1.5 or 1.6 "magnification"; isn't it really just cropping.

I'm not expert on digicams, but I do read a few mags each
month...it's my understanding that the so-called focal
length increase of 1.5-1.6 doesn't tell the whole story.
First of all, doesn't a small sensor merely crop the image,
so that it "seems" as though the focal length has been
extended? Wouldn't we be better off with a full frame
sensor and crop our own images? There would be more data
coming from a bigger sensor, so cropping an image to match
the area covered by a smaller sensor shouldn't result in
data loss. Right?

Another thing is that the perspective of an image taken with
a certain focal length lens is the same whether used on a
film camera, full framed digicam, OR on an APS sized sensor
equipped digicam. The only difference is that the APS sized
sensor crops the image. Perspective remains identical.


dave

 

Is a 35mm image just a "crop" of a 6x6 medium format image? Is a 6x6
just a crop of a 4x5? Sure - in a sense.

What really matters is the resolution obtained by that sensor. Why waste
resources (ie, $$$'s) on unwieldy big cameras and big lenses to cover a
large sensor area if an APS sized sensor has all the resolution you need
(and eventually, it will)?

There is the issue with perspective, but that's minor IMHO.

YMMV,

B.A.S.

 

That is correct.

I used to feel the same way. I'm no longer so sure.

Sensor size is primarily an issue of technology and costs. A bigger
sensor, with the same number of pixels/sq in as the small sensor would
cost very much more, as evidenced by the cost of the few DSLRs with a
sensor as large as a 35mm frame.

There's no magic to the 35mm frame size. I have my grandfather's
camera that uses 102 roll film. You can make a 4x6 contact print from
a negative from that camera. 35mm developed over the years as the
acceptable compromise between film, lens, camera, ergonomic and
economic concerns. Smaller film formats such as APS and 126 have been
tried but did not displace 35mm. Larger film sizes, notably 120, are
regularly used. Cameras using 120 file are notably larger than 35mm
cameras.

Everyone talks about 35mm as "full frame" because that's what we are
most used to. DSLRs are aimed at the same market as traditional 35mm
cameras were aimed at, so the terminology is used. However, the
technology is different. As long as the resolution can be as good as
film (it's not yet IMHO), I don't care if the sensor is the same size
as 35mm or not.

The trend of computer and chip technology over the last 40 years has
been "more and faster stuff in a smaller space." I expect that trend
will continue for a while, resulting in increased resolution.

Guy

 

Correct, Dave. It is just a crop, any magnification is limited to the
increased enlargement needed to create a print of the same size as a
full-frame camera's image.
That enlargement does have an impact on depth of field but not, as you say,
on perspective.

 

Yes on all counts. Welcome to the world of marketing digital cameras
to the general public.

There are arguments that can be made here, depending on your
definition of 'perspective'. By the dictionary definition, perspective
remains the same from a fixed vantage point, regardless of camera *or*
focal length used. It is the process of foreshortening objects with greater
distance.

And if you *really* want to be technical, perspective never changes,
period. Objects decrease in apparent size by the same percentage with
distance, no matter what.

What changes is the appearance they have in the final image, which
renders perspective into a flat plane. While it doesn't change with focal
length, the overall affect on the image does, so it can (arguably) be said
to change with focal length. The other issues that come into play are field
curvature and rendered field-of-view, where a wide-angle lens may distort
the image more towards the edges, and will also render a wide field-of-view
into a narrower angle on the print - the fisheye effect. Yes, you get
greater curvature towards the edges, and yes, cropping the image will
reduce the apparent effect of this. "Perspective", in and of itself, never
changes, but the image is radically altered nonetheless. There just hasn't
been a good term put forth yet to describe this.


- Al.

 

Yes.. The best way to understand would be to take a 35mm
film camera. Mount it on a tripod and take two identical
frames. You'll wind up with two frames of film containing
the identical image.

Now cut an APS sized square from the middle of one of the
frames. Then print both the full frame and the cropped square
at 6" x 4"

You have to enlarge the APS sized square more than the full
35mm frame. As a result, the image made with the cropped piece
of film will have a narrower field of view. It will appear to
have been taken by a lens with a longer focal length.

Other than that, nothing else about the image changes.

Sure.. That's no problem.. All you have to do is buy a digital
camera with a full 35mm sensor. They make them. They start
at around $7000.00 US Here's one:

http://www.bhphotovideo.com/bnh/controller/home?O=productlist&A=details&Q=&sku=266188&is=REG

Currently, large sensors are expensive to manufacture. Regardless
of the size, the sensor adds the greatest expense to the cost of
any digital camera. That's one reason why even simple P&S digital
cameras are at least 3X the cost of a 35mm compact P&S

 

And where can I buy a camera with an APS sized sensor that performs the same
as a Canon 1Ds? Even the sensor in the 1D mark II is bigger than APS.

 

Bay Area Dave wrote:

You have the right idea, but ,let me add one additional thing to your
thought process.

All else being equal (same sensor quality and MP) the smaller sensor
(1.6) would not only crop the image a little more, but it would also be like
using a finer grain film since it can resolve the same total amount of
information as the larger sensor.

There does seem to be a tendency for larger sensors to have less problem
with "noise" that is the stuff that looks a lot like grain in a high speed
film, under low light and the seem to have a tendency to have fewer problems
in some other areas, so over all the larger one may produce a better image.

 

Al, I think this is misleading. The *relative* size of two objects at
different distances from the observer *will* change if he moves closer
or further away.

Example: If you have 2 identical objects, one at 10 metres and the other
at 20 metres, the latter will "appear" 0.5 times the size of the closer
one. If the observer then moves back 20 metres, the further object will
appear 0.75 times the size of the closer one.

Also the apparent angle of a line between two objects not in a straight
line with the observer will change on a change in the position of the
observer.

These effects are what is generally understood by "perspective", it
changes with change of position, and *only* with change of position.
Anything done by way of changes of lens focal length at the same
position is merely a change of magnification.

I realise you said the same thing in your first paragraph, but your
second paragraph seemed to be contradicting the first, and to be
incorrect.

 

There seems to me be some lack of clarity in saying, on the one hand,
"same sensor quality" and on the other saying the smaller one "can
resolve the same amount of image as the larger sensor.

It seems to me that the "quality" of a sensor (ignoring important issues
such as colour and contrast) is related to pixel *density* rather than
pixel *number*. To equate quality with total pixel number is to confuse
quality with quantity. Nor does "higher density" always mean "higher
quality" - see below.

Yes, I have gained the same impression (which again reinforces the view
that quality is not just pixel numbers). I suspect that the smaller
pixels used on many digital cameras will never match the image quality
of a similar number of pixels on a larger sensor with lower density, for
quantum reasons. It is likely that "sensor quality", like optimum
resolution for lenses, is fated for reasons of physics to rise to a peak
at a given density and then decline.

OTOH, it could be that having such a large investment in 35mm lenses I
am just desperate to convince myself that full-frame sensors at
reasonable prices will arrive in the not-too-distant future. Otherwise I
will be condemned to be deprived of decent wide angle in all my digital
work....

 

Perspective is controlled by *location* -- where you're standing
relative to the subjects. The focal length has nothing to do with it.
The classic demonstration of this is the series of photos commonly
provided as counter mats by various lens manufacturers showing a
series of pictures taken with each of their lenses all from the same
position and pointed the same direction. The size relationships
between the objects are the same in all of the pictures (that's what
"perspective" means in this context); they look like they're just
crops out of the center of the widest picture in the set (except, for
the nitpickers, that there's no increase in grain or anything).

 

6x6 film camera is a good example.
Besides perspective, depth of field and lens speed are changed on digital
SLR.

 

Wouldn't it be fair to say, though, that all the various noise
components could conceivably be reduced in small ccds, with each new
generation, but that the best optics and photos tend to come from more
'glass', rather than smaller diameter lenses, regardless?

 

I would say so. Of course this all leads back to the same conclusion:
It is best to judge the results than the equipment used to create them.

 

No.

There is no reason why optics smaller than current 35mm ones should not
perform as well, and indeed lots of practical indication that they
perform better (until, presumably, one gets so small that interference
effects kick in).

Sensors, OTOH, are already using technology which has dimensions close
to the wavelength of the light.

 

Well said, Guy. I've been coming to the same conclusions.

B.A.S.

 

[]

No.

- there are ultimately quantum limits to the noise, which will favour
larger chips. How large depends on how much noise you can tolerate, and is
therefore subjective.

- more glass is bad, because it causes more light loss, more internal
reflections and increases manufacturing complexity. Lenses for formats
smaller than "35mm" can be just as good, providing diffraction does not
become an issue. So again, there is an engineering and subjective
trade-off between the performance and the lens complexity.

Cheers,
David

 

http://www.outbackphoto.com/artofraw/raw_05/essay.html

 

Yes, but resolution tests are done on that crop, so the whole story is
told.

You just figured out the Canon 10D vs 1Ds. The 1Ds actually has
(slightly) lower resolution than the 10D, shown here...

http://www.outbackphoto.com/artofraw/raw_05/essay.html

Right, Sigma makes it quite easy to understand with their "sports
finder." The sports finder is a great feature, a big advantage over
all the other D&SLRs, but truth be told its a money saver. All they
did was darken the cropped-out part of the SA-9's (film SLR) full
frame view finder. When you put any lens on the SDs, you see exactly
what is happening, and how it gets cropped...

http://www.dpreview.com/reviews/sigmasd10/page3.asp

It's a really great feature because you can compose/anticipate using
cues beyond the imaged frame, why all cropping DLSRs don't do this
baffles me.