[jsherman999]

There are actually ZERO changes in DOF in your examples. DOF is roughly 22in to 23in in all examples.
Perceived sharpness may change as the ratio of in focus v.s. out of focus area becomes larger.
But actual DOF changing? Nope!

You're trying to make the same argument a few angry folks made when Lee Jay originally posted that The DOF in the image is much greater as you crop. As I said earlier, and people point out to seemingly deaf ears, is that what's happening on the sensor doesn't change - the change in DOF comes when you change the display size of the crop. This is why you can't divorce display size, viewing distnace, etc from the question of DOF. This should be very clear to you, and I suspect it is - no-one is saying anything physically changes with the sensor, but the act of cropping and magnifying the crop changes the image DOF. The evidence is right there in front of you.

By the way, other important things change when you display crops at the same size as the original image, like image noise and image DR. Display size is always entirely pertinent - keeping things at the sensor level only is what's missing the point.

.

---------- Post added 03-31-14 at 11:26 AM ----------

WTF? Field of view does not affect DOF in any way what-so-ever!

You're not getting what I (and Bob Atkins) was saying - the change in FOV affects the image - it makes it a radically different image, although the DOF stays the same. But if you frame it the same on FF from the same distance, with the same FOV and F-stop (presumably to keep the same subject in the frame with the same shutter speed,) your FF image will have 1.3 stops less DOF (but will be taken with a FL 1.5x longer to retain the same FOV.)

The point I was making was that "DOF does not change if you use the same lens on both formats from the same distance", which is what your link seemed to want to prove, isn't a point that ever was in dispute.

The rest of the statements do not even make sense without qualifying which parameters are staying the same or changing.
Simply including lens and sensor/body type is insufficient for any calculation and are completely meaningless on their own.

The parameters that matter, which are not fleshed out in all the bullet points: keeping F-stop the same, and distance to subject the same.

For a practical example:

33mm f/1.8 aps-c =~ 50mm f/2.8 FF. (in terms of FOV, DOF)


.

---

Last edited by jsherman999; 03-31-2014 at 10:31 AM.

[amoringello]

Your DOF is changing because you are trying to keep the subject looking the same in Crop v.s. Full frame.
That requires either a change in focal length or distance to subject.
Of course DOF is going to change!

If you change any of; aperture, focal length or distance to subject, naturally DOF changes!
But none of that is dependent on the sensor size.

Keeping aperture, focal length and distance the same (and to some extend CoC), the DOF will not change, but your view does.

If you go so far as to include the ratio of DOF to the area in view, you are now talking about another subject altogether.

[jsherman999]

Your DOF is changing because you are trying to keep the subject looking the same in Crop v.s. Full frame.
That requires either a change in focal length or distance to subject.
Of course DOF is going to change!

More and more I feel like you're arguing in circles here, for some reason - I'm presenting something to refute what you said and you respond with "well, if you say it like that, of course you're right." I guess I'm glad you agree.

Also, I'm left wondering what you wanted to prove with that link - it doesn't refute anything anyone said. Are you maybe arguing against a straw man there, or arguing against what you think someone said? Why did you post that link?

.

If you change any of; aperture, focal length or distance to subject, naturally DOF changes!
But none of that is dependent on the sensor size.

Of course not - no-one said any differently. But you need to change focal length to keep the framing the same when you change formats. When you're shooting a wedding portrait, for example, do you get the head/shoulders of the couple in the frame on FF, switch to aps-c and then accept only one of the couple in the frame? No, you typically frame the same, and that's going to involve zooming to a different FL than what you would have shot on the other format. (or moving forward/backward, if you're willing to change perspective too.)



.

[amoringello]

More and more I feel like you're arguing in circles here, for some reason - I'm presenting something to refute what you said and you respond with "well, if you say it like that, of course you're right." I guess I'm glad you agree.

Quote possible.

Of course not - no-one said any differently. But you need to change focal length to keep the framing the same when you change formats. When you're shooting a wedding portrait, for example, do you get the head/shoulders of the couple in the frame on FF, switch to aps-c and then accept only one of the couple in the frame? No, you typically frame the same, and that's going to involve zooming to a different FL than what you would have shot on the other format. (or moving forward/backward, if you're willing to change perspective too.)

You are talking about framing, not about DOF. View and framing definitely change with sensor size. Neither have a direct effect on DOF.
You cannot just go around and change parameters defining DOF to fit your desired framing and then expect that the laws of physics bend to suit those desires.

DOF does not change with sensor size. DOF changes when framing is changed to fit the same view across different sensor sizes.
To compensate you may be able to change one of the other parameters to retain the same DOF.



To that extent, I believe you are correct and we agree.
If you read my first post, you will see that is what I stated from the beginning.

[clackers]

what's happening on the sensor doesn't change - the change in DOF comes when you change the display size of the crop.
.

I'm afraid you're conflating DoF with FoV, jsherman.

DoF has never referred to what proportion of an image is in focus.

It's meant the range of physical distance from the lens that's in focus.

[oxidized]

DOF does not change with sensor size. DOF changes when framing is changed to fit the same view across different sensor sizes.

Hence a larger senor (a) makes you reframe (b) which makes you have shallower depth of field (c). a causes b which causes c, hence a causes c....

[amoringello]

Hence a larger senor (a) makes you reframe (b) which makes you have shallower depth of field (c). a causes b which causes c, hence a causes c....

wow. just wow.
... and thus the flow of misinformation and misunderstanding propagates again.

[clackers]

Hence a larger senor (a) makes you reframe (b) which makes you have shallower depth of field (c). a causes b which causes c, hence a causes c....

You're changing the variable of distance to subject, Oxidized.

Check out any DoF calculator and you'll see that frame size definitely affects the range.

[jsherman999]

I'm afraid you're conflating DoF with FoV, jsherman.

Not at all. I'm pointing out there how cropping an image, and then enlarging the crop to the same display size as the original uncropped image also magnifies the area 'in acceptable focus' as a percentage of the image height. There's nothing new or revolutionary about what I'm saying.

DoF has never referred to what proportion of an image is in focus.

It's meant the range of physical distance from the lens that's in focus.

It's the range of measurement behind and in front of the plane of focus which looks acceptably sharp, or 'in focus'. This range itself (absolute linear measurement from plane of focus or lens) never changes because of a crop (of course - how could it?) but the act of cropping and enlarging the crop to the same display sizes changes the DOF (range of 'in focus') in the resulting image. There's lots, lots (& lots) of discussion on this, and some people have a hard time accepting it because they get hung up on thinking that someone is saying that cropping changed the physical distances in the image somehow - no-one is saying that.

Here are a few typical discussion threads on this, if you have some idle time and want to see how these arguments usually play out: link1, and link2.


.

[oxidized]

You're changing the variable of distance to subject, Oxidized.

i felt that is fairly obviously when i said "reframe" it is impossible to reframe without getting closer...

[jsherman999]

You are talking about framing, not about DOF.

No, I'm talking about how it all fits together, and in this sub-discussion I'm pointing out how your DOF changes when you keep the framing and F-stop the same, and shoot from the same position on different formats.

Example:

50mm f/2.8 FF vs. 35mm f/2.8 aps-c, from same position (forgive vertical framing gaffe)

...To that extent, I believe you are correct and we agree.
If you read my first post, you will see that is what I stated from the beginning.

Except you went down a rabbit hole along the way by maintaining that cropping/enlarging has no effect on image DOF (for a while)

EDIT: I do see in your original post in this thread how you do talk about some of these issues in the same way as me, my original response to you was in relation to something you said to Steve that was misleading.

.

---------- Post added 03-31-14 at 03:16 PM ----------

i felt that is fairly obviously when i said "reframe" it is impossible to reframe without getting closer...

Well, you could use a different FL to re-frame, which is what most folks usually do, at least most zooming type folks

[Lowell Goudge]

One thing for every one to consider with respect to depth of field

It was initially defined as "acceptably sharp" for a point source when blown up from a 35mm frame to 8 x 10 inches.

Acceptably sharp was an infinitely small point being represented ass less than 1/100 of an inch.

So let's keep all discussion related to an 8x10 print.

How the eye perceives larger prints when viewed from a proportional distance i will leav to others argued, but I suspect if the image size to viewing distance remains constant to an 8x10 print the DOF remains unchanged.

Cropping in, changes that argument, as you are enlarging more therefore reducing depth of field, That is what a crop sensor does when you enlarge the cropped portion to the same 8 x 10 print

[Boris_Akunin]

One thing for every one to consider with respect to depth of field

It was initially defined as "acceptably sharp" for a point source when blown up from a 35mm frame to 8 x 10 inches.

Acceptably sharp was an infinitely small point being represented ass less than 1/100 of an inch.

Well, it's easy enough to translate that to DOF on the sensor, we just need to adjust for the crop factor.

[Lowell Goudge]

Well, it's easy enough to translate that to DOF on the sensor, we just need to adjust for the crop factor.

It equates to 20 microns CofC for APS-C and 30 microns for full frame when using the full image, any other crop factor would be scaled accordingly.

But you still need to look at the place on a sensor where an infinitely small point becomes a circle of 20 or 30 microns

[amoringello]

Except you went down a rabbit hole along the way by maintaining that cropping/enlarging has no effect on image DOF (for a while)

Still true. Words do not change the laws of physics. (It is not just a suggestion, its the law! )
Since you're changing your distance, you are not maintaining an equivalent image circle projected from the lens.
Thus DOF cannot be maintained... . Pure mathematics.

Open your aperture 1 stop and your two images will have nearly identical background blur because you have mathematically balanced and maintained DOF.

If you don't understand what the image circle is that is being projected from the lens, you will never understand DOF.