Telephoto Reach and Digital Cameras - lens comparison

On Sat, 18 Dec 2010 08:24:53 -0800, Paul Furman
wrote:

Ofnuts wrote:
Superzooms Still Win wrote:

As for your physics proving everything, it FAILS because it does NOT take
into account the figure of the lenses. The lenses on the superzoom camera
can and ARE polished to diffraction-limited quality.

Yes, they are polished to diffraction-limited quality at the top of a
high moutain by virgin girls under the full moon. They are then laid on
silk cushions and brought back in the valley using buffalo carts, and
the tiny moves of the lenses on the cushions caused by the gentle
rocking of the cart finishes the polishing to perfection.

Panasonic's Leica lenses are finished off wrapped in silk between Swiss
virgin's breasts trotting on horseback down the mountain as they sing
the praises of their fine accuracy and innovative engineering;
instilling emotional warmth, impact, and subtlety.
http://www2.panasonic.com/webapp/wcs...tGroupId=24999

The text makes a virtue out of necessity:

"The DC Vario-Summicron is Leica's preeminent lens. Its exceptional
optics give photos an exquisite soft-focus effect, ... "




Eric Stevens

Roger N. Clark (change username to rnclark) wrote:
Paul Furman wrote:

Now some more abstract comparisons, I suspect the reason you don't see
super fast P&S lenses with shallow DOF is the exact same reason they
cost a lot in 35mm format. P&S are designed for economy and it would be
too expensive to grind lenses that small to such tolerance;
unmarketable. Now I'm talking '35mm equivalent aperture' which is a
concept with little acceptance but it's valid if you want shallow DOF.
If you just scale down a 50mm f/1.4 lens to fit a P&S, it would still be
f/1.4 but there would be more DOF which means you'd need to redesign it
as something like 10mm f/0.7,

Paul,
If the crop multiplier is 5x, so the 10mm lens on the P&S is the
same as the 50 mm lens on the 35mm sized sensor, the aperture
diameter must be maintained for the same depth of field. Thus
the 50 mm f/1.4 becomes 1.4/5 = f/0.28 on the 10 mm lens, which
is pretty much impossible to build.

Thanks, I am not to be trusted with math g.

As I wrote my rant/lecture above, one thing that helped me make sense of
why DOF changes with smaller formats at the same f-stop is this
visualization: Imagine instead of shrinking the lens and camera, that
the world grew larger by a factor of five. Or imagine a studio setting
scaled up five times if that's too much of a stretch.

The enlarged studio scene and subjects would be five times as far away
from the camera. DOF gets shallower the closer you focus: macro shooting
has very shallow DOF whereas subjects out near infinity will be rendered
with more 'thick' DOF.

Or, to get a shallow DOF look from a P&S, you could make a scale model
at 1/5th the size and be in macro range, if you could get a 10mm f/1.4
lens. There are c-mount video lenses that come close like maybe 12mm f/1.8.

Of course, if you want more DOF, none of this matters but I thought it's
an interesting way to put things into perspective.

Val Hallah wrote:
On Dec 6, 5:20 pm, "David J Taylor"david-
wrote:
A comparison of fixed focal length lenses& consumer zooms on DSLRs when
photographing distant objects:

http://www.clarkvision.com/articles/telephoto_reach/

including a comparison with a super-zoom P&S.

http://www.clarkvision.com/articles/...vs.dslr.compar...

A longer zoom may not be better than a shorter fixed focal length lens.

better off with a spacecraft though...

http://www.dailymail.co.uk/sciencete...ed-detail.html

Tell me about it. Spacecraft do a lot more than simply get better
resolution. See:

http://www.sciencemag.org/content/326/5952/562.abstract

or the Moon as never before seen:
http://www.lpi.usra.edu/meetings/lpsc2010/pdf/2302.pdf
(I'm working on the full resolution, full coverage results now.)

Spacecraft are great, if you have a few hundred million dollars!

The cost from superzoom to DSLR telephotos is roughly proportional
to the resolution increase. The jump to orbital spacecraft at the
Moon is a huge jump in cost for the resolution increase.
And even so, a spacecraft in orbit around the Moon makes a lousy
camera for sports and wildlife photography on Earth!

Roger

On 12/18/2010 08:58 PM, Eric Stevens wrote:
On Sat, 18 Dec 2010 08:24:53 -0800, Paul
wrote:

Ofnuts wrote:
Superzooms Still Win wrote:

As for your physics proving everything, it FAILS because it does NOT take
into account the figure of the lenses. The lenses on the superzoom camera
can and ARE polished to diffraction-limited quality.

Yes, they are polished to diffraction-limited quality at the top of a
high moutain by virgin girls under the full moon. They are then laid on
silk cushions and brought back in the valley using buffalo carts, and
the tiny moves of the lenses on the cushions caused by the gentle
rocking of the cart finishes the polishing to perfection.

Panasonic's Leica lenses are finished off wrapped in silk between Swiss
virgin's breasts trotting on horseback down the mountain as they sing
the praises of their fine accuracy and innovative engineering;
instilling emotional warmth, impact, and subtlety.
http://www2.panasonic.com/webapp/wcs...tGroupId=24999

The text makes a virtue out of necessity:

"The DC Vario-Summicron is Leica's preeminent lens. Its exceptional
optics give photos an exquisite soft-focus effect, ... "


Yes, made me smile too.

--
Bertrand

On 12/18/10 1:20 PM, in article ,
"Superzooms Still Win" wrote:

On Sat, 18 Dec 2010 02:12:50 -0800, bobwilliams wrote:

Superzooms Still Win wrote:
On Fri, 17 Dec 2010 01:06:43 -0800, bobwilliams wrote:

I admire Roger's contributions to this newsgroup and take all his
data, images and conclusions very seriously.
Bob

Then you are a fool. He poses BAD SCIENCE as a way to get controversy so
people visit his website where he then tries to sell his ****-poor tourists
crapshots. He's that desperate for anyone to see his photography because
nobody wants to. It's his only motive and method. His calculations have
been proved wrong so many times in the past that people just stopped
bothering to correct all his errors. Then idiots like you fall for his
bull**** song and dance.

If am a fool, then I am in good company in this NG.
Roger posts a lot of Quantitative data that are not available anywhere
else. I use them a lot.

Some of his photos are awesome.
This shot of an egret won first prize against International competition
in NATURE'S BEST Magazine. This magazine is the premier Nature
Photography Magazine in the world. I have been a subscriber for many
years and you will not find any touristy snapshots in this Magazine.
See:
http://www.spamvision.com/galleries/....2003.img_5113
.egret-flight.f-600.html


BFD, even a stopped-clock is right twice a day. Quite frankly I'm suspect
of whoever judged that improperly exposed and poorly composed zoo snapshot.
I've posted ones better than that as throw-aways here just for examples to
prove all these trolls wrong on issues unrelated to marketable photography.

Uhhhh, and just where did you post them, Tonto?

I stopped entering international competitions when I found out how easy
they were to win. There's no challenge in winning awards. One time I
entered a purposely crappy photo just as a joke, to show all my
photographer friends just what kind of crap-photography will win an award.

"ALL of my photographer friends..."

Now that's funny. I don't care WHO you are, that there's FUNNY!

bobwilliams wrote:

Roger posts a lot of Quantitative data that are not available anywhere
else. I use them a lot.

Some of his photos are awesome.

Thanks Bob.

I don't enter many contests (little time). I did enter Natures Best this year.
This one was selected as Highly Honored in this year's contest:

http://www.clarkvision.com/galleries...130.f-900.html

It should be in the fall issue of Natures Best that should be out
shortly if not already.

Roger

Dudley,

The relevant equation above is

Diffraction spot diameter = 2 * 1.22 w * f / D =

2.44 * w * f_ratio,

whe
w = wavelength
f = focal length
D = aperture diameter
f_ratio = the f/ratio of the optical system

So if you specify wavelength of light in microns, the
diffraction
spot diameter is in microns.

The diffraction spot size for green light of 0.53 microns

(5300 angstroms) is:

f/2: 2.6 microns,
f/4: 5.2 microns,
f/8: 10.3 microns.

The above from:
http://www.clarkvision.com/articles/...il/index.html#
diffraction

So you see that cameras with small pixels (P&S cameras
typically
have pixels under 2 microns, and superzoom cameras with f/4
and
slower lenses thus have pixels much smaller than the
diffraction
diameter.

But more important is resolution on the subject. With more
real
focal length, the diffraction spot becomes a smaller angular
diameter
if you keep f/ratio the same. That is because the diameter of
the
aperture is getting larger. So the longer focal lengths

found in DSLR telephotos provides proportionally greater

resolution on a subject.

Resolution on a subject is given by:
resolution = constant / lens diameter,
where the constant is determined by the wavelength.


The Dawes limit is the limit at which no more detail on a
subject
can be resolved, or in other words, zero contrast on closely

spaced detail, like line pairs in a test target, or hairs on
an
animal.

e.g.:
http://en.wikipedia.org/wiki/Dawes'_limit

resolution in arc-seconds = 4.56 /D
where D is the clear aperture in inches
and the 4.56 is for a particular wavelength (green light).

With D on the bottom of the equation, the finest detail one
can
resolve goes down as the lens aperture goes up. That dictates

the fundamental difference between P&S superzoom cameras and

DSLRs, and the field of view lens multiplier is now a great

confusing factor among photographers selecting digital cameras.


Roger


Thanks, Roger, that gives me a fair bit to think about.


After a quick scan, I just have a couple of supplemental
questions:

1. When you refer to the Diffraction spot diameter, I'm
having a tough time wrapping my head around what's occuring at
the pixel site. Does the diameter represent the area where
photons (of finite size) get focused, with individual photons
diverging somewhat due to diffraction? (ie: the diffraction
spot diameter represents some sort of a scatter zone., Or,
does the energy charge of the photon somehow get spread out
over a larger area, due to the wave-like nature of light?


If I knew a bit more about the nature of photons, this
probably would not be so mystifying.

Anyway, the second question is a bit less theoretical, I hope:


2. When you say that the crop factor of the small sensored
DSLR's is causing a lot of confusion, is it because of the
diffraction spot diameter? I mean, if one takes a regular
lens and puts it on the smaller-sensored body, one would
expect the diffraction spot diameter to remain the same size
as it would be on a full-frame body. But, of course, the
smaller sensor would have more resolving power, due to the
larger number of tinier pixels packed tighter, so the level of
detail should suffer as long as the lens's focal length is not
altered. Correct?

But, because of the crop factor, the resulting image would
look more like a pic taken with a longer lens, which on a full-
frame body would actually result in smaller diffraction spot
diameters for the longer focal length, so the degradation of
detail would actually be compounded. All in all, the crop
sensored body should exhibit considerably less detail in a
picture if the lens from a full-frame camera is used.


Take Care,
Dudley

Dudley Hanks wrote:

Thanks, Roger, that gives me a fair bit to think about.


After a quick scan, I just have a couple of supplemental
questions:

1. When you refer to the Diffraction spot diameter, I'm
having a tough time wrapping my head around what's occuring at
the pixel site. Does the diameter represent the area where
photons (of finite size) get focused, with individual photons
diverging somewhat due to diffraction? (ie: the diffraction
spot diameter represents some sort of a scatter zone., Or,
does the energy charge of the photon somehow get spread out
over a larger area, due to the wave-like nature of light?

If I knew a bit more about the nature of photons, this
probably would not be so mystifying.

Actually, it is complex. It has to do with the dual nature of
light, acting both as a particle and a wave. Consider the diffraction
spot in the focal plane before the pixel. A point source, like a star,
will form an approximately shaped Gaussian distribution
with ringe around it. The diameter I was referring to is where
the intensity drops from the high in the center to zero before rising a
little in the first ring. But in an extended image, like in a everyday
photo, there are numerous overlapping diffraction disks,
with the effect of smearing the fine detail and reducing contrast
of the fine detail.

The lens focuses the light, but diffraction diverges the angles
of the light rays bent by the lens, so the lens can't make perfect
focus. You could think of the diffraction spot as kind of
a "scatter zone."

Wikipedia has a pretty good write-up of the dual nature of light:
http://en.wikipedia.org/wiki/Wave–particle_duality


Anyway, the second question is a bit less theoretical, I hope:


2. When you say that the crop factor of the small sensored
DSLR's is causing a lot of confusion, is it because of the
diffraction spot diameter?

No, not just DSLR, but P&S cameras too. Some people think
that the crop sensor gives them more telephoto reach. For example,
put the same lens on a Canon 5D mark II 21 megapixel full-frame camera
and then on a Canon 30D 8-megapixel 1.6x crop camera and they think
the 1.6 crop camera gives them more telephoto reach, meaning detail
on a distant subject, like the Moon, or a distant bird would be better.

But the Canon 5DII and 30D have the same pixel pitch, so get the same
numbers of pixels on a subject. It is pixel size and true focal
length that controls pixels on subject, not sensor size. It is the
lens quality and diffraction that limits the detail that the pixels
sample.

Camera manufacturers have added a lot of confusion in this area too.
They often specify focal length without saying it is 35-mm equivalent
field of view focal length.

Significant confusion also occurs in the f/ratios. For example, Ziess
makes a spotting scope with a small sensor build in and claims the
the scope, with an 85 mm clear apertu

"As a digital camera, the PhotoScope provides a truly superior
telephoto lens: photographically the equivalent of a 600mm @ f4.0
to 1800mm @ f5.6 zoom lens on a 35mm camera"
http://www.zeiss.com/c1256bcf0020be5...25755c006de445

Claiming f/ratio is completely bogus. An 1800 mm f/5.6 lens would
have an aperture of 321 mm and collect 14.3 times more light,
and have a very different depth-of-field than the 85 mm telescope.
And, of course, they ignored the detail limiting effects of
diffraction.

I mean, if one takes a regular
lens and puts it on the smaller-sensored body, one would
expect the diffraction spot diameter to remain the same size
as it would be on a full-frame body.

Yes, but in the digital image might appear different if
the pixel spacing is different. The actual diffraction
would be the same, but the pixel sampling could be different
for different pixel pitches.

But, of course, the
smaller sensor would have more resolving power,due to the
larger number of tinier pixels packed tighter, so the level of
detail should suffer as long as the lens's focal length is not
altered. Correct?

That is not necessarily true. For example, the Canon 5DII
and 30D have the same pixel sampling yet are different
sized sensors. It has to do entirely with pixel pitch
(spacing) and not crop factor. Sensor size affects total
field of view, of course.

But, because of the crop factor, the resulting image would
look more like a pic taken with a longer lens, which on a full-
frame body would actually result in smaller diffraction spot
diameters for the longer focal length, so the degradation of
detail would actually be compounded. All in all, the crop
sensored body should exhibit considerably less detail in a
picture if the lens from a full-frame camera is used.

Not at all. You are saying the commonly believed ideas, but
crop factor does not control detail on a subject. A cropped sensor
does not magically give more telephoto reach because of the crop.
Take any image and crop it. Do you get more detail on a subject? No.
This is illustrated in Figure 2 at:
http://www.clarkvision.com/articles/telephoto_reach/
The 1D Mark II image on the left is a 1.3x crop sensor, and shows less
detail than the 5D Mark II full frame image next to the 1DII image,
yet both were taken with the same lens. Pixel pitch is controlling
the detail, not sensor size. And that is illustrated by the right-most
image in the figu a Canon 7D 1.6x crop sensor shows the most detail
because the pixel pitch is smallest, not because of the crop.
And for this lens, diffraction is smaller than the all the pixel
sizes.

See:
http://www.clarkvision.com/articles/cropfactor/

Roger

"Roger N. Clark (change username to rnclark)" username@qwest.
net wrote:
Dudley Hanks wrote:

Thanks, Roger, that gives me a fair bit to think about.


After a quick scan, I just have a couple of supplemental
questions:

1. When you refer to the Diffraction spot diameter, I'm
having a tough time wrapping my head around what's occuring at
the pixel site. Does the diameter represent the area where
photons (of finite size) get focused, with individual photons
diverging somewhat due to diffraction? (ie: the diffraction
spot diameter represents some sort of a scatter zone., Or,
does the energy charge of the photon somehow get spread out
over a larger area, due to the wave-like nature of light?

If I knew a bit more about the nature of photons, this
probably would not be so mystifying.

Actually, it is complex. It has to do with the dual nature of
light, acting both as a particle and a wave. Consider the diffraction
spot in the focal plane before the pixel. A point source, like a star,
will form an approximately shaped Gaussian distribution
with ringe around it. The diameter I was referring to is where
the intensity drops from the high in the center to zero before rising a
little in the first ring. But in an extended image, like in a everyday
photo, there are numerous overlapping diffraction disks,
with the effect of smearing the fine detail and reducing contrast
of the fine detail.

The lens focuses the light, but diffraction diverges the angles
of the light rays bent by the lens, so the lens can't make perfect
focus. You could think of the diffraction spot as kind of
a "scatter zone."

Wikipedia has a pretty good write-up of the dual nature of light:
http://en.wikipedia.org/wiki/Wave–particle_duality


Thanks, I appreciate the info. As I was reading your summary,
I got to wondering if there is more to the cause of
diffraction than just the lens bending the light.


It occured to me that there are a few other opportunities for
diffraction: 1) The light passing through the lens would no
doubt encounter obstacles, at least on an atomic level, i.e.,
the constituent particles of the glass. 2) air particles
surrounding lens elements could knock a few photons off course,
and 3) for all I know, some of the diffraction could occur as
photons bump into each other as they cross paths after being
focused and head towards their inverted image destination.


Any thoughts on whether or not these could contribute to
diffraction?


Anyway, the second question is a bit less theoretical, I hope:


2. When you say that the crop factor of the small sensored
DSLR's is causing a lot of confusion, is it because of the
diffraction spot diameter?

No, not just DSLR, but P&S cameras too. Some people think
that the crop sensor gives them more telephoto reach. For example,
put the same lens on a Canon 5D mark II 21 megapixel full-frame camera
and then on a Canon 30D 8-megapixel 1.6x crop camera and they think
the 1.6 crop camera gives them more telephoto reach, meaning detail
on a distant subject, like the Moon, or a distant bird would be better.

But the Canon 5DII and 30D have the same pixel pitch, so get the same
numbers of pixels on a subject. It is pixel size and true focal
length that controls pixels on subject, not sensor size. It is the
lens quality and diffraction that limits the detail that the pixels
sample.

Camera manufacturers have added a lot of confusion in this area too.
They often specify focal length without saying it is 35-mm equivalent
field of view focal length.

Significant confusion also occurs in the f/ratios. For example, Ziess
makes a spotting scope with a small sensor build in and claims the
the scope, with an 85 mm clear apertu

"As a digital camera, the PhotoScope provides a truly superior
telephoto lens: photographically the equivalent of a 600mm @ f4.0
to 1800mm @ f5.6 zoom lens on a 35mm camera"
http://www.zeiss.com/c1256bcf0020be5...25755c006de445

Claiming f/ratio is completely bogus. An 1800 mm f/5.6 lens would
have an aperture of 321 mm and collect 14.3 times more light,
and have a very different depth-of-field than the 85 mm telescope.
And, of course, they ignored the detail limiting effects of
diffraction.




I mean, if one takes a regular
lens and puts it on the smaller-sensored body, one would
expect the diffraction spot diameter to remain the same size
as it would be on a full-frame body.

Yes, but in the digital image might appear different if
the pixel spacing is different. The actual diffraction
would be the same, but the pixel sampling could be different
for different pixel pitches.

But, of course, the
smaller sensor would have more resolving power,due to the
larger number of tinier pixels packed tighter, so the level of
detail should suffer as long as the lens's focal length is not
altered. Correct?

That is not necessarily true. For example, the Canon 5DII
and 30D have the same pixel sampling yet are different
sized sensors. It has to do entirely with pixel pitch
(spacing) and not crop factor. Sensor size affects total
field of view, of course.




But, because of the crop factor, the resulting image would
look more like a pic taken with a longer lens, which on a full-
frame body would actually result in smaller diffraction spot
diameters for the longer focal length, so the degradation of
detail would actually be compounded. All in all, the crop
sensored body should exhibit considerably less detail in a
picture if the lens from a full-frame camera is used.

Not at all. You are saying the commonly believed ideas, but
crop factor does not control detail on a subject. A cropped sensor
does not magically give more telephoto reach because of the crop.
Take any image and crop it. Do you get more detail on a subject? No.
This is illustrated in Figure 2 at:
http://www.clarkvision.com/articles/telephoto_reach/
The 1D Mark II image on the left is a 1.3x crop sensor, and shows less
detail than the 5D Mark II full frame image next to the 1DII image,
yet both were taken with the same lens. Pixel pitch is controlling
the detail, not sensor size. And that is illustrated by the right-most
image in the figu a Canon 7D 1.6x crop sensor shows the most detail
because the pixel pitch is smallest, not because of the crop.
And for this lens, diffraction is smaller than the all the pixel
sizes.

See:
http://www.clarkvision.com/articles/cropfactor/

Roger


Indeed, I've been listening to too much of the populist hype
on these groups, when I should have been reading a few more
tech sheets. I guess I just kind of suspected cropped sensor
cams would have tighter packed pixels, i.e. an increase in
pitch. It didn't occur to me they can be on a par with full-
framed bodies.

Yesterday, I did a bit of snooping around, and I came up with:


http://www.the-digital-picture.com/r...eos-rebel-xsi-
450d-digital-slr-camera-review.aspx

I was surprised at how well my XSi stacks up against the 40D
in particular.

I was also surprised that the pixel size of the XSi is in the
same range as some of the full-frame bodies, and not all that
far off the Canon heavy hitters like the 5D2.


I guess I'll have to brush off my math skills, and start
looking into camera internals a bit more, paying a lot more
attention to pixel pitch instead of crop factor.


Thanks, Roger, for taking the time to point me in the right
direction.

Take Care,
Dudley