Crop sensor effect on apertures

[Rick Knepper]

Here's why I work with FF



This could also serve as someone's reason for shooting APS-C.

All of this other stuff makes no difference to me because I will never contemplate another APS-C sensor so no need for the comparison. If I need an FOV crop or shallow DoF etc, I'll buy the optics that produce the effects I need. Buy larger format for the extra detail.

I have comtemplated moving to medium format one of these days. I'll face some of the same considerations as those who ponder FF vs APS-C but my FF will become the APS-C (so to speak) vs the MF sensor and I won't be able to use any of my 35mm lenses. In medium format, 85mm becomes a standard lens. See what I mean? Plenty of 35mm > MF conversion tables on the net. Check this chart out:

http://www.photoethnography.com/ClassicCameras/filmformats.html

Same principles apply going the other way.

Dear All,

I want you opinions in a rather technical matter that has been troubling me.

I have read in a number of threads, that in equal apertures, full frame cameras produce smaller depth of field than the cropped ones. s

According to your experience is that true? Does that mean that an f/2.8 lens on a 7D, behaves like an f/4.5 lens on a 5D?

 

[carlk]

Everything I said in this thread are correct. Only problem is these can not be understood by real stupid people. BTW I take your comment as a compliment.

I have said I will ignore those real stupid ones but I have no problem discuss the subject with people like GB or a few others even when might have a slight different point of view.

Yes I can be blunt sometimes.

Replace "blunt" with "simultaneously stupid and insulting", and you'll be closer to the mark. Cheers.

BTW you can go read up in this very thread where your inaccuracies are exposed yet again. It's past the point of comedy at this point.

I guess you couldn't figure out how to work the "ignore" feature, hmm? Come on, you can admit it.

 

[iljai]

Only one question. Difference in dof is caused by cropping or following resizing which is applied to get required same size of the images.

It's caused by the different enlargement ratios of the differently-cropped images. I guess you could say that's the resizing.

And real question is if we're trying to find what causes what and as result don't care about FOV do we care about different image size and does the same image size required?

If you're not comparing the same FOV and the same final image size, you're comparing different images. Thus, I don't think that's really valid.

I think that comparing at different fov but at the same image size I'm still comparing different images. Yes it's not practical for usual photography but there's goal for doing this. Based on the difference of this two comparisons we could establish that difference in softness is due to different enlargement not cropping.

Doesn't it make change in softness a consequence of our requirement for the same resulting image size. As well as change in FL is consequence of requirement for the same FOV and perspective which in turn causes even more dramatic change in dof that even different enlargement can't compensate?

 

[Lee Jay]

Only one question. Difference in dof is caused by cropping or following resizing which is applied to get required same size of the images.

It's caused by the different enlargement ratios of the differently-cropped images. I guess you could say that's the resizing.

And real question is if we're trying to find what causes what and as result don't care about FOV do we care about different image size and does the same image size required?

If you're not comparing the same FOV and the same final image size, you're comparing different images. Thus, I don't think that's really valid.

I think that comparing at different fov but at the same image size I'm still comparing different images. Yes it's not practical for usual photography but there's goal for doing this. Based on the difference of this two comparisons we could establish that difference in softness is due to different enlargement not cropping.

True, but we don't usually change our final print size as we crop.

Doesn't it make change in softness a consequence of our requirement for the same resulting image size. As well as change in FL is consequence of requirement for the same FOV and perspective which in turn causes even more dramatic change in dof that even different enlargement can't compensate?

Exactly. Smaller sensors have less DOF with the same lens, the same f-stop, and the same final size, but we tend to use a shorter focal length to get the same framing, and that more than compensates for the natural narrower DOF that smaller sensors have and gives you a larger DOF overall at the same FOV and f-stop. This is because zooming out changes both magnification and entrance pupil size at the same f-stop. Either one would compensate for the larger enlargement ratio such smaller sensors require, but doing both overcompensates thus yielding more overall DOF.

--
Lee Jay
(see profile for equipment)

 

[iljai]

When you crop, and display/print the same size , the apparent softness increases (so called COC - a term that I hate). This decreases the DOF.

I see two actions: one is croping, another is resizing to get the same image size. Which of those is responsible for softness change?

It's looks like a loop: method of measuring dof require to do something that changes this parameter.

The enlargement. You have to do it to keep the image size the same because DOF is measured at a reference size (assuming a reference distance, etc.)

Doesn't it make change in DOF a consequence of the requirement for the same resulting image size not cropping? As well as change in FL is consequence of requirement for the same FOV and perspective which in turn causes even more dramatic change in dof that even previous enlargement don't compensate?

To summer it up is it wrong to say that for cropped body the only difference is FOV and only when we try to compensate it not changing perspective on the same resulting image (size) we would be forced to change FL and will resive different DOF?

 

[iljai]

When you crop, and display/print the same size , the apparent softness increases (so called COC - a term that I hate). This decreases the DOF.

I see two actions: one is croping, another is resizing to get the same image size. Which of those is responsible for softness change?

Both, together. Just cropping without resizing doesn't change anything, you will just see a smaller portion of the image.

What actually changes the DOF is the enlargement . Cropping and resizing results in more enlargement, but there are other ways of increasing the enlargement (like simply increasing viewing size, zooming).

When you enlarge an image, all the optical imperfections become more visible. Everything appears less sharp, and a part of what previously appeared as acceptably sharp will now go past the threshold into softness.

That's obviouse. My point was that cropping itself doesn't change the dof it's what you do to compensate it.

 

[graphikal]

Everything I said in this thread are correct.

Oh, really? YOU MEAN LIKE THIS? LOLOL

http://forums.dpreview.com/forums/read.asp?forum=1029&message=37448215

Giving your incorrect "opinions", heavily laden with insults, merely makes you a jackass. That's not discussion; it boils down to self-abuse in the end.

 

[carlk]

DELETED. Replied the wrong post.

There is FOV difference but no DOF difference. DOF will be different ONLY if you try to make FOV the same by: 1) change distance or 2) change focal length. This has been repeated by so many people so many times I really don't know why there are still people who can't get it.

When you crop, and display/print the same size, the apparent softness increases (so called COC - a term that I hate). This decreases the DOF.

I really don't know why there are still people who can't get it.

Here is an example (taken from Bob Atkins DOF but any other DOF calculator will give you the same):

FF, FL=50mm, distance to object = 2m, f/2. DOF = 0.19m
crop, FL=50mm, distance to object = 2m, f/2. DOF = 0.12m

 

[carlk]

Well what I said there is not only correct but it’s so simple it can be easily understood by even slightly retarded person.

Everything I said in this thread are correct.

Oh, really? YOU MEAN LIKE THIS? LOLOL

http://forums.dpreview.com/forums/read.asp?forum=1029&message=37448215

Giving your incorrect "opinions", heavily laden with insults, merely makes you a jackass. That's not discussion; it boils down to self-abuse in the end.

 

[iljai]

Only one question. Difference in dof is caused by cropping or following resizing which is applied to get required same size of the images.

It's caused by the different enlargement ratios of the differently-cropped images. I guess you could say that's the resizing.

And real question is if we're trying to find what causes what and as result don't care about FOV do we care about different image size and does the same image size required?

If you're not comparing the same FOV and the same final image size, you're comparing different images. Thus, I don't think that's really valid.

I think that comparing at different fov but at the same image size I'm still comparing different images. Yes it's not practical for usual photography but there's goal for doing this. Based on the difference of this two comparisons we could establish that difference in softness is due to different enlargement not cropping.

True, but we don't usually change our final print size as we crop.

But it's not that uncommon to look at pictures taken by cell phone on it's screen while pictures taken by dslr usually printed or transfered to computer and watched on big screen. As soon as you ask what picture looks better you're doing exactly this - comparing images of different sizes taken with different systems.

Doesn't it make change in softness a consequence of our requirement for the same resulting image size. As well as change in FL is consequence of requirement for the same FOV and perspective which in turn causes even more dramatic change in dof that even different enlargement can't compensate?

Exactly. Smaller sensors have less DOF with the same lens, the same f-stop, and the same final size, but we tend to use a shorter focal length to get the same framing, and that more than compensates for the natural narrower DOF that smaller sensors have and gives you a larger DOF overall at the same FOV and f-stop. This is because zooming out changes both magnification and entrance pupil size at the same f-stop. Either one would compensate for the larger enlargement ratio such smaller sensors require, but doing both overcompensates thus yielding more overall DOF.

So, does cropped body differ only by FOV? And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

Amen

 

[mattr]

So, does cropped body differ only by FOV? And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

Well, we shouldn't decide based on "convenience"

If you want to compare images taken with cameras with differently sized sensors, you have to compare similar images. That means perspective, subject framing and DoF (and obviously display size and viewing distance) have to be the same. In order to do this you need lenses with different focal length and different f-ratios. There is simply no way around it.

 

[Great Bustard]

This is going to come as a complete shock to you, but Joseph James [that's me] is correct in nearly everything he says.

I can say the same - but I am not an optical engineer, just a boring mathematician.

As an experiment, an engineer, a physicist and a mathematician are placed in separate rooms and left with a can of food but no can-opener. A day later, the rooms are opened one by one.

In the first room, the engineer is snoring, with a battered, opened and emptied can. When asked, he explains that when he got hungry, he beat the can to its failure point.

In the second room, the physicist is seen mouthing equations, with a can popped open beside him. When asked, he explains that when he got hungry, he examined the stress points of the can, applied pressure, and 'pop'!

In the third room, the mathematician is found sweating, and mumbling to himself, 'Assume the can is open, assume the can is open...'

 

[Lee Jay]

So, does cropped body differ only by FOV?

Well, sort of, but only if you leave the final size scaled by sensor size. That would be very odd, in my opinion.

And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

Not convenient - realistic. In reality, I don't choose my final print size based on the size of the sensor used to capture the image.

--
Lee Jay
(see profile for equipment)

 

[Great Bustard]

X-vision, Great Bustard has established his credibility through many years and many incarnations in this forum.

Maybe he lacks a bit of visibility due to the reincarnation process

Ya have to admit, though, that "Great Bustard" is a classic nick! It would be a shame if that one got capped, too.

I remember myself having this sort of reaction, before getting the hand of the reincarnation, "who the heck is this new guy quoting so often Joe Mama?".

Among my favorites:

http://forums.dpreview.com/forums/read.asp?forum=1032&message=36055797

His viewpoints and writings can be thought-provoking and may require one to look at what we already know (or think we know) in a different way, but they ultimately withstand scrutiny. You obviously know a lot of stuff, but you don't understand all the implications of what you know. I recommend that you take his invitation to examine what you know in a different light by studying his link.

People usually have difficulty with paradigm shifts. In this case, changing the old paradigm of f-ratio and exposure with a new paradigm based on aperture and total light:

http://www.josephjamesphotography.com/equivalence/#exposure

 

[Great Bustard]

But it's not that uncommon to look at pictures taken by cell phone on it's screen while pictures taken by dslr usually printed or transfered to computer and watched on big screen. As soon as you ask what picture looks better you're doing exactly this - comparing images of different sizes taken with different systems.

...what conclusion would you make comparing the photos at the different sizes?

So, does cropped body differ only by FOV? And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

http://forums.dpreview.com/forums/read.asp?forum=1029&message=37453545

Carlk's take is not incorrect, it is simply a different perspective on the issue which I, personally, find less useful. That is, I think it's more useful to say that, for example, 200mm f/2 on 1.6x is equivalent to 320mm f/3.2 on FF than it is to say that 200mm f/2 on 1.6x has a framing that is 1.6x tighter than on FF.

 

[Powerdoc]

No : because the DOF is only a matter of aperture, focal and distance to the subject. From the compact camera to the big format, these rules never changes.

Yes, because at equal framing, when you want to shot a portrait at 10 feets (or whatever distance) you will have to use a 100 mm with an APS C camera, and a 160 mm with the FF.

In a physical point of vue, the 160 mm at equal distance from the subject, as smaller DOF than the 100 mm, hence, the 1,3 stop difference in term of DOF.
--
I love the crop factor at the long end, I hate it in the wide range

 

[iljai]

So, does cropped body differ only by FOV?

Well, sort of, but only if you leave the final size scaled by sensor size. That would be very odd, in my opinion.

If I magnify both images by 10 times it would be different?

If I scale them to the same size it's different story but again it's not because of bodies but of what I did to images.

And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

Not convenient - realistic. In reality, I don't choose my final print size based on the size of the sensor used to capture the image.

May be it's exactly in reverse? Knowing what size of image you'll require (for publishing or work, etc) you choose the body?

 

[Lee Jay]

If I scale them to the same size it's different story but again it's not because of bodies but of what I did to images.

But that means you have different sized final images, whether on paper or on screen.

And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

Not convenient - realistic. In reality, I don't choose my final print size based on the size of the sensor used to capture the image.

May be it's exactly in reverse? Knowing what size of image you'll require (for publishing or work, etc) you choose the body?

Usually not. For example, I shoot weddings with a 5D and a 20D. The client doesn't know which was used for any particular shot, and their selections of print sizes are not in any way based on the sizes of the sensor in the camera.

--
Lee Jay
(see profile for equipment)

 

[iljai]

So, does cropped body differ only by FOV? And doesn't the dof, effective FL and effective aperture are the consequences of the way we decided is convenient to compare them?

Well, we shouldn't decide based on "convenience"

If you want to compare images taken with cameras with differently sized sensors, you have to compare similar images . That means perspective, subject framing and DoF (and obviously display size and viewing distance) have to be the same. In order to do this you need lenses with different focal length and different f-ratios. There is simply no way around it.

Who says that I should compare similar images?

Here example: I took few shots with ipod touch 4th gen. On its screen they're acceptable. On the 21 inch monitor - awful. On the next day I took the same image with xti. And obviously I'm looking on them on the monitor. What you're saying that I don't have a right to compare image on ipod with image (from dslr) on monitor. Any reasons why I shouldn't do that?

 

[dsjtecserv]

X-vision, Great Bustard has established his credibility through many years and many incarnations in this forum.

Maybe he lacks a bit of visibility due to the reincarnation process

Ya have to admit, though, that "Great Bustard" is a classic nick! It would be a shame if that one got capped, too.

But you would also have the option of coming back as Otis ( http://en.wikipedia.org/wiki/Great_Bustard ).

Dave
--
http://www.pbase.com/dsjtecserv