Answer this one question!

[Eugene Powers]

what does it mean in terms of the last lens element? Is it the same as on FF lenses or is it smaller? Why does it produce smaller circle?

--
Eugene

The only time a smaller sensor with the same pixel count is superior to a larger sensor (aka higher pixel density) is when you are focal-length limited.

Lee Jay

 

[Eugene Powers]

compare images from different systems but you can't say that one system is better than another in the corners unless you make one common element.

And that element is the lens. And it has to the same physical lens not two copies of the same lens. Otherwise you will have one result and somebody else will have another result.

--
Eugene

The only time a smaller sensor with the same pixel count is superior to a larger sensor (aka higher pixel density) is when you are focal-length limited.

Lee Jay

 

[joe mama]

compare images from different systems but you can't say that one
system is better than another in the corners unless you make one
common element.
And that element is the lens. And it has to the same physical lens
not two copies of the same lens. Otherwise you will have one result
and somebody else will have another result.

You and I differ on the "common element". For me, the "common element" is an equivalent image. I mean, photography is all about the image it produces, right? So, if I'm comparing two systems, I will want to compare the pictures it produces, not the means required to get the pictures, and I will want those pics to be equivalent .

I cannot take the same pics with a 5D + 16-35 / 2.8L and a 30D with the same lens. It simply cannot be done. Well, technically, in the range where the equivalent FLs overlap, but that's just plain silly.

However, in the sprit of compromise, I will not even argue against that. If you feel a valid test would be, for example, the 5D + 16-35 / 2.8L @ 32mm, f / 9 vs the 30D with the same lens at 20mm, f / 5.6, then, sure, I'll agree that test is valid.

Just as an FYI -- there is precious little difference in IQ, if any at all, between the 16-35 / 2.8L at 21mm and the 17-55 / 2.8 IS at 21mm, which is what the test I linked to used. Likewise, there is basically no difference between the 85 / 1.2L and 135 / 2L, either. I can link the PZ tests, if you don't believe me.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[Eugene Powers]

I see now what you mean but I have to disagree with you.

To me the proper test would be this: one lens 16-35mm in this case at the same focal length and same aperture pointed at the brick wall. This way DOF is taken out of equation. So the only difference in this test is the camera and that is exactly what we are trying to compare. The Camera.

--
Eugene

The only time a smaller sensor with the same pixel count is superior to a larger sensor (aka higher pixel density) is when you are focal-length limited.

Lee Jay

 

[Lee Jay]

what does it mean in terms of the last lens element? Is it the same
as on FF lenses or is it smaller? Why does it produce smaller
circle?

You mean the rear-most element? It means it can be slightly smaller.

It produces a smaller image circle because of vignetting. A 500/4 would be a 180° circular fisheye on a large enough format if it weren't for vignetting. Since it's for 35mm-format, it's designed to vignette somewhere a bit outside the 35mm frame.

--
Lee Jay
(see profile for equipment)

 

[UziY]

In your shooting situatuion, would you like a 15-44/1.8L lens? Are
you willing to spend over $2000 for it?

Well because of the clarity that Eugene has stated, no, not at this
time. I have no potential customers or myself complaining about the
DOF I am currently getting. When I control the shooting situation I
can use a 70-200 f2.8L on my 1.6 crop and get excellent DOF by
placing the subject and background where I want.

What I am starting to see the value in with a 1.6 crop when
shooting in situations that I don't control. I get excellent
shutter speed without having to go to ISO3200 when the light is dim.

As I stated in my original question:
If I am using my 24-70 at f2.8 ISO1600 on a 30D so that I get
shutter speeds around 1/125 would it be wise for me to go full
frame?
Now I know that full frame won't help "IN THIS ONE SITUATION". In
my above question, I can not go any faster than f2.8 (lense
limitation) and don't want noise any worse than ISO1600, AND can
not use a shutter slower than 1/125 otherwise the subject will
blur...

... in order to match DOF from FF to crop camera you need to
stop down FF camera.
... if you stop down FF camera you get less light coming in so
you need to increase ISO or slow down shutter to match the same
exposure.
... It is all based on DOF

"Noise any worse than ISO1600" is not well defined - ISO1600 has much less noise on a 5D than on a 30D, which in turn has much less noise than a P&S with a small sensor.

You and Eugene are correct: if you need the DOF, you will need to stop down more on a FF. You can increase ISO without degradation of quality (relative to your 30D), so things will not get worse; however, they will not get any better.

Only reason to consider FF in your case is if you either need a shallower DOF, or at least are willing to accept a shallower DOF in return for less noise and better image quality.
--
Uzi
http://www.pbase.com/uyoeli

 

[Medic Laviña]

I think you guys are not paying attention to each other's points. The way I look at it, Joe's illustration mainly deals with FOV and DOF equivalence (minus exposure level), while the other guy's deal with FOV and exposure level (minus DOF). To add exposure level to Joe's illustration, it would be:

1) 24mm, f/2.8, 1/320sec @ ISO 100 on 1.6x (crop) DSLR
2) 38mm, f/4.5, 1/200sec @ ISO 100 on FF DSLR

The above examples should yield the same equivalent image as far as FOV, DOF and exposure level is concerned. Right?

--
Medic
-----------------------------------------------------

• The camera is mightier than the pen.

 

[joe mama]

I think you guys are not paying attention to each other's points.
The way I look at it, Joe's illustration mainly deals with FOV and
DOF equivalence (minus exposure level),

I most certainly do deal with exposure:

http://forums.dpreview.com/forums/read.asp?forum=1029&message=23294660

Allow me to quote from the above link:

So, to get the same exposure, you either have to reduce the shutter
speed by a factor of 1.6 or multiply the ISO by a factor of 2.56, or
some equivalent inbetween combination.

and again:

http://forums.dpreview.com/forums/read.asp?forum=1029&message=23304161

Here's another quote from the above link:

If you shoot f / 2.8, ISO 100 on 1.6x and I shoot f / 4.5, ISO 250 on
FF, we will get the same shutter speed and exposure.

I probably said it in other posts, too, but I believe I've made my point.

To add exposure level to Joe's illustration, it would be:

1) 24mm, f/2.8, 1/320sec @ ISO 100 on 1.6x (crop) DSLR
2) 38mm, f/4.5, 1/200sec @ ISO 100 on FF DSLR

The above examples should yield the same equivalent image as far as
FOV, DOF and exposure level is concerned. Right?

Aaaargh! I messed up -- again! You have to divide the shutter speed by the square of the crop factor, not just the crop factor. My bad! I so, so, so apologize for this error!

Thus, it should be:

1) 24mm, f/2.8, 1/320sec @ ISO 100 on 1.6x (crop) DSLR
2) 38mm, f/4.5, 1/125sec @ ISO 100 on FF DSLR

However, rather than changing the shutter speed, you can up the ISO instead:

1) 24mm, f/2.8, 1/320sec @ ISO 100 on 1.6x (crop) DSLR
2) 38mm, f/4.5, 1/320sec @ ISO 250 on FF DSLR

I can't believe I made such an error. I feel like a total idiot! My most sincere apologies!

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[Lee Jay]

I see now what you mean but I have to disagree with you.
To me the proper test would be this: one lens 16-35mm in this case
at the same focal length and same aperture pointed at the brick
wall. This way DOF is taken out of equation. So the only difference
in this test is the camera and that is exactly what we are trying
to compare. The Camera.

I've done that test with my 17-40L. The full-frame image wins on most of the available focal lengths and f-stops, a few are close.

--
Lee Jay
(see profile for equipment)

 

[joe mama]

I see now what you mean but I have to disagree with you.

Heh! : )

To me the proper test would be this: one lens 16-35mm in this case
at the same focal length and same aperture pointed at the brick
wall. This way DOF is taken out of equation. So the only difference
in this test is the camera and that is exactly what we are trying
to compare. The Camera.

But, that test does not reflect real life usage of the lens. Who takes pics where DOF is not in the equation? I mean, lots of people don't think about DOF, but DOF is a factor in the pic, pretty much every single time. Except for those that just take pics of flat objects, such as brick walls. : )

I'll agree that the "brick wall test" is a good test, but you must still stop the lens down 1 1/3 stops on the FF camera to make it valid. However, the tests that you were saying were invalid, were of books on a bookshelf, which is, of course, the same thing as shooting a brick wall.

On the side, let me address the issue of using the same lens. I see now what you mean about copy variation. Let's say we were using the 24-105 / 4L IS on the 5D and 17-55 / 2.8 IS on the 30D (an ideal comparison). But let's say that one of the lenses was a sharp copy whereas the other was a dud. Point taken. When comparing the two systems with their respective equivalent lenses, we must first ascertain that both lenses are good copies. I totally agree. Thus, when a test is performed comparing the two systems with the two different lenses, we cannot be certain that the results are conclusive until we know about performance of the two lenses.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[NeilCUK]

http://forums.dpreview.com/forums/read.asp?forum=1029&message=23300792

If I may quote the opening sentence from the above link:

A FF sensor has 2.56 (1.6 ^ 2) times the area of a 1.6x sensor. Thus,
it collects 2.56 times as much light for the same FOV (framing) and f-
ratio.

forgive me, as i had never heard of an "f-ratio" until now, which i believe is some calculation to do with f-stop and something else (is it focal length divided by physical aperture?)

it's all because "f-stop" is NOT an absolute value for the size of the
aperture

That is the critical point that so many seem to miss.

I agree, and i think it's because nobody is aware of that when going into these 'debates'... and instead of asking, they simply think they know it all already

I had no idea that the physical aperture size was different on each camera... and to be honest, for the general photographer, i'm not sure it's something you need to know

I mean, calculations are all well and good, but at the end of the day you can be a fantastic photographer without knowing exactly how everything is calculated and works out

here's another question for you:

there's some other argument going round about how cropped cameras are sharper at the edges because they use a 'sweeter' part of the lens

from what we've learnt here, we realise that the smaller sensors are not just a "crop" of the image at all (i.e. there is no wasted light falling outside of the sensor area), but it is simply the SAME image but seen from further away (with the same framing)

therefore the lens itself is not at a "sweeter" point at all, because it is not the case that only part of the lens is being used ... is this correct?

i.e. if on full frame, the whole of a 100mm lens "is used"

on a 50% smaller sensor, a 100mm lens is NOT using just 50% of the lens

OR, is it the case that yes, only part of a 100mm lens is "being used" on a crop camera, because the physical size of the aperture does not allow light to be collected from parts of the lens

i.e. on a 50% smaller sensor, because of the physical aperture size, only 50% of the lens elements are used (the outside-most 50% is presumably not used) ?

so, which is true?

and my main question - irrespective of which is true, what is the argument put forward for EF-S lenses? how do EF-S lenses "perform better" (or at least have the potential to) on smaller sensors?

--
-----
Neil C
http://www.homelands.me.uk/gallery/

 

[Eugene Powers]

I see now what you mean but I have to disagree with you.

Heh! : )

To me the proper test would be this: one lens 16-35mm in this case
at the same focal length and same aperture pointed at the brick
wall. This way DOF is taken out of equation. So the only difference
in this test is the camera and that is exactly what we are trying
to compare. The Camera.

But, that test does not reflect real life usage of the lens. Who
takes pics where DOF is not in the equation? I mean, lots of
people don't think about DOF, but DOF is a factor in the pic,
pretty much every single time. Except for those that just take
pics of flat objects, such as brick walls. : )

Well, most cars are tested for all kinds of things but in real life most people don't drive them at maximum speed or maximum cornering.

I'll agree that the "brick wall test" is a good test, but you must
still stop the lens down 1 1/3 stops on the FF camera to make it
valid. However, the tests that you were saying were invalid, were
of books on a bookshelf, which is, of course, the same thing as
shooting a brick wall.

Invalid because two totally different lenses were used.

I still don't understand why FF camera needs to be stopped down.
DOF is out so the same aperture should be used.

On the side, let me address the issue of using the same lens. I
see now what you mean about copy variation. Let's say we were
using the 24-105 / 4L IS on the 5D and 17-55 / 2.8 IS on the 30D
(an ideal comparison). But let's say that one of the lenses was a
sharp copy whereas the other was a dud. Point taken. When
comparing the two systems with their respective equivalent lenses,
we must first ascertain that both lenses are good copies. I
totally agree. Thus, when a test is performed comparing the two
systems with the two different lenses, we cannot be certain that
the results are conclusive until we know about performance of the
two lenses.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my
photos. If you wish to use any of my photos for any purpose other
than editing in these forums, please ask.

--
Eugene

The only time a smaller sensor with the same pixel count is superior to a larger sensor (aka higher pixel density) is when you are focal-length limited.

Lee Jay

 

[joe mama]

forgive me, as i had never heard of an "f-ratio" until now, which i
believe is some calculation to do with f-stop and something else
(is it focal length divided by physical aperture?)

An "f-ratio" is the aperture divided by the focal length. For example, a 100mm lens with an f-ratio of f / 4 will have an aperture of 25mm.

I had no idea that the physical aperture size was different on each
camera... and to be honest, for the general photographer, i'm not
sure it's something you need to know

The physical aperture is not different. Here's the deal. Let's say you are using 100mm on 1.6x. Then you would use 160mm for the same framing on FF. If you had an f-ratio of f / 4 with the 100mm lens on the 1.6x camera, the physical aperure measures 25mm. To get the same physical aperture with the 160mm FL of the FF camera, you need to use f / 6.4.

I mean, calculations are all well and good, but at the end of the
day you can be a fantastic photographer without knowing exactly how
everything is calculated and works out

Absolutely. But when you are comparing two systems, you need to be aware of these differences in order to make a valid comparison. If you're just out there taking good pics, then who cares? : )

here's another question for you:

there's some other argument going round about how cropped cameras
are sharper at the edges because they use a 'sweeter' part of the
lens

from what we've learnt here, we realise that the smaller sensors
are not just a "crop" of the image at all (i.e. there is no wasted
light falling outside of the sensor area), but it is simply the
SAME image but seen from further away (with the same framing)

Cropped cameras are a crop of the total image circle. But the whole of the glass in the lens is used to create all parts of that circle, both the part the falls on the sensor and the part that lies off the sensor.

therefore the lens itself is not at a "sweeter" point at all,
because it is not the case that only part of the lens is being used
... is this correct?

The cropped camera is using the "sweet part" of the image circle, but the FF camera is stopped down for an equivalent image, so this balances out at the edges, but actually gives FF a significant advantage in the center.

i.e. if on full frame, the whole of a 100mm lens "is used"

on a 50% smaller sensor, a 100mm lens is NOT using just 50% of the
lens

No. The whole of the lens is used for the whole of the image. It's just that the image quality is higher in the center and degrades as you move towards the edges, and a cropped camera uses the central part. But, again, as I said above, this only gives 1.6x the edge at the same f-ratio. Since you use 1.6x the f-ratio on FF, the FF image circle is actually equal on the edges, but better in the center.

OR, is it the case that yes, only part of a 100mm lens is "being
used" on a crop camera, because the physical size of the aperture
does not allow light to be collected from parts of the lens

No.

i.e. on a 50% smaller sensor, because of the physical aperture
size, only 50% of the lens elements are used (the outside-most 50%
is presumably not used) ?

so, which is true?

The whole of the lens is used to create the whole of the image. It is not true that the center of the lens creates the center of the image and the edges of the lens creates the edges of the image.

and my main question - irrespective of which is true, what is the
argument put forward for EF-S lenses? how do EF-S lenses "perform
better" (or at least have the potential to) on smaller sensors?

Here ya go:

http://forums.dpreview.com/forums/read.asp?forum=1029&message=23311239

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[Lee Jay]

I still don't understand why FF camera needs to be stopped down.
DOF is out so the same aperture should be used.

Right...that's why - to keep the aperture (not the f-stop) the same.

--
Lee Jay
(see profile for equipment)

 

[joe mama]

But, that test does not reflect real life usage of the lens. Who
takes pics where DOF is not in the equation? I mean, lots of
people don't think about DOF, but DOF is a factor in the pic,
pretty much every single time. Except for those that just take
pics of flat objects, such as brick walls. : )

Well, most cars are tested for all kinds of things but in real life
most people don't drive them at maximum speed or maximum cornering.

I'm missing your point.

I'll agree that the "brick wall test" is a good test, but you must
still stop the lens down 1 1/3 stops on the FF camera to make it
valid. However, the tests that you were saying were invalid, were
of books on a bookshelf, which is, of course, the same thing as
shooting a brick wall.

Invalid because two totally different lenses were used.

But if both lenses have equal performance, why do you feel that is invalid? It is completely valid. Even if the lenses were not equal, then, barring a defective lens, so be it. Whichever lens is better speaks to the advantage of the format that lens is being used on.

I still don't understand why FF camera needs to be stopped down.
DOF is out so the same aperture should be used.

Because we are talking about equivalent images. DOF is not out -- DOF is central . I mean, if you take DOF out of the equation, you are taking f-ratio out of the equation, so I may as well just put the camera on whatever f-ratio I please.

To that end, if you wish to compare the two systems at their maximum sharpness, that is, set the f-ratio on each system to whatever f-ratio delivers the best image, then I'll agree to that, although by doing that, you are saying that all you care about is sharpness, and that DOF is irrelevant to you, which, methinks, is absurd.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[NeilCUK]

An "f-ratio" is the aperture divided by the focal length. For
example, a 100mm lens with an f-ratio of f / 4 will have an
aperture of 25mm.

surely that's 100/4 - focal length / aperture?

f/ratio = focal length / aperture from http://www.northern-stars.com/tel_magnification.htm

I had no idea that the physical aperture size was different on each
camera... and to be honest, for the general photographer, i'm not
sure it's something you need to know

The physical aperture is not different. Here's the deal. Let's
say you are using 100mm on 1.6x. Then you would use 160mm for the
same framing on FF. If you had an f-ratio of f / 4 with the 100mm
lens on the 1.6x camera, the physical aperure measures 25mm. To
get the same physical aperture with the 160mm FL of the FF
camera, you need to use f / 6.4.

Getchya, thank you

... and this explains why DoF is smaller for FF cameras for an equivalent field of view i assume

Cropped cameras are a crop of the total image circle. But the
whole of the glass in the lens is used to create all parts of
that circle, both the part the falls on the sensor and the part
that lies off the sensor.

ok... so with a smaller sensor, there is wasted light that falls outside the sensor area?

The cropped camera is using the "sweet part" of the image circle,
but the FF camera is stopped down for an equivalent image, so this
balances out at the edges, but actually gives FF a significant
advantage in the center.

gotchya

presumably, stopping down gives a similar DoF also ?

and my main question - irrespective of which is true, what is the
argument put forward for EF-S lenses? how do EF-S lenses "perform
better" (or at least have the potential to) on smaller sensors?

Here ya go:

http://forums.dpreview.com/forums/read.asp?forum=1029&message=23311239

ta

--
-----
Neil C
http://www.homelands.me.uk/gallery/

 

[joe mama]

An "f-ratio" is the aperture divided by the focal length. For
example, a 100mm lens with an f-ratio of f / 4 will have an
aperture of 25mm.

surely that's 100/4 - focal length / aperture?

f/ratio = focal length / aperture from
http://www.northern-stars.com/tel_magnification.htm

Either way. On many lenses, you will see the minimum f-ratio written as 1 : 2.8 instead of 2.8 : 1.

The physical aperture is not different. Here's the deal. Let's
say you are using 100mm on 1.6x. Then you would use 160mm for
the same framing on FF. If you had an f-ratio of f / 4 with the 100mm
lens on the 1.6x camera, the physical aperure measures 25mm. To
get the same physical aperture with the 160mm FL of the FF
camera, you need to use f / 6.4.

Getchya, thank you

... and this explains why DoF is smaller for FF cameras for an
equivalent field of view i assume

Smaller for the same FOV and f-ratio.

Cropped cameras are a crop of the total image circle. But the
whole of the glass in the lens is used to create all parts of
that circle, both the part the falls on the sensor and the part
that lies off the sensor.

ok... so with a smaller sensor, there is wasted light that falls
outside the sensor area?

Of course! Even with a FF camera. I mean, the lens casts an image circle , not a rectangle. However, a 1.6x sensor has only 39% the area of a FF sensor, so a lot more light falls outside a 1.6x sensor than a FF sensor.

The cropped camera is using the "sweet part" of the image circle,
but the FF camera is stopped down for an equivalent image, so this
balances out at the edges, but actually gives FF a significant
advantage in the center.

gotchya

presumably, stopping down gives a similar DoF also ?

If you stop the FF lens down 1 1/3 stops, then the DOF is identical.

ta

I'm here to serve. : )

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[joe mama]

I still don't understand why FF camera needs to be stopped down.
DOF is out so the same aperture should be used.

Right...that's why - to keep the aperture (not the f-stop) the same.

I overlooked that part of Eugene's post. Once again, to put some numbers to it:

30D at 100mm, f / 4: aperture = 100mm / 4 = 25mm.
5D at 160mm, f / 6.4: aperture = 160mm / 6.4 = 25mm.

Same FOV, same aperture, same DOF. Different FLs and f-ratio.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

Please feel free to criticize, make suggestions, and edit my photos. If you wish to use any of my photos for any purpose other than editing in these forums, please ask.

 

[Eugene Powers]

--
Eugene

The only time a smaller sensor with the same pixel count is superior to a larger sensor (aka higher pixel density) is when you are focal-length limited.

Lee Jay

 

[Lee Jay]

f-stop (f-number) = focal length / aperture

Solving for aperture

aperture = focal length / f-stop

When you change formats and keep the field of view constant (constant framing) you change focal length. When you change focal length you change f-number to keep constant aperture. If you keep constant f-stop, you change aperture, and aperture (not f-stop) is what controls DOF and the total amount of light the sensor receives.

--
Lee Jay
(see profile for equipment)