Sign in with
Sign up | Sign in
Your question

DSLR Depth of field

Last response: in Digital Camera
Share
Anonymous
February 7, 2005 6:34:56 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

I'm considering buying a DSLR, cheaply. This would mean a not-quite
current model.

Someone mentioned that because the image sensor is smaller than 35mm
film, the depth of field will be greater than I might be used to in
similar shooting conditions. I'm curious to know if this is true, and
(briefly, if possible) why.

Does this mean I'll have trouble blurring the background when I want
to? To complicate things further, the standard lenses that seem to
come with these cameras don't seem to be very "fast", although I
suppose that's probably because they are zooms. This would make it
even harder to limit depth of field.

Informed advice please.

Greg Guarino

More about : dslr depth field

February 7, 2005 6:34:57 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"Greg G" <gdguarino@verizon.net> wrote in message
news:usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com...
> I'm considering buying a DSLR, cheaply. This would mean a not-quite
> current model.
>
> Someone mentioned that because the image sensor is smaller than 35mm
> film, the depth of field will be greater than I might be used to in
> similar shooting conditions. I'm curious to know if this is true, and
> (briefly, if possible) why.
>
> Does this mean I'll have trouble blurring the background when I want
> to? To complicate things further, the standard lenses that seem to
> come with these cameras don't seem to be very "fast", although I
> suppose that's probably because they are zooms. This would make it
> even harder to limit depth of field.
>
> Informed advice please.
>
Depth of field doesn't change. If you have a 50mm lens it will have the same
depth of field on a SLR or a dSLR. What does change is the Field of View
(FOV) so to match the FOV of my Pentax LX with a 50mm lens on my Pentax *ist
D I would need a lens around 33mm. The depth of field is deeper on a wide
angle. This is what causes confusion. You would have to calculate or use a
DOF table to adjust the aperture to match the DOF. You would have to use the
shorter lens opened up more on the dSLR to match the equivalent lens on a
35mm SLR
Anonymous
February 7, 2005 6:34:58 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Darrell commented courteously ...

> Depth of field doesn't change. What does change is the
> Field of View (FOV)

Because a digital's lens is closer to the "focal plane"
(sensors) than a 35mm film, correct?

It interests me that there are probably few people today
who know that, mathematically, there is no such thing as
"depth of field".

The lens is only in focus at one narrowly defined point.
It was the development of the concept of "circle of
confusion", where photographers could take advantage of
the fact that human eyes aren't all that good at resolving
fine detail (we tend to resolve spots out of focus as
little blurry circles), that enabled the concept of depth
of field to be developed and successfully used until this
day.

The standard for COC we use today was agreed to many, many
years ago (I'd have to dust off my old Nikon Handbook from
1969 to verify the date and the other relavent facts
here). And, neat thing that it is, if you want more DOF,
just agree to a bigger COC!

--
ATM, aka Jerry
Related resources
Anonymous
February 7, 2005 6:34:59 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

All Things Mopar <usenetMAPS123@comcast.net> wrote:

> Darrell commented courteously ...
>
> > Depth of field doesn't change. What does change is the Field of View
> > (FOV)
>
> Because a digital's lens is closer to the "focal plane" (sensors) than a
> 35mm film, correct?

Nope. The lens' focal length is the distance between the focal plane and
lens' optical center. The whole idea of interchangeable lens mounts (and
thus the SLR concept) is to guarantee the lens focal length, between
cameras and even manufacturers. So a K-mount (I shoot Pentax) is a
K-mount is a K-mount in terms of focal length, whether it's on an ME or
an *istD.

The field of view of the lens itself doesn't actually change, either.
Just the recorded portion of it. :-) The sensor is smaller, not closer
to the lens.
February 7, 2005 6:34:59 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"All Things Mopar" <usenetMAPS123@comcast.net> wrote in message
news:Xns95F65426B466ReplyToken@216.196.97.131...
> Darrell commented courteously ...
>
> > Depth of field doesn't change. What does change is the
> > Field of View (FOV)
>
> Because a digital's lens is closer to the "focal plane"
> (sensors) than a 35mm film, correct?
>
In a dSLR the sensor is the same distance as the film, otherwise the lens
would NOT focus to infinity. The image area is just cropped. The only lenses
that may be closer to the sensor are the Canon EF-S lenses. The fact all
other lenses from the makers 35mm SLR cameras function is because the snesor
and film is at the same distance.
Anonymous
February 7, 2005 10:23:39 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

If you are shooting with a 50 MM lens on a film camera and the same lens on
a digital then the DOF will be less with the digital as the lens is acting
as a 75MM on the digital.

Ken


"Darrell" <dev/null> wrote in message
news:o JidndJxRYdOcpvfRVn-pg@rogers.com...
> "Greg G" <gdguarino@verizon.net> wrote in message
> news:usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com...
>> I'm considering buying a DSLR, cheaply. This would mean a not-quite
>> current model.
>>
>> Someone mentioned that because the image sensor is smaller than 35mm
>> film, the depth of field will be greater than I might be used to in
>> similar shooting conditions. I'm curious to know if this is true, and
>> (briefly, if possible) why.
>>
>> Does this mean I'll have trouble blurring the background when I want
>> to? To complicate things further, the standard lenses that seem to
>> come with these cameras don't seem to be very "fast", although I
>> suppose that's probably because they are zooms. This would make it
>> even harder to limit depth of field.
>>
>> Informed advice please.
>>
> Depth of field doesn't change. If you have a 50mm lens it will have the
> same
> depth of field on a SLR or a dSLR. What does change is the Field of View
> (FOV) so to match the FOV of my Pentax LX with a 50mm lens on my Pentax
> *ist
> D I would need a lens around 33mm. The depth of field is deeper on a wide
> angle. This is what causes confusion. You would have to calculate or use a
> DOF table to adjust the aperture to match the DOF. You would have to use
> the
> shorter lens opened up more on the dSLR to match the equivalent lens on a
> 35mm SLR
>
>
Anonymous
February 7, 2005 11:48:27 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

No....
The lens is still a 50mm lens with a 50mm focal length and the same DoF.
You are just using the centre two-thirds of the image - i.e it is being
cropped.

Guy

IMKen wrote:
> If you are shooting with a 50 MM lens on a film camera and the same
> lens on a digital then the DOF will be less with the digital as the
> lens is acting as a 75MM on the digital.
>
> Ken
>
>
> "Darrell" <dev/null> wrote in message
> news:o JidndJxRYdOcpvfRVn-pg@rogers.com...
>> "Greg G" <gdguarino@verizon.net> wrote in message
>> news:usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com...
>>> I'm considering buying a DSLR, cheaply. This would mean a not-quite
>>> current model.
>>>
>>> Someone mentioned that because the image sensor is smaller than 35mm
>>> film, the depth of field will be greater than I might be used to in
>>> similar shooting conditions. I'm curious to know if this is true,
>>> and (briefly, if possible) why.
>>>
>>> Does this mean I'll have trouble blurring the background when I want
>>> to? To complicate things further, the standard lenses that seem to
>>> come with these cameras don't seem to be very "fast", although I
>>> suppose that's probably because they are zooms. This would make it
>>> even harder to limit depth of field.
>>>
>>> Informed advice please.
>>>
>> Depth of field doesn't change. If you have a 50mm lens it will have
>> the same
>> depth of field on a SLR or a dSLR. What does change is the Field of
>> View (FOV) so to match the FOV of my Pentax LX with a 50mm lens on
>> my Pentax *ist
>> D I would need a lens around 33mm. The depth of field is deeper on a
>> wide angle. This is what causes confusion. You would have to
>> calculate or use a DOF table to adjust the aperture to match the
>> DOF. You would have to use the
>> shorter lens opened up more on the dSLR to match the equivalent lens
>> on a 35mm SLR
Anonymous
February 7, 2005 12:23:00 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Paul Mitchum commented courteously ...

> Nope. The lens' focal length is the distance between
> the focal plane and lens' optical center. The whole
> idea of interchangeable lens mounts (and thus the SLR
> concept) is to guarantee the lens focal length,
> between cameras and even manufacturers.

Maybe I wasn't clear enough. I know that non-SLR digitals
have lenses closer to the sensors, so focal lengths have
to be shorter. Hence, most manufacturer's give you the
35mm equivalent so that old film photographers can relate.

I'd like to respectfully disagree with your statement that
focal length is the distance between the FP and the lens'
optical sensor. Just a single example may suffice. I used,
but never owned, Nikon's excellet 50-300mm zoom back in
the 1970's. It was expensive, and it was a beast. But, it
was also close to twice the length of the 110-300 off-
brand lens for my Nikon FTN.

If you meant actual focal length rather than
magnification, I'd agree with you - i.e., 300mm giving 6X
magnification compared to 50mm

--
ATM, aka Jerry
February 7, 2005 2:29:36 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

>>
> Depth of field doesn't change. If you have a 50mm lens it will have the
> same
> depth of field on a SLR or a dSLR. What does change is the Field of View
> (FOV) so to match the FOV of my Pentax LX with a 50mm lens on my Pentax
> *ist
> D I would need a lens around 33mm. The depth of field is deeper on a wide
> angle. This is what causes confusion. You would have to calculate or use a
> DOF table to adjust the aperture to match the DOF. You would have to use
> the
> shorter lens opened up more on the dSLR to match the equivalent lens on a
> 35mm SLR
>
>

I have a problem here, I agree with your reasoning that changing the sensor
size shouldn't change the depth of field just the FOV , all we are doing is
cropping the area, but if I use the online DOF calculatoar at
http://www.dofmaster.com/dofjs.html it doesn't seem to agree.

ie enter 50mm 'actual' lens at f2.8, distance 30ft and when using a 35mm
format the DOF is 20ft and using a 10D it's 12.2 ft.

Do I misunderstand, is the program wrong or some thing else ??:o )

Thanks
Anonymous
February 7, 2005 2:29:37 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

dylan <no@nospam.com> wrote:

> > Depth of field doesn't change. If you have a 50mm lens it will have the
> > same depth of field on a SLR or a dSLR. What does change is the Field of
> > View (FOV) so to match the FOV of my Pentax LX with a 50mm lens on my
> > Pentax *ist D I would need a lens around 33mm. The depth of field is
> > deeper on a wide angle. This is what causes confusion. You would have to
> > calculate or use a DOF table to adjust the aperture to match the DOF.
> > You would have to use the shorter lens opened up more on the dSLR to
> > match the equivalent lens on a 35mm SLR
>
> I have a problem here, I agree with your reasoning that changing the
> sensor size shouldn't change the depth of field just the FOV , all we are
> doing is cropping the area, but if I use the online DOF calculatoar at
> http://www.dofmaster.com/dofjs.html it doesn't seem to agree.
>
> ie enter 50mm 'actual' lens at f2.8, distance 30ft and when using a 35mm
> format the DOF is 20ft and using a 10D it's 12.2 ft.
>
> Do I misunderstand, is the program wrong or some thing else ??:o )

The depth of field of a lens setting is determined by the circle of
confusion, not the size of the film. If the COC of both 35mm film and
the digital sensor were the same, then the DOF would be the same with
the same lens. But they're not. :-)
February 7, 2005 2:29:38 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"Paul Mitchum" <usenet@mile23.c0m> wrote in message
news:1grlhat.1a4ipn6tw8s35N%usenet@mile23.c0m...
> dylan <no@nospam.com> wrote:
>
>
> The depth of field of a lens setting is determined by the circle of
> confusion, not the size of the film. If the COC of both 35mm film and
> the digital sensor were the same, then the DOF would be the same with
> the same lens. But they're not. :-)

What is the CoC for an APS-C/DX sized sensor? I jhave figures for most film
formats, but I can't recall seeing the dSLR numbers.
Anonymous
February 7, 2005 3:57:12 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

All Things Mopar wrote:

> Darrell commented courteously ...
>
>
>>Depth of field doesn't change. What does change is the
>>Field of View (FOV)
>
>
> Because a digital's lens is closer to the "focal plane"
> (sensors) than a 35mm film, correct?
>
> It interests me that there are probably few people today
> who know that, mathematically, there is no such thing as
> "depth of field".

Most people don't even get the notion of depth-of-field, never mind graduate to
DOF being subjective viewing of the final print or projection.

> The lens is only in focus at one narrowly defined point.

Plane (which I'm sure you meant). Although for most lenses it will get softer
at the edges/corners but this is due to lens underperformance, not absense of a
focus plane.

<snipped>

> The standard for COC we use today was agreed to many, many
> years ago (I'd have to dust off my old Nikon Handbook from
> 1969 to verify the date and the other relavent facts
> here). And, neat thing that it is, if you want more DOF,
> just agree to a bigger COC!

Yes, however a larger CoC implies a further viewing distance.

Here's some empirical stuff to get you reaquainted with the numbers in use.

http://www.nikonlinks.com/unklbil/dof.htm


--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 7, 2005 4:04:02 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

IMKen wrote:

> If you are shooting with a 50 MM lens on a film camera and the same lens on
> a digital then the DOF will be less with the digital as the lens is acting
> as a 75MM on the digital.

The final constraint being: when you print it (view it) at a given size.


--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 7, 2005 4:38:26 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Mon, 7 Feb 2005 08:48:27 -0000, "Bigguy" <Bigguy@nowhere.com>
wrote:

>No....
>The lens is still a 50mm lens with a 50mm focal length and the same DoF.
>You are just using the centre two-thirds of the image - i.e it is being
>cropped.
>
>Guy
>

DOF DOES change between DLSRs and SLRS because the circle of confusion
is different.

DOF IS NOT AOV

I'll post my discussion on this ONCE AGAIN....

I use a lot of Acronyms here, so briefly:

CoC - Circle of Confusion
DoF - Depth of Field
NL - Near Limit (Closest subject still in focus)
FL - Far Limit (Furthest subject still in focus)
HF - Hyperfocal Distance. (Optimum focus point to maximize DoF depth)
AoV - Angle of View

CoC on a DSLR with a smaller-than-35mm sensor is different to regular
35mm. For a typical 1.5 to 1.6 sensor factor your DSLR DOF at f8 will
appear similar to an f11 shot on a 35mm camera if you adjust lenses to
give similar Angle of view..

i.e. DSLR's have a deeper DOF than 35mm SLRs when AoV is matched
between the cameras.

Comparing an SLR to an DSLR with a 1.5 sensor ratio and fictitious
lenses to give identical AoVs using a f5.6 aperture lens focused at
8m:

A SLR with a 108mm NL=7.233m FL=8.949m (Depth: 1.716m)
A DSLR with a 72mm NL=6.830m FL=9.653m = (Depth: 2.823m)

Even if you stick with the same lens on both cameras,
say a 50mm f4 focused at 6m, the DoF's differ significantly:

SLR 50mm f4. HF=22.3m NL=4.73m FL=8.18 (Depth: 3.45m)
DSLR 50mm f4. HF=31.3m NL=5.04m FL=7.41 (Depth: 2.37m)

You can use any DOF calculator that allows you to enter a custom CoC
to represent your sensor size to learn the differences.

Here's how you can look-up your camera's Coc:
http://dfleming.ameranet.com/digital_coc.html

And here is a DOF page + calculator:
http://www.conent.com/ConAdv/Encyclopaedia/Photography/...

--
Owamanga!
Anonymous
February 7, 2005 4:38:27 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Owamanga wrote:

> On Mon, 7 Feb 2005 08:48:27 -0000, "Bigguy" <Bigguy@nowhere.com>
> wrote:
>
>
>>No....
>>The lens is still a 50mm lens with a 50mm focal length and the same DoF.
>>You are just using the centre two-thirds of the image - i.e it is being
>>cropped.
>>
>>Guy
>>
>
>
> DOF DOES change between DLSRs and SLRS because the circle of confusion
> is different.

Yep.


--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 7, 2005 4:47:09 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

In article <1grlhat.1a4ipn6tw8s35N%usenet@mile23.c0m>,
usenet@mile23.c0m (Paul Mitchum) wrote:

> The depth of field of a lens setting is determined by the circle of
> confusion, not the size of the film. If the COC of both 35mm film and
> the digital sensor were the same, then the DOF would be the same with
> the same lens. But they're not. :-)

No.. in fact a smaller chip has LESS DoF because the CoC is smaller.
But, because you need a shorter focal length to get the same image
projected onto the smaller chip, DoF in the final result will be larger.
(the FL factor is more important)

Lourens
Anonymous
February 7, 2005 5:54:51 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Sun, 6 Feb 2005 08:49:44 -0500, "Darrell" <dev/null> wrote:

>
>"Paul Mitchum" <usenet@mile23.c0m> wrote in message
>news:1grlhat.1a4ipn6tw8s35N%usenet@mile23.c0m...
>> dylan <no@nospam.com> wrote:
>>
>>
>> The depth of field of a lens setting is determined by the circle of
>> confusion, not the size of the film. If the COC of both 35mm film and
>> the digital sensor were the same, then the DOF would be the same with
>> the same lens. But they're not. :-)
>
>What is the CoC for an APS-C/DX sized sensor? I jhave figures for most film
>formats, but I can't recall seeing the dSLR numbers.

Here are some popular ones (not all APS-C/DX):

Nikon D1H, D1X, D2H, D70, D100 all have a CoC of 0.020
Canon 1D - 0.023
Canon 1DS - 0.030
Canon D30 - 0.019
Canon Digital Rebel - 0.019
Sigma SD-10, SD-9 are 0.018
Olympus E-1 - 0.015

--
Owamanga!
Anonymous
February 7, 2005 6:13:58 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Lourens Smak <smak@wanadoo.nl> wrote:

> In article <1grlhat.1a4ipn6tw8s35N%usenet@mile23.c0m>,
> usenet@mile23.c0m (Paul Mitchum) wrote:
>
> > The depth of field of a lens setting is determined by the circle of
> > confusion, not the size of the film. If the COC of both 35mm film and
> > the digital sensor were the same, then the DOF would be the same with
> > the same lens. But they're not. :-)
>
> No.. in fact a smaller chip has LESS DoF because the CoC is smaller. But,
> because you need a shorter focal length to get the same image projected
> onto the smaller chip, DoF in the final result will be larger. (the FL
> factor is more important)

You're right. I'm right. We're both right. :-)

My point was that lenses don't have circles of confusion, and COC are
the fundamental currency of DOF.
Anonymous
February 7, 2005 7:49:05 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Alan Browne commented courteously ...

[snip]
> Plane (which I'm sure you meant).
[snip]

Good Grief! What did I say "plain"? Yes, of course it is
"plane"...

Thanks for the link on DoF

--
ATM, aka Jerry
Anonymous
February 7, 2005 9:26:00 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

In article <usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com>,
Greg G <gdguarino@verizon.net> wrote:
>I'm considering buying a DSLR, cheaply. This would mean a not-quite
>current model.
>
>Someone mentioned that because the image sensor is smaller than 35mm
>film, the depth of field will be greater than I might be used to in
>similar shooting conditions. I'm curious to know if this is true, and
>(briefly, if possible) why.
>
>Does this mean I'll have trouble blurring the background when I want
>to? To complicate things further, the standard lenses that seem to
>come with these cameras don't seem to be very "fast", although I
>suppose that's probably because they are zooms. This would make it
>even harder to limit depth of field.
>
>Informed advice please.


Having mis-calculated this once, and been called on it, I think I
might be able to offer a helpful contribution.

Don't confuse yourself with all these CoC & AOV numbers, etc.

Your eventual aim, presumably, is to end up with a given final image
size (either image dimension in pixels, or print size in inches,
and with your own chosen composition of the image. If that is the case,
then the *only* variable that affects your DOF is the actual physical
diameter of the aperture on your lens. That's it. All the other things
(the film/sensor size, the angle of view of the full image, the amount
of enlargement & cropping you have to do for composition) cancel out.

An example; If you take a full-frame image on a 35mm camera
using a 300/f2.8 lens, and a full-frame image of the same scene
using your 1.5 crop-factor DSLR and a 200/f2.8 lens, and print
them both up to the same final size, you'll have more DOF on the
print from the DSLR; it will match the DOF from a 300/f4 lens on
the 35mm body (all lenses assumed to be used at maximum aperture).

Not only that; it will roughly match the DOF from, and angle-of
view of, a 600/f8 on a medium-format body, or even a 1000/f14
lens on a 4x5 camera (allowing for the different aspect ratio).
February 7, 2005 9:54:17 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Mon, 07 Feb 2005 03:34:56 GMT, Greg G <gdguarino@verizon.net> wrote:

>I'm considering buying a DSLR, cheaply. This would mean a not-quite
>current model.
>
>Someone mentioned that because the image sensor is smaller than 35mm
>film, the depth of field will be greater than I might be used to in
>similar shooting conditions. I'm curious to know if this is true, and
>(briefly, if possible) why.
>
>Does this mean I'll have trouble blurring the background when I want
>to? To complicate things further, the standard lenses that seem to
>come with these cameras don't seem to be very "fast", although I
>suppose that's probably because they are zooms. This would make it
>even harder to limit depth of field.
>
>Informed advice please.

Let me tell you - you will have NO TROUBLE blurring the backgrounds!

I took some portraits of people, I was about 4 feet from them, they were 8 feet
from a wall - the wall was totally blurred out at F5.

With closeups of flowers, I've had only one flower in focus, all the background
blurred, at F11.

I usually want MORE DOF!!

>Greg Guarino
Anonymous
February 8, 2005 2:16:32 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Mon, 7 Feb 2005 18:26:00 +0000 (UTC), johnf@panix.com (John
Francis) wrote:

>In article <usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com>,
>Greg G <gdguarino@verizon.net> wrote:
>>I'm considering buying a DSLR, cheaply. This would mean a not-quite
>>current model.
>>
>>Someone mentioned that because the image sensor is smaller than 35mm
>>film, the depth of field will be greater than I might be used to in
>>similar shooting conditions. I'm curious to know if this is true, and
>>(briefly, if possible) why.
>>
>>Does this mean I'll have trouble blurring the background when I want
>>to? To complicate things further, the standard lenses that seem to
>>come with these cameras don't seem to be very "fast", although I
>>suppose that's probably because they are zooms. This would make it
>>even harder to limit depth of field.
>>
>>Informed advice please.
>
>
>Having mis-calculated this once, and been called on it, I think I
>might be able to offer a helpful contribution.
>
>Don't confuse yourself with all these CoC & AOV numbers, etc.
>
>Your eventual aim, presumably, is to end up with a given final image
>size (either image dimension in pixels, or print size in inches,
>and with your own chosen composition of the image. If that is the case,
>then the *only* variable that affects your DOF is the actual physical
>diameter of the aperture on your lens. That's it. All the other things
>(the film/sensor size, the angle of view of the full image, the amount
>of enlargement & cropping you have to do for composition) cancel out.
>
>An example; If you take a full-frame image on a 35mm camera
>using a 300/f2.8 lens, and a full-frame image of the same scene
>using your 1.5 crop-factor DSLR and a 200/f2.8 lens, and print
>them both up to the same final size, you'll have more DOF on the
>print from the DSLR; it will match the DOF from a 300/f4 lens on
>the 35mm body (all lenses assumed to be used at maximum aperture).
>
>Not only that; it will roughly match the DOF from, and angle-of
>view of, a 600/f8 on a medium-format body, or even a 1000/f14
>lens on a 4x5 camera (allowing for the different aspect ratio).

I asked the original question, and I didn't know it would get such a
variety of answers. But I did a "thought experiment" today and I think
I know the answer now.

I agree that if all you change is the "recording device" from film to
a digital sensor, the circle of confusion (terrible name in my
opinion) projected on it will be exactly the same size. However, since
the digital sensor is smaller, we will need to enlarge the picture
recorded on it more than we would the 35mm film.

This leads me to conclude that the DOF for a given aperture should be
what I would expect from a lens with the same field of view, i.e. a
longer lens on a 35mm. I think the poster above probably just goofed,
but this would be LESS DOF on the digital camera for a given focal
length as the field of view is smaller.

Similarly, if we did crop down the 35mm image to show the same field
of view as the digital sensor, we would also increase the visibility
of the CoC by having enlarged the image more, yielding the same
apparent DOF.

Greg Guarino
Anonymous
February 8, 2005 2:16:33 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

I don't know why everyone had to turn this into a science project. If you
use a 100mm lens on a 35mm camera, it will produce a certain depth of field
based on the aperture it's set at when you take the photo. Wide angle
lenses tend to produce more depth of field, while telephotos produce less
depth of field. Also, the further away the image is from the lens, the more
depth of field you will get.

Take that same lens and put it on a DSLR camera and you might be working
with a magnification ratio of 1.5. This is because the image sensor is
smaller than a 35mm frame, and the lens acts exactly the same, except that
it "appears" to be a 150mm lens because the image sensor is cropping the
image you would get on a 35mm camera. Otherwise, all things are equal.

Now, to get the same image size captured in the viewfinder you will have to
move back, placing your image further from the lens, thereby creating a bit
more depth of field. So, you have to think about three things when you are
determining depth of field. Wide angle lenses = more in focus at any given
aperture. Telephoto lenses = less. A larger aperture will give you less
depth of field, and the further away your subject gets the more depth of
field you will have.

It's exactly the same as using a 35mm camera, but you now have a cropped
image in a DSLR (same lens), and you often have to move the image back to
frame the same scene. Otherwise, the characteristics of any given focal
length lens is the same.

The best way to find all this out is to experiment, as this works better
than using a depth of field preview, since it just darkens the image. Hey,
on a DSLR processing costs you nothing. Experiment all you want.


"Greg G" <gdguarino@verizon.net> wrote in message
news:stsf01dfiqjes2g7t72kfeep60sngsbi05@4ax.com...
> On Mon, 7 Feb 2005 18:26:00 +0000 (UTC), johnf@panix.com (John
> Francis) wrote:
>
>>In article <usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com>,
>>Greg G <gdguarino@verizon.net> wrote:
>>>I'm considering buying a DSLR, cheaply. This would mean a not-quite
>>>current model.
>>>
>>>Someone mentioned that because the image sensor is smaller than 35mm
>>>film, the depth of field will be greater than I might be used to in
>>>similar shooting conditions. I'm curious to know if this is true, and
>>>(briefly, if possible) why.
>>>
>>>Does this mean I'll have trouble blurring the background when I want
>>>to? To complicate things further, the standard lenses that seem to
>>>come with these cameras don't seem to be very "fast", although I
>>>suppose that's probably because they are zooms. This would make it
>>>even harder to limit depth of field.
>>>
>>>Informed advice please.
>>
>>
>>Having mis-calculated this once, and been called on it, I think I
>>might be able to offer a helpful contribution.
>>
>>Don't confuse yourself with all these CoC & AOV numbers, etc.
>>
>>Your eventual aim, presumably, is to end up with a given final image
>>size (either image dimension in pixels, or print size in inches,
>>and with your own chosen composition of the image. If that is the case,
>>then the *only* variable that affects your DOF is the actual physical
>>diameter of the aperture on your lens. That's it. All the other things
>>(the film/sensor size, the angle of view of the full image, the amount
>>of enlargement & cropping you have to do for composition) cancel out.
>>
>>An example; If you take a full-frame image on a 35mm camera
>>using a 300/f2.8 lens, and a full-frame image of the same scene
>>using your 1.5 crop-factor DSLR and a 200/f2.8 lens, and print
>>them both up to the same final size, you'll have more DOF on the
>>print from the DSLR; it will match the DOF from a 300/f4 lens on
>>the 35mm body (all lenses assumed to be used at maximum aperture).
>>
>>Not only that; it will roughly match the DOF from, and angle-of
>>view of, a 600/f8 on a medium-format body, or even a 1000/f14
>>lens on a 4x5 camera (allowing for the different aspect ratio).
>
> I asked the original question, and I didn't know it would get such a
> variety of answers. But I did a "thought experiment" today and I think
> I know the answer now.
>
> I agree that if all you change is the "recording device" from film to
> a digital sensor, the circle of confusion (terrible name in my
> opinion) projected on it will be exactly the same size. However, since
> the digital sensor is smaller, we will need to enlarge the picture
> recorded on it more than we would the 35mm film.
>
> This leads me to conclude that the DOF for a given aperture should be
> what I would expect from a lens with the same field of view, i.e. a
> longer lens on a 35mm. I think the poster above probably just goofed,
> but this would be LESS DOF on the digital camera for a given focal
> length as the field of view is smaller.
>
> Similarly, if we did crop down the 35mm image to show the same field
> of view as the digital sensor, we would also increase the visibility
> of the CoC by having enlarged the image more, yielding the same
> apparent DOF.
>
> Greg Guarino
Anonymous
February 8, 2005 5:19:26 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

In article <stsf01dfiqjes2g7t72kfeep60sngsbi05@4ax.com>,
Greg G <gdguarino@verizon.net> wrote:
>On Mon, 7 Feb 2005 18:26:00 +0000 (UTC), johnf@panix.com (John
>Francis) wrote:
>
>>In article <usnd01tg9r4m1hh2p5518mah1vslhvl71k@4ax.com>,
>>Greg G <gdguarino@verizon.net> wrote:
>>>I'm considering buying a DSLR, cheaply. This would mean a not-quite
>>>current model.
>>>
>>>Someone mentioned that because the image sensor is smaller than 35mm
>>>film, the depth of field will be greater than I might be used to in
>>>similar shooting conditions. I'm curious to know if this is true, and
>>>(briefly, if possible) why.
>>>
>>>Does this mean I'll have trouble blurring the background when I want
>>>to? To complicate things further, the standard lenses that seem to
>>>come with these cameras don't seem to be very "fast", although I
>>>suppose that's probably because they are zooms. This would make it
>>>even harder to limit depth of field.
>>>
>>>Informed advice please.
>>
>>
>>Having mis-calculated this once, and been called on it, I think I
>>might be able to offer a helpful contribution.
>>
>>Don't confuse yourself with all these CoC & AOV numbers, etc.
>>
>>Your eventual aim, presumably, is to end up with a given final image
>>size (either image dimension in pixels, or print size in inches,
>>and with your own chosen composition of the image. If that is the case,
>>then the *only* variable that affects your DOF is the actual physical
>>diameter of the aperture on your lens. That's it. All the other things
>>(the film/sensor size, the angle of view of the full image, the amount
>>of enlargement & cropping you have to do for composition) cancel out.
>>
>>An example; If you take a full-frame image on a 35mm camera
>>using a 300/f2.8 lens, and a full-frame image of the same scene
>>using your 1.5 crop-factor DSLR and a 200/f2.8 lens, and print
>>them both up to the same final size, you'll have more DOF on the
>>print from the DSLR; it will match the DOF from a 300/f4 lens on
>>the 35mm body (all lenses assumed to be used at maximum aperture).
>>
>>Not only that; it will roughly match the DOF from, and angle-of
>>view of, a 600/f8 on a medium-format body, or even a 1000/f14
>>lens on a 4x5 camera (allowing for the different aspect ratio).
>
>I asked the original question, and I didn't know it would get such a
>variety of answers. But I did a "thought experiment" today and I think
>I know the answer now.
>
>I agree that if all you change is the "recording device" from film to
>a digital sensor, the circle of confusion (terrible name in my
>opinion) projected on it will be exactly the same size. However, since
>the digital sensor is smaller, we will need to enlarge the picture
>recorded on it more than we would the 35mm film.
>
>This leads me to conclude that the DOF for a given aperture should be
>what I would expect from a lens with the same field of view, i.e. a
>longer lens on a 35mm. I think the poster above probably just goofed,
>but this would be LESS DOF on the digital camera for a given focal
>length as the field of view is smaller.
>
>Similarly, if we did crop down the 35mm image to show the same field
>of view as the digital sensor, we would also increase the visibility
>of the CoC by having enlarged the image more, yielding the same
>apparent DOF.


Well, at least you got one thing right. The DOF you get from enlarging
the central portion of a 35mm frame is, indeed, exactly the same DOF
that you would get from using the same lens on a 1.5 crop-factor DSLR
(and creating prints and/or digital images of indentical dimensions);
DOF doesn't depend on the technology.

Where you are getting wrong in your "thought experiment" is assuming
that if you have frame-filling images (such as with a 300mm on a 35mm,
and a 200mm on a 1.5 crop-factor DSLR) you get the same DOF when both
lenses are set to the same relative aperture (such as, say, f2.8).

That's not the case. In fact you don't even need a DSLR to show this.
If you mount a 200mm lens on your film camera, set it to f2.8, and
make a print from the central portion of the resulting exposure, the
resulting image will match the angle of view of a full-frame print
made with a 300mm lens. But if you make the prints the same size,
the DOF in your first print won't match the DOF of a print made with
the 300mm lens set to f2.8, even though you are magnifying that first
exposure more during printing; you need to stop the 300mm lens down
to f4 (increasing the depth of field) before the DOF will match.

As I said, what you need to keep the same is the *absolute* aperture,
not the relative aperture. Forget the "thought experiments"; get
outside and perform the experiments for real - it's easy enough. You
should soon convince yourself I didn't "just goof" - I got it right.
Anonymous
February 8, 2005 5:24:35 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

In article <iPSdnfdtmsnCipXfRVn-pQ@comcast.com>,
Sheldon <sheldon@XXXXXXXXsopris.net> wrote:
>I don't know why everyone had to turn this into a science project. If you
>use a 100mm lens on a 35mm camera, it will produce a certain depth of field
>based on the aperture it's set at when you take the photo. Wide angle
>lenses tend to produce more depth of field, while telephotos produce less
>depth of field. Also, the further away the image is from the lens, the more
>depth of field you will get.
>
>Take that same lens and put it on a DSLR camera and you might be working
>with a magnification ratio of 1.5. This is because the image sensor is
>smaller than a 35mm frame, and the lens acts exactly the same, except that
>it "appears" to be a 150mm lens because the image sensor is cropping the
>image you would get on a 35mm camera. Otherwise, all things are equal.
>
>Now, to get the same image size captured in the viewfinder you will have to
>move back, placing your image further from the lens, thereby creating a bit
>more depth of field.

No, no, no. Don't introduce yet another complication. All images should
be taken from the same spot, and cropped and enlarged appropriately to keep
the same final angle of view in the resulting print.

Changing the position from which the shot is taken alters the perspective.
Anonymous
February 8, 2005 1:42:09 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

All Things Mopar wrote:

> Alan Browne commented courteously ...
>
> [snip]
>
>>Plane (which I'm sure you meant).
>
> [snip]
>
> Good Grief! What did I say "plain"? Yes, of course it is
> "plane"...

Worse than that, you said "point".

Cheers,
Alan.

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 8, 2005 1:42:10 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Alan Browne commented courteously ...

> Worse than that, you said "point".

Guess I had a bad hair day. I vaguely remember talking
about "focus point" but thought I used "focal plane" in
the right way. I may have mixed up my terms. Better start
proofreading, Huh?

Thanks, and you have a great day, hear?!

--
ATM, aka Jerry
Anonymous
February 8, 2005 1:50:10 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Sheldon wrote:

> I don't know why everyone had to turn this into a science project. If you

At some point words fail to find the truth. Only by actually running the
numbers and testing it can comparisons be made truthfully.


--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 8, 2005 3:33:07 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Mon, 07 Feb 2005 12:57:12 -0500, Alan Browne
<alan.browne@freelunchVideotron.ca> wrote:

>> The lens is only in focus at one narrowly defined point.
>
>Plane (which I'm sure you meant). Although for most lenses it will get softer
>at the edges/corners but this is due to lens underperformance, not absense of a
>focus plane.
>

Genuine question: is this "softening" because the "focal plane" should
strictly be the inside of a (shallow) "focus sphere"?

Regards,
Graham Holden (g-holden AT dircon DOT co DOT uk)
--
There are 10 types of people in the world;
those that understand binary and those that don't.
Anonymous
February 8, 2005 3:33:08 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Graham Holden wrote:

> On Mon, 07 Feb 2005 12:57:12 -0500, Alan Browne
> <alan.browne@freelunchVideotron.ca> wrote:
>
>
>>>The lens is only in focus at one narrowly defined point.
>>
>>Plane (which I'm sure you meant). Although for most lenses it will get softer
>>at the edges/corners but this is due to lens underperformance, not absense of a
>>focus plane.
>>
>
>
> Genuine question: is this "softening" because the "focal plane" should
> strictly be the inside of a (shallow) "focus sphere"?

Focus sphere? Is that above the tropopause?

A plane is a plane. Flat. Parallel with the film surface.

A nice well behaved lens should focus that plane evenly across the film plane.
But, esp at larger apertures, the edges/corners are softer than the middle which
might be confused as out of focus.

I hope that answers it ('cause I don't really understand your question).

Cheers,
Alan

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 8, 2005 5:08:53 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"Sheldon" <sheldon@XXXXXXXXsopris.net> wrote in message
news:iPSdnfdtmsnCipXfRVn-pQ@comcast.com...
>I don't know why everyone had to turn this into a science project.
>If you use a 100mm lens on a 35mm camera, it will produce a certain
>depth of field based on the aperture it's set at when you take the
>photo. Wide angle lenses tend to produce more depth of field, while
>telephotos produce less depth of field.

Unless you adjust the aperture to the same physical size. The F-number
is aperture, relative to focal length. However if both lenses have the
same physical aperture size, the light cone that focusses on the
sensor elements is approx. equal and thus DoF is equal.

Since smaller sensors commonly use shorter focal lengths to keep a
similar field of view, the physical aperture size is smaller (at the
same F-number) and thus DoF is larger.

Bart
Anonymous
February 8, 2005 5:43:29 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

All Things Mopar wrote:

> Alan Browne commented courteously ...
>
>
>>Worse than that, you said "point".
>
>
> Guess I had a bad hair day. I vaguely remember talking
> about "focus point" but thought I used "focal plane" in
> the right way. I may have mixed up my terms. Better start
> proofreading, Huh?

Nah. We need some entertainment.

>
> Thanks, and you have a great day, hear?!

Yo!


--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 8, 2005 7:32:44 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Greg G <gdguarino@verizon.net> wrote:

> I agree that if all you change is the "recording device" from film to
> a digital sensor, the circle of confusion (terrible name in my
> opinion) projected on it will be exactly the same size.

No, it won't: CoC is usually defined as something close to the length
of the diagonal divided by 1750.

Andrew.
Anonymous
February 8, 2005 8:24:08 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Tue, 08 Feb 2005 10:48:12 -0500, Alan Browne
<alan.browne@freelunchVideotron.ca> wrote:

>Graham Holden wrote:
>
>> Genuine question: is this "softening" because the "focal plane" should
>> strictly be the inside of a (shallow) "focus sphere"?
>
>Focus sphere? Is that above the tropopause?
>
>A plane is a plane. Flat. Parallel with the film surface.
>
>A nice well behaved lens should focus that plane evenly across the film plane.
>But, esp at larger apertures, the edges/corners are softer than the middle which
>might be confused as out of focus.
>
>I hope that answers it ('cause I don't really understand your question).
>
>Cheers,
>Alan

What I meant (I think) is something like:

(a) You set up a lens to accurately focus an image at the centre of the
sensor (or film). There is a certain distance from the "centre of the
lens" (not sure what the right term is) to the centre of the sensor.

(b) Because the edges/corners of the sensor/film lie on a plane, then the
distance from them to the "centre of the lens" will be slightly longer than
above, so you might expect them to be slightly out of focus (i.e. softer).

(c) If the sensor was very slightly concave, so that all points on it (even
the edges/corners) were exactly the same distance from the "centre of the
lens", then would all the image be in focus?

Something like this. At the moment, I'm working from barely remembered
O/A-level physics. I probably ought to dig out a primer on basic optics.


Regards,
Graham Holden (g-holden AT dircon DOT co DOT uk)
--
There are 10 types of people in the world;
those that understand binary and those that don't.
Anonymous
February 8, 2005 8:24:09 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Graham Holden wrote:

> What I meant (I think) is something like:
>
> (a) You set up a lens to accurately focus an image at the centre of the
> sensor (or film). There is a certain distance from the "centre of the
> lens" (not sure what the right term is) to the centre of the sensor.

I believe it's "nodal point", but I may be mistaken.

>
> (b) Because the edges/corners of the sensor/film lie on a plane, then the
> distance from them to the "centre of the lens" will be slightly longer than
> above, so you might expect them to be slightly out of focus (i.e. softer).

The lines going to the corners of the film should be in proportional length to
the lines going the corresponding corner of the scene at the plane of focus.

>
> (c) If the sensor was very slightly concave, so that all points on it (even
> the edges/corners) were exactly the same distance from the "centre of the
> lens", then would all the image be in focus?

You're assuming that the focus 'plane' (subject) is bent onto a spherical
surface at constant radius from the lens. But it ain't so. (I used to think
this as well).

Go set up a shot of a brick wall. Get the lens perfectly perpendicular to the
wall (film plane parallel to wall). Light it from the sides such that the
texture is clear. Focus manually as sharp as you can. Shoot wide open and f/8
- f/11. Try with a 50mm, a 28 to 35 and a 100mm. (Don't use a zoom lens).

Cheers,
Alan
--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 8, 2005 9:16:35 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

In article <cub59i$l5s$1@inews.gazeta.pl>,
Alan Browne <alan.browne@freelunchVideotron.ca> wrote:
>Graham Holden wrote:
>
>> What I meant (I think) is something like:
>>
>> (a) You set up a lens to accurately focus an image at the centre of the
>> sensor (or film). There is a certain distance from the "centre of the
>> lens" (not sure what the right term is) to the centre of the sensor.
>
>I believe it's "nodal point", but I may be mistaken.
>
>>
>> (b) Because the edges/corners of the sensor/film lie on a plane, then the
>> distance from them to the "centre of the lens" will be slightly longer than
>> above, so you might expect them to be slightly out of focus (i.e. softer).
>
>The lines going to the corners of the film should be in proportional length to
>the lines going the corresponding corner of the scene at the plane of focus.

Agreed that this is the *desired* behavior. However, the closer
to this the more complex the lens for a given focal length. Simple
lenses tend to have a more nearly spherical focal "plane".

>> (c) If the sensor was very slightly concave, so that all points on it (even
>> the edges/corners) were exactly the same distance from the "centre of the
>> lens", then would all the image be in focus?
>
>You're assuming that the focus 'plane' (subject) is bent onto a spherical
>surface at constant radius from the lens. But it ain't so. (I used to think
>this as well).
>
>Go set up a shot of a brick wall. Get the lens perfectly perpendicular to the
>wall (film plane parallel to wall). Light it from the sides such that the
>texture is clear. Focus manually as sharp as you can. Shoot wide open and f/8
>- f/11. Try with a 50mm, a 28 to 35 and a 100mm. (Don't use a zoom lens).

Get a simple single-element lens, and adapt it to the camera
body and take the same shot. (The shorter the focal length, the more
extreme the behavior.) Your lenses are expensive and complex. A lot of
work has been done to minimize this behavior, which is called "spherical
aberration", IIRC.

But if you want it even more minimized, get an apochromat lens.
They are not fast lenses. No autofocus or auto diaphragm. These are
designed for one purpose only -- photographing a totally flat piece of
artwork onto a totally flat sensor (large format film, typically). I
worked with a pair of these in a previous life (back when I was still
working). A 2" (50.8mm) for making negatives to produce hybrid
integrated circuits, and a much longer one (about 6" in a view camera on
a lathe bed optical bench) for making negatives to produce printed
circuit boards.

Enjoy,
DoN.
--
Email: <dnichols@d-and-d.com> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---
February 10, 2005 10:30:56 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

>
> The depth of field of a lens setting is determined by the circle of
> confusion, not the size of the film. If the COC of both 35mm film and
> the digital sensor were the same, then the DOF would be the same with
> the same lens. But they're not. :-)

Can you explain a bit further why they are not ?. Is it the qualities of the
CCD/CMOS sensor ?
Anonymous
February 10, 2005 10:30:57 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

dylan wrote:

>>The depth of field of a lens setting is determined by the circle of
>>confusion, not the size of the film. If the COC of both 35mm film and
>>the digital sensor were the same, then the DOF would be the same with
>>the same lens. But they're not. :-)
>
>
> Can you explain a bit further why they are not ?. Is it the qualities of the
> CCD/CMOS sensor ?

DOF is a subjective quality of printing related to the Circle-of-Confusion at
'taking' time. The "real" COC is always the same for a given focal length and
aperture. The COC used in computations (such as lens markings) is based on
empirical numbers for printing at about 8x10 and viewing at a given distance (10
inches I believe) with an expected visible resolution of so many printed line
pairs per mm (which varies as well 'tween 5 and 8).

The fulcrum here is the ratio of the size of the sensor and the size of the print.

See:

http://www.nikonlinks.com/unklbil/dof.htm

--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 10, 2005 11:05:54 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

On Thu, 10 Feb 2005 19:30:56 -0000, "dylan" <no@nospam.com> wrote:

>
>> The depth of field of a lens setting is determined by the circle of
>> confusion, not the size of the film. If the COC of both 35mm film and
>> the digital sensor were the same, then the DOF would be the same with
>> the same lens. But they're not. :-)
>
>Can you explain a bit further why they are not ?. Is it the qualities of the
>CCD/CMOS sensor ?

Hmm. He's a little confusing there. *if* the sensor was 35mm, CoC
would be the same as a 35mm film, and therefore DoF would be the same.

CCD/CMOS sensors however, are usually smaller than 35mm. It's got
nothing to do with any other attribute than physical size.

--
Owamanga!
Anonymous
February 10, 2005 11:05:55 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Owamanga <nomail@hotmail.com> wrote:

> On Thu, 10 Feb 2005 19:30:56 -0000, "dylan" <no@nospam.com> wrote:
>
> >> The depth of field of a lens setting is determined by the circle of
> >> confusion, not the size of the film. If the COC of both 35mm film and
> >> the digital sensor were the same, then the DOF would be the same with
> >> the same lens. But they're not. :-)
> >
> >Can you explain a bit further why they are not ?. Is it the qualities of the
> >CCD/CMOS sensor ?
>
> Hmm. He's a little confusing there.

Yes, I could have worded it better.

> *if* the sensor was 35mm, CoC would be the same as a 35mm film, and
> therefore DoF would be the same.

The only reason you can reliably scale the CoC between differently-sized
sensors (or film) is because it's assumed that the image from either
source will be printed out at the *same* size. That is, both sources'
CoC are computed relative to an 8x10 print.

In practice, this distinction won't matter that much for most people.
But that doesn't make the distinction untrue. :-)

> CCD/CMOS sensors however, are usually smaller than 35mm. It's got nothing
> to do with any other attribute than physical size.

No, CoC isn't a given based on the size of the film (or sensor); it's
based on the output.

You can see this in effect if you take a photo you shot that's maybe
less than sharp, load it into photoshop and shrink it down to, say 25%.
It will seem sharper at the smaller size. Look at it at 100%, however,
and all the sharpness problems pop right out at you. That's because you
effectively change the circle of confusion by shrinking the *output*;
your eye can't see the problems in the smaller version. And since DoF is
mostly a function of the CoC, you are also changing the depth of field
of the image by shrinking it.

Here's an example: My attempt to capture Grand Junction, CO from within
Colorado National Monument using twilight and a car window mount. Shrunk
down to 25% it looks adequate, in terms of sharpness (JPEG artifacts
aside), because shrinking it effectively enlarges the circle of
confusion to cover the sharpness problems. The image's DoF is very wide,
from the bush in the foreground to the mountains on the other side of
the valley...

<http://www.mile23.com/misc/co_natl.jpg&gt;

But the 100% image effectively shrinks the circle of confusion back to
reveal every detail of how truly unsharp this image really is. Its DoF
is essentially zero due to camera shake. Here's a crop...

<http://www.mile23.com/misc/co_natl_crop.jpg&gt;

Both images were taken with the same camera, lens, and sensor (since
they're the same image), but the apparent sharpness changes a great deal
with different output sizes. Apparent sharpness and the circle of
confusion are basically the same thing, so changing the output size
changes the circle of confusion, and thus DoF, all other things being
equal.

(If only I had used a tripod instead...)
February 11, 2005 12:43:16 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

..
>
>> *if* the sensor was 35mm, CoC would be the same as a 35mm film, and
>> therefore DoF would be the same.
>
> The only reason you can reliably scale the CoC between differently-sized
> sensors (or film) is because it's assumed that the image from either
> source will be printed out at the *same* size. That is, both sources'
> CoC are computed relative to an 8x10 print.
>
> In practice, this distinction won't matter that much for most people.
> But that doesn't make the distinction untrue. :-)
>
>> CCD/CMOS sensors however, are usually smaller than 35mm. It's got nothing
>> to do with any other attribute than physical size.
>
> No, CoC isn't a given based on the size of the film (or sensor); it's
> based on the output.
>

So I now understand that it's nothing to do with the sensor characteristics
or it's size, that's what baffled me because I couldn't understand how
changing the sensor size altered the DOF, it would be like putting a 35mm
back on a MF camera, does that change the DOF, or the DOF should change
towards the centre of a 35mm image (maybe it does and I haven't noticed)?

So it's related to printing the outputs to the same size (ie 8x10), sounds
possible but surely once the captured image has recorded the scene with in
and out of focus areas it doesn' matter how big or small you print it.

Sorry to labour this but I'd like to understand.
Anonymous
February 11, 2005 12:43:17 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

dylan <no@nospam.com> wrote:

[..]
> >> CCD/CMOS sensors however, are usually smaller than 35mm. It's got
> >> nothing to do with any other attribute than physical size.
> >
> > No, CoC isn't a given based on the size of the film (or sensor); it's
> > based on the output.
>
> So I now understand that it's nothing to do with the sensor
> characteristics or it's size, that's what baffled me because I couldn't
> understand how changing the sensor size altered the DOF, it would be like
> putting a 35mm back on a MF camera, does that change the DOF, or the DOF
> should change towards the centre of a 35mm image (maybe it does and I
> haven't noticed)?
>
> So it's related to printing the outputs to the same size (ie 8x10), sounds
> possible but surely once the captured image has recorded the scene with in
> and out of focus areas it doesn' matter how big or small you print it.

Sure it does. The boundary between in and out of focus is, by
definition, the depth of field. If you can't see that boundary, then the
dept of field is large. If you can see it, then it's small. It all boils
down to what you can *see,* not what's on the film.

Pick up your camera, set the aperture way open, focus on something, then
pull the focus off just a tiny bit. Pull the object out of focus just
enough that it *appears* to be barely out of the depth of field you see
through the lens.

Take a picture. Download that image into the computer and look at it at
100%: It's a little bit out of focus. Shrink it down to 15% or so, and
it'll look like it's in focus (or at least, more in focus than at 100%).
You just changed the depth of field of the image by making it smaller,
making it *appear* that more of the image is in focus. You did this
after the exposure, without altering the focal length or aperture of
your lens. The smaller you go, the more the image *appears* to be in
focus. Which means the object in the image is within the depth of field,
even though you deliberately set it out of focus.

Try it. :-)

HTH.

--
"Eighty percent of Republicans are just Democrats who don't know what's
going on." -- Robert Kennedy, Jr.
February 11, 2005 2:57:20 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

>
> Take a picture. Download that image into the computer and look at it at
> 100%: It's a little bit out of focus. Shrink it down to 15% or so, and
> it'll look like it's in focus (or at least, more in focus than at 100%).
> You just changed the depth of field of the image by making it smaller,
> making it *appear* that more of the image is in focus. You did this
> after the exposure, without altering the focal length or aperture of
> your lens. The smaller you go, the more the image *appears* to be in
> focus. Which means the object in the image is within the depth of field,
> even though you deliberately set it out of focus.
>
> Try it. :-)
>
> HTH.

Thanks, I understand your argument but still not 100% convinced that's the
whole story because how does the DOF scale work on a lens (not that the
latest models have them) , it would need to be related to the final print
size ?

Cheers :o )
Anonymous
February 11, 2005 3:22:58 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

dylan <no@nospam.com> wrote:

> > Take a picture. Download that image into the computer and look at it at
> > 100%: It's a little bit out of focus. Shrink it down to 15% or so, and
> > it'll look like it's in focus (or at least, more in focus than at 100%).
> > You just changed the depth of field of the image by making it smaller,
> > making it *appear* that more of the image is in focus. You did this
> > after the exposure, without altering the focal length or aperture of
> > your lens. The smaller you go, the more the image *appears* to be in
> > focus. Which means the object in the image is within the depth of field,
> > even though you deliberately set it out of focus.
> >
> > Try it. :-)
> >
> > HTH.
>
> Thanks, I understand your argument but still not 100% convinced that's the
> whole story because how does the DOF scale work on a lens (not that the
> latest models have them) , it would need to be related to the final print
> size ?

Yup. The DOF scale on the lens is based on an assumption that the output
will be 8x10, 10 inches from your face. It's a safe assumption, too.

--
"Eighty percent of Republicans are just Democrats who don't know what's
going on." -- Robert Kennedy, Jr.
Anonymous
February 11, 2005 5:50:44 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"Paul Mitchum" <usenet@mile23.c0m> wrote in message
news:1grrs0g.1vgw0kr1louttgN%usenet@mile23.c0m...

> Apparent sharpness and the circle of
> confusion are basically the same thing, so changing the output size
> changes the circle of confusion, and thus DoF, all other things being
> equal.
>

Here's what I understand about CoC.

From what I have read, the generally accepted numbers for the resolving
power of the average human eye range from about 5 to about 7
lines/millimeter at a distance of 25 cm. If we equate the acceptable
diameter of a blur circle to the resolving power, then we could accept a CoC
of 0.2 mm at 25 cm, or one mm at 1250 mm using the 5 lines/mm or one mm at
1750 mm using 7 l/mm. These correspond to visual resolving angles of 0.8
milliradians and 0.6 mr, respectively.

It is conventional to define the "standard" viewing distance for a print as
its diagonal measure, and at this distance you would like for everything to
appear sharp that is intended to appear sharp. For an 8X10 inch print, that
is about 32.5 cm, a little larger than the 25 cm reference value noted
above. However, 5 l/mm at 32.5 cm is equivalent to about 7 l/mm at 25 cm,
so all the numbers are in the same ball park. So, using the 7 l/mm visual
acuity number, the blur circle, or CoC, should be not greater than 0.14 mm
on the 8X10 print. [ Note: For a 10 ft X 20 ft billboard, an acceptable
blur circle for limiting sharpness would be about 0.2 inches for a viewing
distance of about 20 ft. (of course, no billboard is that sharp)].

To see what the CoC on the camera image plane is required to produce the
desired CoC on a given size print, you apply the magnification factor. A
35mm frame has dimensions of 24 X 36 mm. Magnifying the 24mm dimension to 8
inches (cropping the ends) is a factor of 8.5. Dividing the 0.14mm CoC on
the 8X10 print by 8.5, we get 0.016mm as the required maximum size of the
CoC on the image plane (film or digital).

Let's come at it from another direction. A Canon 10D has a sensor length of
22.5 mm which is covered by 3072 pixels. This gives a pixel pitch of
0.00732 mm. If we define the CoC as two pixel spacings (centering the blur
circle on one pixel and letting it extend to the adjacent sensor in each
direction), then we get 0.015 mm. Given the discrete pixels, there is not
much to be gained by trying to resolve finer than this.

This all implies that a Canon 10D will realize the full limits of average
human visual acuity on prints viewed at their "normal" viewing distance or
greater. This can be limited by lens abberations and diffraction, of
course.

I believe that Canon claims a CoC of 0.035 mm on the sensor, which is twice
the numbers I came up with above. I don't know why, maybe it has to do with
the Bayer arrangement of the photosensors. Also, it could be that the
degree of unsharpness that they define as the limit of Depth Of Field is
less than the maximum sharpness achieveable by the sensor.

You can see that there are uncertainties in all these numbers and choices to
be made in how you define things, so don't get carried away with too much
precision here.

Stan
Anonymous
February 11, 2005 9:21:02 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

In article <cui6j2$d32$1@news7.svr.pol.co.uk>, dylan <no@nospam.com> wrote:
>>
>> Take a picture. Download that image into the computer and look at it at
>> 100%: It's a little bit out of focus. Shrink it down to 15% or so, and
>> it'll look like it's in focus (or at least, more in focus than at 100%).
>> You just changed the depth of field of the image by making it smaller,
>> making it *appear* that more of the image is in focus. You did this
>> after the exposure, without altering the focal length or aperture of
>> your lens. The smaller you go, the more the image *appears* to be in
>> focus. Which means the object in the image is within the depth of field,
>> even though you deliberately set it out of focus.
>>
>> Try it. :-)
>>
>> HTH.
>
>Thanks, I understand your argument but still not 100% convinced that's the
>whole story because how does the DOF scale work on a lens (not that the
>latest models have them) , it would need to be related to the final print
>size ?

Yes, it would. And it was. As, in fact, is the standard DOF definition.
(something to bear in mind when checking using your DOF preview feature).

But, let's face it, how many people would complain that their standard
3x5 or 4x6 prints had too much depth of field?
February 11, 2005 10:09:46 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

>
> Yes, it would. And it was. As, in fact, is the standard DOF definition.
> (something to bear in mind when checking using your DOF preview feature).
>
> But, let's face it, how many people would complain that their standard
> 3x5 or 4x6 prints had too much depth of field?
>
>

Thanks.
Anonymous
February 11, 2005 10:44:24 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"Stan Prevost" <sprevost@knology.net> wrote in message
news:110q6l4os77d199@corp.supernews.com...
>
Oops. Due to being interrupted by such trivialities as having to leave for
a doctor's appointment, being in a hurry, and being somewhat stupid, I made
some errors in my previous post. The errors affect my conclusions. Here is
my latest version.

Here's what I understand about CoC.

From what I have read, the generally accepted numbers for the resolving
power of the average human eye range from about 5 to about 7
lines/millimeter at a distance of 25 cm. If we equate the acceptable
diameter of a blur circle to the resolving power, then we could accept a CoC
of 0.2 mm at 25 cm, or one mm at 1250 mm using the 5 lines/mm or one mm at
1750 mm using 7 l/mm. These correspond to visual resolving angles of 0.8
milliradians and 0.6 mr, respectively.

It is conventional to define the "standard" viewing distance for a print as
its diagonal measure, and at this distance you would like for everything to
appear sharp that is intended to appear sharp. For an 8X10 inch print, that
is about 32.5 cm, a little larger than the 25 cm reference value noted
above. However, 5 l/mm at 32.5 cm is equivalent to about 7 l/mm at 25 cm,
so all the numbers are in the same ball park. So, using the 7 l/mm visual
acuity number, the blur circle, or CoC, should be not greater than 0.14 mm
on the 8X10 print. [ Note: For a 10 ft X 20 ft billboard, an acceptable
blur circle for limiting sharpness would be about 0.2 inches for a viewing
distance of about 20 ft. (of course, no billboard is that sharp)].

To see what the CoC on the camera image plane is required to produce the
desired CoC on a given size print, you apply the magnification factor. A
35mm frame has dimensions of 24 X 36 mm. Magnifying the 24mm dimension to 8
inches (cropping the ends) is a factor of 8.5. Dividing the 0.14mm CoC on
the 8X10 print by 8.5, we get 0.016mm as the required maximum size of the
CoC on the image plane (film or digital).

Rather than a full-frame 35mm, let's consider a Canon 10D, which has sensor
dimensions of 15.1 mm X 22.7 mm. The magnification ratio to 8X10 as defined
earlier is 13.5. The CoC on the sensor required to produce the desired CoC
on a given size print is .14/13.5 = 0.01mm. The 1.6X sensor factor can be
seen here.

Let's come at it from another direction. A Canon 10D has a sensor length of
22.7 mm which is covered by 3072 pixels. This gives a pixel pitch of
0.00739 mm. If we define the CoC as two pixel spacings (centering the blur
circle on one pixel and letting it extend to the adjacent sensor in each
direction), then we get 0.015 mm. Given the discrete pixels, there is not
much to be gained by trying to resolve finer than this.

This all implies that a Canon 10D may not provide the full degree of
sharpness that average
human visual is capable of perceiving on prints viewed at their "normal"
viewing distance. The sensor will limit the resolution to a value that is
about 50% greater than the eye is capable of perceiving. This can be
further degraded by lens abberations and diffraction, of
course. But all will be well if you view the 8X10s at 52 cm rather than
32.5 cm. :-)

I use a Canon 10D, and I can't tell that the digital sensor degrades the
image at all when viewing a print at normal distance. However, my eyes are
not the best. A 20D gets a lot closer.

I believe that Canon claims a CoC of 0.035 mm on the 10D sensor, which is
twice
the numbers I came up with above. I don't know why, maybe it has to do with
the Bayer arrangement of the photosensors. Also, it could be that the
degree of unsharpness that they define as the limit of Depth Of Field is
less than the maximum sharpness achieveable by the sensor.

You can see that there are uncertainties in all these numbers and choices to
be made in how you define things, so don't get carried away with too much
precision here.

Stan
Anonymous
February 12, 2005 4:10:42 PM

Archived from groups: rec.photo.digital.slr-systems (More info?)

Stan Prevost wrote:

> "Stan Prevost" <sprevost@knology.net> wrote in message
> news:110q6l4os77d199@corp.supernews.com...
>
> Oops. Due to being interrupted by such trivialities as having to leave for
> a doctor's appointment, being in a hurry, and being somewhat stupid, I made
> some errors in my previous post. The errors affect my conclusions. Here is
> my latest version.
>
> Here's what I understand about CoC.
>
> From what I have read, the generally accepted numbers for the resolving
> power of the average human eye range from about 5 to about 7

Just link to a well edited discourse next time:
http://www.nikonlinks.com/unklbil/dof.htm



--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: there's no such thing as a FreeLunch.
Anonymous
February 13, 2005 1:48:38 AM

Archived from groups: rec.photo.digital.slr-systems (More info?)

"Alan Browne" <alan.browne@freelunchVideotron.ca> wrote in message
news:culgqu$98p$2@inews.gazeta.pl...
> Stan Prevost wrote:
>
>> "Stan Prevost" <sprevost@knology.net> wrote in message
>> news:110q6l4os77d199@corp.supernews.com...
>>
>> Oops. Due to being interrupted by such trivialities as having to leave
>> for a doctor's appointment, being in a hurry, and being somewhat stupid,
>> I made some errors in my previous post. The errors affect my
>> conclusions. Here is my latest version.
>>
>> Here's what I understand about CoC.
>>
>> From what I have read, the generally accepted numbers for the resolving
>> power of the average human eye range from about 5 to about 7
>
> Just link to a well edited discourse next time:
> http://www.nikonlinks.com/unklbil/dof.htm
>

Not quite sure what you mean. None of the errors I referred to have
anything to do with the article you link to. What that article says is in
agreement with the first part of my post. The calculations I initially
hurriedly did (and later corrected) were outside the scope of that article
and related the pixel spacing of a digital sensor to the CoC.
Related resources
!