Fisheye?... - DPChallenge Forums

02/14/2007 09:50:21 PM · #1

Are the fields of view the same for a normal 16mm lens and a 16mm fisheye lens? (minus the distortion?)

Canon EF 70-200mm f/2.8L IS USM

02/14/2007 09:54:42 PM · #2

.

Message edited by author 2007-02-14 22:24:00.

skylercall

Canon EOS-7D

02/14/2007 10:08:52 PM · #3

Would you mind posting your findings here? Others might like to know.

02/14/2007 10:16:42 PM · #4

Sure thing! =]

//www.bobatkins.com/photography/technical/field_of_view.html

Someone correct me if I'm wrong, but from what I'm reading it looks like a 16mm rectilinear is wider than 16mm fisheye, by as much as 3-5mm.

Is that right?

Message edited by author 2007-02-14 22:23:52.

02/14/2007 10:22:52 PM · #5

Actually, a 16mm fisheye is wider than a 16mm rectilinear. Hold on a minute, I'll post some calculated numbers...

Edit:

For 16mm fisheye on EOS 10D:
Diagonal FOV = 100.8Â°
Horz. FOV = 83.1Â°
Vert. FOV = 54.6Â°

For 16mm rectilinear on EOS 10D:
Diagonal FOV = 81.1Â°
Horz. FOV = 70.6Â°
Vert. FOV = 51.4Â°

Message edited by author 2007-02-14 22:26:57.

Canon EF-S 17-55mm f/2.8 IS USM

02/14/2007 10:30:29 PM · #6

So... I guess I got the numbers backwards... =\

Thanks kirbic!

idnic

Canon EOS-60D

02/14/2007 10:31:59 PM · #7

How wide is wide? This is the 16mm fisheye (I call him fishey)

hankk

Canon EOS-5D

Sigma 24-70mm f/2.8 EX Aspherical DF for Canon

02/15/2007 11:05:26 AM · #8

Check out this link, it shows the difference between a 15mm fisheye and several rectilinear lenses //forums.dpreview.com/forums/read.asp?forum=1029&message=11982207
You have to click on the picture to see the whole thing.

02/15/2007 11:10:15 AM · #9

Wow. Very cool. And very helpful. Thanks! =]

02/15/2007 12:23:55 PM · #10

Originally posted by tryals15:

Are the fields of view the same for a normal 16mm lens and a 16mm fisheye lens? (minus the distortion?)

There are several things I find fascinating about this question...

1-That a 16mm lense could be considered 'normal' (rectilinear).

2-There is a big difference between full framed cameras - 35mm or larger format cameras - vs. today's 1.6 factored DSLRs related to shorter focal lengths yet manufacturers ignor that.

3-How physical optics with depth-of-field(DOF) and field-of-view(FOV) plays into the mix.

'Fisheye' is a lense distortion where any lines that do not pass through the center of focus are rendered curved and, generally speaking, fisheye lenses have a field-of-view (FOV) 80 degrees or wider. Rectilinear lenses, using multiple optical components, attempt to correct for that and make all lines 'strait' within the FOV.

In the 'old' 35mm film days 20mm lenses were considered 'fisheye' since they covered an FOV > than 80 degrees. In those days linear correction was pretty much unheard of for a 20mm lens.

But with today's 1.6 factored smaller chips you have to have a lense with a focal length range 12-13mm to have the same FOV of a fisheye of the past. That confuses the issue.

What complicates the situation is that camera manufactures like to state focal lengths in 35mm equivalents as if that is an apples to apples comparison. It is not. Why? Because DOF is a lot narrower for a 12mm lens vs. a 20mm lens and therein lies the rub.

You may have the same FOV with a 12mm lens on a 1.6 factored camera chip as a 20mm on a 35mm camera but you have almost half the DOF. That can make all the difference in the world when composing a picture.

02/15/2007 12:44:41 PM · #11

I'm not sure how to interpret your tone; I'm hoping that by fascinating you mean just that, and not ignorant.

1 - Is it not, though?

2 - I think it's fairly simple to assume that I wanted a comparison between the lenses on a like platform, which happens to be a cropped sensor (thanks for the info, kirbic)

3 - I am aware of this, but I am not currently concerned about the variations in DoF.

02/15/2007 02:19:19 PM · #12

Originally posted by tryals15:

My fascination comes from the fact that I come from an era where there were few or no rectilinear lenses at 16mm focal length and I have a 28mm lens considered 'wide angle' on a 35mm format camera for FOV. On a 1.6 crop lense its FOV is barely wider than 'normal' lens from the past but still has a much wider DOF.

Manufacturers continue to erroneously express focal lengths in terms of '35mm equivalents' in many consumer cameras. They have confused the meaning of a focal length. My feeling, unrelated to your question, is that this is false advertising because it misrepresents focal lengths both in terms on FOV and in terms of the amount of detail data they are able to collect. I find that fundamentally deceptive.

I realize that in DSLRs the focal lengths of lenses are expressed as true but they still ignore the significance of both FOV and DOF reduction as if it is meaningless for a 1.6 crop factored camera. It isn't.

In all likelihood the reason the FOV and DOF issues are not bothersome to you is because you've always shot with 1.6 crop factor cameras or never really noticed the difference in DOF before. Nothing wrong with that, but some people have shot with 35mm and do notice.

Message edited by author 2007-02-15 14:19:58.

02/15/2007 02:31:22 PM · #13

I grew up shooting film through a Pentax P3, I am aware of the difference in shooting FF or a cropped sensor as it relates to both DOF and FOV; That, however, was not what I was concerned about yesterday.

Mr_Pants

SiPix Blink

02/15/2007 02:38:27 PM · #14

Originally posted by stdavidson:

I realize that in DSLRs the focal lengths of lenses are expressed as true but they still ignore the significance of both FOV and DOF reduction as if it is meaningless for a 1.6 crop factored camera. It isn't.

Wouldn't the DOF actually be increased using a cropped sensor, as subject distance would be greater?

If I've misinterpreted what you mean, then I apologise in advance.

02/15/2007 02:59:35 PM · #15

Originally posted by Mr_Pants:

Wouldn't the DOF actually be increased using a cropped sensor, as subject distance would be greater?

If I've misinterpreted what you mean, then I apologise in advance.

The subject of how DoF is affected by changing various variables is much more complex than it looks on the surface. Depending on your assumptions, you can coherently argue that one condition yields greater DoF, then turn around and just as coherently argue that there's less DoF. It all has to do with your "bounding assumptions." It's particularly thorny when we try to compare different formats. The questions that come into play are:
- How do we define DoF?
- Do we change FL to keep the same framing or change subject distance and hold FL constant?
- Are we comparing images at 100% on screen, or at specific print size? Note: this changes several assumptions
There are a few universal truths:
- AOTBE (All Other Things Being Equal), reducing focal length increases DoF
- AOTBE, increasing f number (smaller aperture size) increases DoF
- AOTBE, increasing subject distance increases DoF

For a real-life lesson in the complexity of the subject of DoF, go here and calculate the hyperfocal distance for your camera and one of your lenses at a selected aperture. Then go outside, mount that lens, set that aperture, focus on somethign at the hyperfocal distance (or set it on the distance scale if yu have one) and take a shot. Now tell me why objects at infinity are fuzzy, when the calculator says they should be sharp. Hint: the calculator isn't wrong, per se, it's operating with some assumptions that are likely different than yours.

Message edited by author 2007-02-15 15:04:21.

Mr_Pants

SiPix Blink

02/15/2007 04:08:39 PM · #16

Originally posted by kirbic:

Originally posted by Mr_Pants:
Wouldn't the DOF actually be increased using a cropped sensor, as subject distance would be greater?

If I've misinterpreted what you mean, then I apologise in advance.

The subject of how DoF is affected by changing various variables is much more complex than it looks on the surface. Depending on your assumptions, you can coherently argue that one condition yields greater DoF, then turn around and just as coherently argue that there's less DoF. It all has to do with your "bounding assumptions." It's particularly thorny when we try to compare different formats. The questions that come into play are:
- How do we define DoF?
- Do we change FL to keep the same framing or change subject distance and hold FL constant?
- Are we comparing images at 100% on screen, or at specific print size? Note: this changes several assumptions
There are a few universal truths:
- AOTBE (All Other Things Being Equal), reducing focal length increases DoF
- AOTBE, increasing f number (smaller aperture size) increases DoF
- AOTBE, increasing subject distance increases DoF

For a real-life lesson in the complexity of the subject of DoF, go here and calculate the hyperfocal distance for your camera and one of your lenses at a selected aperture. Then go outside, mount that lens, set that aperture, focus on somethign at the hyperfocal distance (or set it on the distance scale if yu have one) and take a shot. Now tell me why objects at infinity are fuzzy, when the calculator says they should be sharp. Hint: the calculator isn't wrong, per se, it's operating with some assumptions that are likely different than yours.

What I should have added to my initial post was that I was considering the case where the camera-to-subject distance remains the same. Using a cropped sensor, one would use a shorter FL, hence greater DOF

02/15/2007 04:13:42 PM · #17

Originally posted by Mr_Pants:

Originally posted by stdavidson:

I realize that in DSLRs the focal lengths of lenses are expressed as true but they still ignore the significance of both FOV and DOF reduction as if it is meaningless for a 1.6 crop factored camera. It isn't.

Wouldn't the DOF actually be increased using a cropped sensor, as subject distance would be greater?

If I've misinterpreted what you mean, then I apologise in advance.

If at a constant distance, the subject DOF would remain the same regardless of cropping factor whether captured with a so-called full framed 35mm or 1.6 cropped CCD detectors. Therefore DOF, per se, would not be increased in this case. DOF always calculates the same and is based on aperature, distance and focal length unrelated to detector size. FOV varies based on detector size.

But if what you are suggesting is that it might appear to be narrower DOF due to reduced FOV for a 1.6 crop factored detector then you could be right since much of a OOF part of the image would be excluded from the frame because of a narrow FOV for a less than 35mm detector size.

Mr_Pants

SiPix Blink

02/15/2007 04:16:34 PM · #18

Originally posted by stdavidson:

Originally posted by Mr_Pants:

Originally posted by stdavidson:

I realize that in DSLRs the focal lengths of lenses are expressed as true but they still ignore the significance of both FOV and DOF reduction as if it is meaningless for a 1.6 crop factored camera. It isn't.

Wouldn't the DOF actually be increased using a cropped sensor, as subject distance would be greater?

If I've misinterpreted what you mean, then I apologise in advance.

If at a constant distance, the subject DOF would remain the same regardless of cropping factor whether captured with a so-called full framed 35mm or 1.6 cropped CCD detectors. Therefore DOF, per se, would not be increased in this case. DOF always calculates the same and is based on aperature, distance and focal length unrelated to detector size. FOV varies based on detector size.

But if what you are suggesting is that it might appear to be narrower DOF due to reduced FOV for a 1.6 crop factored detector then you could be right since much of a OOF part of the image would be excluded from the frame because of a narrow FOV for a less than 35mm detector size.

If one is using both a 'full frame' camera and a 1.6 crop camera from the same position and taking photos of the same subject, then the focal length of the lens used on the 1.6 crop camera, to achieve the same framing, will be shorter, therefore the DOF will be increased. Witness any discussion about achieving shallow DOF with a point and shoot camera.

02/15/2007 05:08:54 PM · #19

Originally posted by Mr_Pants:

I think I see what you are saying.

You are right. To achieve the same FOV using a 160mm lens taken with a 35mm format camera you would use a shorter 100mm lens on a 1.6 cropped CCD detector. Yes, that would have a larger DOF due to the shorter 100mm focal length, given that the distance to subject and f/stop remained the same between the two images. That is because DOF is calculated from the focal length, distance and f/stop.

Manufacturers love to have us think that way. They want us to think a 100mm lens on a 1.6 crop factored DSLR is just like a 160mm on a 35mm camera because they have the same FOV. Unfortunately, the truth is that the typical 35mm with the 160mm lens will capture a lot more light and detail than than 100mm lens under the same FOV conditions. Therein lies the rub.

Message edited by author 2007-02-15 17:10:04.

Sigma 24-70mm f/2.8 EX Aspherical DF for Canon

02/15/2007 05:13:11 PM · #20

Mr_Pants is correct.
Let's do an example... we're shooting a subject at a distance of 10 feet. We have an APS-C (1.6-crop) and 35mm (1.0 crop) DSLRs. We mount a 50mm lens on the APS-c cam, and an 80mm lens on the 35mm cam. Now, since 80mm is 1.6*50mm, the FoV will be the same, so the subject framing will be identical for the same subject distance. We set both to f/8.
Now that we've specified our test set-up, we'll need to agree on how to define DoF. For the purposes of our test, we'll assume we want to view our images on screen at 100% magnification, and that we'll gauge what's within the DoF as what looks sharp at 100%. Thus, we'll need to maintain our Circle of Confusion at the sensor to equal or less than twice the pixel pitch (the Nyquist limit of the sensor). We'll set it to 0.013mm for the APS-C cam, and 0.016mm for the 35mm cam.
When we run the calculations, we find that the DoF should be 2.53 feet for the APS-C camera, and 1.82 feet for the 35mm camera.
Another way to look at it is to assume a constant print size, and this case gives different results. If we assume a 8x12 inch print, and a 0.01 inch CoC at the print, then the CoC at the sensor will be 0.03mm for the 35mm camera, and for the APS-C cam 0.019mm. For this case, the 35mm cam will have a DoF of 2.25 feet, compared to 3.77 feet for the APS-C cam.

hankk

Canon EOS-5D

02/16/2007 11:52:53 AM · #21

Originally posted by kirbic:

Then the FoV also depends on the photosite pitch, at least in this example. The 300D, 350D and 400D would each have a different DoF, because while each has an APS-C sized sensor; they each have a different number of photosites on the sensors, and so each has a diferent photosite pitch.

Message edited by author 2007-02-16 11:53:47.