Prism Diagonals Pros & Cons
George Curtis
Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
uses them while Televue does not (that I know of).
Duane Sand
"George Curtis" <gcu...@pacbell.net> wrote:
>Anyone have any comments on prism diagonals vs. mirror. I notice
>Celestron uses them while Televue does not (that I know of).
In small sizes, good prisms are cheaper than good mirror star diagonals.
But in scaling up to larger sizes, the thin mirror becomes cheaper
than a big chunk of clear quality glass.
Prisms' surfaces are easier to maintain; mirrors eventually need
recoating.
If given a low-cost reflective coating, a mirror loses more
light than a prism with multi antireflection coatings on its
in & out facets. A mirror with high-reflection coatings delivers
more light than the best prism.
Prisms must be made out of BAK4 or similar high-refractive index
glass in order to fully reflect light arriving at slight angles,
just as in binocs. Current catalogs never say what the prisms are
made of; I suppose all the astro brand stuff is good and all the
department store stuff is bad.
For very fast scopes having wide actual fields of view, I suspect
that prisms may lose some light because the incident angles exceed
the angles handled by BAK4's refractive index. I haven't yet done
the calculations to test that hunch.
Optically, prisms act like a very thick plate of glass with parallel
front and back faces. The shear thickness of the path thru the
glass can cause separation of focal length for angled red light rays
versus angled blue light rays versus near-perpendicular rays,
in fast scopes. This is a non-problem for slow focal-ratio scopes
because none of the rays have much angle to be exagerrated by the
glass's slow speed of light.
Another factor in diagonals is whether the prism or mirror is well-seated
in a way that collimates well and maintains good wavefront accuracy.
45-degree and Amici prisms apparently introduce some diffraction scatter
in their images. Okay for finders but not so good for the main scope.
Plus they have multiple reflection surfaces to lose light.
Jay Reynolds Freeman
In article <58osjr$q...@dfw-ixnews11.ix.netcom.com> Duane Sand <d.s...@ix.netcom.com> writes:
> [nice summary]
Typical prisms which are not BAK-4 are beginning to vignette the
edge of the objective for the on-axis image of an f/4 binocular. At
slower f-numbers, no problem. And unless designed in, prisms should
not be used in telescopes slower than about f/7 or f/8 just because
they introduce spherical and chromatic aberration. So while a BAK-4
star diagonal might be good marketing hype, I doubt that anybody
really needs one.
--
Jay Reynolds Freeman -- fre...@netcom.com -- I speak only for myself.
cro...@airmail.net
"George Curtis" <gcu...@pacbell.net> wrote:
>Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
>uses them while Televue does not (that I know of).
TeleVue offers both a prism and a mirror, at least for the
Ranger/Pronto; one (your choice) is provided with the scope. The
mirror provides a 90 degree bend, while the prism is 45 degrees.
Clearly, the former will be more comfortable for most astronomical
viewing, while the latter would be more appropriate for nature and
scenery. The greater loss of light with the prism shouldn't be a
concern since most scenery/nature viewing (if not all) is done during
the daytime when there's light to burn.
Andy Domonkos
I personally prefer the mirror diagonal, it appears to give better images,
less color, sharper, than my prism diagonal. Of course, no diagonal works
even better but can be a pain in the neck.
Andy
John Ongtooguk
George Curtis (gcu...@pacbell.net) wrote:
: Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
: uses them while Televue does not (that I know of).
A good mirror diagonal should be better as a prism diagonal
has two air/glass surfaces and a thick section of glass in
addition to the reflective surface that a mirror has. Other
than hearing about defective secondary mirrors on Newtonians
I haven't heard of any complaints with either.
John Ongtooguk (jo...@vcd.hp.com)
J. B. Kuster
Jay Reynolds Freeman <fre...@netcom.com> wrote:
> Typical prisms which are not BAK-4 are beginning to vignette the
> edge of the objective for the on-axis image of an f/4 binocular. At
> slower f-numbers, no problem. And unless designed in, prisms should
> not be used in telescopes slower than about f/7 or f/8 just because
> they introduce spherical and chromatic aberration. So while a BAK-4
> star diagonal might be good marketing hype, I doubt that anybody
> really needs one.
>
All prisms I have seen for sale and used in a star diagonal are of BK7
glass. For example in the Edmund Sci catalog... A standard 2 wave, BK7
glass prism of the size used in 1.25" diagonal is $75 coated. A high
toelrance 1/4 wave coated prism is $97. A BAK4 glass prism of the size
needed in 1.25" star diagonal is not even listed and would likely be very
expensive if it was. As Mr. Freeman says nobody really needs one.
J Blake
Harvey Mace
> Typical prisms which are not BAK-4 are beginning to vignette the
>edge of the objective for the on-axis image of an f/4 binocular. At
>slower f-numbers, no problem. And unless designed in, prisms should
>not be used in telescopes slower than about f/7 or f/8 just because
^
Don't you mean faster? BTW, do you know if this effect is a fnuction
of just the f ratio, or whether it also depends on the magnification
used? ( I've been thinking of buying the Ranger which has f/6.8)
>they introduce spherical and chromatic aberration.
Regards, Harvey
Andy Domonkos
Although I prefer a quality mirro diagonal I am comforted by knowing I
have a good quality prism diagonal in my accy case for the day I realize
my mirror diag needs re-coating.
Andy
:
:
Tim Gillespie
John Ongtooguk wrote:
>
> George Curtis (gcu...@pacbell.net) wrote:
> : Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
> : uses them while Televue does not (that I know of).
>
> has two air/glass surfaces and a thick section of glass in
> addition to the reflective surface that a mirror has. Other
> than hearing about defective secondary mirrors on Newtonians
> I haven't heard of any complaints with either.
>
> John Ongtooguk (jo...@vcd.hp.com)
John:
Those are good points, but I wonder. Since a mirror scatters so much
more light than a refractive medium, and owing to the 86% or so
reflectivity of most mirrors, I wonder how the differences stack up,
both theoretically and empirically. Any one out there have any hard data
comparing the two?
I was always under the (probably misguided) impression that prisms were
better than mirrors but were seldom used for reasons of both cost and
weight. A prism requires three optically flat surfaces (a mirror
obviously only requires one) two of which must be very close to
perfectly perpendicular. Also, a prism would be significantly heavier
than a comparable mirror.
As for the type of glass used, I recall (long ago) an optics professor
telling my class that *any* glass had a high enough refractive index to
cause total internal reflection in a prism. Oh, wait, that means there
is reflection in both systems. Hmmm, I guess my agrument about
scattering for a mirror is invalid then.
OK, both systems scatter due to reflection. So, the comparison then
seems to reduce to transmission loss at the surfaces for the prism
against reflectance loss of the mirror (I am assuming that the prism
suffers no reflectance loss, i.e. TOTAL internal reflection). If the two
surfaces of the prism are multicoated, transmission loss can be
minimized. Also, since the prism is located at the eyepiece end, and
since the surfaces are flat, incident light angles will be *very* close
to perpendicular, further enhancing the coatings' effect. So, as I see
it (probably incorrectly), if the prism can produce better than 86%
transmission, then it would outperform a (non-overcoated) mirror. Please
feel free to point any flaws in my reasoning, this is (almost) pure
speculation.
Tim Gillespie
John Ongtooguk
Tim Gillespie (tgil...@mail.tds.net) wrote:
: >
: > A good mirror diagonal should be better as a prism diagonal
: > has two air/glass surfaces and a thick section of glass in
: > addition to the reflective surface that a mirror has. Other
: > than hearing about defective secondary mirrors on Newtonians
: > I haven't heard of any complaints with either.
: >
: > John Ongtooguk (jo...@vcd.hp.com)
: Those are good points, but I wonder. Since a mirror scatters so much
: more light than a refractive medium, and owing to the 86% or so
: reflectivity of most mirrors, I wonder how the differences stack up,
: both theoretically and empirically. Any one out there have any hard data
: comparing the two?....
:
: OK, both systems scatter due to reflection. So, the comparison then
: seems to reduce to transmission loss at the surfaces for the prism
: against reflectance loss of the mirror (I am assuming that the prism
: suffers no reflectance loss, i.e. TOTAL internal reflection). If the two
: surfaces of the prism are multicoated, transmission loss can be
: minimized. Also, since the prism is located at the eyepiece end, and
: since the surfaces are flat, incident light angles will be *very* close
: to perpendicular, further enhancing the coatings' effect. So, as I see
: it (probably incorrectly), if the prism can produce better than 86%
: transmission, then it would outperform a (non-overcoated) mirror. Please
: feel free to point any flaws in my reasoning, this is (almost) pure
: speculation.
Well, I was speculating too. I think that an apples to apples
comparison would be an enhanced coated mirror and a multi-coated
prism, where I don't think that there would much to choose
between the two regarding transmission, so that leaves the
additional scattering at the two air/glass surfaces on the
prism, the impact of a nominal thick plane of glass, and
the parallelism of that plane of glass which is controlled
by the tolerances on the prism. The only data point that I
can come up with is that some types of scopes use two or even
three mirrors and can still perform decently, but two prisms
such as those used in image correction assemblies never do.
Porro or roof prisms require more reflections though so it's
not really a good comparison.
John Ongtooguk (jo...@vcd.hp.com)
Jay Reynolds Freeman
> > Typical prisms which are not BAK-4 are beginning to vignette the
> > edge of the objective for the on-axis image of an f/4 binocular. At
> > slower f-numbers, no problem. And unless designed in, prisms should
> > not be used in telescopes slower than about f/7 or f/8 just because
> BTW, do you know if this effect is a fnuction
> of just the f ratio, or whether it also depends on the magnification
> used? ( I've been thinking of buying the Ranger which has f/6.8)
focal ratio, and is linear, I believe, in the total distance that the
beam travels within the prism; that is, a 2-inch prism star diagonal
would be worse than a 1.25-inch one. As for magnification, the way
to think about it is, that the spherical and chromatic aberration are
present at the focal plane, but conceivably you may not be using
enough power to see them.
I suspect that a telescope as fast as the Ranger is on the ragged
edge of unacceptable when used with a 1.25-inch prism diagonal.
mbe
jo...@vcd.hp.com (John Ongtooguk) wrote:
>George Curtis (gcu...@pacbell.net) wrote:
>: Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
>: uses them while Televue does not (that I know of).
> A good mirror diagonal should be better as a prism diagonal
> has two air/glass surfaces and a thick section of glass in
> addition to the reflective surface that a mirror has. Other
> than hearing about defective secondary mirrors on Newtonians
> I haven't heard of any complaints with either.
> John Ongtooguk (jo...@vcd.hp.com)
Hi,
Because I am a very enthousiastic variable star observer, my concern
is always limiting magnitude and highest possible light transmission
from a given telescope.
In this respect I tested diagonals (prism and mirrors)
Conclusion: the Celestron 1.25 inch prism diagonal with M/C
(multi-coating) gives much more light transmission than a 1.25 or 2
inch mirror diagonal of TeleVue. If you compare these two the
difference is immediately visible. It looks like you loose about 1
inch of aperture on a Celestron 8. So the difference is at least 10 %
I expect.
rgds
Mbe
John Ongtooguk
mbe (m...@glo.be) wrote:
: Conclusion: the Celestron 1.25 inch prism diagonal with M/C
: (multi-coating) gives much more light transmission than a 1.25 or 2
: inch mirror diagonal of TeleVue. If you compare these two the
: difference is immediately visible. It looks like you loose about 1
: inch of aperture on a Celestron 8. So the difference is at least 10 %
: I expect.
The 10% number is about right if the TV uses standard coatings
which seem to run a bit over 85% reflectance, compared to
multi-coated refractive surfaces where the transmission might
be up to 98% or higher. Enhanced coatings on mirrors seem to be
in the 95% range, so if the TV uses enhanced coatings something
else is going on. In any case TV seems to be one of the few
companies that offer mirror diagonals so it's perhaps a moot
point. Pocono advertises something like a 1/20th wave enhanced
coated diagonal but as I recall it's kind of pricey. If you
have both types it'd be interesting to hear if there are any
other differences besides transmission, on slower and faster
scopes.
John Ongtooguk (jo...@vcd.hp.com)
ed_turco
I've been watching this thread for while and have to hop in. Prism
diagonals, even when made *perfectly*, contribute aberrations to the
final image, mainly coma and spherical... While it is true that above
f/10 or so, it doesn't seem to make much difference; it is very
difficult in a fine instrument to draw the line at what is allowable.
For this reason, my advice is to never use prisms.
Why not play it safe? Mirror diagonals can be made as perfectly as
possible and *never* contribute these kinds of problems to the final
image. A prism made to these same standards would beyond the reach of
any sane person's budget, and still degrade the image as described
above.
I have to add that it is totally unconscionable for a manufacturer to
provide a prism diagonal for a system of less than f/10 or
thereabouts. This is why you'll find that Al Nagler doesn't provide
prism diagonals with his products.
ed_turco
jbl...@aol.com
>Conclusion: the Celestron 1.25 inch prism diagonal with M/C
>(multi-coating) gives much more light transmission than a 1.25 or 2
>inch mirror diagonal of TeleVue. If you compare these two the
>difference is immediately visible. It looks like you loose about 1
>inch of aperture on a Celestron 8. So the difference is at least 10 %
I find the above conclusions interesting. I have only recently become a
little discouraged with the star diagonal that came with my Celestar 8.
Recently I viewed some objects using the diagonal, then with no diagonal.
The view with no diagonal was noticeably better, especially on Saturn.
This observation was on a night of very good seeing and a limiting mag of
6+. I have always seen a slight improvement with no diagonal, but this
time it was a whole magnitude of difference.
I have always heard that a mirror diagonal beats a prism one. So I have
been thinking about getting a mirror model. However, the above comments
give me pause to question that. Maybe, the diagonal that came with my
scope is not multi-coated. However, the prism surface does appear to have
that slight violet coloration as other coated surfaces. Do all Celestron
star diagonals come coated or is the multi-coated model an upgrade item?
And.. would not a properly coated mirror reflect more light than a coated
prism pass light?
Any input or comments from the group would be helpful.
J Blake
Gary Forbus
lwz...@uriacc.uri.edu (ed_turco) wrote:
Stuff clipped....
> I have to add that it is totally unconscionable for a manufacturer to
>provide a prism diagonal for a system of less than f/10 or
>thereabouts. This is why you'll find that Al Nagler doesn't provide
>prism diagonals with his products.
>ed_turco
Ed,
The Ranger comes with the choice of a 90 deg. mirror diagonal or a 45
deg. erecting prism. The literature claims that the mirror diagonal
is "best optically and required for high magnification viewing."
See
http://www.rahul.net/resource/regular/products/tele-vue-optics/pg21.htm
Lionel Sandell
ed_turco wrote:
>
> I've been watching this thread for while and have to hop in. Prism
> diagonals, even when made *perfectly*, contribute aberrations to the
> final image, mainly coma and spherical... While it is true that above
> f/10 or so, it doesn't seem to make much difference; it is very
> difficult in a fine instrument to draw the line at what is allowable.
> For this reason, my advice is to never use prisms.
>
Regards, Lionel
Ross Brunetti
I ordered my Ranger with the 45° erect image diagonal (wanted the erect
image for birdwatching, and I already had a Lumicon mirror diagonal).
Well, the 45° was (and is) fine in daytime, but at night...any bright
object shows diffraction bands to the left and right. When sending in my
warranty registration, I added a note about this, and asked if Mr.
Nagler could design an erect image diagonal without diffraction. Imagine
my surprise when TeleVue returned my note with an addendum from Al
himself, stating that it "just isn't possible."
Guess we'll just have to live with it. But, I sure would like an erect
image, angled view device.
Disadvantage to mirror diagonals? The mirror coating will, eventually,
deteriorate. My Lumicon, after 10 years, shows signs of such.
Ross Brunetti
Duane Sand
The 1.25" prism diagonal I got from a telescope specialty store
when I purchased a Celestron RF80 f/8 80mm spotting scope appears
to not have any anti-reflection coatings. There's no brand info
on this diagonal. Sighting thru just the diagonal indoors, it
was interesting to see the loss of internal reflection beyond
certain angles. The fall-off of light at the transition is
nearly total, with a narrow band of color fringe. I had thought
that perhaps the % of reflection would diminish gradually beyond
the critical angle, but instead it is quite sudden. In a fast
scope, I think the result would be a vignetting of one edge
(top?) of the circular objective aperture, giving it somewhat
of a downwards D shape. That should give a distinctive distortion
to diffraction patterns.
On the question of glass types, it occurred to me that in scopes
which are slow enough to rely on BK7's reflection effects, BK7
prisms would be superior to BAK4 prisms because BK7's lower
index of refraction would lessen the longitudinal color dispersion
caused by the prism.
Clive Gibbons
In article <32B376...@thegrid.net>,
Ross Brunetti <ro...@thegrid.net> wrote:
>Well, the 45° was (and is) fine in daytime, but at night...any bright
>object shows diffraction bands to the left and right. When sending in my
>warranty registration, I added a note about this, and asked if Mr.
>Nagler could design an erect image diagonal without diffraction. Imagine
>my surprise when TeleVue returned my note with an addendum from Al
>himself, stating that it "just isn't possible."
Actually Ross, there is a way to make an image erector which doesn't
produce diffraction spikes off of bright objects. It's called a Porro
Prism erector, but instead of bending light through a 90-degree or
45-degree angle, the eyepiece axis stays inline with the main scope. The
diffraction effect you're noticing is caused by the "roof angle" in the
roof prism of the 45-degree diagonal. If Nagler used a porro-prism
erector this, image defect would disappear, but pointing the scope
skyward to view would be less comfortable on your neck!
Cheers,
--
Clive Gibbons "I wasn't expecting the
Technician, Spanish Inquisition..."
McMaster University.
Larry Manuel
jo...@vcd.hp.com (John Ongtooguk) wrote:
>George Curtis (gcu...@pacbell.net) wrote:
>: Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
>: uses them while Televue does not (that I know of).
> A good mirror diagonal should be better as a prism diagonal
> has two air/glass surfaces and a thick section of glass in
> addition to the reflective surface that a mirror has. Other
> than hearing about defective secondary mirrors on Newtonians
> I haven't heard of any complaints with either.
> John Ongtooguk (jo...@vcd.hp.com)
I have it on quite good authority [Barry Arnold, Edmonton, AB] that
for f/ ratios longer than f/8, a prism is much to be preferred. The
refractive surfaces scatter significantly less light than a reflective
surface. I believe the hypotenuse of the prism acts more like a
refractive surface than a reflective one - as far as light scatter is
concerned. If coated, no amount of roughness in the coating metal can
scatter light, since it is the glass surface which functions as the
reflective surface. Remember - reflectively coated prisms often have
paint over the reflective coating to protect it almost forever.
A good prism is very expensive, and refractors [indeed Newtonians] f/8
and longer are considered passe by the masses who buy such creatures.
I think this explains why prisms are scorned.
Larry.
Larry Manuel, lma...@telusplanet.net, Calgary, Alberta, Canada
Robert Kemper
Does TeleVue have a 1 1/4" MIRROR diagonal? I have a TeleVue 1 1/4"
diagonal I got from Pocono, and it's a prism. All the catalogs I
have refer to TeleVue 90 degree diagonals as:
2" mirror diagonal OR
1 1/4" diagonal (no reference to mirror or prism)
It's very nicely constructed, but definitely a prism.
: >Anyone have any comments on prism diagonals vs. mirror. I notice Celestron
: >uses them while Televue does not (that I know of).
: TeleVue offers both a prism and a mirror, at least for the
: Ranger/Pronto; one (your choice) is provided with the scope. The
: mirror provides a 90 degree bend, while the prism is 45 degrees.
--
........................................................
~~~~~~~~~~~~~ Robert A. Kemper
~~~~~ ~~~~~ Pacific-Sierra Research Corporation
~~~ ~~~ 1400 Key Blvd., Suite 700
~ ~ Arlington, Virginia 22209
~~~~~~~~~~~~~ Voice : (703) 516-6394
~~~~~~~~~ FAX : (703) 524-2420
~~~~~ E-mail: rke...@sed.psrw.com
~ http://www.psrw.com/~rkemper/
John Ongtooguk
Thomas Back (am...@cleveland.Freenet.Edu) wrote:
: Indeed. a prism will add its own aberration (overcorrected spherical
: and color), but until you raytrace a system with a prism (I have -- with
: ZEMAX), your just guessing at the aberrational residuals. I own a 2"
: multi-coated prism (Badder Planetarium) and at f/9, it works superbly.
: With 99% transmission, and the lack of light scatter, this diagonal beats
: any mirror diagonal I've ever used. Of course, a well made prism will
: cost more, but until interference coated supersmooth diagonals are
: available, the high quality prism for moderate to long focal length
: refractors will reign supreme.
Well, I guess I haven't kept track of all of the problems with
mirrors. I understand the 1/2 vs 2x difference in how refractive
and reflective surfaces react to defects on the glass but one
would assume no difference between a mirror and prism polished
to the same level of roughness as they are both a reflective
surface. So reflective coatings on the mirror would responsible
for difference ?
John Ongtooguk (jo...@vcd.hp.com)
Jay Reynolds Freeman
> Thanks for clearing up the many misconceptions -- in this thread --
> about prisms, when used in astronomical instruments. I won't go over
> the scatter problem with mirrors again (God forbid: If they haven't
> learned by now, they never will!), nor the BaK-4 nonsense. And the
> answers in this thread about color and spherical being increased by
> the use of a prism is way over-simplistic.
>
> Indeed. a prism will add its own aberration (overcorrected spherical
> and color), but until you raytrace a system with a prism (I have -- with
> ZEMAX), your just guessing at the aberrational residuals.
It was I who started the complaint about prisms in fast systems. I
did not say that it was simple, and I submit that I was not guessing
-- though I didn't check Selby's calculations, as given in ATM. (I
forget which volume, and as usual, the relevant books are at the other
terminal.) I used his formulae to evaluate the spherical and
chromatic aberration of a conventional crown glass prism (I forget
which common crown he used), of side 35 mm (eyeball for what's in my
diagonals), at f/5.6 (my 63 mm "Baby Brandon"), and found that the
system had more than a quarter-wave of spherical aberration just from
the prism, and similarly severe longitudinal chromatic aberration.
The formulae depended on the objective through f-number alone --
aperture and focal length entered in only through their ratio -- and
it was algebraically obvious that these aberrations increased rapidly
with decreasing f-number. Thus I inferred that for any system this
fast or faster, prism star diagonals of the sizes generally used in
1.25-inch hardware are a no-no unless designed in (that is, unless the
rest of the system has countervailing opposite aberrations).
The matter is particularly important because of the popularity of
fast high-quality systems like the Genesis, the Ranger/Pronto, and
much of the recent Astro-Physics stuff, which are all fast enough to
enter the range in which Selby's work indicates that prisms are
marginal, particularly two-inch ones (aberration increasing with prism
OPD). It's not at all clear to me that an extra quarter wave or more
of spherical aberration is preferable to the scattering off a
conventional first-surface mirror.
If I am wrong, I would of course like to know about it; I have not
bothered to ray-trace, as I have no handy package and would have to
chug through it by hand in Lisp, and I had no cause to mistrust Selby.
Is it your gut feeling that these numbers are wrong? Perhaps you
could check with something more reliable than my envelope-back GUI?
In any case, my little Brandon performs noticeably better without
any of the several prism diagonals that I own, all of which work more
or less alike, and all of which give satisfactory service in my 90 mm
f/9 Vixen fluorite or my 102 mm f/9.8 Vixen conventional doublet,
which are slow enough that prisms should be no problem, again
according to Selby. I have been too lazy to go out and chase down a
mirror diagonal for the Brandon, since mostly I use it at low
magnification anyway.
The thrust of Selby's original article, incidentally, was not prisms
in star diagonals -- fast well-corrected refractors being rather scare
in the 1930s -- but prisms as Newtonian diagonals. Amateurs were
beginning to make Newtonians that were fast enough for prisms to be a
bother, according to Selby's calculations, particularly with the even
bigger prisms that large amateur Newtonians required.
If memory serves, all Selby did was express moderately well-known
formulae for spherical and chromatic aberration introduced by a
plane-parallel plate of glass, in terms of the f-number of the
incoming beam.
Jay Reynolds Freeman
I previously discussed some of Horace Selby's calculations about
the effects of prism diagonals in fast telescopes; now I am at home
with _Amateur_Telescope_Making_ in front of me.
The relevant section is toward the end of Selby's chapter on
eyepieces in ATM III; it's not very well indexed -- in my edition, the
text starts on page 505. If anyone wants to look this up, note that
the formula for L sub sph in the middle of page 506 is typeset
incorrectly; the "square root" symbol is supposed to stop at the end
of the first fraction, and the symbol "alpha" is missing in two
places. It should read like this, with due allowance for ASCII:
Lsph = T * sqrt{(1 - sin**2 alpha)/(N**2 - sin**2 alpha)} - (T/N),
which incidentally simplifies still further, for calculator work, to
Lsph = T * [sqrt{(1 - sin**2 alpha)/(N**2 - sin**2 alpha)} - (1/N)].
What Selby presents is calculations of marginal spherical
aberration (versus the paraxial ray), and of longitudinal chromatic
aberration for marginal rays in two colors; his formulas allow the
user to select choice of wavelengths (given refractive indexes). He
tabulates results for a crown with NC = 1.514, ND = 1.517, and Ng =
1.526, taking chromatic aberration between C and g lines.
I misremembered his tolerance for longitudinal aberration; he
recommends, not the Rayleigh limit of 1/4 wave, but a much more
demanding figure of a third of that. My Brandon is marginal at that
much more severe level, with a 35 mm prism, which means that it is
probably okay for spherical aberration. For chromatic aberration,
however, the situation is as bad as I remembered, and in fact much
worse (assuming Selby is right): At f/5.6, a 35 mm prism gives my
Brandon roughly 2.5 times as much chromatic aberration as the full
Rayleigh limit for that aberration. At f/8, a 35 mm prism causes
negligible spherical aberration but chromatic aberration approximately
equal to the Rayleigh limit. At f/6, by my calculation, spherical
aberration is still far less than the stringent tolerance that Selby
recommends, but chromatic aberration is approximately twice the
Rayleigh limit.
Jay Reynolds Freeman
I have been in EMail correspondence with Thomas Back about prisms
in fast systems, and with his permission here post his comment to me.
(I am not posting it because he said something flattering of me in it;
rather, because it contains relevant information from a knowledgeable
source.)
########################################################################
Hi Jay,
I didn't mean to say you were over-simplifying the aberration of
prisms, in fact, you were the only one that had a grasp of the
problem.
It is true that a prism can be expressed as a plane-parallel piece
of glass. The important thing to remember is this only applies in a
simple mathematical formula for a mirror system. In an achromatic or
apochromatic lens, the variations of spherochromatic aberration and
secondary and tertiary color make the use of a prism very interesting
indeed. Raytracing in ZEMAX -- various APO designs with 25mm and 50mm
prisms with f/ratio's of f/6 to f/12 -- shows that the OPD spherical
levels for different wavelengths change in different amounts,
sometimes improving a systems at some wavelength at the cost of
others. The ONLY way to know what is really happening is to know the
exact design, raytrace the total system, or star test the system, and
check to see if the contrast is higher with the prism, or the mirror
diagonal. But even with these spherical and color variations, the
aberration levels do get excessive at around f/7 or f/8. I agree that
the use of a prism with your f/5.6 Brandon is not the best choice,
unless, the lens has just the right amount of undercorrected spherical
and undercorrected color to compensate for the prism, which I very
much doubt. I will say that with my 7.1" f/9, there is a slight
increase of violet in the defocus star test image. But none of this is
visible at focus.
Best Regards,
Thomas M. Back
########################################################################
abzmou...@googlemail.com
RichA
> Hi,I've always used celestron stardiagnol,, recently I purchased a celestron 90° erect image correcting prism,I've not yet used it.can anyone tell me what differences to expect and the advantage and disadvantage, if there is any,thanks.
Martin Brown
degree prism.
https://en.wikipedia.org/wiki/Amici_roof_prism
Its claim to fame is that the image is the same way round as reality.
--
Regards,
Martin Brown
Razzmatazz
Razzy
Chris.B
If you can't tell the difference then what does it matter?
Except that you've saved some money for magical hi-fi cables. ;-)