# Should 4/3 lenses be half the size of full frame lenses?

Discussion in 'Digital Cameras' started by bob, Jun 13, 2011.

1. ### bobGuest

4/3 camera has a sensor that is about half the length and height of a full
frame sensor (1/4 the area).

Does that mean a 4/3 camera lense would be half the diameter of an
equivalent full frame lense?

bob, Jun 13, 2011

2. ### David J TaylorGuest

... and half the length, and one eighth of the weight! Doesn't seem to be
so in practice, though. You also use the word "equivalent". To be
equivalent in light gathering power an f/2.8 full-frame lens would need to
be replaced by an f/1.4 half-frame lens. Somewhat more costly to make, if
possible at all.

Cheers,
David

David J Taylor, Jun 13, 2011

3. ### David J TaylorGuest

[]
Yes, perhaps the OP might like to consult the archives.
Which is why I asked the OP to clarify what they meant by "equivalent".
For the same number of photons on the sensor, what I say is correct.

I haven't done the depth of field sums - that was likely Roger Clarkson.

Cheers,
David

David J Taylor, Jun 13, 2011
4. ### David J TaylorGuest

[]
Lenses with the same f/number (more strictly, T/number) will deliver the
same number of photons per unit area, but as the full-frame sensor has
four times the sensitive area of the half-frame sensor, it gathers four
times as much light - four times as many photons. To get the same number
of photons, you need a lens with twice the physical opening, four times
the aperture area, and one half the numeric f/number.

It's why small sensor cameras need a higher light level to get the same
signal-to-noise ratio or, put another way, why they are noisier at higher
ISOs. Their smaller sensor captures fewer photons for a given f/number
and light level.

David J Taylor, Jun 13, 2011
5. ### David Dyer-BennetGuest

You're assuming the same pixel count, which is regrettably frequently
true, but is not an inherent aspect of sensor size. I think a lot of
the
confusion arises from your relating the issue to sensor size
rather than pixel size (okay, sensel size).

Same number of photos per unit area is what matters for exposure.
"Everybody knows" this and it's true, as true as anything is anyway.
Your statements always start out looking like they're trying to
deny this, which gets everybody up in arms, because this really
*is* true.
Hmmm; this feels like a back-door way of coming at the problem, and
also
doesn't give me any help in minimzing noise -- both my cameras have
a base ISO of 200, so I can't turn down the ISO in either case.

But one of them is a Nikon D700, and one is an Olympus E-PL2. I know
which one is noisier at high ISO! It's the one with the smaller
pixels,
which in this case is the one with the smaller sensor. (They're also
the
same number of pixels, so that makes all sorts of comparisons easy.)

Signal-to-noise ratio isn't a stated spec or anything we have a
standard
for measuring in cameras (that's widely used and understood, anyway).

It's generally true that, for any given technology, smaller pixels
will be
noisier at ANY ISO (than larger pixels at that same ISO). Won't
argue
against that for a moment.

But despite having been around this course before, I still wasn't
able to read that out of your initial statements. It'd maybe be good
to work on how you present this issue, to avoid the initial confusion
and opposition.

David Dyer-Bennet, Jun 13, 2011
6. ### David J TaylorGuest

[]
Pixel count actually has nothing to do with it. It's the sensor area
which matters. Pixel count sets how you trade-off spatial resolution for
signal-to-noise ratio at the pixel level.
Yes, it is true.
More sensitive area reduces noise at the same ISO when more photons are
captured, other things being equal.
But a newer 4/3 sensor, one with higher quantum efficiency, might just
beat an older APS-C sensor. That's when all things /aren't/ equal.
It takes time to think about these issues, as they are not initially
obvious.

Cheers,
David

David J Taylor, Jun 14, 2011
7. ### bobGuest

To me, an 4/3 lense that is equivalent to a 35mm lense should have the same
maximum f number and angle of view.

This way, a 35mm camera and a 4/3 camera can shoot the same scene using the
same exposure (shutter speed and aperture) and end up with more or less
identical photos (except DOF, noise, and maybe resolution).

bob, Jun 14, 2011
8. ### David J TaylorGuest

Yes, that's a perfectly fair way of looking at it - depending on your
definition of "identical"! Given that, why aren't the lenses half the
linear dimension, a quarter of the area, and an eighth of the weight,
particularly for micro-4/3 where back-focus requirements are less?

Cheers,
David

David J Taylor, Jun 14, 2011
9. ### Neil EllwoodGuest

The glass might scale but the mounting bits containing the glassware do
not scale, look at scale models of loco engines, wagons and carriages and
aircraft for an instance or more.

Neil Ellwood, Jun 14, 2011
10. ### Paul FurmanGuest

With 1/4 the sensor size so 1/4 the print size.

Paul Furman, Jun 14, 2011
11. ### BruceGuest

That's only true if the Four Thirds sensor has a quarter as many
pixels as the full frame sensor.

If the number of pixels is the same, the print size will be the same
for the same ppi at the printing stage. All that matters at the
printing stage is the number of pixels. The printer has no idea
whether those pixels came from a P&S, Four Thirds, APS-C, full frame
or medium format.

Bruce, Jun 14, 2011
12. ### Neil HarringtonGuest

That makes no sense whatever.

An f/2.8 lens is an f/2.8 lens, regardless of what format it's meant to
cover. Its "light gathering power" is the same as any other lens of similar
f-number as far as photographic purposes are concerned, disregarding
differences in coating, etc.

Even if you were right about that light-gathering business you'd be wrong in
your conclusion: a half-frame format does not have half the diagonal of full
frame. (It's 30mm vs about 43mm.)

Neil Harrington, Jun 14, 2011
13. ### GuestGuest

yes it does.
it's light gathering power per unit area is the same (exposure) but a
larger sensor has a larger area so the total amount of light collected
is higher.
4/3rds is approximately 1/4 the area of full frame, or two stops, thus
an f/2.8 lens on full frame is equivalent to an f/1.4 lens on 4/3rds.

Guest, Jun 14, 2011
14. ### Neil HarringtonGuest

Which is all that matters, so you could stop right there.
Why would you need or want four times as many photons if you're going to
throw 3/4 of them off the sensor? If instead you concentrate them all on the
quarter-size sensor, then you've gained two full stops of exposure. So how
on earth do you make that "equivalent" to the different f-stop.

Regardless of format size, f/2.8 is f/2.8. If we needed to concern ourselves
with physical aperture size, we wouldn't need f-stops in the first place.
It's the *ratio* of focal length to aperture size that matters, hence the
f-number.
Regardless of format size, *per unit area* the sensor captures the same
number of photons at the same f-number.

Neil Harrington, Jun 14, 2011
15. ### GuestGuest

for exposure yes, for total light no.
but the area is bigger, so the total is higher.

Guest, Jun 14, 2011
16. ### Neil HarringtonGuest

Yes. Stop right there.
You didn't stop where I told you. Larger sensor = larger area = more light
collected = *same exposure*, no difference as long as the f-number is the
same.
*Whoa!* where do you get the "or two stops" business?
Absolutely not. f/2.8 on full frame is equivalent to *f/2.8* on 4/3, on
medium format, on Minox, on roll film, on 16mm, on Super 8, or on anything
else. f/2.8 is f/2.8.

Neil Harrington, Jun 14, 2011
17. ### GuestGuest

that's because there's more to say.
however, more light means better signal/noise ratio.

if you match the noise to obtain the *same* image quality (by raising
the iso) you can use smaller f/stops on the larger sensor. that's where
the difference is.

a good explanation is here:
1/4 the area is 1/4 the light.

one stop is half the light and two stops is 1/4. therefore, a sensor
that has 1/4 the area is a two stop difference.

Guest, Jun 14, 2011
18. ### dj_nmeGuest

You really should have stopped with "For exposure, yes."
That is "total light" as far as taking photographs is concerned.
So what?
The same f-stop setting gives the same exposure at the same shutter
speed regardless of format size.

You seem to have not grasped the reason why some people use the term
"equivalent f-stop" when taking 4/3, APS and FF.

Using FF as the baseline, all the others have greater Depth of Field
(DoF) with lenses of the same angle of view (IE: 35mm Equivalent Focal
Length [35mm EFL]) with the same f-stop setting.
"equivalent f-stop" is using a aperture (f-stop) setting that gives an
equivalent DoF to 35mm at their 35mm EFL, not the same exposure value at
the same shutter speed.
To get the same DoF on a 4/3 as a FF the f-stop setting must be twice
the size (two f-stops); IE: set the 4/3 to f/1.4 @ 25mm FL (50mm in 35mm
EFL) to get the same Dof as FF set to f/2.8 @ 50mm FL.
For the same DoF on APS as FF the f-stop setting must be 50 percent
bigger (one f-stop); IE: set the APS to f/2.0 @ 35mm FL (50mm in 35mm
EFL) to get the same Dof as FF set to f/2.8 @ 50mm FL.
It gives a much greater exposure setting at the same shutter setting:
the shutter speed must be set higher (shorter time) to compensate for
more light let in by the wider (lower f-stop setting) aperture.

dj_nme, Jun 14, 2011
19. ### dj_nmeGuest

Of course it makes no sense: it is dead wrong.
If "bob" wrote that for the same 35mm equivalent FL that he was after
the same DoF, then two stops wider aperture (for 4/3, in this case) will
get the same DoF as FF (35mm full frame) at the same angle of view lens
(35mm equiv focal length).
It does not work for exposure, only DoF.
In this case (for 4/3), it is really a quarter-frame format (about 22mm
Not that I really believe that the facts will "get in the way" of "bob"s
argument.

dj_nme, Jun 14, 2011
20. ### GuestGuest

nope. exposure is light per unit area. if you have more area (a larger
sensor), you have more total light distributed over the larger area,
which ends up being the same for a given area.
but not the same result. the noise and depth of field are different.
actually you haven't grasped it, since you contradict yourself below.
true, but smaller formats will have more noise (all other things being
equal) so you don't get the same image.

right, which for 4/3rds, is 2 stops wider.
you can also raise the iso, which has the effect of an equivalent s/n
ratio, and then you can use the same shutter speed with the equivalent
f/stop.

that contradicts what you said earlier.

Guest, Jun 14, 2011