Digital 17mm Is Not Equivelent 27mm on 35mm Film

[a]

Archived from groups: rec.photo.digital (More info?)

Just realised that even though manufactures give 35mm equivalents, it is not
really true.

If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
still get distorted pics just like 17mm lenses on 35mm film cameras.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"A" <nospam@nospam.com> wrote in message
news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...
> Just realised that even though manufactures give 35mm equivalents, it is
not
> really true.
>
> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
> still get distorted pics just like 17mm lenses on 35mm film cameras.

The "equivalents" really only apply to *field of view,* since you're really
just cropping out the middle portion of the normal 35mm film camera's image.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Mark²" <mjmorgan(lowest even number here)@cox..net> wrote in message
news:FL3xd.59115$ka2.18959@fed1read04...
>
> "A" <nospam@nospam.com> wrote in message
> news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...
>> Just realised that even though manufactures give 35mm equivalents, it is
> not
>> really true.
>>
>> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor),
>> you
>> still get distorted pics just like 17mm lenses on 35mm film cameras.
>
> The "equivalents" really only apply to *field of view,* since you're
> really
> just cropping out the middle portion of the normal 35mm film camera's
> image.
>
I assume most of you know this, but I didn't until I read it in a book. If
you tilt the camera up when taking a tall building, the resulting 'falling
back effect" can be corrected using the deform tool in a photo editor, with
some reduction of picture content. Sort of like having an old swing and tilt
plate camera (which I bet would do a better job).
Dave Cohen

 

[Bob]

Archived from groups: rec.photo.digital (More info?)

"Dave Cohen" <dave@example.net> wrote in
news:1103417385.c5fa8a62fb815db7e0b4afea6ffddb7d@teranews:

> Sort of like having an old swing and tilt
> plate camera (which I bet would do a better job).
>

Yeah. Somehow using a view camera with it's movements produces a different
picture than taking a digital photo and manipulating it. My wild guess is
it might have something to do with lens distortion.

I have noticed, at least with some images, if you do a simple perspecive
change in software, some of the lines that should be parallel with others
end up pointing in different directions.

Bob

--
Delete the inverse SPAM to reply

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"A" <nospam@nospam.com> wrote in message
news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...
> Just realised that even though manufactures give 35mm equivalents, it is
> not
> really true.
>
> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
> still get distorted pics just like 17mm lenses on 35mm film cameras.


It's the biggest misconception in digital photography. People think a
smaller sensor somehow changes the optics of the lens. Film or digital --
it's the same light projected through the same lens, you just use a smaller
piece of it with the digital.

You used the term "crop factor." That's what it's all about.

Good shooting,
Bob Scott

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

<< It's the biggest misconception in digital photography. People think a
smaller sensor somehow changes the optics of the lens. Film or digital --
it's the same light projected through the same lens, you just use a smaller
piece of it with the digital. >>

Bob-

Funny you should mention that. Another related misconception is that sensor
size is not as important as the number of megapixels. In other words, it is OK
to have a small sensor if you have enough megapixels.

Thirty years ago lenses were rated in "lines per millimeter". I don't fully
understand the more modern modulation transfer ratio (MTR), but I can relate
lines to pixels. So, above a certain number, no matter how many pixels you
have, the image resolution is limited by the lens resolving power. Fewer
sensor millimeters means fewer "lines" in the resulting image. This is another
flaw in the claim of 35 mm equivalency since you would divide lines per
millimeter by the cropping factor.

Perhaps it wasn't a significant factor when sensors were no more than one
megapixel, but is certainly is in today's eight to ten megapixel world.

Fred

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

Dave Cohen wrote:

> I assume most of you know this, but I didn't until I read it in a book. If
> you tilt the camera up when taking a tall building, the resulting 'falling
> back effect" can be corrected using the deform tool in a photo editor, with
> some reduction of picture content. Sort of like having an old swing and tilt
> plate camera (which I bet would do a better job).
> Dave Cohen

Not many people know that movements on a LF camera (swing and tilt) introduces
distortion into the image, by altering the aspect ratio of the subject. The
converging - sometimes diverging - lines can be corrected with movements, but at
the expense of the building or whatever being rendered taller, and the
foreground over large.

Perspective correction can be done in Photoshop with Transform tools, and the
image aspect ratio can be corrected as well with Image Size by un-ticking
Constrain Proportions, and adjusting either height or width as appropriate.
It's up to the operator how he decides when it is correct, but it can be done.

Colin

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"A" <nospam@nospam.com> wrote:

> Just realised that even though manufactures give 35mm equivalents, it is
not
> really true.
>
> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
> still get distorted pics just like 17mm lenses on 35mm film cameras.

No, you get exactly the same image you would have with a 27mm on 35mm. The
way the lens projects the image onto the sensor scales exactly. The only
thing that changes is the DOF, which is increased by the ratio of the
formats in the smaller format.

As Fred M. points out, the mognification from the sensor to the print is
larger, so for the same resolution lens, you also may have a lower
resolution print, but most people find that film's MTF and noise are so
grody that an 8MP 1.6x sensor captures about the same apparent detail as ISO
100 film in 35mm.

David J. Littleboy
Tokyo, Japan

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"A" <nospam@nospam.com> wrote in message
news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...

> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
> still get distorted pics just like 17mm lenses on 35mm film cameras.

I don't believe it. Can you post some pictures to prove it? Take a picture
with a 17mm lens on a digital camera, and take an equivalent picture with a
28mm lens on a film camera. Then post both images side by side and show us
why you think they're so different.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

On Mon, 20 Dec 2004 17:20:41 GMT, "Andrew Koenig" <ark@acm.org> wrote:

>"A" <nospam@nospam.com> wrote in message
>news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...
>
>> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
>> still get distorted pics just like 17mm lenses on 35mm film cameras.
>
>I don't believe it. Can you post some pictures to prove it? Take a picture
>with a 17mm lens on a digital camera, and take an equivalent picture with a
>28mm lens on a film camera. Then post both images side by side and show us
>why you think they're so different.

This will never happen, because the image will be practically
identical.

--
Owamanga!

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Owamanga" <nomail@hotmail.com> wrote in message
news:7b3es0h8fk16ki3820ats023ku563sbvl1@4ax.com...

> This will never happen, because the image will be practically
> identical.

That's what I think too, which is why I'm asking for evidence to the
contrary.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

In article <7b3es0h8fk16ki3820ats023ku563sbvl1@4ax.com>,
Owamanga <nomail@hotmail.com> wrote:
>On Mon, 20 Dec 2004 17:20:41 GMT, "Andrew Koenig" <ark@acm.org> wrote:
>
>>"A" <nospam@nospam.com> wrote in message
>>news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...
>>
>>> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
>>> still get distorted pics just like 17mm lenses on 35mm film cameras.
>>
>>I don't believe it. Can you post some pictures to prove it? Take a picture
>>with a 17mm lens on a digital camera, and take an equivalent picture with a
>>28mm lens on a film camera. Then post both images side by side and show us
>>why you think they're so different.
>
>This will never happen, because the image will be practically
>identical.

Well, the depth of field might change, that depends on the apertures used in
each case.

There is a myth that the choice of focal length has some sort of magic
effect on the perspective of a scene, and I think that's where a lot of this
squabbling over terms comes from. I can only suppose that those who relate
focal length to perspective this way have read a simplified explanation of
the effects of telephoto and wide angle lenses in composition, and think
that the differences they see are due to the focal length, rather than the
distance to their subject. Thinking about it carefully for a few minutes
should be enough to dispell this myth, but some insist on clinging to it.

Addressing the original point, the distortion from a wide-angle lens is a
direct consequence of the field of view represented in the image. Since the
17mm lens on the smaller format digital gives the same field of view as a
28mm lens (I'm assuming we're talking about so-called 1.6x digicams here) on
a larger format 35mm camera, the resultant images will show the same amounts
of rectilinear distortion.

To those who still don't get this, I suggest you think for a moment about
the focal lengths used in zoom compact digital cameras. These ofetn have
lenses that don't even reach 10mm. However, these cameras are perfectly
capable of taking telephoto images which do not show the sort of distortion
associated with ultra-wide angle photography with rectilinear lenses.

People who shoot multiple different film formats tend to understand this
stuff properly (they have to), and know, for example, that a 110mm lens on
6*9 medium format gives pretty much exactly the same results as a 50mm lens
on 35mm, or a 28mm lens on an EOS 20D, the only differences being in terms
of the detail in the final image, and the aperture they'll need to use to
get the same depth of field.

Once you understand this, you'll see why it doesn't make a blind bit of
difference whether you think about this phenomenon as a "crop factor", a
"focal length magnifier", or anything else. They're all equivalent.

Personally, I don't see why it's important to see everything in terms of
35mm equivalents anyway. For me, with my 10D, a 28mm lens is standard, a 50
is a short telephoto, suitable for portraits, a 20 is a wide-angle, and a 14
is an ultrawide.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

In article <8k4xd.3728$yK.3408@newsread3.news.atl.earthlink.net>,
Robert Scott <desmobob@earthlink.net> wrote:
>
>"A" <nospam@nospam.com> wrote in message
>news:cq2g8h$nop$1@newsg2.svr.pol.co.uk...
>> Just realised that even though manufactures give 35mm equivalents, it is
>> not
>> really true.
>>
>> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
>> still get distorted pics just like 17mm lenses on 35mm film cameras.
>
>
>It's the biggest misconception in digital photography.

Indeed it is, and if you, like the original poster, think that you'll get
the same area distortions from rectilinear projection with a given lens,
then I'm afraid you're suffering from said misconception yourself. ;-)

Think about where the distortion comes from. Imagine a 50mm simple lens
focusing an image on to a piece of paper below it, the paper underneath the
centre of the lens will be exactly 50mm from the optical centre of the lens.

Now with a little basic schoolboy triganometry, we can get to the bottom of
this "crop factor" business. Let's do the maths:

The paper 10mm out from the centre point will be 51mm from the optical
centre of the lens (by Pythagoras - the light is coming along the hypotenuse
of a right-angled triangle 50mm by 10mm by 51mm). The paper 20mm out from
the centre will be 54mm from the optical centre of the lens.

This means that as we get further and further from the middle, we are
"stretching" the image, causing a distortion. This is essentially the same
problem that cartographers have when they want to map the curved surface of
the Earth on to a flat map. If you represent the area accurately, the
relative angles get distorted. If you represent the angles accurately, the
area gets distorted. You can't keep both.

We have the same choice in photography - you can use a rectilinear lens (the
angles stay the same, but the areas get distorted), or you can use a fisheye
lens (the areas stay the same, but the angles get distorted). You can't keep
both, and this is where distortion comes from.

*but* the amount of distortion varies according to the distance from the
optical centre of the lens, as we've already seen. With a 35mm camera, a
point at the centre of the short edge is 18mm from the centre of the frame.
With a camera like an EOS 20D, the point at the edge of that same line is
11.25mm from the centre of the frame.

Now think about what that extra distance represents. If you think of a light
ray coming through the centre of the lens and hitting the edge of the scene,
that ray will come from an angle off to the side of arctan(18/50) = 20
degrees (draw a diagram if you can't see why).

We'll get the same level of distortion from a lens that makes a 20 degrees
angle 11.25 mm out from the centre on out 20D (since that's where the edge
of the frame is).

A bit of triganometry says that the focal length is 11.25 / tan(20) (again,
draw a diagram if you can't see why), which is 31.25mm.

So we get the same amount of distortion from a 31.25mm focal length on a 20D
as we do from a 50mm focal length on 35mm.

50 / 31.25 is, surprise surprise, 1.6, so it all works out nicely.

>People think a smaller sensor somehow changes the optics of the lens.

And, as we know, that's impossible. However, as we've seen above, the
smaller sensor *requires* different optical properties to get the same
level of distortion. Since we can't change the optics of the lens, we need
to change the lens to one that has the optics we need, that being one with a
1.6 times shorter focal length.

QED

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Chris Brown" <cpbrown@ntlworld.no_uce_please.com> wrote in message
news:8nkj92-jsn.ln1@narcissus.dyndns.org...
SNIP
> Well, the depth of field might change, that depends on the apertures
> used in each case.

Obviously, since the aperture number is the physical size related to
the focal length, the larger focal length also has a larger physical
aperture size than the shorter focal length lens has at the same
aperture value.

> There is a myth that the choice of focal length has some sort of
> magic effect on the perspective of a scene, and I think that's where
> a lot of this squabbling over terms comes from.

IMO it is all due to some people changing *two* parameters at the same
time, both focal length *and* shooting position to achieve the same
field of view in the cropped image. Changing the shooting position
changes the magnification factor between near and far objects, and
thus perspective. Cropping is just that, cropping.

SNIP
> Once you understand this, you'll see why it doesn't make a blind
> bit of difference whether you think about this phenomenon as a "crop
> factor", a "focal length magnifier", or anything else. They're all
> equivalent.

But calling it a "focal length magnifier" is what started the whole
confusion, so I'd avoid it. "Crop factor" (compensation) is
technically correct and it describes what happens. To compensate for
the crop factor, without changing perspective/shooting position, a
lens with a wider FOV needs to be used.

Bart

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

In article <41c737f6$0$1746$e4fe514c@news.xs4all.nl>,
Bart van der Wolf <bvdwolf@no.spam> wrote:
>
>But calling it a "focal length magnifier" is what started the whole
>confusion, so I'd avoid it. "Crop factor" (compensation) is
>technically correct and it describes what happens.

Only in relationship to 35mm, which is somewhat arbitrary, as the lenses
project a circle, not a 36*24mm rectangle, so all images from them are
"cropped" to a varying extent.

It's useful if you're coming from a 35mm background, and are using the exact
same lenses, but apart from that there's nothing magical about 35mm. For
example, if I was able to mount my 1938 Zeiss Tessar on my 10D, what would
the crop factor be? That lens has never been used infront of 35mm film.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Chris Brown" <cpbrown@ntlworld.no_uce_please.com> wrote in message
news:85oj92-e2o.ln1@narcissus.dyndns.org...
> In article <8k4xd.3728$yK.3408@newsread3.news.atl.earthlink.net>,
> Robert Scott <desmobob@earthlink.net> wrote:

>>It's the biggest misconception in digital photography.
>
> Indeed it is, and if you, like the original poster, think that you'll get
> the same area distortions from rectilinear projection with a given lens,
> then I'm afraid you're suffering from said misconception yourself. ;-)


Your math certainly seems sound and your explanation is compelling enough to
convince me I had the wrong idea about what was happening.

But I just need to do a simple test to prove it to myself. I'm reserving
further comment on this thread until I take a pair of identical images at
18mm with the D70 and 28mm with the F4S.



Thanks for taking the time to lay it out the way you did. That was very
enlightening.

Happy Holidays everyone!

Bob
the stubborn cynic

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Bart van der Wolf" <bvdwolf@no.spam> wrote:
>
> SNIP
> > Once you understand this, you'll see why it doesn't make a blind
> > bit of difference whether you think about this phenomenon as a "crop
> > factor", a "focal length magnifier", or anything else. They're all
> > equivalent.
>
> But calling it a "focal length magnifier" is what started the whole
> confusion, so I'd avoid it.

No! This whole confusion is because the original poster didn't realize that
lenses function differently on different formats. His claim was that the
"focal length multiplier" didn't apply!

Since "focal length multiplier" tells you (or should tell you<g>) that the
lens _functions_ differently, it's the better term.

> "Crop factor" (compensation) is
> technically correct and it describes what happens. To compensate for
> the crop factor, without changing perspective/shooting position, a
> lens with a wider FOV needs to be used.

Thinking of it as a "crop factor" is what caused both the original poster's
error and the common error in which people think the DOF stays the same for
the same lens. That's because it focuses people's minds on a given lens
being used on both cameras and not realizing that the same lens is functions
differently on the different cameras.

The best term would be "format conversion factor", since that's what it is.
That would make people realize that the 1.6x cameras are a different format
from 35mm and allow them to think about the different photographic functions
of a given lens on the different formats.

David J. Littleboy
Tokyo, Japan

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

In article <FtMxd.6263$yK.3597@newsread3.news.atl.earthlink.net>,
Robert Scott <desmobob@earthlink.net> wrote:
>
>"Chris Brown" <cpbrown@ntlworld.no_uce_please.com> wrote in message
>news:85oj92-e2o.ln1@narcissus.dyndns.org...
>> In article <8k4xd.3728$yK.3408@newsread3.news.atl.earthlink.net>,
>> Robert Scott <desmobob@earthlink.net> wrote:
>
>>>It's the biggest misconception in digital photography.
>>
>> Indeed it is, and if you, like the original poster, think that you'll get
>> the same area distortions from rectilinear projection with a given lens,
>> then I'm afraid you're suffering from said misconception yourself. ;-)
>
>
>Your math certainly seems sound and your explanation is compelling enough to
>convince me I had the wrong idea about what was happening.
>
>But I just need to do a simple test to prove it to myself. I'm reserving
>further comment on this thread until I take a pair of identical images at
>18mm with the D70 and 28mm with the F4S.

Go for it, and do keep a link to the images. This does, as you pointed out,
cause a lot of confusion, and seeing is believing.

>Thanks for taking the time to lay it out the way you did. That was very
>enlightening.

Glad you found it useful.

>Happy Holidays everyone!

Likewise

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Mark²" <mjmorgan(lowest even number here)@cox..net> wrote:
>
> "A" <nospam@nospam.com> wrote in message
>>
>> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor), you
>> still get distorted pics just like 17mm lenses on 35mm film cameras.
>
> The "equivalents" really only apply to *field of view,* since you're really
> just cropping out the middle portion of the normal 35mm film camera's image.

Is it also true that if you take a portrait of a person with a big nose
using a 50mm lens on your DSLR instead of an 85mm(*) lens on your SLR,
the nose will look bigger in the DSLR picture?

(*) The two aren't quite equivalent, 80mm would be.

 

[Guest]

Archived from groups: rec.photo.digital (More info?)

"Bill Tuthill" <can@spam.co> wrote in message news:41c86c74@news.meer.net...
"Mark²" <mjmorgan(lowest even number here)@cox..net> wrote:
>
> "A" <nospam@nospam.com> wrote in message
>>
>> If you shoot tall buildings at 17mm on digital (with 1.6 crop factor),
you
>> still get distorted pics just like 17mm lenses on 35mm film cameras.
>
> The "equivalents" really only apply to *field of view,* since you're
really
> just cropping out the middle portion of the normal 35mm film camera's
image.

Is it also true that if you take a portrait of a person with a big nose
using a 50mm lens on your DSLR instead of an 85mm(*) lens on your SLR,
the nose will look bigger in the DSLR picture?

(*) The two aren't quite equivalent, 80mm would be.

It would give you the same perspective because you would likely back away to
the same distance you would have if you used the 80mm on a film camera.
Remember, the nose:face size ratio (big or smaller nose) is NOT determined
by the focal length, but instead, the relative difference in distance to the
nose tip compared with the distance to the rest of the face.

This is easier to conceptualize if you imagine an EXTREME close-up.
Imagine yourself taking a picture of someone's face only one inch away from
a person's nose. If you did this, the nose would be one inch away, and the
rest of their face would be perhaps two inches and further away...TWICE the
distance away that their nose was from you.

This will greatly exaggerate the size appearance of the nose because there
is a nose:face distance ratio of 1:2. This difference will remain true
regardless of what focal length you have mounted.

Now imagine backing away from the nose/face. As you back away to a full one
foot (12 inches) from the nose, the ratio of distance from nose:face has now
changed to a far more similar 12:13 ratio (12 inches from the nose, and 13
inches from the face). As you move back to 8 feet away (for example), the
ratio of distance from nose vs. distance from face will become tiny (96:97).
You can see from this extreme example how the farther you back up, and less
out of whack the ratio of distance between (distance to) nose and (distance
to) face will become.

**This ratio will remain consistent regardless of what lens you use. What
WILL change is how much of your subject is visible within the frame.
So...when you've got your 50mm lens attached, and you add the 1.6 crop
factor, you will have to back up to a similar distance than you would with
your ~80mm lens.
--This will give you the same perspective.

It's all about relative distance to the subject's closest attribute (in this
case, the nose) compared with the distance to the rest of the subject.

You've probably seen those funny snap-shots taken of cows, or dogs, where it
seems like their nose stretches right out to you? -Those look that way
simply because they were taken at such close range that the distance ratio
was extreme...and that they used such a wide-angle lens, that you could
still see the entire face/head to appreciate the distorted difference.
-Mark