Full Frame Fallacy?

[rennie12]

that is, the larger individual photocells made possible by a larger sensor give less noise and a "cleaner" image.

To some degree this can be offset in a sense by better software, etc - but only in a sense, since the same better software applied to a larger sensor will still give better results.

That said, the 35 mm so called full frame is an accident of history, related to no print size, and invariable results in more cropping than needed.

(The original 35mm camera, which morphed - due to demand - into the Leica - was designed as a tester for 35mm motion picture film (so that different runs of film could be "matched" by processing. A still camera made testing batches of film cheaper and much easier.

But the size we know as 35mm full frame is simply TWICE the size of the 35mm motion picture frame of the day - entirely an accidental size.

The ideal larger sensor is probably what is called medium format as per Hasselblad, etc. Next to this the most logical sensor to use with 35mm legacy lenses is between APS and 35mm full frame - the largest between these two that would be a usable format - like 4/3 or 8x10 approximate proportions.

However, of course, this will never happen since the 35 mm full frame has been unoficcialy canonized.
--
bill wilson

 

[Gearóid Ó Laoi Garry Lee]

They are indeed, but the cameras are bigger (and usually heavier) and the lenses are also a lot bigger and heavier.

I have both systems. Olympus and FF canon. If the disadvantages of the Canon were not significant, I would get rid of the Olympus, but I don't and in fact use it more.

 

[Arnold P]

They are indeed, but the cameras are bigger (and usually heavier) and
the lenses are also a lot bigger and heavier.
I have both systems. Olympus and FF canon. If the disadvantages of
the Canon were not significant, I would get rid of the Olympus, but I
don't and in fact use it more.

I did not notice 5D was bigger than E-3. Neither does it appear so with the new prototype from Sony. The imaging benefits are obvious, but not in comparing size and weight of each model.

AP

 

[Arnold P]

huge full frame lenses sometimes match 4/3 lens quality,but will
never be better [nt]

Last time you told us there could be no front/back focusing problems with Olympus cameras and I have lost count since then. Better is a very subjective term and may tell us nothing abut the final quality of the image. Same as huge. What is huge for you may feel like very small and tiny for the other. Try to be more specific in your next posts, if you can.

AP

 

[Ehrik]

Because of smaller and lighter shutter and mirror,
4/3rds can potentially offer

(a)
higher framerate

This would only be true when the larger sensor has more pixels.

As I said initially, I was considering the mass of the mirror. A lighter mirror
can react faster. Especially in a body of constant weight/size. Data
throughput may be the ilmitation. Then you can use a high speed crop
like the D2X.

(b)
quiter operation

Possibly. My understanding is that the Canon 1D series cameras have
a "quite mode" where the shutter is silent, although that almost
definitely comes at the expense of frame rate.

If you don't accept a speed penalty then a smaller mirror can be made to
work quiter.

(c)
less vibrations

Only if the cameras have the same mass.

And if they don't the 4/3rds camera instead has a weight advantage.

(d)
higher flash synch speed

I know nothing about flash. : )

It's related to the speed the shutter blades are moving and smaller blades
can move faster and there is less distance to move. Larger heavier shutters
with a long travel have to resort to the "running gap" already at slower
shutter speeds and then the sensor will not be evenly luminated by the
flash.

The 1D MkI had a high flash synch speed though, thanks to the electronic
shutter of its Matsushita-made CCD. As did some Nikons using CCDs. But CMOS
doesn't seem to have this capability.

(e)
shorter blackout

For the same shutter speed, isn't the blackout time the same? I
don't know.

Blackout is mirror up time + shutter activation time (which can be
longer than the shutter speed) + mirror down time.
All three can be made shorter in a 4/3rds camera because the masses are
less and the distances are shorter.

(f)
more effective sensor IS

To be seen. Go Sony! : )

Sony will not have an easy time to make its SSS work as effectively on FF
as it does on smaller sensors, due to sensor mass. Maybe they will settle
for a stop less.

Did I do well? : ) Of course, as I always say, while it's
interesting to talk about the potential of different systems, when
choosing a system, one is best served by what it can do for you
today.

In which case 4/3rds scores points for the only proven anti-dust filter,
user-upgradable lens firmware, lightest and smallest body and kit lenses,
lightest and smallest antishake body, world's fastest AF(?), world's
fastest contrast detect AF on a DSLR (?), lightest 600mm and 840mm
equiv. setup that still does AF, compatibility with the largest number of
legacy mounts and cheapest weather sealed bodies to name a few.

For example, let's pretend someone actually wanted a FF DSLR.
There choices are the 5D with a slow frame rate and no weather
sealing, or the huge and expensive Nikon D3 or Canon 1Ds. So, while
a weather sealed FF DSLR with a high frame rate could be made with
the dimensions of the 5D, there currently is no such machine, so that
doesn't help. One could then make a compelling argument that the IQ
of the Olympus E3 is easily "good enough" and the better choice for
many, if not most, people.

Ok, so it seems we agree there.

Just my two oere
Erik from Sweden

 

[Jonas B]

They are indeed, but the cameras are bigger (and usually heavier) and
the lenses are also a lot bigger and heavier.
I have both systems. Olympus and FF canon. If the disadvantages of
the Canon were not significant, I would get rid of the Olympus, but I
don't and in fact use it more.

I did not notice 5D was bigger than E-3. Neither does it appear so
with the new prototype from Sony. The imaging benefits are obvious,
but not in comparing size and weight of each model.

And the E-3 and the 5D actually weights the same as well. But then again - I don't see the point in comparing them as they are made for very different uses.

Sticking to the size and weight it is clear to me that if I upgrade to an E-3 (from the E-510) my main kit will be heavier than the corresponding 5D kit. That's possible as I am not interested in lenses over about 100mm (35mm equiv focal length).

FF "a lot bigger and heavier"? Indeed not. I'm a little disappointed about that at the same time as I can understand why it is the way it is. I hope the A900 (?) will be good, and smaller...

For anyone needing really long reach there are advantages with the 4/3 system.

--
Jonas

 

[joe mama]

As I said initially, I was considering the mass of the mirror. A
lighter mirror can react faster. Especially in a body of constant weight/size.

I think the mirror can react as quickly as the engineers want to make it react, but a larger and heavier mirror will have more mirror slap, which will result in more shake unless the body is also heavier. Of course, this issue can be greatly ameliorated by reducing the frame rate with a "quiet shutter" or using "live-view". This I believe to be reasonable, because I'm not sure you often need a high framerate in situations where the shake caused by mirror slap is an issue.

If you don't accept a speed penalty then a smaller mirror can be made to
work quieter.

I've often wondered about that. Does there really need to be a mirror slap at all? Can't the mirror be stopped by another method than hitting a stop?

less vibrations

Only if the cameras have the same mass.

And if they don't the 4/3rds camera instead has a weight advantage.

For sure. I'm not sure why FF DSLRs are so large to begin with -- film DSLRs aren't, and it doesn't seem like they should require much more electronics, certainly not as much as the size differential suggests. I've often wondered if the size of larger sensor cameras has to do with spacing things out more for less electronic interference and better heat dissapation, rather than the actual bulk of the eqipment.

higher flash synch speed

I know nothing about flash. : )

It's related to the speed the shutter blades are moving and smaller
blades can move faster and there is less distance to move. Larger heavier
shutters with a long travel have to resort to the "running gap" already at
slower shutter speeds and then the sensor will not be evenly luminated by
the flash.

What are shutter blades and how do they work? I thought there was only a mirror on a DSLR. There's no reason a big mirror can't move as fast as a small mirror, but, as discussed above, that implies greater mirror slap.

shorter blackout

For the same shutter speed, isn't the blackout time the same? I
don't know.

Blackout is mirror up time + shutter activation time (which can be
longer than the shutter speed) + mirror down time.

Hmm. So the mirror is up, and there's still a delay before the image is captured? That's odd. Why is that?

All three can be made shorter in a 4/3rds camera because the masses are
less and the distances are shorter.

Again, I think a larger mirror can move just as quickly as a smaller mirror, just with more shutter slap.

--
--joe

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

 

[joe mama]

Sony will not have an easy time to make its SSS work as effectively
on FF as it does on smaller sensors, due to sensor mass. Maybe they will
settle for a stop less.

I would disagree with this, as in-lens IS moves elements of different sizes depending on the lens, yet the effectiveness in stops is not affected.

Did I do well? : ) Of course, as I always say, while it's
interesting to talk about the potential of different systems, when
choosing a system, one is best served by what it can do for you
today.

In which case 4/3rds scores points for the only proven anti-dust filter,
user-upgradable lens firmware, lightest and smallest body and kit
lenses, lightest and smallest antishake body, world's fastest AF(?), world's
fastest contrast detect AF on a DSLR (?), lightest 600mm and 840mm
equiv. setup that still does AF, compatibility with the largest number of
legacy mounts and cheapest weather sealed bodies to name a few.

Definitely a lot of points for 4/3, for sure, but I'll take "world's fastest AF" with a grain of salt, for now. : )

For example, let's pretend someone actually wanted a FF DSLR.
There choices are the 5D with a slow frame rate and no weather
sealing, or the huge and expensive Nikon D3 or Canon 1Ds. So, while
a weather sealed FF DSLR with a high frame rate could be made with
the dimensions of the 5D, there currently is no such machine, so that
doesn't help. One could then make a compelling argument that the IQ
of the Olympus E3 is easily "good enough" and the better choice for
many, if not most, people.

Ok, so it seems we agree there.

Yeah, I think technology has reached a point where choosing a DSLR system based on the differences in IQ is really not an important consideration for the vast majority, if only because the vast majority do not print large enough to where the differences in IQ are that significant. In my office, I have tons of pics from 8x12 to 20x30 ranging from compacts to the 5D, and while people lavish praise over the pics, sometimes even citing IQ in particular, no one has ever commented on the IQ of one image being "better" than another, as they are all "good enough" to the point that it is the scene, not the IQ, that matters the most, by far.

On the other hand, the number and range of images I get that are "keepers" grows with the IQ of the equipment, and I am more comfortable to attempt shots I usually would have skipped before.

In a recent thread, someone who's used the D3, 5D, and 40D, commented that the 40D AF was significantly better than the other two systems, so I'm really looking forward to the 5DII for a better AF system alone (the 5D is exceptional with its center AF point in any light, and excellent with the outer AF points in good light, but the outer AF points falter as the light falls). So, I'm thinking D3 noise performance with 40D AF performance might be the ticket. On the other hand, I shoot all primes and Canon offers no primes below 300mm with IS, so Sony's FF system might be something I'll really want to look into, assuming their ultra fast primes match Canon's and their AF and sensor tech is just as good. One thing I really envy about many of the smaller sensor DSLR systems is that they have in-camera IS.

--
--joe

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

 

[Rriley]

I've often wondered about that. Does there really need to be a
mirror slap at all? Can't the mirror be stopped by another method
than hitting a stop?

pneumatic shock absorber, E1 had one, its reputed to be the quietest SLR ever

What are shutter blades and how do they work? I thought there was
only a mirror on a DSLR. There's no reason a big mirror can't move
as fast as a small mirror, but, as discussed above, that implies
greater mirror slap.

Nikon was first with titanium shutter blades, they did that for a reason, less mass moves faster

Again, I think a larger mirror can move just as quickly as a smaller
mirror, just with more shutter slap.

eventually mass momentum shapes the speed that it can react at, the energy imparted and the max speed of the system. Im not sure we are at or near that point in any system yet though
--
Riley

I'd give my right arm to be ambidextrous

 

[Rriley]

I would disagree with this, as in-lens IS moves elements of different
sizes depending on the lens, yet the effectiveness in stops is not
affected.

its in trouble here i think, the majority of OIS lenses are F4 or slower, do you know many F2.8? (not really my bag). Clearly bigger glass means less effective OIS, limit the aperture to F4, and things shape up better

the explanation would go, the mass of the optical moving part (especially for FF) weighs more than a 4/3rds (possibly even APSC) sensor and cradle. The speeds are so high that the ability to react quickly is very much affected. It takes in the area of 50x the power to move a FF sensor at the same rate as a 4/3rds. Glass components may weigh less than a FF sensor, but not less than a 4/3rds sensor

on the same, since Panasonic's Leica D OIS lenses can be fitted to Oly cameras with IS, tests done both ways reveal an advantage to IS over OIS. This was the state of play before E3 and its 5 stops IS
--
Riley

I'd give my right arm to be ambidextrous

 

[joe mama]

I would disagree with this, as in-lens IS moves elements of different
sizes depending on the lens, yet the effectiveness in stops is not
affected.

its in trouble here i think, the majority of OIS lenses are F4 or
slower, do you know many F2.8? (not really my bag). Clearly bigger
glass means less effective OIS, limit the aperture to F4, and things
shape up better

The 70-200 / 2.8L IS is the most "famous" of Canon's IS lenses. In addition, the 300 / 2.8L IS is reportedly the fastest focusing (although it's difficult for me to believe that it's faster than the 100 / 2). There's also the 400 / 2.8L IS, but I've never heard anything about its AF speed one way or another. On its way is the 200 / 2L IS (the only prime below 300mm that would have IS), and it's AF is "supposed" to be quick, so we shall see.

the explanation would go, the mass of the optical moving part
(especially for FF) weighs more than a 4/3rds (possibly even APSC)
sensor and cradle. The speeds are so high that the ability to react
quickly is very much affected. It takes in the area of 50x the power
to move a FF sensor at the same rate as a 4/3rds. Glass components
may weigh less than a FF sensor, but not less than a 4/3rds sensor

Really? A 35mm FF sensor has 4 times the area of a 4/3 sensor, and even if it were twice as thick, that would be 8 times the mass. Is the power required to shake it proportional to the square of the mass?

on the same, since Panasonic's Leica D OIS lenses can be fitted to
Oly cameras with IS, tests done both ways reveal an advantage to IS
over OIS. This was the state of play before E3 and its 5 stops IS

I saw some video about that on you-tube a while ago. I'd really like to see a systematic comparison of the effectiveness of IS of different systems at a variety of FLs. For example, it's oft claimed that IS is less effective at shorter FLs than longer FLs. Let's say that for a particular shooter, the 1 / FL rule applies at 100mm. Some claim that this rule is too generous at, say, 400mm, but overly cautious at 24mm.

But 5 stops seems rather generous. That means that I could take a shot of a static subject at, say, 100mm, 1/100s without IS (which I know I can easily do), and do just as well at 100mm, 1/3s with IS. Seems too incredible to me. Shoot, even two stops would be amazing. And, man, if that Sony FF DSLR has two stops worth of in-camera IS, then, shoot, it's gonna be hard to stay with Canon! : )

--
--joe

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

 

[joe mama]

I've often wondered about that. Does there really need to be a
mirror slap at all? Can't the mirror be stopped by another method
than hitting a stop?

pneumatic shock absorber, E1 had one, its reputed to be the quietest
SLR ever

Now that's a wicked good feature I'd love! Did they keep it in the E3? If not, that was foolish.

What are shutter blades and how do they work? I thought there was
only a mirror on a DSLR. There's no reason a big mirror can't move
as fast as a small mirror, but, as discussed above, that implies
greater mirror slap.

Nikon was first with titanium shutter blades, they did that for a
reason, less mass moves faster

I still don't know what shutter blades are, though. Are they what hold the mirror?

Again, I think a larger mirror can move just as quickly as a smaller
mirror, just with more shutter slap.

eventually mass momentum shapes the speed that it can react at, the
energy imparted and the max speed of the system. Im not sure we are
at or near that point in any system yet though.

For sure. Eventually, the smaller mirror will have the advantage, and I think we can agree that by 60 fps, it will never matter if one is better than the other or not -- would certainly keep the inside of the camera cool, though. : )

But, yeah, I've no idea how much of an engineering challenge it is in terms of the mirror vs the electronics moving all that data.

--
--joe

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

 

[Kurt Petersen]

What are shutter blades and how do they work? I thought there was
only a mirror on a DSLR. There's no reason a big mirror can't move
as fast as a small mirror, but, as discussed above, that implies
greater mirror slap.

Nikon was first with titanium shutter blades, they did that for a
reason, less mass moves faster

I still don't know what shutter blades are, though. Are they what
hold the mirror?

Wow. I can't quite beleive the author of all these long posts and articles on equivalence doesn't know what a shutter is. You aren't pulling our leg are you?
http://en.wikipedia.org/wiki/Focal_plane_shutter

Not trying to flame or ridicule you Joe, just very suprised.

 

[rkhpedersen]

the one which is best for the job.

There is no "one hammer does it all" as there is no "one camera does it all".

 

[Rriley]

The 70-200 / 2.8L IS is the most "famous" of Canon's IS lenses. In
addition, the 300 / 2.8L IS is reportedly the fastest focusing
(although it's difficult for me to believe that it's faster than the
100 / 2). There's also the 400 / 2.8L IS, but I've never heard
anything about its AF speed one way or another. On its way is the
200 / 2L IS (the only prime below 300mm that would have IS), and it's
AF is "supposed" to be quick, so we shall see.

see i knew there was something mom forgot to mention

Really? A 35mm FF sensor has 4 times the area of a 4/3 sensor, and
even if it were twice as thick, that would be 8 times the mass. Is
the power required to shake it proportional to the square of the mass?

mass momentum ^5

I saw some video about that on you-tube a while ago. I'd really like
to see a systematic comparison of the effectiveness of IS of
different systems at a variety of FLs. For example, it's oft claimed
that IS is less effective at shorter FLs than longer FLs. Let's say
that for a particular shooter, the 1 / FL rule applies at 100mm.
Some claim that this rule is too generous at, say, 400mm, but overly
cautious at 24mm.

ive shot at 1/5th sec on 22mm, its bound to be harder when speeds are so low
longer lenses would benefit, thats my take anyway

But 5 stops seems rather generous. That means that I could take a
shot of a static subject at, say, 100mm, 1/100s without IS (which I
know I can easily do), and do just as well at 100mm, 1/3s with IS.
Seems too incredible to me. Shoot, even two stops would be amazing.
And, man, if that Sony FF DSLR has two stops worth of in-camera IS,
then, shoot, it's gonna be hard to stay with Canon! : )

5 stops, look i own the thing and even i dont know
im not into those sorts of chances and havnt had the time to experiment
--
Riley

I'd give my right arm to be ambidextrous

 

[Raist3d]

does gets old in this forum. Very old.

Here's 4.5 stops of real IS with the E-3:

http://forums.dpreview.com/forums/read.asp?forum=1022&message=25665477

Obviously not all shots will be in this high of a result but they do exist and the camera can do it. Olympus also clearly advertises "up to 5 stops" not "you get 5 stops no matter what."

But 5 stops seems rather generous. That means that I could take a
shot of a static subject at, say, 100mm, 1/100s without IS (which I
know I can easily do), and do just as well at 100mm, 1/3s with IS.
Seems too incredible to me. Shoot, even two stops would be amazing.
And, man, if that Sony FF DSLR has two stops worth of in-camera IS,
then, shoot, it's gonna be hard to stay with Canon! : )

Better buy that E-3 now!

--
--joe

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

--
Raist3d (Photog. Student & Tools/Systems/Gui Games Developer)
Andreas Feininger (1906-1999) 'Photographers — idiots, of which there are
so many — say, “Oh, if only I had a Nikon or a Leica, I could make great
photographs.” That’s the dumbest thing I ever heard in my life. It’s
nothing but a matter of seeing, and thinking, and interest. That’s what
makes a good photograph.'

 

[Rriley]

I've often wondered about that. Does there really need to be a
mirror slap at all? Can't the mirror be stopped by another method
than hitting a stop?

pneumatic shock absorber, E1 had one, its reputed to be the quietest
SLR ever

Now that's a wicked good feature I'd love! Did they keep it in the
E3? If not, that was foolish.

yep it was foolish, its still quite quiet though, comparing to 40D anyway

What are shutter blades and how do they work? I thought there was
only a mirror on a DSLR. There's no reason a big mirror can't move
as fast as a small mirror, but, as discussed above, that implies
greater mirror slap.

Nikon was first with titanium shutter blades, they did that for a
reason, less mass moves faster

I still don't know what shutter blades are, though. Are they what
hold the mirror?

i uploaded this for Joseph Wisniewski before



although thats the Copal M8 shutter off a Leica M8, in principle they are all the same. The blades move both up and down to expose the sensor, the faster they move the more shutter speeds you have. I think the M8 is going to 1/4000th, E3 has 1/8000th already. Blades moving at 1/4000th used to be the cutting edge of technology. It couldnt be done with shutter curtains (another variety with a fabric shutter) or steel or ally, so they went to titanium for high strength/weight. Its important they remain quite rigid too, things tend to get very plastic at those speeds

Again, I think a larger mirror can move just as quickly as a smaller
mirror, just with more shutter slap.

eventually mass momentum shapes the speed that it can react at, the
energy imparted and the max speed of the system. Im not sure we are
at or near that point in any system yet though.

For sure. Eventually, the smaller mirror will have the advantage,
and I think we can agree that by 60 fps, it will never matter if one
is better than the other or not -- would certainly keep the inside of
the camera cool, though. : )

Evil will do it better, the future is video

But, yeah, I've no idea how much of an engineering challenge it is in
terms of the mirror vs the electronics moving all that data.

i dont think we approach the video capability found elsewhere, like dual pipeline video cards with 512Mb memory, but rightly so, its a challenge

--
Riley

I'd give my right arm to be ambidextrous

 

[Raist3d]

talking to people who "know" about equipment they don't own, gets very very old in this forum (dpreview clipped it on re-edit).

--
Raist3d (Photog. Student & Tools/Systems/Gui Games Developer)
Andreas Feininger (1906-1999) 'Photographers — idiots, of which there are
so many — say, “Oh, if only I had a Nikon or a Leica, I could make great
photographs.” That’s the dumbest thing I ever heard in my life. It’s
nothing but a matter of seeing, and thinking, and interest. That’s what
makes a good photograph.'

 

[Rriley]

does gets old in this forum. Very old.

Here's 4.5 stops of real IS with the E-3:

http://forums.dpreview.com/forums/read.asp?forum=1022&message=25665477

Obviously not all shots will be in this high of a result but they do
exist and the camera can do it. Olympus also clearly advertises "up
to 5 stops" not "you get 5 stops no matter what."

very reassuring, thanks...

Better buy that E-3 now!

again ? ....
--
Riley

I'd give my right arm to be ambidextrous

 

[baddboy]

I know I'm in the minority here, but here's what I think until someone explains to me where I've gone wrong..

I have a problem with the oft-stated factoid that "larger sensors have less noise because they gather more light". This doesn't really apply if you compare lens + sensor systems, rather than just sensors, and if you allow for the fact that the lens focal length needs to be adjusted if the sensor area is changed.

Another way of thinking of it is to think about what happens at the front of the lens, rather than the rear. The amount of light gathered by the lens depends on the area of the front piece of glass, and the the field of view of the lens (and the exposure details and the scene brightness). The focal length of the lens determines whether this quantity of light is put onto a smaller or larger image circle, but to a first approximation, the quantity of light for each pixel is set by lens area and is independent of sensor area (again, if the lens is matched to the sensor).

So, low light response is not fundamentally compromised by a smaller sensor.

What IS given up by a smaller sensor is the number of photons which the photosites can absorb at the bright end, in other words, dynamic range. This is because the photosite well capacity is determined by its area. But this is highlight response, not low-light response.

The guys who designed 4/3 understood that the lens gathers the light, and the role of the sensor is to read it. They saw value in making the sensor as small as practical within the diffraction and DR limits, so that they could grab the benfits of telecentricity. They must have seen that the compromise, given their (or any other) sensor size was between the number of Mp and the DR, and if they've got anything wrong it's that they failed to see the renewed demand for more and more Mp.

The other bone I'd like to pick with the orthodoxy is this idea that the smaller sensor causes the camera to produce pictures with a greater depth-of-field. Everyone knows that shorter lenses have a greater dof - we all learnt that in the film age. BUT if you allow for the greater resolution required on a smaller sensor (if you want to preserve the pixel count), the dof shortens. I've done calculations (using the equations on Wikipedia) and they've told me that for a given Mp count, dof is virtually identical for a 35mmFF sensor and a 4/3 sensor, if the Mp count and the front lens element diameter are the same.

Cheers,

Baddboy.