Prev: P & S Raw?
Next: How to determine if you can properly choose a camera
From: Paul Furman on 26 Feb 2010 18:04
Henry Olson wrote:
> Diffraction
> size is more revealed by and proportional to distance. The smaller
> focal-lengths required on smaller sensors don't reveal as much diffraction
> as a longer focal-length on a larger sensor.

I won't even ask what this was supposed mean 'cause it's nonsense. If
anything vaguely the opposite of reality but too jumbled to make sense of.
From: MikeWhy on 27 Feb 2010 03:10
"Bill W D" <billwd971610(a)apopularisp.net> wrote in message
news:fd3fo5daik78fcrmk5bmp7g4fc3avbvm32(a)4ax.com...
> The ONLY reason diffraction is less visible at larger apertures (in
> diffraction limited glass) is that the greater amount of light devoted to
> the center of the airy-disk (the real information) overwhelms the dimmer
> amount of light dispersed into the diffraction. That diffraction dispersal
> width and intensity which never changes. Smaller apertures don't CAUSE
> more
> diffraction, they only allow it to become more visible because there is
> less light to focus into the center of the airy-disk. It is the exact same
> width of dispersion no matter how large or small the aperture if the
> distance is retained.

I'll tell you what. I'll go dig out my college physics texts and review
them. Then I'll come back and we'll have this discussion. In the meantime,
none of the above changes anything, and none of it differs from naive
observation and measurement. When I make the hole smaller beyond a certain
point, and I do know with good precision where that point is, the image gets
progressively softer. Endless examples on the web with EXIF, and controlled
ISO lens chart shots confirm those numbers. I don't have to be an organic
chemist to light my Coleman stove. It lights just the same, and while it
illuminates my world, it seems to only darken yours.

From: Pete on 27 Feb 2010 03:59
MikeWhy wrote:
> "Henry Olson" <henryolson(a)nospam.org> wrote in message
> news:7n7fo514807qknv9kt25c80viaqc9v0ecb(a)4ax.com...
>> ...
>> No, we're not saying the same things. And "tiny" does not mean poor
>> optics.
>> Tell that to my diffraction-limited quality plan-apochromatic 100x
>> oil-immersion phase-contrast microscope objective; which delivers nice
>> images even when pushed to 1200x. According to your rudimentary way of
>> thinking about diffraction I should see nothing but diffraction through
>> that objective lens. Educate yourself. The free tutor you're getting on
>> the
>> internet doesn't seem to be working in your favor. You have to at least
>> know enough to know what resources on the net are misinformation
>> fabrications or genuine information.
>
> The minute they start using your apo 100x lens in a P&S, or you use the
> P&S lens on your microscope, I'll acknowledge the relevance. Until then,
> let's just figure that the Hubble isn't gratuitously big for big's sake,
> and that its f/24 optics had some influence on the sensor array's 15
> micron pixels. OTOH, if small is good, smaller would be even better, and
> Hubble's optics and 16k pixels would fit in a thimble, but only for
> convenience's sake so we could find it if we should drop it.

An oil-immersion phase-contrast microscope objective will get more
resolution at the expense of image accuracy, as explained here:
http://www.microscopyu.com/tutorials/java/mtf/spatialvariation/

That would be useful for a P&S, not.

Hubble? Obviously designed on the back of a cigarette packet and at the last
minute some kind company donated a 15 micron pitch sensor. The f/24 optical
path is just a coincidence :-)

Pete


From: Paul Furman on 27 Feb 2010 12:34
Henry Olson wrote:
> No, we're not saying the same things. And "tiny" does not mean poor optics.
> Tell that to my diffraction-limited quality plan-apochromatic 100x
> oil-immersion phase-contrast microscope objective; which delivers nice
> images even when pushed to 1200x.

The smallest feature that visible light can reveal is about half a
micron because lightwaves have width. Enlarge that 1000x as you suggest
onto a 2.5 micron pixel as we are discussing and that feature is going
to reach across about 200 pixels, eating up about 5,000 pixels worth of
sensor space. Not that this has any relevance to vaguely normal
photography but...

The only 1200x microscope objectives out there are toys, because it
doesn't work.

An f2.8 lens used at 20X has an effective aperture of about f/58. You
need a larger format camera to make use of that. Small sensors are
generally better for microscope objectives because of the smaller image
circles but by the time you reach 100x, you need big pixels to work
efficiently:
http://www.microscopyu.com/tutorials/java/digitalimaging/pixelcalculator/index.html

Microscope objectives are faster than photographic camera lenses, you
mostly need very small pixels to resolve all the objectives offer, and
you need a bare sensor, not a super-zoom on the back of the scope. There
are some old macro lenses for film but they were designed for 35mm up to
4x5 format so not as sharp on an AP-S sensor or 4/3 and really would do
best on MF.
From: MikeWhy on 28 Feb 2010 05:10
"Henry Olson" <henryolson(a)nospam.org> wrote in message
news:6nqho5luoghia7pquf4i5hba1qphb019g1(a)4ax.com...
> On Sat, 27 Feb 2010 08:59:35 -0000, "Pete"
> <available.on.request(a)aserver.com> wrote:
>
>>MikeWhy wrote:
>>> "Henry Olson" <henryolson(a)nospam.org> wrote in message
>>> news:7n7fo514807qknv9kt25c80viaqc9v0ecb(a)4ax.com...
>>>> ...
>>>> No, we're not saying the same things. And "tiny" does not mean poor
>>>> optics.
>>>> Tell that to my diffraction-limited quality plan-apochromatic 100x
>>>> oil-immersion phase-contrast microscope objective; which delivers nice
>>>> images even when pushed to 1200x. According to your rudimentary way of
>>>> thinking about diffraction I should see nothing but diffraction through
>>>> that objective lens. Educate yourself. The free tutor you're getting on
>>>> the
>>>> internet doesn't seem to be working in your favor. You have to at least
>>>> know enough to know what resources on the net are misinformation
>>>> fabrications or genuine information.
>>>
>>> The minute they start using your apo 100x lens in a P&S, or you use the
>>> P&S lens on your microscope, I'll acknowledge the relevance. Until then,
>>> let's just figure that the Hubble isn't gratuitously big for big's sake,
>>> and that its f/24 optics had some influence on the sensor array's 15
>>> micron pixels. OTOH, if small is good, smaller would be even better, and
>>> Hubble's optics and 16k pixels would fit in a thimble, but only for
>>> convenience's sake so we could find it if we should drop it.
>>
>>An oil-immersion phase-contrast microscope objective will get more
>>resolution at the expense of image accuracy, as explained here:
>>http://www.microscopyu.com/tutorials/java/mtf/spatialvariation/
>>
>>That would be useful for a P&S, not.
>>
>
> By referencing a principle distantly related to phase contrast microscopy
> and then babbling nonsense about it? You're right, that's not helpful at
> all to anything being discussed. But you go right ahead and juggle your
> red-herrings. If you can't dazzle them with brilliance try to baffle
> everyone with random-website bullshit, right?


He obviously meant to post this link:
http://www.microscopyu.com/tutorials/java/digitalimaging/pixelcalculator/index.html.

Have fun.

> The point being made that small optics are not automatically "bad" just
> because they are small nor are they prone to more diffraction. The shorter
> focal-lengths being used for smaller sensors lessens any problems from
> diffraction faced by lenses with longer focal-lengths on larger sensors.

If that's your misapprehension, allow me set it straight. I never once said
otherwise about small lenses, but see my further comments below. I addressed
only the pixel pitch of the sensor, and its early susceptibility to the
detail robbing effects of diffraction. For the same set of physics, and the
same diffraction for the given conditions, the smaller pixel will resolve
diffraction rings that the larger pixel won't even notice. Is that finally
clear enough?

....
> How much did it cost them to get the Hubble Telescope mirror
> diffraction-limited? $450,000,000. When it costs that much in time and
> manpower to create just ONE diffraction-limited curve on a 2.4 meter

<... huge, empty snip ...>

Tiny makes sense because it's cheap, not because it produces better or even
equally good images. But, come to think of it, since I never said so before
I'll say so now. Those cheap Fuji bastards! Despite the low development and
manufacturing cost of even mediocre, tiny optics, they hobbled that POS with
an f/5.6 lens on the 2.4 micron pitch sensor. What kind of idiots do they
think their customers are? Dollars to donuts, what do you bet those same
penny pinching theives bothered to build a lens that delivers 50% constrast
line pairs to match the 2.4 micron pixel pitch? Recalling that they're cheap
and easy to make, that is.


First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4
Prev: P & S Raw?
Next: How to determine if you can properly choose a camera