
same MTF Profile or would there be a diffrence ? If there would be a
diffrence, than why and what would be diffrent ? As you always agree
the Energyconcentration is the point who shows the real quality, would
it be the same ? If your answeres are yes, would you agree to give me 2
such systems for retesting with the laser and wavelenght you recomment ?
thanks for your answere
Markus
All of them leave nothing to
> wish better, but they are different in off-focus images. You will be
> not able to differ them looking on planets or double stars.
> I can add for a whole picture, that their 3-rd order spherical
correction
> typically better than 1/20 wave.
>
Blandp1
Hash: SHA1
In article <38cd4376$1@news>, Bryan Greer
<bgreer@*takeout*fpi-protostar.com> writes:
>On the Mak, for instance, every time I change a surface
>(R1) that improves the star test symmetry, the PSF value also
>slightly improves.
The real question is, if you take a 1/10 wavefront unstartest
optimised system and compare the extra focal patterns to a 1/8
or 1/6 wavefront system with optimized patterns, does the 1/10
wavefront system test look worse?
Philip J. Blanda III
-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.1 for non-commercial use <http://www.pgp.com>
iQA/AwUBOM4tm8qM3UPhf8W7EQKK7ACg1CwL7uQh2BAtNKvs3FOxhjcvoV8AoI6u
kn1EGgDUbtl/BdfBqkf/X1OG
=uRrb
-----END PGP SIGNATURE-----
PGP Public key available
Chris1011
Newtonian mirror in practice, but a complex Maksutov system with
aspheric curves is a different story.
>>
Actually, it is a Mak-Cass with all spherical surfaces that would be more
likely to produce thes unsymmetrical results. When complex aspherics are added,
then the pattern would start to look more and more symmetrical.
The question is would a more symmetrical pattern be a better performer? Another
question is at what level do we have essentially a perfect amateur system. Is
that level 1/10 wave or does it need to be better than that? Judging by
reaction to some of the new telescopes now becoming available, it would seem
that 1/6 wave is quite acceptable.
Roland Christen
ASTRO-PHYSICS
ar...@selena.kherson.ua
Derek Wong <daw...@earthlink.net> wrote:
Hello Derek,
> > At this moment we are quite busy manufacturing 10" F/14 Mak-cass
> > systems for AP . I bet, you will be not able to judge which system is
> > better looking on their interferorgams.
>
> I'm sure you are right with this point, but...
No any "but" , please. :-))))
> > All of them leave nothing to
> > wish better,
>
> Wrong. I want a 12" Mak with a 20-23% obstruction which weighs close to
> 50 pounds :-)
>
> And yes, the seeing here justifies it.
No problem at all. Drop your name to Roland's notification list
for such 12" model. We will for sure will make them too just after
the 10" project will stabilize. This will be soon, I belive.
According to Roland's estimations, this 12" will weight close to
55 pounds. Exactly as you need.
Clear sky,
Valery.
ar...@selena.kherson.ua
lude...@my-deja.com wrote:
> In article <8ak9vj$t05$1...@nnrp1.deja.com>,exact> >
> > Derek,
> >
> > This is simple misleading, that perfect from practical point of view
> > system, MUST has a symmetrical diffraction patterns. Simple myth.
> > systems for AP . I bet, you will be not able to judge which system is
> > better looking on their interferorgams.
>
>
> a question: Lets estimate you are able as you say to produce same size,
> same focallenght ,same central obstruction MCT with same RMS and same
> P.t.v. but one with symmetrical image and one with asymmetrical image,
> would you see in an testreport made by an indepentent company show for
> both systems the same RMS, same P.t.v., same energyconcentrationand
> same MTF Profile or would there be a diffrence ? If there would be a
> diffrence, than why and what would be diffrent ? As you always agree
> the Energyconcentration is the point who shows the real quality, would
> it be the same ? If your answeres are yes, would you agree to give me 2
> such systems for retesting with the laser and wavelenght you recomment ?
>
> thanks for your answere
>
> Markus
Markus,
The answer on your question is very simple. Almost ALL scopes with
same , say, 98% Strehl, does has a different wave front maps. So,
they will show you a different off-focus images.
About MCT selection you may ask Roland. We making these
systems exclusively for him. I can't break my words given to him.
You are right about that the main factor is the energy concentration.
ar...@selena.kherson.ua
dcha...@gvi.net (Dan Chaffee) wrote:
> Dear Valery,
>
> You have given a specific example of the well corrected Mak Cass's
> asymmetry in the star test, but before this lapses into another
> possible hotbed of confusion, (if not debunking of "the book
> from the devil's library":-), let me ask you this question:
>
> Have you ever seen a well collimated, equalibriated newtoinan
> with near perfect star test symmetry(including smooth surface and
> no sign of astigmatism) be _out performed_ by another newtonian
> of equal aperture/obstruction/thermally stable/collimated that shows
> stronger differences in the star test asymmetry---on planets?(assuming
> equally good seeing, eyepieces,etc of course) I'm speaking of pure
> newtonian, not mak-newt, or a newtonian with optical window.
>
> What's missing in Nil's Olof's simulation is a real life comparison
> of the actual damage done to a low contrast, extended object. So it's
> not clear from this that the less symmetrical star test could actually
> be the better performer(or as good) --even with what is theoretically
> the same rms error.
>
> Dan Chaffee
> Kansas City
Hello Dan,
I don't remember all my youth-time optical works when I was
involved in such simple optics manufacturing as Newtonians.
On your question I can say, that in my opinion there is a
possibility for some kind of correction with better symmetry
to be less good than another systems with less symmetry.
All depends of RMS wave front error, or more correct - from
Nils Olof Carlin
In an earlier thread it was suggested that some "star-testers" would declare
their opinions of telescopes using the "secondary breakout points", that is
the defocusing needed to see the secondary shadow emerge in the diffraction
pattern, each side of focus. This is described in Suiter's "Star testing
Astronomical telescopes", where it is used with a fixed obstuction of 1/3
diam, and applied to determine the "correction", i.e. the 3rd order
spherical aberration. It is NOT valid for higher-order aberration.
My idea was that the asymmetry of the "breakout" would be largely dependent
on the "flatness" of the wavefront near the obstruction, and if this
isolated area focused close to the total best focus, the "breakout" would be
fairly symmetric.
To test this idea, I made up (using Zernike polynomials) a rather "perverse"
example of two wavefronts with identic 3rd and 5th order deviations, but of
opposite sign - as Derek showed, the RMS deviations (and P-V for that
matter) are equal, but one is "flat" around the center and "sloping" at the
edge compared to best focus, the other is the opposite.
The examples do not rigidly prove anything, but strongly suggests that the
(perverted) "breakout" test would indeed give very different (and
misleading) results in the two cases.
The defocused images are more or less asymmetric, but the particular aspect
of asymmetry I have tested is the "breakout", but on the other hand the
"flat center" wavefront shows the asymmetric edge of a poor edge instead.
I haven't proved that the in-focus images are identical, even if I suspect
they may be. With 1/42 wave RMS, they would give a great performance if
ported to the real world ;-)
A few comments to some posts:
Bryan Greer said: "On the Mak, for instance, every time I change a surface
(R1) that improves the star test symmetry, the PSF value also slightly
improves...... But keep in mind, these exceptions to the rule (if they
exist) won't come into play for the vast majority of people who are trying
to use the star test. The star test is straightforward for the majority of
telescope designs typical amateurs come across.
"
To get something similar to my "perverted" results, you probably have to
change something to the other side of the optimum - if indeed possible in a
real case. But apparently Maks in particular can have higher order sph.abb.
thatcannot be evaluated like Suiter says about pure 3rd order.
Derek: "I'm going to talk to Nils Olof about Roland's other statement--that
a Mak with a big central obstruction can "hide" significant aberrations and
have a symmetric star test."
At least I would guess it can give better symmetry of the "breakout"! But I
haven't tried.
Dan Chaffee: "What's missing in Nils Olof's simulation is a real life
comparison of the actual damage done to a low contrast, extended object."
The Strehl ratio is the same, but if the in-focus images are different, I do
not think the difference would be conspicuous in real life. The simulation
is hard to do at true focus, but very close (1 mm off, compared to 6 mm in
the images on Derek's page) showed very similar images in all cases.
Dan again: "At best , this example of Zernike polynomial expansion only
indicates
that a less symmetrical star test _may_ be an equal performer to one close
to symmetry, not that a scope testing closer to symmetry would produce worse
images than one less symmetrical. Yet, a paraboloid's star test lacking in
symmetry may very well (and always has been in my experience) be a poorer
performer to one more similar either side of focus."
This makes perfect sense to me, but again my simulation tried to illustrate
the effects on "breakout" symmetry, not of symmetry of the defocused pattern
in general. And the aberration is a *very special* mix of 3rd and 5th
terms - if likely at all, more likely to be realized with a Mak than with a
Newt - I don't know about refractors.
Nils Olof
Bryan Greer
> Bryan Greer wrote:
>
> > Hello Mr. Wong,
>
> Bryan, kind of formal for someone you met at RTMC a couple years ago and
> who bought one of your fine diagonals :-)
Yes! I apologize for the mental lapse, Derek. I believe I met you and Akana that
day. The long hours are getting to me.
> Ah, but this is exactly what was shown here--the answer to your last
> question is no. Both "Peak" and "Flat" have exactly the same size
> aberrations, 3rd and 5th order and rms values. "Flat" has a much better
> star test than "Peak", so a better star test does NOT always mean better
> optical performance. This is a simple Newtonian, although again I don't
> know if these mirrors would ever be produced in the real world.
I'm sorry, but I guess I don't understand exactly what Nils did here (I thought I
did). I thought he simply showed that higher order S.A. can introduce additional
asymmetry that can fool someone in a star test. Could you please privately resend
me your original posting? Please send it to bgr...@fpi-protostar.com. I've
deleted it, and I should have reviewed it more carefully the first time.
> Of course, although I wouldn't want to tweak a 1/42 wave rms scope!
Neither would I! :-)
> 1. It is POSSIBLE to have an excellent scope (1/10 wave p to v
> wavefront with 1/42 wave rms) with a very asymmetric star test.
I agree, and Suiter acknowledges this as well. Again, the question I have is does
aspherizing to improve the symmetry also improve the in-focus PSF.
BTW, if you or Nils will be attending any of the major starparties this year,
please let me know. I'd like to meet you again, and Nils for the first time. I
know it's a long trip for Nils. :-)
WHALEN44
I think for the majority of observers, a very smooth 1/6 wave scope (around
1/40 wave RMS or better) would be very acceptable. I would guess less than 5%
of amatuers out there have owned a very smooth 1/6 wave (or better) scope.
However there is as you well know, another group that is within that 5%, that
wants more. Perhaps we should only be talking RMS, but I still think P-V values
also have some meaning.
I think high end scopes such as your MCT's need to be at least 1/8 wave P-V,
and 1/50 wave RMS in the larger and more expensive sizes. Otherwise what's the
point?
I would guess the "perfect" amatuer system to be around 1/15 wave P-V, with an
RMS around 1/75 wave. However I don't think we will ever see this in a
production optic set. What do you think?
Roland wrote> Another question is at what level do we have essentially a
perfect amateur system. Is that level 1/10 wave or does it need to be better
than that? Judging by reaction to some of the new telescopes now becoming
available, it would seem that 1/6 wave is quite acceptable. >
>Roland Christen
>ASTRO-PHYSICS
Richard Whalen
whal...@aol.com
"Time spent observing the heavens is not deducted from your lifespan"
RAnder3127
>systems exclusively for him. I can't break my words given to him.
Ironic, isn't it? Years ago, mfg's boasted
about their optic's corrections because they meant nothing and they had no
basis in reality. Now that they HAVE to mean something, they have suddenly
become
proprietary?
-Rich
Al Gore: "I don't care if gas hits $4 per gallon. The American people have
had
it too cheap for too long. How will we ever move to clean fuels like
methanol if gasoline remains inexpensive? $5000/yr is ok to get
cleaner air." :)
Chris1011
I think for the majority of observers, a very smooth 1/6 wave scope (around
1/40 wave RMS or better) would be very acceptable. I would guess less than 5%
of amatuers out there have owned a very smooth 1/6 wave (or better) scope.
>>
I've used a 1/6 wave SCT, but with rather rough and abrupt zonal errors. This
scope did not have good contrast as compared to other, smoother systems. Was
your 5.7" Ceravolo a 1/6 wave smooth, or was it better? How did it perform
compared to your new 8" TEC?
I agree that for high end larger systems the RMS smoothness should exceed 1/50
RMS. Systems like that really snap in even when the seeing is not so good - and
that's all I've experienced around here for the last 3 months - not so good
seeing.
Roland Christen
ASTRO-PHYSICS
WHALEN44
The HD was around 1/6 wave and very smooth, though I think but for one very
small spot it would have been around
1/9 wave. It's the only scope I have ever owned that star tested significantly
better than it tested by interferometer.....
Comparing the TEC to the HD is tough, as they are so different in about every
respect.
5.7" verses 8", f6 verses f 15.5, different eyepiece positions etc.
I do prefer the TEC, as it has even better optics in a larger aperture along
with more image scale and better observing position.
I do miss the wider views of the HD once in a while, though I really never used
it that much for wide field observing. I do miss its lightweight and
portability though.
What the TEC has done I guess is give me more versatility with it's added
aperture and contrast.
The contrast of this design is simply amazing. As one very experienced
observer said at Highlands Stargaze, " black does not get any blacker" . Also I
am amazed at how well the Maksutov Cassegrain design handles high
magnification. While my MNT topped out at about 80x per inch, at that ratio the
MCT is just starting to get interesting.
I am 200% sold on high quality MCT's...
<Was your 5.7" Ceravolo a 1/6 wave smooth, or was it better? How did it perform
compared to your new 8" TEC?>
Nils Olof Carlin
have not seen it on the list.
-------------------
ar...@selena.kherson.ua
whal...@aol.com (WHALEN44) wrote:
> Hi Roland,
>
> The HD was around 1/6 wave and very smooth, though I think but for one very
> small spot it would have been around
> 1/9 wave. It's the only scope I have ever owned that star tested significantly
> better than it tested by interferometer.....
Hi Rich!
Above your info is next in a row to show us that such ancient
definition as P-V should be forgotten and as maximum can be
used to describe a 3-order spherical aberration correction. But
this can mislead one easilty. So, the best way is to forget the
P-V. The RMS is quite enough to describe the quality of any
given system.
I am glad t hear that you already 200% sold for MCT with small
c.o.
Valery Deryuzhin
ARIES INSTRUMENTS Co.
lude...@my-deja.com
> Hi Rich!
>
> Above your info is next in a row to show us that such ancient
> definition as P-V should be forgotten and as maximum can be
> used to describe a 3-order spherical aberration correction. But
> this can mislead one easilty. So, the best way is to forget the
> P-V. The RMS is quite enough to describe the quality of any
> given system.
Hi Valery,
you are 200% right here, it was the reason for Zeiss not to print
anymore the p.t.v and even Todd Gross cannot believe that his 4" have
,after startesting was done, only 1/5 wave, which is a smaller local
error.
Markus
Bob May
Firebird is better! Lets give up the silly number's game on the
newsgroup.
--
Bob May
I don't read attachments to posts as they may give me a
virus If I expect an attachment from you I will open it..
You may have a brilliant thought but if you put it into an
attachment I won't read it and thus both you and I lose.
I don't like to say it but unfortunatly, there are those who
insist upon being nasty to the rest of us. Bob May
ar...@selena.kherson.ua
lude...@my-deja.com wrote:
> In article <8anetp$4kk$1...@nnrp1.deja.com>,>
> > Hi Rich!
> >
> > Above your info is next in a row to show us that such ancient
> > definition as P-V should be forgotten and as maximum can be
> > used to describe a 3-order spherical aberration correction. But
> > P-V. The RMS is quite enough to describe the quality of any
> > given system.
>
> Hi Valery,
>
> you are 200% right here, it was the reason for Zeiss not to print
> anymore the p.t.v and even Todd Gross cannot believe that his 4" have
> ,after startesting was done, only 1/5 wave, which is a smaller local
> error.
>
> Markus
Of course, I am right. :-)
Anekdot:
Now we will wait star-testers experts estimations given in RMS
and Strehl format !! Hey, guys, imagine, please, a follow picture:
Markus and his best friend(s) will walk around, say, AstroFest ,
telescope field. Then they look trough their competitor scope(s)
and speak:
Markus (M) :" I found it to be 0.028RMS in a green light."
Friend(F) : "I think it was slightly better - 0.025 in a geen light".
M: "OK, how you found it? "
Markus, no offence, please. Just joke with a good part of
truth. To remain as test expert, soon, you will need to take
an appropriate equipmet on all star-parties you will attend
(perfect flat, interferometer, electro generator, installation kits).
This will be the results is you will refuse from P-V and still
wish to be an expert in testing. :-)))))))
Valery.
ZodiacMan
Derek Wong <daw...@earthlink.net> wrote:
> Please note that errors are possible, and comments are welcome. Can
> someone simulate this in another program like Zemax to confirm? Nils
> Olof knows a lot more about optics and the DIFFRACT program than I do,
> so I put his e-mail at the bottom of the web page. The "ugly" math
> discussion is below for those of you not sleeping :-)
> Derek
>
I am lost here. What are you all trying to prove or to teach us in
this thread?
I believe Derek has read "Telescope Optics: Evaluation and Design" by
Rutten and Venrooij. But, unlike me, I think Derek actually understood
what he was reading.
I read Suiter's "Star Testing" book. I didn't understand 90 percent of
it, but I did like the pictures. <grin> The useful part of the book
for me came from the pictures, and when he summarized at the beginning
or at the end of the chapter what point he was trying to get across to
amateurs like me.
Regards,
--
ZodiacMan
lude...@my-deja.com
arggghhhhhhh,
Valery, it seems you can never stop , he ?
Markus
> Anekdot:
> Now we will wait star-testers experts estimations given in RMS
> and Strehl format !! Hey, guys, imagine, please, a follow picture:
> Markus and his best friend(s) will walk around, say, AstroFest ,
> telescope field. Then they look trough their competitor scope(s)
> and speak:
>
> Markus (M) :" I found it to be 0.028RMS in a green light."
> Friend(F) : "I think it was slightly better - 0.025 in a geen light".
> M: "OK, how you found it? "
>
> Markus, no offence, please. Just joke with a good part of
> truth. To remain as test expert, soon, you will need to take
> an appropriate equipmet on all star-parties you will attend
> (perfect flat, interferometer, electro generator, installation kits).
> This will be the results is you will refuse from P-V and still
> wish to be an expert in testing. :-)))))))
>
> Valery.
>
Bryan Greer
I re-read Nils's original analysis more carefully (thanks for resending this
info to me). His conclusion that the shadow breakout is sensitive to the wave
front profile near the central obstruction is interesting. In fact, I came
across a real telescope two years ago that seems to support this. A friend's
medium sized, thin mirror, Dob was giving a conflicting star test. Per the
shadow breakout test, it appeared undercorrected, but it also had fuzziness
inside of focus characteristic of overcorrection (or maybe a very broad turned
edge). He was still in the process of figuring it, so it was important that we
really understand what was going on. We masked the mirror with different sized
masks to figure it all out. We determined that the mirror was undercorrected
from the center out to maybe 50%, where it became close to null. The extreme
outer zone was a little overcorrected, and this was causing all the fuzz inside
focus. (I also computer simulated star test images for the surface we thought
we had, and it did match closely what we were seeing.) Based on this, I
wondered if the behavior of the breakout was more closely tied to the inner
zones. Nils's experiment seems to indicate it does to some extent.
In any case, this is good reason to heed Suiter's warning against using the
breakout test as the sole method of determining S.A. Small paraboloids probably
are much more likely to have simple 3rd order S.A. dominate, but clearly these
larger (and thinner) mirrors can have some exotic blends. :-)
Nils wrote:
> I haven't proved that the in-focus images are identical, even if I
> suspect they may be.
>
I agree. I would expect the in-focus performance to be the same too.
In regards to the original issue of whether a better star test always means
better in-focus performance, I don't think Nils's analysis addresses this
directly. (I understand it wasn't Nils' intent to prove/disprove that, so this
is not a criticism.) Unless I'm missing something (quite possible), it seems to
show that two optical systems with the same rms and P-V errors can show
different star tests. I don't think anyone will dispute this. After all, there
are an infinite number of surfaces that can have the same rms and P-V values.
Dan Chaffee wrote:
> At best , this example of Zernike polynomial expansion only
> indicates that a less symmetrical star test _may_ be an equal
> performer to one close to symmetry, not that a scope testing
> closer to symmetry would produce worse images than one less
> symmetrical.
>
This is my point, too. I've spent a little time trying to simulate such a
critter, and haven't been able to yet. I realize that's not proof, but it
reinforces my suspicions at least. :-) And as Nils fairly pointed out, this was
not the intent of his experiment anyway.
Dan also wrote:
> Yet, a paraboloid's star test lacking in symmetry may very well
> (and always has been in my experience) be a poorer performer
> to one more similar either side of focus.
>
My experience too. :-)
ar...@selena.kherson.ua
lude...@my-deja.com wrote:
> In article <8apbnl$iu4$1...@nnrp1.deja.com>,> >>
>
> arggghhhhhhh,
> Valery, it seems you can never stop , he ?
>
> Markus
Markus! Do you really loose your humor sense?
BTW. I will easily pay 20 baks if I ever see this scene
in reality. :-)
So, I think that P-V is still important, however.
Nils Olof Carlin
ZodiacMan wrote:
>I am lost here. What are you all trying to prove or to teach us in
>this thread?
The essential message is that star testing (contrary to the opening
statement of Suiter's book) isn't so easy - and it seems from previous
threads that there are people that judge telescope optics without deep
enough understanding of the fundamentals, and that their judgments sometimes
are highly misleading. Also it was stated that it is possible to "tweak" the
optical figure of a telescope (Maksutov) to satisfy these star-tersters'
ideas, without actually improving the performance in focus.
So my idea was to make up two "simulated mirrors" that would have similar if
not identical performance in focus, but would appear very different when
star-tested by the inappropriate criteria - demonstrating that such tests
may mean nothing at all IF the aberration is a combination of different
orders of spherical aberration.
It also demonstrates the power of the internet - though we are separated by
continents and oceans, with the exchange of a few emails, Derek Wong could
supply the mathematic proofs that were needed to lead my original hunch to a
meaningful conclusion! And for that matter, the program used to make the
images was a result of such idea exchange with another amateur (Jim
Burrows).
>I read Suiter's "Star Testing" book. I didn't understand 90 percent of
>it, but I did like the pictures. <grin> The useful part of the book
>for me came from the pictures, and when he summarized at the beginning
>or at the end of the chapter what point he was trying to get across to
>amateurs like me.
It is a marvellous book, but it is not easy ;-) The use of the
orthogonality of the Zernike polynomials was directly inspired from this
book - I don't see how else it could have been done.
Regards,
Nils olof
Regards,
Nils Olof
Nils Olof Carlin
Bryan Greer wrote
>I re-read Nils's original analysis more carefully (thanks for resending
this
>info to me). His conclusion that the shadow breakout is sensitive to the
wave
>front profile near the central obstruction is interesting. In fact, I came
>across a real telescope two years ago that seems to support this. A
friend's
>medium sized, thin mirror, Dob was giving a conflicting star test. Per the
>shadow breakout test, it appeared undercorrected, but it also had fuzziness
>inside of focus characteristic of overcorrection (or maybe a very broad
turned
>edge).
This illustrates the difficulties of doing eye-ball Fourier transforms -
particularly when you do not even have the phase of the function to
transform ;-)
This is what star testing actually is about - obtaining the Fourier
transforms of aberrated waveforms that are intentionally enhanced by adding
various amounts of 1st order spherical aberration - and without the phase
information!
I don't know enough maths to guess if this is generally possible - but some
aberrations do indeed produce very characteristic patterns that may help you
decide the nature, and to some extent the amount, of aberrations.
But considering the complexity of the (fourier transformed) patterns makes
the notion of "symmetric star test" (or the opposite) somewhat suspicious to
me - or "better" star test! The degree of apparent symmetry may be
misleading, just as the narrower concept of secondary breakout symmetry
seems to be.
>He was still in the process of figuring it, so it was important that we
>really understand what was going on. We masked the mirror with different
sized
>masks to figure it all out. We determined that the mirror was
undercorrected
>from the center out to maybe 50%, where it became close to null. The
extreme
>outer zone was a little overcorrected, and this was causing all the fuzz
inside
>focus.
It kind of srikes me that a good old-fashioned Foucault test would have
showed the same thing even more simply ;-)
>......Based on this, I
>wondered if the behavior of the breakout was more closely tied to the inner
>zones. Nils's experiment seems to indicate it does to some extent.
When star-testing, we move into the domain where geometric optics starts to
be meaningful, even if only to a limited degree. But this suggested to me
that the behaviour of the outer parts of the defocused image would be
largely determined by the edge, and the inner parts likewise by the central
parts of the aperture (near the obstruction, if any). So far, this idea
seems to make sense.
>In any case, this is good reason to heed Suiter's warning against using the
>breakout test as the sole method of determining S.A. Small paraboloids
probably
>are much more likely to have simple 3rd order S.A. dominate, but clearly
these
>larger (and thinner) mirrors can have some exotic blends. :-)
I guess this is valid for spherical surface optics in general - you could
pseudo-star test Maks, but I suspect the refractor owners won't silently
accept stick-on "obstructions" for the break-out tests.
>Unless I'm missing something (quite possible), it seems to
>show that two optical systems with the same rms and P-V errors can show
>different star tests.
You didn't miss....
>> At best , this example of Zernike polynomial expansion only
>> indicates that a less symmetrical star test _may_ be an equal
>> performer to one close to symmetry, not that a scope testing
>> closer to symmetry would produce worse images than one less
>> symmetrical.
>>
>This is my point, too. I've spent a little time trying to simulate such a
>critter, and haven't been able to yet.
My "flat" image, at the set defocusing, appears more "symmetric" than the
"peak" one, but again I am suspicious of comparing "symmetry" in general.
Nils Olof
N 56 deg 26 min, E 13 deg 50 min
lude...@my-deja.com
> Markus! Do you really loose your humor sense?
I think you know me long time enough, yes ?
>
> BTW. I will easily pay 20 baks if I ever see this scene
> in reality. :-)
this is a great offer from you , let me think about.
> So, I think that P-V is still important, however.
not for you as we know, but for us startester of course
Markus
lude...@my-deja.com
still important, however.
>
> Valery.
I forgot one thing: Do you agree that a smooth optics can have an
smooth turned edge , where the RMS still can be very good ? Do you
agree that a smooth zone can still provide a good RMS ? So we need the
p.t.v. too of course, otherwise you have wild free live to do what you
want, without an customerchance of claim
Markus
lude...@my-deja.com
"Nils Olof Carlin" <nilsolo...@telia.com> wrote:
>
Hi Nils and Brian,
I feel, something is fully missed here. Everybody must think now, that
a startest is only done by comparing inside and outside focus image at
some disctance from the focus, correct ? Is that all the startest what
experienced startesters are doing ? Of course not, or you doing the
startest in an wrong way.
The real startest can be made in 3 diffrent ways:
Nr1, as everybody is doing it mostly: going a little deeper inside and
outside of focus and comparing the images. Thats the way everybody
discuss it
Nr.2, doing the startest infocus. A scope with a spherical aberration
showing and brighter thicker diffraction ring , than a scope with no or
nearly no spherical aberration. Here it is easy to check an 1/4 wave
from an 1/6 or 1/8 wave scope, if you have enough experience and other
prooved scope to compare with
Nr3, doing the startest at highest power just going a very little bid
in and out of focus. Most scopes which looks superb at method Nr1,
showing here real face by the method Nr3, a soft outside image and
sharpinside image
There is no way out to say that Method Nr2 and Nr3 are not a correct
method to check quality, even aspherized optics must pass method Nr.2
and Nr.3, or they have an problem.
So doing the startest correctly, still makes sense and give you the
right impression about the quality of the scope
Markus
bratis...@my-deja.com
> Now we will wait star-testers experts estimations given in RMS
> and Strehl format !! Hey, guys, imagine, please, a follow picture:
> Markus and his best friend(s) will walk around, say, AstroFest ,
> telescope field. Then they look trough their competitor scope(s)
> and speak:
>
> Markus (M) :" I found it to be 0.028RMS in a green light."
> Friend(F) : "I think it was slightly better - 0.025 in a geen light".
> M: "OK, how you found it? "
Jokes aside, I'm quite prepare to take a bet that many people CAN
distinguish Strehl to within 0.1 or even better (yours truly included).
I can't have it to three decimal places, but I will always distinguish
between 0.8, 0.9 and 0.95+, high order spherical present or not. Just
by using extrafocal and in-focus star test. Seeing of course has to be
excellent, at least 8 or 9 Pickering for aperture in question (I have to
be able to see diffraction rings).
I'm also sure that Thomas, Markus and few other people would easily
take on your bet too.
And I don't know about others, for me if a wavefront entering my eye
deviates Strehl less than 5%, I don't care if it is .96, 0.97 or
0.9999 .
Bratislav
valery_...@my-deja.com
lude...@my-deja.com wrote:
> I forgot one thing: Do you agree that a smooth optics can have an
> smooth turned edge , where the RMS still can be very good ? Do you
> agree that a smooth zone can still provide a good RMS ? So we need the
> p.t.v. too of course, otherwise you have wild free live to do what you
> want, without an customerchance of claim
>
> Markus
Good question. I prefer optics with smooth zone, no matter
what P-V is (of course, in a smart limits) but with low enough
RMS (somewhat 0.02 wave) vs flat wave front say, 1/10
P-V and poor smoothness and 0.03 RMS.
Valery Deryuzhin.
ARIES..
valery_...@my-deja.com
lude...@my-deja.com wrote:
> In article <8arehp$3ds$1...@nnrp1.deja.com>,>
> > Markus! Do you really loose your humor sense?
>
> I think you know me long time enough, yes ?
>
I see now that I know you a long time, but still not good enough.
Before, you always looks quite humoric guy. But now....
What really happens? I am a bit worry about you. :-)
Valery.
ar...@selena.kherson.ua
A ga. Especially if we add 33% c.o. (it will take its participation
in Strehl decreasing to somewhat 0.8+ ) and you will say -" this is 96%
Strehl optics." Quite funny.
valery_...@my-deja.com
Even after such detailed explanations of your own experience,
the Star Test credibility still not looks convincing, however.
Valery.
john
It would be interesting to run MTF tests on two such mirrors to see which
preserves greatest contrast.
John Lawler
RAnder3127
simply too much time required to
produce optics which are for all intents
and purposes, perfect. I'm sure it's possible, but is it even remotely
affordable? The current price for a 2"
lens with a complex aspheric surface from
an industry-oriented lens mfg. is something like $8000. That's for two
surfaces.
bratis...@my-deja.com
> A ga. Especially if we add 33% c.o. (it will take its participation
> in Strehl decreasing to somewhat 0.8+ ) and you will say -" this is
96%
> Strehl optics." Quite funny.
Valery,
you should know better. 33% c.o drops Strehl to about 90%, not 80%.
If you don't trust me, try your Zemax.
Anyway, yes, I'll estimate final Strehl with or without obstruction
(as effect of obstruction is easy to calculate and apply to final
Strehl - you just multiply them. And drop in SR because of obstruction
can very closely be approximated with area loss. Again, if you don't
believe me, try Zemax. Or solve the integral :-)
Bratislav
valery_...@my-deja.com
Bratislav,
I belive you. I just don't have to keep in mind such unimportant
exact figures. I wrote .8+ where the + is undefined and keeping
in mind that a systems typically far from perfect, my fast .8 + is
better describes the reality.
You impress me by your ability to make express image analysis
and results reduction just in memory.
OK, now I know that I was mistaken in my inoffensive anekdot
and I see that it was not so far from reality and I need to add one
more personage to a story about the most modern star test . :-)))
I can only imagine one's (novice!) confusion when he will be told:
" Sir, your scope is somewhat 0.035 RMS + c.o. participation."
What he must feel?
~( %^ ( Nightmare!!!
My best regards,
Valery.
valery_...@my-deja.com
rande...@aol.com (RAnder3127) wrote:
> Maybe we should ask if it there is
> simply too much time required to
> produce optics which are for all intents
> and purposes, perfect. I'm sure it's possible, but is it even remotely
> affordable? The current price for a 2"
> lens with a complex aspheric surface from
> an industry-oriented lens mfg. is something like $8000. That's for two
> surfaces.
> -Rich
You right Rich. I only can add that such small thing like 2"
with complex aspherics is more difficult to make than 6" with
same kind of aspherics.
But this is slightly off-topic. Don't you think?
Dan Chaffee
On Thu, 16 Mar 2000 20:11:20 GMT, "Nils Olof Carlin"
<nilsolo...@telia.com> wrote:
>> We determined that the mirror was
>undercorrected
>>from the center out to maybe 50%, where it became close to null. The
>extreme
>>outer zone was a little overcorrected, and this was causing all the fuzz
>inside
>>focus.
>
>It kind of srikes me that a good old-fashioned Foucault test would have
>showed the same thing even more simply ;-)
Well, Bryan didn't say but, it
depends on the focal ratio, don't you think? Anything under f/6 isn't
as readily quantified, especially between center zone and 50%zone.
Actually, I find it rather a tough call for optics considerably slower
under the Foucault for these zones. For us 'low tech' amateurs,
nothing beats cross-referencing knife edge with star test patterns,
IMHO.
>
>
> >......Based on this, I
>>wondered if the behavior of the breakout was more closely tied to the inner
>>zones. Nils's experiment seems to indicate it does to some extent.
>
and On Thu, 16 Mar 2000 12:18:17 -0500, Bryan Greer
<bgreer@*takeout*fpi-protostar.com> wrote:
>
>>In any case, this is good reason to heed Suiter's warning against using the
>>breakout test as the sole method of determining S.A. Small paraboloids
>probably
>>are much more likely to have simple 3rd order S.A. dominate, but clearly
>these
>>larger (and thinner) mirrors can have some exotic blends. :-)
I too have difficulties with trying to be sure of a quantitative
evaluation with shadow breakout., and generally it is of less
consequence to me when evaluating the diffraction patterns
than looking at the overall structure and trying to envision
the cross section of the caustic horn. Yeah, I know, thumbs down
for geometric optics...<g>
>
>>Unless I'm missing something (quite possible), it seems to
>>show that two optical systems with the same rms and P-V errors can show
>>different star tests.
But can two (of the same kind) optical systems showing the same
startest have different rms & P-V errors?
Dan Chaffee
Nils Olof Carlin
Dan ,
>>It kind of srikes me that a good old-fashioned Foucault test would have
>>showed the same thing even more simply ;-)
>
>Well, Bryan didn't say but, it
>depends on the focal ratio, don't you think? Anything under f/6 isn't
>as readily quantified, especially between center zone and 50%zone.
Depends - "naked" pinstick may not be much good here but with Couder mask
you lose no accuracy, I am sure. The accuracy is determined by the zonal
width and only indirectly on the f/ratio - or more accurately the focal
length (you can check my ideas on the subject at www.atmsite.org - they
might be somewhat against conventional wisdom).
>Actually, I find it rather a tough call for optics considerably slower
>under the Foucault for these zones.
I take it you mean unmasked Foucault? In my opinion, masked should be safe
from the potential errors of unmasked test (others may claim the
opposite...).
> For us 'low tech' amateurs,
>nothing beats cross-referencing knife edge with star test patterns,
>IMHO.
No doubt - but in the present case, I suggested the other way round -
cross-referencing star test patterns with knife edge testing ;-)
>I too have difficulties with trying to be sure of a quantitative
>evaluation with shadow breakout., and generally it is of less
>consequence to me when evaluating the diffraction patterns
>than looking at the overall structure and trying to envision
>the cross section of the caustic horn. Yeah, I know, thumbs down
>for geometric optics...<g>
>
>>
>>>Unless I'm missing something (quite possible), it seems to
>>>show that two optical systems with the same rms and P-V errors can show
>>>different star tests.
>
>But can two (of the same kind) optical systems showing the same
>startest have different rms & P-V errors?
Not *exactly* - but the phase information in the startest image is not
recoverable, and what similar looks would mean, I dare not guess. Star test
images can be misleading......
Nils Olof
lude...@my-deja.com
Valery,
do we have now 2 types of strehl ? One real tested and one recalculated
?
For sample your 10" MCT with 23% or 25% C.O. will be tested in your
interferometer with 1/50 RMS, will it have an print out on the
testdocument of 98,4 % or will it be printed out as about 90% ? If the
print out will be 98,4%, than I dont understand , what you try to say.
If it not will be print out 98,4% but 90%, than you call Zeiss and
other instituts who makes honest testreport on Telescopes with an C.O.
liers, yes ?
I know well of the recalculation by removing the secondary influence
and therefore the decreasing of the strehl, but this have nothing to do
with the optical quality. The recalculation is only made to be able to
compare it with an nonobstructed telescope, the Telescope with C.O.
will still have the 98.4% strehl.
Markus
Markus
ar...@selena.kherson.ua
lude...@my-deja.com wrote:
> In article <8asltl$vd7$1...@nnrp1.deja.com>,> >
>
> Valery,
>
> do we have now 2 types of strehl ? One real tested and one recalculated
> ?
>
> For sample your 10" MCT with 23% or 25% C.O. will be tested in your
> interferometer with 1/50 RMS, will it have an print out on the
> testdocument of 98,4 % or will it be printed out as about 90% ? If the
> print out will be 98,4%, than I dont understand , what you try to say.
> If it not will be print out 98,4% but 90%, than you call Zeiss and
> other instituts who makes honest testreport on Telescopes with an C.O.
> liers, yes ?
> I know well of the recalculation by removing the secondary influence
> and therefore the decreasing of the strehl, but this have nothing to do
> with the optical quality. The recalculation is only made to be able to
> compare it with an nonobstructed telescope, the Telescope with C.O.
> will still have the 98.4% strehl.
>
> Markus
Again liers? What with you?
Secondly. What do you think why we try to make the c.o. in
our optics as smaller as possible ( 23% ) ? Why Intes try to make
their MN with small c.o. ?
I hope all is clear.
APM
>
>Again liers? What with you?
Valery,
could you please explain, what you mean with above :Again Liers ?
>
>Secondly. What do you think why we try to make the c.o. in
>our optics as smaller as possible ( 23% ) ? Why Intes try to
make
>their MN with small c.o. ?
to get the better contrast of course
waiting your answere
Markus
>I hope all is clear.
>
>
>Valery.
>
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.
>
>
* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!
Bryan Greer
Nils wrote:
> But considering the complexity of the (fourier transformed)
> patterns makes the notion of "symmetric star test" (or the
> opposite) somewhat suspicious to me - or "better" star test!
>
From a theoretical standpoint, perfect symmetry _always_ means a perfect,
spherical, waveform. Asymmetry _always_ means there is some sort of aberration
present. Forgetting the telescope for a moment, this is just how a spherical
waveform collapses as it approaches the diffraction minima. Even a surface that
has an overall perfect correction, but localized roughness, will have slightly
different in and out patterns (in theory), though no one will be able to see it
as anything other than decreased contrast in the rings.
Nils also wrote:
> The degree of apparent symmetry may be misleading, just
> as the narrower concept of secondary breakout symmetry
> seems to be.
>
I completely agree, and your experiment showed on example of this. There
certainly _is_ room for debate as to how much perceived asymmetry results in a
given amount of degradation to the image. I think this is really at the core of
this (endless <g>) discussion. The "hardware" required to produce these
wavefronts is, of course, SURFACES. The more surfaces that are involved, and/or
the more complex (i.e. aspheric) they are, the greater chance of having higher
order aberrations present. Apparently, this is why the Mak-Cass and fast APO
designs show some asymmetry in the star test. Here is where I have to bow out
to those who have more experience with these types of telescopes (Roland, et
al). I have no trouble believing that a design that shows a modest amount of
asymmetry can still have a good in-focus Strehl ratio. The only question
remaining in my mind is whether further manipulation of their surfaces to
improve symmetry would also improve the in-focus performance. I suspect it
does, but that it's neither practical nor necessary to do it. Bratislav sent me
a design of a Gregory Mak-Cass that had fantastic performance even with a bit of
an asymmetrical star test (not much, though <g>).
Another good example of this is astigmatism. Astigmatism is particularly
damaging to contrast on extended objects such as planets, and yet it is best
detected fairly close to focus. Some people don't even notice modest
astigmatism in the star test unless it's pointed out to them, and yet it can
really hurt the view.
Nils wrote:
> It kind of srikes me that a good old-fashioned Foucault test would
> have showed the same thing even more simply ;-)
>
He was figuring per the knife-edge test by day. However, there were a few
reasons to confirm things by night under the stars. First, it was an f/4.5, and
the Foucault isn't the best tool at that f-number. He was also fighting
astigmatism, and as you know, the Foucault is almost completely insensitive to
that. Finally, this was a large, thin, chunk of Pyrex (not the best material in
the world). The knife-edge readings tended to fluctuate during the daytime
figuring since he didn't allow the 2 to 3 hours for proper cooling between
readings. Despite all this, there was reasonable agreement between the
knife-edge and the star test. Of course, we believed the star test more. <g>
Nils wrote:
> but I suspect the refractor owners won't silently accept stick-on
> "obstructions" for the break-out tests.
>
You're probably right, but I do. <g> (http://www.fpi-protostar.com/dummy.jpg)
Dan Chaffee wrote:
> But can two (of the same kind) optical systems showing the same
> startest have different rms & P-V errors?
>
No, and therein lies the beauty of the point spread function (a.k.a. the "star
test"). The point spread function (PSF) is one (inverse) Fourier transform
removed from the actual complex pupil function (i.e. the wavefront). In theory,
you could take a PSF image, do an inverse Fourier transform to it, and end up
with the original wavefront map! In practice, this can't be done with one PSF
image since, as Nils points out, the phase information isn't present in the
image (only the scalar intensity is). However, it can be done if you have more
than one PSF image of precisely known distances from the best focus.
Mathematically, you can re-create the wavefront map. The neat thing about doing
this is that it even recreates a precise silhouetted image of the secondary
spider support (if the scope has one)! Even that information is in the PSF. <g>
Bryan Greer
> Hi Nils and Brian,
>
> I feel, something is fully missed here. Everybody must think now, that
> a startest is only done by comparing inside and outside focus image at
> some disctance from the focus, correct ? Is that all the startest what
> experienced startesters are doing ? Of course not, or you doing the
> startest in an wrong way.
> The real startest can be made in 3 diffrent ways:
>
Hi Markus,
Yes, I agree! I said the same thing in the similar thread last month.
Studying the symmetry of a star is just a subset of the whole "star test".
There's something to be learned from studying the in-focus Airy disk, the
ring structure further from focus, and the motion in the rings (whether it
be from atmosphere or "local seeing").
bratis...@my-deja.com
> do we have now 2 types of strehl ? One real tested and one
> recalculated
In fact we DO have "two types" of Strehl. Most packages that come with
interferormeter will calculate Strehl Ratio directly from the measured
RMS wavefront using one of approximated formulae (either common
1-(2*pi*sigma[rms])^2 or Mahajan's e^-(2*pi*sigma[rms])^2, both
approximating the cosine terms -or real part of the integral- squared).
But story doesn't finish there, and while obstruction DOESN'T affect
RMS (as there are no phase errors), it will affect the energy
distribution (which is what Strehl Ratio is all about). So we have to
apply the 'corrected Strehl' afterwards for the obstructed systems,
which is roughly proportional to the AREA loss (i.e. obstruction of
0.15 will reduce Strehl by about 2.2%) - for reasonable obstructions,
that is. This bit, as far as I know, is never done by companies that
make obstructed optics.
Some will argue that Strehl is too lenient to obstructions, as it only
looks at center of the Airy disc, and will actually look at some other
methods to assess the damage (like Suiter's method of calculating all
the energy within the first minimum, EER(0) ).
Bratislav
PS there is a THIRD component of Strehl ratio, which is seeing related.
That one of course is never given as potential customers of expensive
optics would have a heart attack ;-)
lude...@my-deja.com
bratis...@my-deja.com wrote:
> In article <8atiug$jnj$1...@nnrp1.deja.com>,
> lude...@my-deja.com wrote:
Hi Bratislav,
who brought out the strehlratio name ? I dont know it, maybe you know
it, and maybe you know the explanation of strehlratio from the person
who created this name. All I have is from the Book Astrooptik written
by the Zeiss Jena APQ and Zeiss Jena Meniscas designer. It is very
difficult explained, so difficult that I cannopt prober translate it.
Maybe in this groupd somebody is out who could translate the zeiss
definition of strehl if I post it in german language ? I think Roland
could translate correctly, but I am not shure, that he would be willing
to do it for me.
thanks
Markus
However let me try to translate, but accept if my translation is not
fully correct translated:
Definitionshelligkeit (german word for strehlratio): Waveoptical
Imagecriterium is the relativly intensity of the pointimage in one
connectionpoint ( maximal or geometricaloptical imagepoint).It can be
closly calculated over the zernike-coefficient ( valuerange 0.8-1),
another possibility of the calculation is possible from the
waveabberation. Diffractionlimited ( Strehl bigger than 0.8), which
means geometrical optical diagonalabberation lays in the size of the
diffractionunsharpness.The strehlratio decrease due
secondarymirrors.Since the secondary shadow have only an small part of
the optical area, relativly to the edgearea, is an nearing in the
calculation of the effektive strehl tolerable and acceptable.
ar...@selena.kherson.ua
APM <ludesN...@ur.pils.de.invalid> wrote:
> In article <8auf71$9p3$1...@nnrp1.deja.com>, >> Markus
> >
> >Again liers? What with you?
>
> Valery,
>
> could you please explain, what you mean with above :Again Liers ?
> waiting your answere
>
> Markus
You wrote:
<than you call Zeiss and other instituts who makes honest
testreport on Telescopes with an C.O. liers, yes ?>
I did mean you AGAIN see a liers around.
May be this is paranoya? - only liers, liers and liers around.
Is not enough? When you will see in peoples something another
than liers?
Valery.
bratis...@my-deja.com
the definition of Strehl Ratio will be the same everywhere. It is the
ratio of intensity at the center of the diffraction disc expressed as
fraction of a theoretical maximum for a given aperture. So for a perfect
scope it will be 1/1=100%, for less perfect ones progressively smaller.
Some people may use theoretical maximum for a given CONFIGURATION
(i.e. taking into the account obstruction upfront) but that is in
my opinion wrong.
As soon as wavefront departs from perfectly spherical, or there is an
obstruction, or image motion ("seeing"), the intensity at the center
of the Airy disc will drop. For more details see Born & Wolf, Mahajan,
or for kind of a light reading, Schroeder's collection of articles
in SPIE milestone series or Suiter's book. in fact, all of those books
together cost less than a single premium eyepiece !
Bratislav
PS I don't really know why SR of 0.8 happens to be accepted as
'diffrection limited'. Could it be for historical reasons, as Lord
Rayleigh's "diffraction limit" third order spherical of 1/4 wave
happens to sit almost exactly at SR of 0.8? Does anyone know for
a fact?
RAnder3127
>fraction of a theoretical maximum for a given aperture. So for a perfect
>scope it will be 1/1=100%, for less perfect ones progressively smaller
Is it possible to get more than 84% of
the light into the diffracton disc or is
16% always going to go into the rings?
Zane
>>ratio of intensity at the center of the diffraction disc expressed as
>>fraction of a theoretical maximum for a given aperture. So for a perfect
>>scope it will be 1/1=100%, for less perfect ones progressively smaller
>
>Is it possible to get more than 84% of
>the light into the diffracton disc or is
>16% always going to go into the rings?
>-Rich
It's possible to apodize an aperture so that no light goes into diffraction
rings. This reduces the total amount of light to the focal plane by a good
amount and results in a somewhat bigger central spot. The Strehl ratio,
though, can never be higher than a plain old circular aperture.
Zane