How does sensor size produce a shallower DOF?...

[spt_gb]

You think that throwing away over half of your captured image ISN'T
loosing quality? If you only wanted an APS sized negative then you
shoot with APS film, not with 35mm and a pair of scisors.
The more you enlarge the more you show up flaws in lens and film.
This is true of any photographic process (digital or film). You
keep trying to equalise enlargement rather than print size and this
is just plain wrong for DOF comparisons.

It looks like you've been suckered into discussing a red herring. A
very red one.

I agree. He was spouting some gibberish about needing to equalise quality and I was trying to point out that you can't do this anyway (not and compare DOF at a fixed print size).

 

[Petteri Sulonen]

ioannis wrote:
[snip]

This is the best answer I ve read here:

Which is a pity, because it's wrong on a number of critical points.

Petteri
--
[ http://www.prime-junta.net/ ]
[ http://p-on-p.blogspot.com/ ]

 

[Leon Wittwer]

A 1.6 cropped camera has to magnify or enlarge 1.6 times more than a FF camera to get the same sized image like an 8x10. That is how it works.

on viewing the same sized print at the same distance. If you want
to define a new concept based in part on equivalent image quality,
go ahead but call it something else so others will not be confused.

Full frame cropped to APS is the same enlargement size. But
digital zoom is not optical zoom.

--
Leon
http://homepage.mac.com/leonwittwer/landscapes.htm

 

[Leon Wittwer]

Anyways, bottom line is that the DOF differences between the
different sensor sizes is over-rated.

if you are trying to get shallow DoF. The smaller format camera, at some point, probably can not get a lens with small enough f number to match the larger format camera. As an extreme example, the FF guy shooting with a f/1.4 lens is going to be tough to match with a cropped camera.

I agree, though, that this probably does not affect the majority of photographers.

--
Leon
http://homepage.mac.com/leonwittwer/landscapes.htm

 

[Leon Wittwer]

Depth of field was defined early in the last century and the concept has served photographers well for a long time. To those who don't understand the purpose of DoF or want to define another concept, please use another name like Depth of Forest or Depth of Stream, etc to avoid confusing the rest of us. Thank you in advance. ;-)
--
Leon
http://homepage.mac.com/leonwittwer/landscapes.htm

 

[Olivier GALLEN]

This is the best answer I ve read here:

Which is a pity, because it's wrong on a number of critical points.
Petteri

Could you please elaborate? (I agree with Donald Cooper explanation for non Macro photography).

Olivier

 

[Lee Jay]

Yes you said you must to enlarge the APS format to a greater degree
and that quality doesn't matter.

It only matters to the point that you have enough pixels to resolve the CoC. If you don't you have pixelation, not more DOF.

Then later you said quality does matter because resolution matters.
But you disregard that notion when enlarging the cropper more using
digital zoom.

The reason a smaller sensor has less DOF has nothing whatsoever to do with quality. If I shoot a 6MP image from my S3IS at 72mm and a 6MP image from a 10D at 72mm, both sharp and in-focus, the S3IS will still have greatly less DOF if they were both shot at the same f-stop and both are printed at the same final size.

Smaller sensors (or film) have less DOF at the same f-stop, focal length and final print size because they have smaller CoCs. This has nothing whatsoever to do with quality loss.

Bob Atkins on the subject:

"• Using the same lens on a EOS 10D and a 35mm film body, the 10D image has 1.6x LESS depth of field than the 35mm image would have (but they would be different images of course since the field of view would be different)" -- http://www.bobatkins.com/photography/technical/digitaldof.html

DOF master on the subject (the assumption is that the dSLR isn't full-frame):

"Using the same lens on a 35mm SLR and a DSLR, we will see less DOF in the DSLR photograph. Although the subject distance is the same for both cameras, the photographs produced from the two recorded images will be different. The DSLR photograph appears to have been taken with a longer lens." -- http://dofmaster.com/dof_dslr.html

DPreview on the subject (actually from Vincent Bockaert):

"For a smaller format (e.g. a sensor with FLM of 1.5X is a smaller format than 35mm film), the maximum permissible CoC is smaller (2), so DOF will be smaller.." -- http://www.dpreview.com/learn/?/Glossary/Optical/Focal_Length_Multiplier_01.htm

--
Lee Jay
(see profile for equipment)

 

[Petteri Sulonen]

I've gone over this so many times over the time I've been here, that I just don't have the energy to do it yet again. Read it for yourself here: [ http://photo.net/learn/optics/dofdigital/ ]

Petteri
--
[ http://www.prime-junta.net/ ]
[ http://p-on-p.blogspot.com/ ]

 

[Lee Jay]

You are very, very lost. You don't even seem to understand what the claim is. So I'll state it for you once again.

If you shoot two shots, both from the same location with the same focal length and f-stop, but you use two different formats (film or digital) and then print the entire image from each (no cropping), both at the same final size (say, 8x10), and if you have sufficient resolution for that size in both images, then the smaller format's print will have smaller DOF than the larger format's print, and the relationship is linear with format linear dimension.

If you deny that simple fact, then too bad for you because you are wrong. As I posted above, Bob Atkins, DPreview, DOF master and a host of others, not to mention the exact equations and the photographic evidence I provided, show the above statement to be true in theory and in practice.

--
Lee Jay
(see profile for equipment)

 

[mgates]

No need to back up. You can always use the same lens at the same
distance to produce an APS negative of the same quality from a full
frame negative. Whether backing up or changing focal length or
comparing at different quality levels might affect lens properties
is a different question. The question was whether film affects lens
properties.

Erm, no you can't. The same lens, at the same distance on APS and
35mm means that one of the two is incorrectly framed. If it's the
APS camera that is incorrectly framed then you've lost the edges of
the scene forever and nothing can bring them back. If it's the 35mm
camera that is incorrectly framed then all you are saying is that a
35mm frame cropped to 1.6 looks just like a 1.6 crop frame (DUH!).

That is all there is to it. Most professionals would "duh" right
back at you for thinking that digital zoom is the equivalent of
optical zoom.

Most professionals would be slapping you about the head for the
idea of framing a 35mm or larger image as if it was APS sized,
cropping off the edges off and shouting "behold, it looks the same
as APS".

Only if they think digital zoom is optical zoom. Don't know any who do.

It doesn't matter if you call it digital zoom, optical zoom,
enlargement or what have you.

I'm not advocating using digital zoom. You are without realizing what you are doing. I am advocating not moving away from the subject or changing lenses when you already have the same picture as a cropper at the same quality level inherent to the full frame picture. It is clear that few understand the inherent limits of using smaller film or a dSLR cropper.

If you have enough detail to make a
quality print at 8x10 then no arguments about how you got there are
important.

Ok but you need to realize that you induced a change in DOF as a direct result of lowering quality. You can do that with any photo and the change in DOF can be coerced in either direction.

At the same quality level their is never a change in lens DOF due to film choice. Only if you enlarge one picture more than the other. In that case the singular change in DOF is due to a corresponding singular difference in enlargement size.

If you take a 5D image and print it at 8x10 you get
about 360DPI (you need to crop it to 5:4). The same shot from a 30D
gives you about 298DPI. The human eye can only resolve about 300dpi
and this is the normal upper limit for pro grade printers, so in
both cases the final output quality is the same, but the APS image
has been enlarged more. This extra enlargement reduces the DOF from
the 30D shot. If you stick to the same lens and distance the the 5D
shot will show extra details from around the borders that the 30D
didn't capture and will look to have more DOF.

Exactly right. You changed the DOF by reducing the quality. If you compare at the same final enlargement size and without reducing the qualtiy then DOF is always the same.

The enlargement size is the same when you crop the film to APS.
THATS the point. You DONT crop 35mm to APS and use the same
enlargement factor.

The APS camera already did that for you.

It doesn't seem like you realize the key to all of this. If you put a 50mm prime on an APS film camera and snap. Then put the 50mm on a FF film camera and snap. The center of the FF image is already identical to the APS image by all measures. You don't need to move away or change lenses to equalize both the FOV and the image quality. You only need a pair of scissors.

DOF equations were designed to compare lens DOFs using full frame optics to include changes in full frame formats. They were never indended to compare scissor crops from the same print to itself and then conlclude that every lens has an infinite number of different DOFs all at the same time.

You have to understand the assumptions before using any equation.

You enlarge what you have to your standard
size. If you take your method a step further do you realy want to
shoot an 8x10 plate, crop it to APS size and say that DOF looks the
same? Would you not rather make a contact print from it
(enlargement 1:1) in which case you've got hugely more apparent DOF.

 

[Robert 55]

Could you define for us how you define quality in regards to DOF?
It seems to be something about resolution / pixels / pixel density

 

[mgates]

Yes you said you must to enlarge the APS format to a greater degree
and that quality doesn't matter.

It only matters to the point that you have enough pixels to resolve
the CoC. If you don't you have pixelation, not more DOF.

Then later you said quality does matter because resolution matters.
But you disregard that notion when enlarging the cropper more using
digital zoom.

The reason a smaller sensor has less DOF has nothing whatsoever to
do with quality. If I shoot a 6MP image from my S3IS at 72mm and a
6MP image from a 10D at 72mm, both sharp and in-focus, the S3IS
will still have greatly less DOF if they were both shot at the same
f-stop and both are printed at the same final size.

There you are comparing lens DOF using two full frame formats. That is your first correct use of the DOF equations.

If you say scissors can change the DOF in your examples above then you can easily induce a change the DOF of your S3 to be less than the 10D.

Smaller sensors (or film) have less DOF at the same f-stop, focal
length and final print size because they have smaller CoCs. This
has nothing whatsoever to do with quality loss.

Bob Atkins on the subject:

"• Using the same lens on a EOS 10D and a 35mm film body, the 10D
image has 1.6x LESS depth of field than the 35mm image would have
(but they would be different images of course since the field of
view would be different)" --
http://www.bobatkins.com/photography/technical/digitaldof.html

Yeah. That's dead wrong. It appears he knows something is wrong since he says you have "different images." If you wind up with different FOVs then clearly this is all non-applicable.

The question is do you allow the dSLR to crop a portion of the FF film to produce its FOV but not allow the FF camera to do the same. If you don't allow cropping to produce a given FOV than the dSLR image should not have been considered in the first place. It cropped already.

DOF master on the subject (the assumption is that the dSLR isn't
full-frame):

"Using the same lens on a 35mm SLR and a DSLR, we will see less DOF
in the DSLR photograph. Although the subject distance is the same
for both cameras, the photographs produced from the two recorded
images will be different. The DSLR photograph appears to have been
taken with a longer lens." -- http://dofmaster.com/dof_dslr.html

Also wrong. There are plenty of folks who do not understand the difference between cropping (digital zoom) to decrease FOV and increasing full frame lens focal length to decrease FOV. I don't argue that.

DPreview on the subject (actually from Vincent Bockaert):

"For a smaller format (e.g. a sensor with FLM of 1.5X is a smaller
format than 35mm film), the maximum permissible CoC is smaller (2),
so DOF will be smaller.." --

See note 2. It says the above taken out of context is wrong.

 

[Lee Jay]

There you are comparing lens DOF using two full frame formats.
That is your first correct use of the DOF equations.

That's the claim I've been making all along! Sheesh!

Read my first post in this thread:
http://forums.dpreview.com/forums/read.asp?forum=1032&message=20147487

And my response to you here:
http://forums.dpreview.com/forums/read.asp?forum=1032&message=20154006

Which includes the statement "for any given final presentation size". That means if you print a negative 8x10, then trim the edges of the negative with scissors and print the resultant smaller negative also at 8x10, you end up with smaller depth-of-field in the final print that came from the smaller negative.

If you say scissors can change the DOF in your examples above then
you can easily induce a change the DOF of your S3 to be less than
the 10D.

Smaller sensors (or film) have less DOF at the same f-stop, focal
length and final print size because they have smaller CoCs. This
has nothing whatsoever to do with quality loss.

Bob Atkins on the subject:

"• Using the same lens on a EOS 10D and a 35mm film body, the 10D
image has 1.6x LESS depth of field than the 35mm image would have
(but they would be different images of course since the field of
view would be different)" --
http://www.bobatkins.com/photography/technical/digitaldof.html

Yeah. That's dead wrong. It appears he knows something is wrong
since he says you have "different images." If you wind up with
different FOVs then clearly this is all non-applicable.

No one claimed that a different format with the same lens focal length has the same FOV. We just claimed that, if you accept that obvious fact, then you will have less DOF with the smaller format.

--
Lee Jay
(see profile for equipment)

 

[mgates]

You think that throwing away over half of your captured image ISN'T
loosing quality?

What do you think APS film does?

If you only wanted an APS sized negative then you
shoot with APS film, not with 35mm and a pair of scisors.
The more you enlarge the more you show up flaws in lens and film.
This is true of any photographic process (digital or film). You
keep trying to equalise enlargement rather than print size and this
is just plain wrong for DOF comparisons.

DUH again! By definition whenever you crop you lose quality.

Really? Please explain how cropping a piece of full frame film to
APS size makes it "lower quality" than APS film.

 

[mgates]

I agree. He was spouting some gibberish about needing to equalise
quality and I was trying to point out that you can't do this anyway
(not and compare DOF at a fixed print size).

Yes you can. Crop the FF image like the APS camera already did. Now you have the same print size at the same quality level. Why do you allow the dSLR to crop but not the FF camera? That makes no sense.

Justify changing both the focal length and enlargement ratio when you already have the same APS image captured within the FF image at the same quality level? That is a misuse of the DOF equaltions.

Lens DOF equations are for comparing full frame formats and prints. They were never intended to be used in such a way that each hardware combination produces an infinite number of DOF answers because scissors exist.

 

[rtendyk]

OK. here's the deal.

It doesn't. Well, actuall it does. Well, what I mean is this.

The f number is a ratio of the focal length of the lens to the aperture diameter. So, if you have a smaller sensor, then to get the same picture you use a shorter lens, that is, one with a shorter focal length.

So if you are shooting at f 5.6 with a full sensor, and at f 5.6 with a smaller sensor, the diameter of the aperture is smaller for the camera with the smaller sensor.

Now, DOF depends on the diameter of the aperture. Smaller diameter yields greater DOF. Therefore, greater DOF for the camera with the smaller sensor under similar picture taking circumstances.

You see this effect comparing a 35 mm film camera with, say, a 2 1/4 size camera at the same f number. The larger camera will yield a shallower DOF.

Look at the photos taken by Richard Avedon, for example. When he uses an 8x10 film camera on a portrait, DOF can be as shallow as an inch, maybe less.

By the way, ignore all that stuff about circle of confusion. The best definition of circle of confusion that I heard was that it was a bunch of photographers discussing DOF

 

[mgates]

You are very, very lost. You don't even seem to understand what
the claim is. So I'll state it for you once again.

If you shoot two shots, both from the same location with the same
focal length and f-stop, but you use two different formats (film or
digital) and then print the entire image from each (no cropping),
both at the same final size (say, 8x10), and if you have sufficient
resolution for that size in both images, then the smaller format's
print will have smaller DOF than the larger format's print, and
the relationship is linear with format linear dimension.

As long as both images are full frame then you have correctly applied the DOF equations. Yes there can be differences resulting from format.

If, on the onther hand, you have used scissors to produce a change in FOV, then all you need to do is use scissors on the other. Presto DOF is the same.

If you deny that simple fact, then too bad for you because you are
wrong. As I posted above, Bob Atkins, DPreview, DOF master and a
host of others, not to mention the exact equations and the
photographic evidence I provided, show the above statement to be
true in theory and in practice.

There are a lot of people who are confused by digital zoom. They think it is optical zoom but it isn't.

 

[mgates]

There you are comparing lens DOF using two full frame formats.
That is your first correct use of the DOF equations.

That's the claim I've been making all along! Sheesh!

Read my first post in this thread:
http://forums.dpreview.com/forums/read.asp?forum=1032&message=20147487

And my response to you here:
http://forums.dpreview.com/forums/read.asp?forum=1032&message=20154006

Which includes the statement "for any given final presentation
size".

Presentation size is unrelated to format.

That means if you print a negative 8x10, then trim the
edges of the negative with scissors and print the resultant smaller
negative also at 8x10, you end up with smaller depth-of-field in
the final print that came from the smaller negative.

If you say scissors can change the DOF in your examples above then
you can easily induce a change the DOF of your S3 to be less than
the 10D.

Smaller sensors (or film) have less DOF at the same f-stop, focal
length and final print size because they have smaller CoCs. This
has nothing whatsoever to do with quality loss.

Bob Atkins on the subject:

"• Using the same lens on a EOS 10D and a 35mm film body, the 10D
image has 1.6x LESS depth of field than the 35mm image would have
(but they would be different images of course since the field of
view would be different)" --
http://www.bobatkins.com/photography/technical/digitaldof.html

Yeah. That's dead wrong. It appears he knows something is wrong
since he says you have "different images." If you wind up with
different FOVs then clearly this is all non-applicable.

No one claimed that a different format with the same lens focal
length has the same FOV. We just claimed that, if you accept that
obvious fact, then you will have less DOF with the smaller format.

Both full frame, yes you do.

 

[mgates]

A 1.6 cropped camera has to magnify or enlarge 1.6 times more than
a FF camera to get the same sized image like an 8x10. That is how
it works.

As I said. Many think digital zoom is optical zoom. It is not.

All things equal, an APS camera snips quality to produce apparent focal length increase. That is why, all things equal, the price is lower. That isn't optical zoom (full frame magnification) it is digital zoom (cropping away quality).

I have to say. I am surpised that so many in the 1/5 forum don't know the difference between digital and optical zoom.

 

[mgates]

Anyways you can see the relationship with a DOF calculator.

DOF calculators are wrong in the instance of cropping a smaller
negative from a larger one. There is no need to change focal
length or subject distance to do that. You have to understand the
assumptions before blindly accepting someone else's work.

Check this out:

http://www.dofmaster.com/dofjs.html

Actual lens focal length: 50mm; selected f-stop: f/2.8; subject
distance: 10 ft

DOF for 35mm film: 2.06 ft
DOF for APS film: 1.71 ft

The only change above is in the different film size; focal length,
subject distance, and aperture are all the same. This, therefore,
is the same as cropping the film. So, please tell us how the DOF
calculator is wrong, or what assumptions we're missing.

Best regards,

That DOF calculator is fine but you are misusing it. The APS film choice is for 25mm lenses used full frame not 35mm lenses after using scissors. That is EF-S (full frame) vs EF (full frame).

Notice that, "Want to use scissors to change this answer?" is not an option.