Equivalent focal lengths and crop factors...

Haven't read here for a while. Wondered when all this would surface
here. Well, it has and it won't go away, so maybe some clarification is
in order.

This whole subject is what happens when the tech geeks and the arty
types fail to communicate, things get lost in the cracks, important
things. We have focal lengths, equivalent focal lengths, angles of
view, image area, depth of field, frame composition, perspective...
real hodge podge!

Two things that weren't considered, and are important, a

1) The interface between photographer and camera, the view finder.

The one thing that does not ever change is the natural field of vision
of an individual, the inherent angle of view. That's the bench mark.
Now, an individual looks throught the viewfinder of a camera. If it's
an SLR, the viewfinder presents a perceived angle of view that does not
change. It does not allow the viewer to perceive the change in the
lens' angle of view.

Look through a rangefinder with interchangeable lenses. Different
marked off fields of view, so the viewer can perceive the changed angle
of view. Since the viewfinder is not part of the lens' optical train,
the perceived angle of view *of the viewfinder* does not change, and the
sense of the differences between the "perspectives" of different focal
length lenses is not apparent.

What happens with the SLR is that the eye is fooled into thinking that
what he/she sees has changed, when it has in fact not changed at all.
How that happens is obvious: A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train. One can often shoot with both
eyes open using a standard lens, without suffering the effects of
difference perceived by each eye.

Can't do that with a telephoto or wide angle. Well, at least I can't...

So what is going on here is that the camera user comes to identify the
perceived effect of different focal lengths, and comes to associate
particular differences with particular focal length lenses. And this is
why the persistent popularity of the "equivalent focal length".

Are the tech geeks right? Yep, they are. Is that important? Yep, it
is. Why? Because some things don't change, and one of them is DOF.
Another is relative distortion of the scene image. Both of these are
very important, of course.

Which leads to the second thing.

2) Perspective.

It turns out that there are two different definitions in common usage.
Now never mind the dictionaries, we know this without consulting them.
The first and literal definition is "what you see from where you are".
Here, we're talking about what you can and cannot see from a point of
view, which changes when the point of view is changed. You change your
viewing position, and some things that were not visible become so, and
some things that were visible, now are not.

That's not the same as the second definition, which has to do with the
relative size of viewed objects. The problem here is that this second
definition can easily be assumed to be a subset of the first: Things
get larger as one gets closer; given the natural angle of view, moving
closer makes an object take up more room in the field of vision.

However, the relative size of objects in the scene can be made
independent of the point of view created perspective. That is easily
accomplished with optics (and recorded in camera) by distorting the
scene optically. Artists do this all the time to achieve the effects
they want. They call it a change in perspective, and their use of the
term is what has established the second definition. Artists use
cameras, and so photography uses both definitions.

What's the problem? Well, there's a "gotcha!!" lurking here. And that
is that a photographer expects the camera to record reality as it is,
because that's what cameras are supposed to do. Right?

Well, that's as may be, but the reality is that no, they do not. What
they do is record an optically produced image, whatever that may be, and
the unwary can really get confused when optical distortion is not
acknowledged.

Here's the reality check: If things change in relative size, expect
different things to be visible. If different things are not visible,
you're looking at an optically produced distortion. And that's how you
tell which definition is being used.

The arty types don't care about all this "semantic detail". Perspective
change via distortion and perspective change via change in position are
deemed equivalent for their purposes, or so it seems. Not being an arty
type I wouldn't know for sure.

The tech geeks insist that they are entirely different things and should
never be confused. Perspective change is never produced by a change in
focal length, so they say, and according to the first definition, they're
right. And for them, apparently, that's all that matters. Not being a
tech geek, I wouldn't know for sure.

The truth is that both are missing an important aspect of this business,
and that is the effect these changes have on the viewer. Trouble is,
both types are also viewers, and because of this lack of complete
understanding, cannot come to an agreement about what's what.

There are some truths in some of the statements made, of course.

You cannot change perspective of the first definition by changing focal
lengths. But it's equally true that you certainly can change
perspective by the second definition by changing focal lengths. What
makes both true is that one makes the change by distortion and the other
does not.

As a camera user of some experience, you can equate the angle of view
provided by a focal length in one format with the perceived angle of
view in another. But the lenses themselves do not change focal length,
nor do they change angle of view. Enter the crop factor.

Incidentally, an optical train is an optical train whether displaying a
recorded image, or recording a displayed image. Optical crop is optical
crop. No, the two trains are not designed alike, but that's irrelevant.
For what its worth, that is...

Maybe if you guys would think all this through, you would realize that
reality does not change, and that differences in opinion are the result
of different perceptions. You want to be completely right? Better
think these things through a bit better. Otherwise, all you guys have
stuff to learn, and are better off listening to each other and trying to
figure out what you've missed.

Myself included, of course. I walk the walk I talk: tell me where I
screwed up here, or what I missed and I'll willingly stand instructed.

Will D.

"Will D." writes:

You cannot change perspective of the first definition by changing focal
lengths. But it's equally true that you certainly can change
perspective by the second definition by changing focal lengths. What
makes both true is that one makes the change by distortion and the other
does not.

The change you are making in 'distortion' are not changing the
perspective. Were this the case, then the objects present in both
frames would have different 'perspectives' (which they do not). The
'perspective' here is created by including other objects around the
edges which were not included in the first frame (or removing them).

B

"A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train."

I don't believe this was true on a 35mm film as the standard is 50mm when
approx a 70mm will give 1:1, but with a 1.6 crop with most digitals it may
now be right !

"Will D." wrote in message
...
Haven't read here for a while. Wondered when all this would surface
here. Well, it has and it won't go away, so maybe some clarification is
in order.

This whole subject is what happens when the tech geeks and the arty
types fail to communicate, things get lost in the cracks, important
things. We have focal lengths, equivalent focal lengths, angles of
view, image area, depth of field, frame composition, perspective...
real hodge podge!

Two things that weren't considered, and are important, a

1) The interface between photographer and camera, the view finder.

The one thing that does not ever change is the natural field of vision
of an individual, the inherent angle of view. That's the bench mark.
Now, an individual looks throught the viewfinder of a camera. If it's
an SLR, the viewfinder presents a perceived angle of view that does not
change. It does not allow the viewer to perceive the change in the
lens' angle of view.

Look through a rangefinder with interchangeable lenses. Different
marked off fields of view, so the viewer can perceive the changed angle
of view. Since the viewfinder is not part of the lens' optical train,
the perceived angle of view *of the viewfinder* does not change, and the
sense of the differences between the "perspectives" of different focal
length lenses is not apparent.

What happens with the SLR is that the eye is fooled into thinking that
what he/she sees has changed, when it has in fact not changed at all.
How that happens is obvious: A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train. One can often shoot with both
eyes open using a standard lens, without suffering the effects of
difference perceived by each eye.

Can't do that with a telephoto or wide angle. Well, at least I can't...

So what is going on here is that the camera user comes to identify the
perceived effect of different focal lengths, and comes to associate
particular differences with particular focal length lenses. And this is
why the persistent popularity of the "equivalent focal length".

Are the tech geeks right? Yep, they are. Is that important? Yep, it
is. Why? Because some things don't change, and one of them is DOF.
Another is relative distortion of the scene image. Both of these are
very important, of course.

Which leads to the second thing.

2) Perspective.

It turns out that there are two different definitions in common usage.
Now never mind the dictionaries, we know this without consulting them.
The first and literal definition is "what you see from where you are".
Here, we're talking about what you can and cannot see from a point of
view, which changes when the point of view is changed. You change your
viewing position, and some things that were not visible become so, and
some things that were visible, now are not.

That's not the same as the second definition, which has to do with the
relative size of viewed objects. The problem here is that this second
definition can easily be assumed to be a subset of the first: Things
get larger as one gets closer; given the natural angle of view, moving
closer makes an object take up more room in the field of vision.

However, the relative size of objects in the scene can be made
independent of the point of view created perspective. That is easily
accomplished with optics (and recorded in camera) by distorting the
scene optically. Artists do this all the time to achieve the effects
they want. They call it a change in perspective, and their use of the
term is what has established the second definition. Artists use
cameras, and so photography uses both definitions.

What's the problem? Well, there's a "gotcha!!" lurking here. And that
is that a photographer expects the camera to record reality as it is,
because that's what cameras are supposed to do. Right?

Well, that's as may be, but the reality is that no, they do not. What
they do is record an optically produced image, whatever that may be, and
the unwary can really get confused when optical distortion is not
acknowledged.

Here's the reality check: If things change in relative size, expect
different things to be visible. If different things are not visible,
you're looking at an optically produced distortion. And that's how you
tell which definition is being used.

The arty types don't care about all this "semantic detail". Perspective
change via distortion and perspective change via change in position are
deemed equivalent for their purposes, or so it seems. Not being an arty
type I wouldn't know for sure.

The tech geeks insist that they are entirely different things and should
never be confused. Perspective change is never produced by a change in
focal length, so they say, and according to the first definition, they're
right. And for them, apparently, that's all that matters. Not being a
tech geek, I wouldn't know for sure.

The truth is that both are missing an important aspect of this business,
and that is the effect these changes have on the viewer. Trouble is,
both types are also viewers, and because of this lack of complete
understanding, cannot come to an agreement about what's what.

There are some truths in some of the statements made, of course.

You cannot change perspective of the first definition by changing focal
lengths. But it's equally true that you certainly can change
perspective by the second definition by changing focal lengths. What
makes both true is that one makes the change by distortion and the other
does not.

As a camera user of some experience, you can equate the angle of view
provided by a focal length in one format with the perceived angle of
view in another. But the lenses themselves do not change focal length,
nor do they change angle of view. Enter the crop factor.

Incidentally, an optical train is an optical train whether displaying a
recorded image, or recording a displayed image. Optical crop is optical
crop. No, the two trains are not designed alike, but that's irrelevant.
For what its worth, that is...

Maybe if you guys would think all this through, you would realize that
reality does not change, and that differences in opinion are the result
of different perceptions. You want to be completely right? Better
think these things through a bit better. Otherwise, all you guys have
stuff to learn, and are better off listening to each other and trying to
figure out what you've missed.

Myself included, of course. I walk the walk I talk: tell me where I
screwed up here, or what I missed and I'll willingly stand instructed.

Will D.

"dylan" writes:

"A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train."

I don't believe this was true on a 35mm film as the standard is 50mm when
approx a 70mm will give 1:1, but with a 1.6 crop with most digitals it may
now be right !

70mm will give 1:1? How can you possibly make this statement given the
huge range of viewfinder magnifications available?

B

Let me try to explain.

If I look through my EOS 3 using a 24-85mm lens and compare the view through
the viewfinder with the view with my eyes I get the same size image direct
or through viewfinder when the lens is 70mm. If I try my EOS-10D the setting
is between 50 and 55mm.

Looking at the new Nikon F6, as an example, they quote viewfinder
magnification as .74 with 50mm and the EOS 1Ds Mk2 is .70 at 50mm.

The diagonal coverage of a 50mm is 40deg and a 70mm is 29deg giving a
maginification of 1.38.

1.38 x 0.7 = .96 and 1.38 x .74 = 1.02

both near enough to 1 for me.

Obviously as you say if there are other magnifications then this doesn't
work, ie Minolta 7 is 0.8 at 50mm, but I've always 70mm to be correct for my
cameras.

Certainly not the standard 50mm !!

Cheers



"Bruce Murphy" wrote in message
...
"dylan" writes:

"A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train."

I don't believe this was true on a 35mm film as the standard is 50mm when
approx a 70mm will give 1:1, but with a 1.6 crop with most digitals it
may
now be right !

70mm will give 1:1? How can you possibly make this statement given the
huge range of viewfinder magnifications available?

B

"dylan" writes:

Let me try to explain.

If I look through my EOS 3 using a 24-85mm lens and compare the view through
the viewfinder with the view with my eyes I get the same size image direct
or through viewfinder when the lens is 70mm. If I try my EOS-10D the setting
is between 50 and 55mm.

Looking at the new Nikon F6, as an example, they quote viewfinder
magnification as .74 with 50mm and the EOS 1Ds Mk2 is .70 at 50mm.

Not a particularly wide selection. The high-end bodies in both ranges
tend to have fairly similar eyepiece magnifications.

The diagonal coverage of a 50mm is 40deg and a 70mm is 29deg giving a
maginification of 1.38.

1.38 x 0.7 = .96 and 1.38 x .74 = 1.02

both near enough to 1 for me.

Obviously as you say if there are other magnifications then this doesn't
work, ie Minolta 7 is 0.8 at 50mm, but I've always 70mm to be correct for my
cameras.

Things vary a bit more with some of the manual focus bodies.

Certainly not the standard 50mm !!

The 'standard' lens is defined in terms of closeness to the diagonal
of the format in question. With such a lens (actually around 45mm on
35mm format) coupled with that format, you'll end up with a print
whose perspective looks natural at a normal viewing distance. In
contrast, shorter lenses will produce prints whose perspective looks
odd unless you're closer, and longer ones odd images unless you are
further away.

B

Will D. wrote:

You cannot change perspective of the first definition by changing focal
lengths. But it's equally true that you certainly can change
perspective by the second definition by changing focal lengths. What
makes both true is that one makes the change by distortion and the other
does not.

The relative size of objects with distance does not change with focal length.
You can easily prove this for yourself with the usual experiments.

Take two pictures from exactly the same point, one with a telephoto and one
with a wide angle, not moving the camera in between. Crop the wide angle
shot so that its framing matches the telephoto shot, and you will see that
they are identical. There is no distortion, and there is no perspective
change by either of your definitions.

--
Jeremy |

In article , dylan wrote:
"A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train."

I don't believe this was true on a 35mm film as the standard is 50mm when
approx a 70mm will give 1:1, but with a 1.6 crop with most digitals it may
now be right !

I guess you grew up with the modern auto-everything cameras, and their tiny
little viewfinders. Try looking through the viewfinder of an older camera
some time; something like a Nikon F2, or even just a Pentax Spotmatic.

Unfortunately when more and more camera automation came along, the makers
had a problem. They were stuck with three choices: to not present details
of what the camera was doing in the viewfinder (unacceptable in those days;
photographers might accept some level of automation, but they knew enough
not to trust the camera without verification); to present the information
alongside the traditional viewfinder (expensive, and difficult to use for
people wearing eyeglasses), or to reduce the size of the viewfinder image
to free up some space for all the additional display.

The modern digitals have, quite often, chosen to shrink the usage even
more (presumably because they share viewfinder optics with a film sibling);
the D70 still only has a magnification of 0.8 with a 50mm lens. Even the
D2X only gets a little better, at 0.86X. Canon are slightly better; the
0.70 of the 1DsII is somewhat understandable because of the larger sensor,
and the 300D matches the D70 at 0.8x. The 10D and 20D are both close to
0.9x, which is better than the far more expensive D2!
But the best viewfinder I've seen in a DSLR is in the Pentax *ist-D/DS.
These both have viewfinder magnifications of 0.95. Considering the fact
that the *ist-DS is in the same price bracket as the D70 and 300D, it's
quite impressive to see a viewfinder that large in a budget(sic) DSLR.

Disclaimer: I own a *ist-D. But I've used several of the alternatives
(including a D1, a 10D, and a D100), so I speak from personal experience.

On 2004-12-22, Jeremy Nixon wrote:
Will D. wrote:

You cannot change perspective of the first definition by changing focal
lengths. But it's equally true that you certainly can change
perspective by the second definition by changing focal lengths. What
makes both true is that one makes the change by distortion and the other
does not.

The relative size of objects with distance does not change with focal length.
You can easily prove this for yourself with the usual experiments.

Take two pictures from exactly the same point, one with a telephoto and one
with a wide angle, not moving the camera in between. Crop the wide angle
shot so that its framing matches the telephoto shot, and you will see that
they are identical. There is no distortion, and there is no perspective
change by either of your definitions.

This is true, but you miss the entire point:

It is the perceived angle of view a) through the viewfinder, and b) when
viewing a print in the normal manner, that creates this effect.

The viewfinder enforces a single angle of view, and so does the standard
print viewing practice. The latter would have you hold the print at
approximately the distance equal to the diagonal of the print itself.

If you want to complete the experiment you cite properly, you should
view the prints at the appropriate distances. For the wide angle shot,
put your nose on the surface of the print, or thereabouts. For the
telephoto shot, stand back across the room. In both cases, the
distortion effect is cancelled. Of course, that might not serve the
intent of the photographer...

But you cannot do this with your viewfinder, and so cannot cancel the
perceived distortion effect at that point in the process.

Again, think this through to the end, and you'll likely answer your own
questions.

Will D.

(repost - server prob.)

dylan wrote:

"A "standard" lens is one that will present
to the viewfinder what the eye would see if the viewfinder were a simple
window, unaltered by an optical train."

I don't believe this was true on a 35mm film as the standard is 50mm when
approx a 70mm will give 1:1, but with a 1.6 crop with most digitals it may
now be right !

Think you could not top post and snip away what is not needed?

The "normal" view with a cropped sensor (1.5) would be about 30 to 35mm, not 70mm.

Cheers,
Alan.

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.