Is there an objective spec that can be used to compare depth of field
for different cameras/lenses? Would hyperfocal distance be able to
serve this purpose?
I'm not an expert on photography. I would like to get a non-SLR camera
that can produce good bokeh (shallow depth of field), but I would like
to be able to compare specs.
bucky3 wrote:
> Is there an objective spec that can be used to compare depth of field
> for different cameras/lenses? Would hyperfocal distance be able to
> serve this purpose?
>
> I'm not an expert on photography. I would like to get a non-SLR camera
> that can produce good bokeh (shallow depth of field), but I would like
> to be able to compare specs.
AFAIK, the Sony R1, Sigma DP1, and Leica M8 are about the only choices
for larger sensor non-slr. There's a recent review of the DP1 on
DPreview.com.
For a typical 1/1.8" sensor compact camera, then f2.8 is equivalent to
about f13 on 35mm for depth of field given same "equivalent" focal
length. It might count out anything but the three cameras above to suit
your needs.
The R1 was probably discontinued as it was no smaller and no less
expensive than dslrs (FWIW, I thought it was quite good). The M8 is
(IMO) very expensive for what it is.
On Jun 6, 4:57 am, bucky3 <buc...@mail.com> wrote:
> Is there an objective spec that can be used to compare depth of field
> for different cameras/lenses? Would hyperfocal distance be able to
> serve this purpose?
>
> I'm not an expert on photography. I would like to get a non-SLR camera
> that can produce good bokeh (shallow depth of field), but I would like
> to be able to compare specs.
Even the hyperfocal computation is not totally objective. All require
a value for acceptable blur circle/coc. That value is still somewhat
subjective.
In order to control depth of field you need some degree of exposure
control, like aperture priority and a very fast lens if you want a
narrow depth of field. As long as you are shooting wide open, the
format size and f/# of the lens pretty well control depth of field, so
it shouldn't vary much from camera to camera with the same format
size, focal length and f/#.
"bucky3" <bucky3@mail.com> schreef in bericht
news:2822c4c2-8d27-415a-be6c-713a087141ec@u6g2000prc.googlegroups.com...
> Is there an objective spec that can be used to compare depth of field
> for different cameras/lenses? Would hyperfocal distance be able to
> serve this purpose?
>
> I'm not an expert on photography. I would like to get a non-SLR camera
> that can produce good bokeh (shallow depth of field), but I would like
> to be able to compare specs.
To compare depth of field, you need to determine how to compare.
On way of doing is, is using a DOF calculator. And use a circle of
not determined by the format, but fixed as compared to the diagonal.
For example 1/1500 of the diagonal. This makes comparing of different
formats with different aperature settings possible.
But a ROUGH way is:
If the size of a sensor is X times the size of another sensor the aperature
number should be divided with the same number X.
Examples:
1 Sensor 1/6 in size of 35 mm format, aperature should be 1/6 of the number
in 35 mm format. 35 mm f 4.0 would be in 1/6 sensor size f 0.666666
Or the other way around 1/6 sensor size with 2.8 as largest aperature. Would
be for 35 mm format 6 times 2.8 is f 16.8
(The calculation is bases on the same angle of view, de same subject
distance
and the same coc where coc is a set fraction of the diagonal)
ben
(The rough method is not totaly accurate, but works for all subject
distances and
for all angles of view and is precise enough in all but the most critical
situations).
On Jun 6, 7:08 am, Don Stauffer in Minnesota <stauf...@usfamily.net>
wrote:
> Even the hyperfocal computation is not totally objective. All require
> a value for acceptable blur circle/coc. That value is still somewhat
> subjective.
OK, but after you pick a value for acceptable COC, then the hyperfocal
calculation would be an objective/consistent comparison?
On Jun 6, 8:46 am, "ben brugman" <b...@niethier.nl> wrote:
> But a ROUGH way is:
> If the size of a sensor is X times the size of another sensor the aperature
> number should be divided with the same number X.
> Or the other way around 1/6 sensor size with 2.8 as largest aperature. Would
> be for 35 mm format 6 times 2.8 is f 16.8
Thanks, that helps. I think it's helpful for me to think: "At f2.8, my
puny point and shoot camera is only achieving the equivalent of f16.8
with a full frame sensor."
BTW, what is the value to use for sensor size? Is is the diagonal
length? Area?
bucky3 <bucky3@mail.com> wrote:
> On Jun 6, 7:08 am, Don Stauffer in Minnesota <stauf...@usfamily.net>
> wrote:
>> Even the hyperfocal computation is not totally objective. All require
>> a value for acceptable blur circle/coc. That value is still somewhat
>> subjective.
>
> OK, but after you pick a value for acceptable COC, then the hyperfocal
> calculation would be an objective/consistent comparison?
If your CoC were chosen as a fraction of format diagonal,
then it would be. One standard is diagonal/1720, which equals
2 minutes of arc when viewed at the same distance as the diagonal
of the format x enlargement.
"bucky3" <bucky3@mail.com> schreef in bericht
news:1f033111-f7dc-4973-89b9-2f21fff25439@r66g2000hsg.googlegroups.com...
> On Jun 6, 8:46 am, "ben brugman" <b...@niethier.nl> wrote:
>> But a ROUGH way is:
>> If the size of a sensor is X times the size of another sensor the
>> aperature
>> number should be divided with the same number X.
>> Or the other way around 1/6 sensor size with 2.8 as largest aperature.
>> Would
>> be for 35 mm format 6 times 2.8 is f 16.8
>
> Thanks, that helps. I think it's helpful for me to think: "At f2.8, my
> puny point and shoot camera is only achieving the equivalent of f16.8
> with a full frame sensor."
>
> BTW, what is the value to use for sensor size? Is is the diagonal
> length? Area?
Because the sensorshapes are different the most practical way is to
compare on diagonal. Most digital sensors are 3:4 where DSLR and
35 mm camera's have 2:3. So you could compare the hight or the width
of the sensor but diagonal is somewhere in the middle of both extremes.
If the sensors have the same shape (both 2:3) it doesn't matter which size
you take because then it works out the same.
I like the 2:3 format better than the 3:4 format so for me I would preferre
to
compare the long side and clip away everything from the 3:4 format to make
it the same.
A way to estimate the chipsize is look at the focal length of the lens on
the barrel
at the wide end. (Often something between 4 and 10 mm). The 35 mm
equivalent is often known. (Often 35, 36 or 38 mm). Divide those two
numbers and you know how smal the chipsize is compared to 35 mm format.
ben, who has a TZ3 which does support 2:3 format and actually has extra
pixels
in the width compared to the 3:4 format. Chipsize of the TZ3 is 1/6 of 35 mm
"Peter Irwin" <pirwin@ktb.net> schreef in bericht
news:g2bt2s$o9h$1@dns.ktb.net...
> bucky3 <bucky3@mail.com> wrote:
>> On Jun 6, 7:08 am, Don Stauffer in Minnesota <stauf...@usfamily.net>
>> wrote:
>>> Even the hyperfocal computation is not totally objective. All require
>>> a value for acceptable blur circle/coc. That value is still somewhat
>>> subjective.
>>
>> OK, but after you pick a value for acceptable COC, then the hyperfocal
>> calculation would be an objective/consistent comparison?
>
> If your CoC were chosen as a fraction of format diagonal,
> then it would be. One standard is diagonal/1720, which equals
> 2 minutes of arc when viewed at the same distance as the diagonal
> of the format x enlargement.
>
hyperfocal calculation would be an objective/consisten comparison,
but it's not very usefull, because it's difficult to interpret hyperfocal
length. Comparing two hyperfocal length's does not realy illustrate the
difference in DOF although it's consistent and objective.
Go with the fraction of the format as written. (1/1720 of the diagonal
is a high standard for DOF, most standards are less.) And compare DOFS
with a DOF calculator. (And offcourse experiment with your camera's).
On Fri, 6 Jun 2008 17:46:19 +0200, "ben brugman" <ben@niethier.nl>
wrote:
>
>"bucky3" <bucky3@mail.com> schreef in bericht
>news:2822c4c2-8d27-415a-be6c-713a087141ec@u6g2000prc.googlegroups.com...
>> Is there an objective spec that can be used to compare depth of field
>> for different cameras/lenses? Would hyperfocal distance be able to
>> serve this purpose?
>>
>> I'm not an expert on photography. I would like to get a non-SLR camera
>> that can produce good bokeh (shallow depth of field), but I would like
>> to be able to compare specs.
>
>To compare depth of field, you need to determine how to compare.
>On way of doing is, is using a DOF calculator. And use a circle of
>not determined by the format, but fixed as compared to the diagonal.
>For example 1/1500 of the diagonal. This makes comparing of different
>formats with different aperature settings possible.
>
>But a ROUGH way is:
>If the size of a sensor is X times the size of another sensor the aperature
>number should be divided with the same number X.
>
>Examples:
>1 Sensor 1/6 in size of 35 mm format, aperature should be 1/6 of the number
>in 35 mm format. 35 mm f 4.0 would be in 1/6 sensor size f 0.666666
>
>Or the other way around 1/6 sensor size with 2.8 as largest aperature. Would
>be for 35 mm format 6 times 2.8 is f 16.8
>
>(The calculation is bases on the same angle of view, de same subject
>distance
>and the same coc where coc is a set fraction of the diagonal)
>
>ben
>
>(The rough method is not totaly accurate, but works for all subject
>distances and
>for all angles of view and is precise enough in all but the most critical
>situations).
I agree with this approach... works surprisingly well for normal
shooting situations.
Note that focal length plays NO ROLE contrary to what many think,
provided the framing of the image is the same.
I also agree that rough methods are eminently useful for DOF
calculations. That's all most of need to know. Furthermore, hardly
anybody can actually compute accurate DOFs because the formulas depend
on the exit pupil magnification, which usually is unknown.
comparing depth of field 
Linear Mode
