DOF with FF v DX - is there a difference?

[digitalDork]

Hi,

One assumption that I've never put to the test is that the smaller the sensor, the deeper the depth of field at any given aperture and equivalent focal length.

My questions are...
1. Is this theory correct?

2. If yes, then is there a noticeable difference in DOF between a portrait shot at 70mm at f2.8 using...
A. the 28-70 on a D3, and
B. the 17-55 on a D300?

If there is a noticeable difference, I would personally rank this as a major reason for getting a FF DSLR, higher than the usual higher ISO / wider angle reasons. This is because I really value tight DOFs in portraits.

Cheers,
dD.

 

[noirdesir]

Very long article, the comparison to between different sensor formats is found under the header: 'Digital depth of field' pretty much at the end.

http://www.vanwalree.com/optics/dof.html

 

[digitalDork]

....although this article does a good job of describing the theory (therefore verifying my theory I guess), it leaves me still unclear if there's any noticeable difference in DOF between the D3/D700 and the D300 when the same photo is taken at the same (or in the case of focal length, equivalent) settings.

 

[maduhbee]

yes. The smaller the sensor, the deeper the DOF. However, if you're at 50mm (not equivalent, but actual 50mm) on both cameras, DOF is the same on both cameras at a given aperture.

I put this theory to test back when I had a Rebel XTi and a 5D. I looked at an object with my naked eye (outside the viewfinder) then I looked inside the view finder to see any changes on a 18-55 lens @ 55mm. Size and distance all remain the same. Took the 5D did the same test, and I get the exact same result, except I can see more around the object. I switched out the a 135mm lens, and the same story. The thing is, a cropped sensor is exactly that: it's cropped. Sure you can start doing "equivalence testing" but I rather stick with what's in front of me. 70mm @f/2.8 on a cropped sensor using the 70-200mm f/2.8 VR lens (for example) is gonna look the same as 70mm @f/2.8 on a full frame. Difference is that you only get a high MP crop of the frame.

Why Do I want full frame? for artistic reasons. If I only want ONE EYE to be in focus, but the rest to be OOF, or something similar in nature, than FF is for me. Obviously that's why I would alway prefer Medium format or Large Format w/ digital backs to a FF sensor.

Otherwise, why waste your time buying Nikon DSLR's when you can go Olympus/Leica/Panasonic with their 4/3's sensor. It's smaller than APS-C, so shouldn't the DOF be even better based on that theory?

 

[Thom Hogan]

....although this article does a good job of describing the theory
(therefore verifying my theory I guess), it leaves me still unclear
if there's any noticeable difference in DOF between the D3/D700 and
the D300 when the same photo is taken at the same (or in the case of
focal length, equivalent) settings.

If you are taking the equivalent photo (same angle of view from same position) and you use the commonly accepted Zeiss approach, yes, by a bit more than a stop difference. Approximately equivalent DOF from same position with same angle of view:

D300, 200mm, f/2
D700, 300mm, f/2.8

--
Thom Hogan
author, Complete Guides to Nikon bodies (18 and counting)
http://www.bythom.com

 

[Steve Ridges]

The technical answer for this is NO. The same lens with the same subject to camera distance at the same aperture will produce the same depth of field.

In practice, however, they differ because the camera to subject distance will be different to compensate for the crop factor or you'll be using a different lens to produce a similar shot.

Example:

I was shooting head shots for a dentist's employees on sight at his office. I shot the first batch with my D2x and 60mm prime. I was using about f/5.6 to get the entire head in focus including nose and ears etc.

I then got my D3 for the second batch of photos. Same conference room, subject in same spot, etc. only this time I needed to use my 70-200 @ 90-95mm to get the head the same size as the previous pictures. At f/5.6, the DOF was too shallow to get everything in focus. I had to shoot f/11 to get the same effect as before.

 

[noirdesir]

it leaves me still unclear
if there's any noticeable difference in DOF between the D3/D700 and
the D300 when the same photo is taken at the same (or in the case of
focal length, equivalent) settings.

The author concludes with this rule of thumb:

"...where R is the ratio of the format dimensions [ie, 1.5 for DX vs FX]. The smaller format employed at an F-number N yields the same DOF as the larger format at an F-number of R × N."

Example: DX vs. FX, R = 1.5
DX at f/2.8 yields the same DOF as 2.8*1.5 ~ 4 on FX

(assuming same distance to subject, same angle of view, with the focal lengths chosen to achieve that same angle of view)
eg, 70 mm f/2.8 on DX yields the same FOV and DOF as 105 mm f/4 on FX

 

[maduhbee]

the answer is, if you like conversion factors, sure, stick with APS-C

 

[digitalDork]

Great, Thom - this is the kind of information I was looking for, and what I suspected.

For me, this is quite a big deal, and as I'm on the verge of buying a D300, this gives me pause for thought...

The D700 may be more expensive than the D300 (less so in a few months time) but you could say that in terms of DOF, the D700 effectively upgrades your lenses by one stop (or from another angle, DX format downgrades your lenses).

If I'm looking to invest in fast optics specifically for shallow DOF, for example the 85 f1.4, then this is a big deal.

Any bokeh-addicts out there who have the same thoughts?

Again, thanks very for the helpful info.

Regards,
dD

....although this article does a good job of describing the theory
(therefore verifying my theory I guess), it leaves me still unclear
if there's any noticeable difference in DOF between the D3/D700 and
the D300 when the same photo is taken at the same (or in the case of
focal length, equivalent) settings.

If you are taking the equivalent photo (same angle of view from same
position) and you use the commonly accepted Zeiss approach, yes, by a
bit more than a stop difference. Approximately equivalent DOF from
same position with same angle of view:

D300, 200mm, f/2
D700, 300mm, f/2.8

--
Thom Hogan
author, Complete Guides to Nikon bodies (18 and counting)
http://www.bythom.com

 

[digitalDork]

Great info - thanks very much.
dD

it leaves me still unclear
if there's any noticeable difference in DOF between the D3/D700 and
the D300 when the same photo is taken at the same (or in the case of
focal length, equivalent) settings.

The author concludes with this rule of thumb:
"...where R is the ratio of the format dimensions [ie, 1.5 for DX vs
FX]. The smaller format employed at an F-number N yields the same DOF
as the larger format at an F-number of R × N."

Example: DX vs. FX, R = 1.5
DX at f/2.8 yields the same DOF as 2.8*1.5 ~ 4 on FX
(assuming same distance to subject, same angle of view, with the
focal lengths chosen to achieve that same angle of view)
eg, 70 mm f/2.8 on DX yields the same FOV and DOF as 105 mm f/4 on FX

 

[digitalDork]

I understand your thoughts, and agree that it's the differing shooting distances that affect the DOF, not the lens itself.

Thanks very much,
dD

yes. The smaller the sensor, the deeper the DOF. However, if you're
at 50mm (not equivalent, but actual 50mm) on both cameras, DOF is the
same on both cameras at a given aperture.

I put this theory to test back when I had a Rebel XTi and a 5D. I
looked at an object with my naked eye (outside the viewfinder) then I
looked inside the view finder to see any changes on a 18-55 lens @
55mm. Size and distance all remain the same. Took the 5D did the same
test, and I get the exact same result, except I can see more around
the object. I switched out the a 135mm lens, and the same story. The
thing is, a cropped sensor is exactly that: it's cropped. Sure you
can start doing "equivalence testing" but I rather stick with what's
in front of me. 70mm @f/2.8 on a cropped sensor using the 70-200mm
f/2.8 VR lens (for example) is gonna look the same as 70mm @f/2.8 on
a full frame. Difference is that you only get a high MP crop of the
frame.

Why Do I want full frame? for artistic reasons. If I only want ONE
EYE to be in focus, but the rest to be OOF, or something similar in
nature, than FF is for me. Obviously that's why I would alway prefer
Medium format or Large Format w/ digital backs to a FF sensor.

Otherwise, why waste your time buying Nikon DSLR's when you can go
Olympus/Leica/Panasonic with their 4/3's sensor. It's smaller than
APS-C, so shouldn't the DOF be even better based on that theory?

 

[digitalDork]

Agree - thanks very much for your reply,
dD.

The technical answer for this is NO. The same lens with the same
subject to camera distance at the same aperture will produce the same
depth of field.

In practice, however, they differ because the camera to subject
distance will be different to compensate for the crop factor or
you'll be using a different lens to produce a similar shot.

Example:
I was shooting head shots for a dentist's employees on sight at his
office. I shot the first batch with my D2x and 60mm prime. I was
using about f/5.6 to get the entire head in focus including nose and
ears etc.

I then got my D3 for the second batch of photos. Same conference
room, subject in same spot, etc. only this time I needed to use my
70-200 @ 90-95mm to get the head the same size as the previous
pictures. At f/5.6, the DOF was too shallow to get everything in
focus. I had to shoot f/11 to get the same effect as before.

 

[digitalDork]

Umm....er...okay...?!?
dD

the answer is, if you like conversion factors, sure, stick with APS-C

 

[noirdesir]

I understand your thoughts, and agree that it's the differing
shooting distances that affect the DOF, not the lens itself.

If you use the same focal length, than changing the distance has an effect on DOF, but normally you use a different focal length for the same purpose on FX compared to DX, that is really what changes your DOF.

 

[tmhw1]

The answer should be NO, for the same lens using the same focus length and aperture.

I'm not exactly an optical engineer but I'm a physicist working in a company that make microscope so I know what I'm talking about. For object off focus, the image will be projected in front of behind the CCD sensor. For a larger aperture hole, the some light will come at a larger angle, when the object is focus in front of the CCD, a larger angle means larger lateral movement hence more blur. It doesn't matter what CCD size you are using, the amount of blur in real space is the same. For the same megapixel, a DX sensor will have a smaller pixel so for the same amount of lateral movement it cover more pixels, so pixelwise, FX sensor will be more blur.

Now when you change lens or focus length then it is a whole new situation. First of all the physical size of the aperture and the position with respect to the CCD iis not the same for different lens. It depends on the lens design. As we discussed earlier the DOF is really depends on the angle of incident, so then those are not fixed you cannot do an good comparison.

However, when you move from dx sensor to a fx sensor, you have to use a longer focal length. Larger focal length tends to get the out of focus image future away from the sensor. Assuming the aperture size is the same, i.e angle the same, with the image future away, you have more lateral movement hence more blur and narrower DOF

Tim

 

[joel9]

If you are taking the equivalent photo (same angle of view from same
position) and you use the commonly accepted Zeiss approach, yes, by a
bit more than a stop difference. Approximately equivalent DOF from
same position with same angle of view:

D300, 200mm, f/2
D700, 300mm, f/2.8

It occurs to me that if this is your choice, there really will be little difference between the D300 and the D700. The extra stop on the lens to get the same DOF will negate the FX sensor's one-stop high-ISO advantage as well as the one-stop diffraction-limit advantage. In the case of these two lenses, there's not even that much difference in weight, cost, or size (although the 200 is a bit cheaper and shorter).

It would seem, then, that the practical difference between the DX choice vs. the FX choice has more to do with the limited selection of lenses (in each format) than with any inherent format advantages. Obviously I'm being a bit simplistic with the math, but am I missing anything important?

Joel

P.S. Those folks who wish Nikon would make an affordable and handholdable professional-quality 100-300mm f/4 lens just need to get the 70-200 f/2.8 for their DX bodies...

 

[digitalDork]

Hi Joel9,

I would agree with you if I were looking to increase DOF, but I'm looking to reduce it.

That is, if I'm forking out £££s for an 85mm f1.4, I want to benefit from that extremely shallow DOF to give portraits that fantastic 3D 'pop'.

But put this lens on a DX camera, and I've increased the DOF at the same aperture + viewing angle compared to the D3/D700, so I get less 'pop'...it's as if I bought an 85mm f1.8.

So, looking at the price difference between the 85's, that D700 price differential aint looking so bad now!

Cheers,
dD

If you are taking the equivalent photo (same angle of view from same
position) and you use the commonly accepted Zeiss approach, yes, by a
bit more than a stop difference. Approximately equivalent DOF from
same position with same angle of view:

D300, 200mm, f/2
D700, 300mm, f/2.8

It occurs to me that if this is your choice, there really will be
little difference between the D300 and the D700. The extra stop on
the lens to get the same DOF will negate the FX sensor's one-stop
high-ISO advantage as well as the one-stop diffraction-limit
advantage. In the case of these two lenses, there's not even that
much difference in weight, cost, or size (although the 200 is a bit
cheaper and shorter).

It would seem, then, that the practical difference between the DX
choice vs. the FX choice has more to do with the limited selection of
lenses (in each format) than with any inherent format advantages.
Obviously I'm being a bit simplistic with the math, but am I missing
anything important?

Joel

P.S. Those folks who wish Nikon would make an affordable and
handholdable professional-quality 100-300mm f/4 lens just need to get
the 70-200 f/2.8 for their DX bodies...

 

[reck]

Since you want to isolate your subject with shallow DOF, your options may improve with the FF. Assuming you want to stay in the same place relative to your subject. Your desired DOF will be available stopped down a stop and may even optimize the IQ of the lens you're using in the bargain. Of course you'll be using a different lens than on the DX.

But if you used the same lens (say, an 85 1.4) and just moved closer, there's no gain.

Here's Thom debating DX vs FX with himself: http://www.bythom.com/d3ord300.htm

You would need to analyze your lenses and your willingness to buy lenses you may not have, to figure out your particular answer.

 

[noirdesir]

D300, 200mm, f/2
D700, 300mm, f/2.8

It occurs to me that if this is your choice, there really will be
little difference between the D300 and the D700. The extra stop on
the lens to get the same DOF will negate the FX sensor's one-stop
high-ISO advantage as well as the one-stop diffraction-limit
advantage.

Yes, but there is no 135 mm f/1.4 one could put on a D300 to get the equivalent of a 200 mm f/2 on a D3. Some combinations can be achieved with both DX and FX, some only with one.

 

[digitalDork]

So, from all the excellent responses, here's what I've gathered:

Take a scenario where...

• I have one lens (let's say the 85mm f1.4) and a D300 and D3
• I want to take a portrait with the minimum DOF possible
• I want to end up with the same field of view for both images (when I use the D300, I know to take a few steps back to compensate for the crop)

Given the above scenario, the D3 image will have 1 stops worth of shallower DOF (desirable) compared to the D300.

So, if shallow DOF is your thing, the price premium of the D700 can be offset against the 1 stop DOF advantage of all your lenses!

Hi,

One assumption that I've never put to the test is that the smaller
the sensor, the deeper the depth of field at any given aperture and
equivalent focal length.

My questions are...
1. Is this theory correct?
2. If yes, then is there a noticeable difference in DOF between a
portrait shot at 70mm at f2.8 using...
A. the 28-70 on a D3, and
B. the 17-55 on a D300?

If there is a noticeable difference, I would personally rank this as
a major reason for getting a FF DSLR, higher than the usual higher
ISO / wider angle reasons. This is because I really value tight DOFs
in portraits.

Cheers,
dD.