Lens musings - Prosumer vs. dSLR

[Tul]

mm, not sure I'm following you there. I agree f/2.8 means the same amount of light per area at sensor, but as one 10D pixel is 6 times bigger than one 717 pixel, it will recieve 6 times as much light at equal f-stop.

Wrong. Sure you have less noise because of a larger sensor, but
much more light will also hit the sensor, meaning more light for
each pixel. So the Sony sensor will gather less light, as the
sensor area is only 1/6 of the Canons.

Perhaps I should clarify.

f/2.8 is f/2.8 REGARDLESS of the lens. A smaller lens is mated to a
smaller sensor or piece of film, hence the EV value at the surface
of the sensor remains the same.

Yes, the 10D sensor has more area, and yes, that's related to S/N.
But the EV is the same for a given area of sensor real estate. If
through some technology the Sony sensor achieved as high a S/N as
the 10D sensor, then the Sony would have a 1 stop advantage (over
my 70-200 f/4) regardless of how much total light was hiting my 10D
sensor.

The future may bring such technology, but right now sensor size is
related directly to S/N.

 

[Tul]

But then the sensor pixel will be half the size, and recieve half as many photons as the larger sensor, thus it will be half as sensitive.

Having read much about how prosumer cams have advantages of small
fast lenses, but very long DOF, I sat down and made a few
calculations on the topic, this is my conclusion:

While the speed of the lens does limit the DOF, there is no
advantage in lens speed/size. A 10D with a lens stopped to, say,
f/4.9 will be exactly as sensitive as a sony F-717 at f/2.0, and
with the same DOF. (at same capture angle)
Because the 717 sensor is smaller it gathers less light and needs a
faster lens, with a shorter focal length. A 10D with a larger
sensor can capture the exact same light using a slower lens with a
longer focal length, which gives exactly the same DOF.

Makes sense to anyone?

--
Some digital cameras, some lenses, 2 eyes

http://www.tom-crowning.com

 

[Tom Crowning]

You have never mentioned pixel size in your original posting.

If you reduce the size of each pixel sensivity decreases, but this is NOT dependent on the sensor size!
Regards,
tc

Having read much about how prosumer cams have advantages of small
fast lenses, but very long DOF, I sat down and made a few
calculations on the topic, this is my conclusion:

While the speed of the lens does limit the DOF, there is no
advantage in lens speed/size. A 10D with a lens stopped to, say,
f/4.9 will be exactly as sensitive as a sony F-717 at f/2.0, and
with the same DOF. (at same capture angle)
Because the 717 sensor is smaller it gathers less light and needs a
faster lens, with a shorter focal length. A 10D with a larger
sensor can capture the exact same light using a slower lens with a
longer focal length, which gives exactly the same DOF.

Makes sense to anyone?

--
Some digital cameras, some lenses, 2 eyes

http://www.tom-crowning.com

--
Some digital cameras, some lenses, 2 eyes

http://www.tom-crowning.com

 

[Jason Hutchinson]

You have never mentioned pixel size in your original posting.
If you reduce the size of each pixel sensivity decreases, but this
is NOT dependent on the sensor size!

Actually, the pixel size doesn't matter much. The sensitivity is more based on the sensor size, which determines the number of photos actually collected. This is why a 1Ds has less noise than the big pixeled 1D for most, if not all the ISOs it is capable of.

Jason

 

[Godfrey]

James,

• An aperture is a hole of a particular size and position in the lens through which light passes. The size of the hole relative to the focal length allows a certain amount of light to pass over a reference amount of time.

• An f/number is a way of expressing that aperture size such that the light transmission through the lens is a standard value, equivalent across different focal lengths, useful in establishing exposure values.

An f/number is calculated by dividing the focal length by physical size of the aperture ... so a 50mm f/2 lens has a maximum aperture of 25mm, a 100mm f/2 has a maximum aperture of 50mm, a 10mm f/2 lens has a maximum aperture of 5mm. Don't think of an f/number as a fraction, however: think of it rather as a number that allows you to know how much light the lens will pass. In all three cases above, setting the lens to f/2 will pass the same amount of light to the film or sensor, per unit area. Larger sensors with more area collect more energy, they operate at quieter amplification levels and have better signal to noise performance.

Depth of Field is most strongly influenced by the physical aperture size and focus distance setting. It's not the f/number or focal length that's important, it's the size of the aperture. One just uses the f/number and focal length to get the aperture size for DoF calculations ... If you look at the equations for DoF, you'll see that the focal length cancels out and the f/number reduces to the aperture size.

Godfrey

I thought that aperature was actually a fraction of focal length.
(i.e. f5 is actually 1/5th the focal length). This means that with
both lenses at f2.0, each lens is gathering 1/2 the light that it
is physically able to collect at each given focal length

If this is true, then what both of you say is true. the sony's lens
at 2.0 will not gather as much light as a canon lens at 2.0 because
it does not need to (nor is it physically able to).

 

[Godfrey]

So what was his point?

Godfrey

There's no real need to keep hashing this around. If there weren't
some advantage to a large sensor, much more expensive camera, no
one would be interested in them.

Godfrey

...Some people think that to equal a 717 lens on a 10D you'd need a
24-120mm f/2.0-2.4, but this is not true. A 24-120 f/5 would be
more like it.

 

[Tul]

So what was his point?

Godfrey

My point was that you dont need a 24-120mm f/2.0-2.4 lens to equal a 717 with a D10, a 24-120 f/5 will give same shutter speeds(for a given ISO) and same DOF.

 

[Tul]

You have never mentioned pixel size in your original posting.
If you reduce the size of each pixel sensivity decreases, but this
is NOT dependent on the sensor size!

Actually, the pixel size doesn't matter much. The sensitivity is
more based on the sensor size, which determines the number of
photos actually collected. This is why a 1Ds has less noise than
the big pixeled 1D for most, if not all the ISOs it is capable of.

Jason

No, reverse that one, a larger pixel will gather more light and thus be more sensitive - proportional to pixel area. Thus to reach ISO 100, less amplification will be needed giving less noise.

 

[Godfrey]

How's that? If I have ISO 100 set on a 10D and an F828, and the exposure requires f/2@1/30 second, you will need that same exposure on both cameras.

If you set the 10D to ISO 400, then the f/4 lens will do at the same aperture/shutter speed setting, but you're not getting the additional quality possible with the 10D. You might be getting equivalent quality, or even a little better due to the Canon's large sensor/low noise characteristic, but not the same as you get at ISO 100.

Godfrey

My point was that you dont need a 24-120mm f/2.0-2.4 lens to equal
a 717 with a D10, a 24-120 f/5 will give same shutter speeds(for a
given ISO) and same DOF.

 

[Tul]

My bad, I shouldn't have mixed ISO into this. I keep losing sight of the fact that an ISO# is a fixed sensitivity...

I guess what I mean is that each pixel recieves the same amount of light when the 717 is at f/2 and the 10D is at f/5. Thus to achieve ISO 100 the 717 needs 6x more amplification.

If you set the 10D to ISO 400, then the f/4 lens will do at the
same aperture/shutter speed setting, but you're not getting the
additional quality possible with the 10D. You might be getting
equivalent quality, or even a little better due to the Canon's
large sensor/low noise characteristic, but not the same as you get
at ISO 100.

Godfrey

My point was that you dont need a 24-120mm f/2.0-2.4 lens to equal
a 717 with a D10, a 24-120 f/5 will give same shutter speeds(for a
given ISO) and same DOF.

 

[Sundance138872]

Depth of Field is most strongly influenced by the physical aperture
size and focus distance setting. It's not the f/number or focal
length that's important, it's the size of the aperture. One just
uses the f/number and focal length to get the aperture size for DoF
calculations ... If you look at the equations for DoF, you'll see
that the focal length cancels out and the f/number reduces to the
aperture size.

--
Focal length is probably the most important factor regarding DOF.

The reason there is so much difference in DOF between a Sony DSC F717 and a 10D with equal focal length (38-190 mm measured up to 35mm film) is because there is a huge difference in actual focal length
Sony = 9,5-47,5 mm
10D = 24-119 mm
By thr way; f/number IS the size of the aperture...

Sundance

 

[Shay Stephens]

The light meter and resulting underexposures does not agree with this theory. The only thing that does apply is the DOF. The DOF with an APS type sensor will be shallower than the 717. But the sensitivity is not greater.

The exposure needed to capture a scene will be the same with the 717 as with the 10D. If the scene called for f/2 and 1/1000 at ISO 100, and you shot the 10D at f/4.9 you would be almost 3 stops underexposed.

You could bump the ISO up (f/4.9 - 1/1000 - ISO 800) to get the shot properly exposed, and it would probably be pretty close noise wise to the ISO 100 of the 717.

--
'Ask not what your camera can do for you, but what you can do with your camera'.

 

[Godfrey]

Right.

The fact is that difference is totally invisible to the photographer, who works in ISO-aperture-shutter time values for exposure, and format-aperture-focus setting values for DoF.

The two cameras are fundamentally different in their imaging qualities, do different jobs, and are not comparable. However, both are excellent cameras and can make superb photographs, within their design constraints.

Godfrey

I guess what I mean is that each pixel recieves the same amount of
light when the 717 is at f/2 and the 10D is at f/5. Thus to achieve
ISO 100 the 717 needs 6x more amplification.

If you set the 10D to ISO 400, then the f/4 lens will do at the
same aperture/shutter speed setting, but you're not getting the
additional quality possible with the 10D. You might be getting
equivalent quality, or even a little better due to the Canon's
large sensor/low noise characteristic, but not the same as you get
at ISO 100.

Godfrey

My point was that you dont need a 24-120mm f/2.0-2.4 lens to equal
a 717 with a D10, a 24-120 f/5 will give same shutter speeds(for a
given ISO) and same DOF.

 

[Tul]

Yup, that's pretty much what it comes down to...If the sensors have similar characteristics, bumbing the ISO 6x higher on the 10D should then give similar amplification and thus noise on both cameras. (Sensitivity is not greater because the 717 amplifies much more to reach ISO 100).

So...At the same 'noise point' (so to speak) the 10D will be much more sensitive. If it is then stopped down to achieve same sensitivity as the 717 again, the DOF will be similar on both cams.

The light meter and resulting underexposures does not agree with
this theory. The only thing that does apply is the DOF. The DOF
with an APS type sensor will be shallower than the 717. But the
sensitivity is not greater.

The exposure needed to capture a scene will be the same with the
717 as with the 10D. If the scene called for f/2 and 1/1000 at ISO
100, and you shot the 10D at f/4.9 you would be almost 3 stops
underexposed.

You could bump the ISO up (f/4.9 - 1/1000 - ISO 800) to get the
shot properly exposed, and it would probably be pretty close noise
wise to the ISO 100 of the 717.

--
'Ask not what your camera can do for you, but what you can do with
your camera'.

 

[peteb]

Interesting to read the technical data on this, but I'm surprised Canon managed to engineer a 67.5mm aperture into the 135mm f2. At it's widest it is only 84mm dia, and I understand the aperture tends to be close(r) to the camera end, where the barrel dia. is significantly less than 84mm. Quite some engineering...

• An aperture is a hole of a particular size and position in the

lens through which light passes. The size of the hole relative to
the focal length allows a certain amount of light to pass over a
reference amount of time.

• An f/number is a way of expressing that aperture size such that

the light transmission through the lens is a standard value,
equivalent across different focal lengths, useful in establishing
exposure values.

An f/number is calculated by dividing the focal length by physical
size of the aperture ... so a 50mm f/2 lens has a maximum aperture
of 25mm, a 100mm f/2 has a maximum aperture of 50mm, a 10mm f/2
lens has a maximum aperture of 5mm. Don't think of an f/number as a
fraction, however: think of it rather as a number that allows you
to know how much light the lens will pass. In all three cases
above, setting the lens to f/2 will pass the same amount of light
to the film or sensor, per unit area. Larger sensors with more area
collect more energy, they operate at quieter amplification levels
and have better signal to noise performance.

Depth of Field is most strongly influenced by the physical aperture
size and focus distance setting. It's not the f/number or focal
length that's important, it's the size of the aperture. One just
uses the f/number and focal length to get the aperture size for DoF
calculations ... If you look at the equations for DoF, you'll see
that the focal length cancels out and the f/number reduces to the
aperture size.

Godfrey

I thought that aperature was actually a fraction of focal length.
(i.e. f5 is actually 1/5th the focal length). This means that with
both lenses at f2.0, each lens is gathering 1/2 the light that it
is physically able to collect at each given focal length

If this is true, then what both of you say is true. the sony's lens
at 2.0 will not gather as much light as a canon lens at 2.0 because
it does not need to (nor is it physically able to).

 

[Sundance138872]

Interesting to read the technical data on this, but I'm surprised
Canon managed to engineer a 67.5mm aperture into the 135mm f2. At
it's widest it is only 84mm dia, and I understand the aperture
tends to be close(r) to the camera end, where the barrel dia. is
significantly less than 84mm. Quite some engineering...

--

Not entirely sure about this, but I think it refers to the physical length of the lens(112mm), not the nominal.

Sundance

 

[Godfrey]

Sundance,

You might have read what I wrote but you didn't understand it at all.

F/numbers are not the same as aperture size... f/numbers are a transmission normalized way of expressing aperture size. You can see this for yourself easily ... f/2 on a 50mm lens opens an aperture 1/2 the size that f/2 on a 100mm lens, just set one of each lens on the camera to f/2 and look into the lens while the shutter is open, you can see the difference.

A lens with focal length=15mm and a lens with focal length=200mm will produce EXACTLY THE SAME DoF if they are focused on the same distance setting and if the physical aperture size is the same, presuming the CoC value is identical. A 15mm lens at f/3.5 has a PHYSICAL aperture size of 4.29mm ... a 200mm lens with a PHYSICAL aperture size of 4.29mm would be set to f/46.

Godfrey

Depth of Field is most strongly influenced by the physical aperture
size and focus distance setting. It's not the f/number or focal
length that's important, it's the size of the aperture. One just
uses the f/number and focal length to get the aperture size for DoF
calculations ... If you look at the equations for DoF, you'll see
that the focal length cancels out and the f/number reduces to the
aperture size.

--
Focal length is probably the most important factor regarding DOF.
The reason there is so much difference in DOF between a Sony DSC
F717 and a 10D with equal focal length (38-190 mm measured up to
35mm film) is because there is a huge difference in actual focal
length
Sony = 9,5-47,5 mm
10D = 24-119 mm
By thr way; f/number IS the size of the aperture...

Sundance

 

[Tul]

a 135mm f/2 lens will have a front element with a diameter of at least 67,5mm. If it is larger, it will be limited further into the lens to maintain f/2. Aperture limiting has to be done in the frequency-plane, otherwise vignetting will occur. This means that using the size of the front element is a perfect limiter for max. aperture, so it very likely is 67,5mm on the 135 f/2.

Interesting to read the technical data on this, but I'm surprised
Canon managed to engineer a 67.5mm aperture into the 135mm f2. At
it's widest it is only 84mm dia, and I understand the aperture
tends to be close(r) to the camera end, where the barrel dia. is
significantly less than 84mm. Quite some engineering...

--
Not entirely sure about this, but I think it refers to the physical
length of the lens(112mm), not the nominal.

Sundance

 

[David Fey]

A very large % of picture takers would never, ever carry several pounds worth of camera and lenses around with them. Smaller, lighter dslr lenses is only the tip of the iceberg.

I see often people trivialize the extra weight.

$.02

 

[Rob Patterson]

The objective diameter of a lens determines how much light will be gathered. This is why the 200 inch Mount Palomar Newtonian reflector was built.

The lens designation f/2.8 is a reciprocal expression. It's the effective focal length of a lens divided by the objective diameter. This is why a 50mm f/1.8 has a smaller diameter than a 50mm f/1.2. The 50mm f/1.2 has an extremely large diameter and much more light gathering power.

If you've never had the pleasure of shooting with an f/1.2 lens as opposed to the same focal length with an f/1.8 value you're missing some incredible performance difference. Shots that'd need a tripod with the f/1.8 can be hand held with a f/1.2 in many, many cases.

One of the advantages of an SLR over a fixed lens camera is the ability to obtain these high light gathering, large diameter lenses. They are expensive though!

It's not that the Sony sensor gathers less light, it's that it has
a lower signal to noise ratio, hence it is more noisy at a given
ISO.

There is less signal because there is less light. The cheap Sony
lens is possible because it is smaller. Because it is smaller, it
gathers less light.
--
David Jacobson