Why is full-frame so important ? / 40D isn't but it's a great camera

Usually, the simple answer is because they compare per-pixel noise
rather than total image noise. The 40D has 23% more pixels than the
20D. With the same sensor size and efficiency, this would represent
1/3 of a stop more per-pixel noise. So, how they got 1/2 - 2/3
stops, I can't comment. I suspect it's because they're wrong, as
I've not heard that claim anywhere else.
If so, then longer reach by more MP would likely a myth ?
Because they are not the same per pixel ?
They capture different "total light" ?
Or sensor efficiency ?
Both are possible, but, if anything, it "should" be in the other
direction.
Sorry Joe, I don't get it here. What direction ?
Is the result will be the same for comparing 400D with 350D too ?
Cannot say.
Well, it's okay.
Because I found my video camcorders,
Canon XL1s & XL2 have different sensitivity too.
If I use the same setting, say f/1.6 1/200 sec gain 0 db,
then XL1s picture look brighter than XL2.
And I compare them in the same monitor.
Their sensors size are the same, but XL2 has more resolution.
Same lenses, too? Dunno -- that's a conundrum, to be sure.
Yes.
I suspect the gain is not equal.
Don't you ever heard that 20D/30D are more sensitive than the true iso ?
iso 100 is actually 125 or something.
If I'm not wrong, I read it in dpreview.
Carpal-Tunnel Syndrome will get me eventually. : )
LOL ;D
 
Usually, the simple answer is because they compare per-pixel noise
rather than total image noise. The 40D has 23% more pixels than the
20D. With the same sensor size and efficiency, this would represent
1/3 of a stop more per-pixel noise. So, how they got 1/2 - 2/3
stops, I can't comment. I suspect it's because they're wrong, as
I've not heard that claim anywhere else.
If so, then longer reach by more MP would likely a myth ?
Because they are not the same per pixel ?
Effective reach is a function of the pixel density, not the number of pixels -- the greater the pixel density, the greater the reach. However, for the same framing without cropping, the image with the greater number of pixels will have the advantage.
They capture different "total light" ?
Or sensor efficiency ?
Both are possible, but, if anything, it "should" be in the other
direction.
Sorry Joe, I don't get it here. What direction ?
The 40D, being more modern, if anything, should have a more efficient sensor and/or more efficient microlens covering, which means that it "should" have less that a 1/3 stop difference in per-pixel noise over the 20D, if anything, and certainly not more.
Because I found my video camcorders,
Canon XL1s & XL2 have different sensitivity too.
If I use the same setting, say f/1.6 1/200 sec gain 0 db,
then XL1s picture look brighter than XL2.
And I compare them in the same monitor.
Their sensors size are the same, but XL2 has more resolution.
Same lenses, too? Dunno -- that's a conundrum, to be sure.
Yes.
I suspect the gain is not equal.
Don't you ever heard that 20D/30D are more sensitive than the true iso ?
iso 100 is actually 125 or something.
If I'm not wrong, I read it in dpreview.
Entirely possible. But if the f-ratio and shutter speed are the same, the images should be exposed the same, regardless of the gain, for the same scene.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/
 
Effective reach is a function of the pixel density, not the number of
pixels -- the greater the pixel density, the greater the reach.
However, for the same framing without cropping, the image with the
greater number of pixels will have the advantage.
You're right.
In the same sensor size, I think more MP mean more density.
They capture different "total light" ?
Or sensor efficiency ?
Both are possible, but, if anything, it "should" be in the other
direction.
Sorry Joe, I don't get it here. What direction ?
The 40D, being more modern, if anything, should have a more efficient
sensor and/or more efficient microlens covering, which means that it
"should" have less that a 1/3 stop difference in per-pixel noise over
the 20D, if anything, and certainly not more.
Yes, I agree. 40D is more modern, so it must be more efficient.
So, do you mean more MP certainly will have worse per-pixel noise,
but sensor efficiency will reduce the gap ?

Is this because light gathering per pixel sensor is different ?
But the same per area of sensor ?
Entirely possible. But if the f-ratio and shutter speed are the
same, the images should be exposed the same, regardless of the gain,
for the same scene.
Yes, I set them in the same f-ratio & shutter speed.
Both cameras have same sensor size.
And gain here is like iso in the still camera.
XL2 is being more modern camera than XL1s.
Too bad, the sensitivity is not the same with the previous model.
So, in low light, XL2 more strugle.

Is it not the time for cramming more MP yet ?

Maybe the sensor technology's growing is not that fast to achieve same sensitivity with more MP ?

Joe,
Foveon has 3 layers, does it capture 3 times more light than bayer one ?

Thanks,

Brian.
 
cons with 5D:
no pop flash(i think)
Actually it is an advantage. The 40D's inbody flash is practically useless unless you want to use the camera as P&S
can only use L lenses
Why is that ?
That's true. You get what you paid for.

The biggest disadvantage of 5D is that it is too old already. As the result it is slower and does not have some latest cool features that 40 offers.
 
no that part of a mistake I guess from the poster. I understood he meant can only use EF lens and not EF-S.
that is if you do understand it.
when i purchased the 40D 2 weeks ago, i had the same concerns. some
of the things i've noticed are

pros with 5D:
far better high iso
more resolution
better dynamic range
better use of L lenses (vs L lenses on 40D)

cons with 5D:
no pop flash(i think)
can only use L lenses
cost

--
thanks,
-rez
--
Cal

Put a Canon to your head, You deserve it....

http://funshots.smugmug.com/
--



http://www.pbase.com/zylen
'Never argue with an idiot, they drag you down to their level and
beat you with experience'
--



http://www.pbase.com/zylen

'Never argue with an idiot, they drag you down to their level and beat you with experience'
 
ok. I thougth that the shot from the 5D would have less DOF due to
the larger sensor but I was wrong. Probably because people keep
comapring similar focal equivalent then or taking into account that
they will frame the subject the as much as possible with either
camera, so getting closer with the FF camera is needed.
It's kinda like the "chicken and the egg" scenario. A larger sensor
how can you compare this scenario to the chicken and the egg? with the chicking and the egg, there is a paradox..which one came first? the chicken or the egg?

it is a paradox because you can't have chicken without the egg and you can't have eggs without chickens.

so..which one came first? :) My only logical explanation for this is that the chickens were made first by God or similar powerfull and intelligent being who can program the DNA to produce the chickens..then those chickens can reproduce by making eggs.

without God, this is just a plain paradox..

as for the FF sensor needing a longer lens, this is another scenario...a scenario of compensation, but I understand that putting a longer lens to get the same framing will give less DOF.
means that you need to use a longer focal length for the same
framing. So, on the one hand, it's the longer focal length that
gives the more shallow DOF for the same perspective, framing, and
f-ratio, that gives the more shallow DOF. On the other hand, the
whole reason you're using the longer focal length is because you are
using a larger sensor, so you can say the more shallow DOF is because
of the larger sensor.

There are two easy ways to state the problem so that there is no
confusion:

1) For the same perspective, framing, f-ratio, and output size,
larger sensors yield a more shallow DOF than smaller sensors.

2) For the same perspective, framing, aperture, and output size, all
systems have the same DOF.

Of course, so many people confuse "aperture" with "f-ratio" and also
don't understand that perspective is the subject-camera distance
(technically, subject-aperture distance, but that's only an issue for
close framing), that even the above simplifications cause confusion.

The reason that DOF is so important is that it is intimitely
connected with corner sharpness, diffraction softening, and
vignetting. For case #1 above, larger sensors will have softer
corners, less diffraction softening, and more vignetting. For case
  1. 2, the differences between systems will be negligible, assuming the
same quality of glass is used. This may be of interest:

http://www.josephjamesphotography.com/equivalence/#diffraction

It's akin to "exposure" and "total light" -- people think they are
the same, but they are not. This is also an important distinction to
make when discussing noise, and is, like DOF, intimitely connected
with aperture. If you're still interested, please take a read:

http://www.josephjamesphotography.com/equivalence/#exposure
thanks for the explanation.
No worries -- I like to type. : )

--
--joe

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



http://www.pbase.com/zylen

'Never argue with an idiot, they drag you down to their level and beat you with experience'
 
Yeah, no pop from my 40D...
Your pictures are great with a 40D. However, try taking a picture of
person (5 foot tall subjects) in a typical home or room size and you
will quickly see that with the 1.6 crop you will need 35mm. You will
also quickly see that 35mm and f2.8 has too much of the background in
focus...
maybe so but your subject is sharper with the 40D..even the hair is soft with the 5D..In his case I do prefer the 40D shot as I think the DOF is too shalow with the 5D.

--



http://www.pbase.com/zylen

'Never argue with an idiot, they drag you down to their level and beat you with experience'
 
in deed, with a 5D this taken with my 400mm lens would have much less detail and would have much less pop as well:

maybe the 5D is the best portrait camera..but when it come to birds....there is nothing like a 40D.







--



http://www.pbase.com/zylen

'Never argue with an idiot, they drag you down to their level and beat you with experience'
 
you're right but I find the pics on the right to be more to my liking. The left photo tends to be soft overall (too narrow a DOF) With that said the exposure looks to be more on the mark but the WB seems shifted slightly to the red. As for (POP) ? not really sure what you mean????
--
Cal


Put a Canon to your head, You deserve it....

http://funshots.smugmug.com/
 
The 40D, being more modern, if anything, should have a more efficient
sensor and/or more efficient microlens covering, which means that it
"should" have less that a 1/3 stop difference in per-pixel noise over
the 20D, if anything, and certainly not more.
Yes, I agree. 40D is more modern, so it must be more efficient.
So, do you mean more MP certainly will have worse per-pixel noise,
but sensor efficiency will reduce the gap ?
Well, a more modern sensor doesn't have to be more efficient, but one would think it would be at least as efficient.

More MP on a sensor with the same size and efficiency will mean worse per-pixel noise. However, if the pixel increase is modest, it's possible to maintain the same per-pixel noise with increased efficiency, but, at this time, we're pretty much at those limits with a Bayer CFA.
Is this because light gathering per pixel sensor is different ?
But the same per area of sensor ?
Yes.
Entirely possible. But if the f-ratio and shutter speed are the
same, the images should be exposed the same, regardless of the gain,
for the same scene.
Yes, I set them in the same f-ratio & shutter speed.
Both cameras have same sensor size.
And gain here is like iso in the still camera.
XL2 is being more modern camera than XL1s.
Too bad, the sensitivity is not the same with the previous model.
So, in low light, XL2 more strugle.
Wish I had an answer for you, but that one has me stumped.
Is it not the time for cramming more MP yet ?
Maybe the sensor technology's growing is not that fast to achieve
same sensitivity with more MP ?
More pixels will yield more detail. And while each pixel will not be receive as much light, the total light will be the same, so total image noise, for the same sensor efficiency, will also be the same.
Foveon has 3 layers, does it capture 3 times more light than bayer one?
In theory, it's possible, but the current implementation of Foveon technology has technological hurdles that prevent it. Under the worst of conditions, a Bayer array will only capture 1/3 of the light that falls on it, but in practice, a Bayer sensor captures about half of the light that falls on it, since there is more green light than the other colors in most circumstances, thus doing better than the 33% efficiency you would think.

Anyway, that means that, at best, future sensors might collect double the light of current sensors for the same size, thus being one stop more efficient in terms of noise -- basically the same gap that currently exists between FF and 1.6x.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/
 
It's kinda like the "chicken and the egg" scenario.
how can you compare this scenario to the chicken and the egg? with
the chicking and the egg, there is a paradox..which one came first?
the chicken or the egg?

it is a paradox because you can't have chicken without the egg and
you can't have eggs without chickens.

so..which one came first? :)
Well, I did say "kinda like" -- at least it was better than saying "it's kinda like why the chicken crossed the road". : )
My only logical explanation for this is
that the chickens were made first by God or similar powerfull and
intelligent being who can program the DNA to produce the
chickens..then those chickens can reproduce by making eggs.

without God, this is just a plain paradox..
OK, let's presume there's a God for a moment that designed the entire Universe (well, now the Multiverse, as we're coming to understand), including chickens. Let's also presume that God had some sort of purpose with this design, rather than "Hey, let's see what happens!"

So, given all that, I have one question: where do mosquitos and flies fit in? I don't give a flying frak what part they serve in an ecological cycle, I want to know why this omniscient God couldn't find a way around that. : )
as for the FF sensor needing a longer lens, this is another
scenario...a scenario of compensation, but I understand that putting
a longer lens to get the same framing will give less DOF.
For the same perspective and f-ratio, yes. Well, there! You got it!
'Never argue with an idiot, they drag you down to their level and
beat you with experience'
Yeah, I hear you. I argue with myself all the time with the same results. : )

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/
 
maybe the 5D is the best portrait camera..but when it come to
birds....there is nothing like a 40D.
I agree completely. I want all the reach I can get. The 40D gets me 60% closer than any FF camera. In fact, I've posted before that if I see cameras all moving to full frame in the future, I'll stock up on some non-FF specimens.
 
maybe the 5D is the best portrait camera..but when it come to
birds....there is nothing like a 40D.
I agree completely. I want all the reach I can get. The 40D gets me
60% closer than any FF camera.
No. The size of the sensor is irrelevant to the resolving power (often called "reach"). It's the size of the pixels that matters. It happens that the smaller sensors tend to have smaller pixels, but not by nearly as much as the 1.6-crop factor would lead you to believe. The larger sensors tend to have more pixels.

450D: 5.2 microns
400D/40D: 5.7 microns
350D/30D/20D: 6.4 microns
1DII/5D: 8.2 microns
1DIII: 7.4 microns
1DsII: 7.2 microns
1DsIII: 6.4 microns

Notice that the 1DII and 5D have the same pixel sizes, and thus the same resolving power at the same focal length. Notice the same about the 350D/20D/30D and the 1DsIII.

Notice also that the 40D is only 12% better than the 1DsIII. Since the 1DsIII can AF at f8, while the 40D can only AF at f5.6, the 1DsIII can handle an extra 1.4x TC and still retain AF. This actually gives it a "reach" advantage over the 40D. In essence, the camera with the longest "reach" (while retaining AF) of the Canon cameras is a full-frame camera.

--
Lee Jay
(see profile for equipment)
 
Hello,

I'm thinking of upgrading my 20D sometime soon, and am thinking of
going with the 40D, or perhaps waiting a few months to see what Canon
shakes out of its camera tree. A successor to the 40D may be amongst
the offerings.
That is not likely. Canon likes to announce cameras at big trade shows. It is expected to announce the 5D MKII in photokina 08. The 45D, or successor to the 40D, will likely be introduced at next year's PMA, held in Feb. 08.
Something I don't understand is the frame size of higher priced DSLRs
like the 5D which offer full frame.

The 20D doesn't. Can somebody explain what the difference is, and
what I actually see or don't see after I take a shot ?
The viewfinder of full frame cameras are bigger than APS-C. And the medium format cameras are bigger still. Bigger viewfinders let you see more of the picture you are taking and also in better detail. That makes for more accurate focusing and framing.
The 20D and 40D are amazing cameras.
Yes they are. And they can shoot much faster than the 5D as well.
Quality is there, and so I
wonder, is a full-frame camera so different in terms of quality ?
The difference is there, but it is not huge. It is not the same order of magnitude as the difference between the 20D and the digicams that have tiny sensors. Noise is lower with the 5D at high ISO settings, but the difference is slight. The main advantage of the 5D is that you can use full frame wide angle lenses without having to buy APS-C only glass. Most APS-C only glass are not as fast as their full frame equivalent in focal length and zoom range. To many people, the 20D and 40D are good enough. For those who are not satisfied with good enough, they would want the 5D. The difference between a Mercedes Benz or BMW and many Japanese cars costing half as much is slight, but the differences are there and they are real. It is up to the buyer to decide whether he/she wants to pay a lot more for a slight difference in performance.
or
is it that what it takes when the shutter is the only thing that is
different ( more area )

Thanks for any explaination gentlemen,

Dave.
The 5D has pixels that are nearly twice as big as those on the 450D. Both have about 12 million pixels. Big pixels mean better signal/noise ratio, better dynamic range and lower noise at high ISO settings. Since the 5D has a sensor that is about twice as big as that of the 40D, the image formed on the 5D sensor does not have to be enlarged quite as much as the same image from the 40D. We know for that a bigger negative produces sharper prints with less noise. This holds true in the digital era as well. Further, it has been known for a long time that the bigger the negative, the less sharp a lens would need to be. That is because the human eye can only resolve about 6 line pairs per mm. A negative that need to be enlarged 10 times would therefore need a lens that must resolve at least 60 lp/mm. A negative that is large enough to produce 10" prints without enlargement, on the other hand, would only need a lens that can resolve no more than 10 lp/mm. So, you don't need a razor shapr lens on a large sensor but you would need them on a smaller sensor. That means an average lens can produce good results on the 5D, whereas you would need L glass on a 40D sensor to produce the same results.
 
No. The size of the sensor is irrelevant to the resolving power
(often called "reach"). It's the size of the pixels that matters.
It happens that the smaller sensors tend to have smaller pixels, but
not by nearly as much as the 1.6-crop factor would lead you to
believe.
Good point, and thanks for Anastigmat's post directly below yours which amplifies the case. It's not exactly 1.6. Plus the lower noise on the FF does allow more uprezzing. But I still love that extra reach right out of the camera. I suppose a clean FF 30 MP camera will be better for me than a 10 MP cropped-sensor camera.

You're both correct in what you write and I appreciate the clarification.
 
what the 5D does is that it will capture the .6 that your camera is
cropping out.

you will have more of the surronding in the shot..and larger file.

Now wether you need all that surronding is depending on your subject.
If you do not need it, you can always crop it out of the final print. But it is impossible to recover what was excluded. Say, you are shooting a bird, and you excluded part of the tail. It is not possible to get the tail back into the final print.
for exemple if you take landscapes, having a FF body lets you use
longer lens to get the same field of view.

so with the 1.6x crop factor camera you would need a 10mm lens to get
the same framing as you would with a 16mm lens and a FF body.

a 16mm lens would be much less expensive than a 10mm lens as well.
Not only that, but some tilt-shift lenses work best when they are on a full frame. A 28mm tilt and shift lens is much more useful than a 45mm tilt and shift.
Lucky with the 40D and EF-S, we do have the excellent 10-22mm so
that'S a moot point.
Yes, but there is no Ef-S tilt and shift lens. An excellent 24-105mm L lens is a much less useful 38mm zoom on the 20D.
the larger pixels mean having less noise at higher ISO.
Precisely. It is important when shooting wildlife and sports to have low noise at high ISO settings.
But if you shoot birds like I do or wildlife, you will get much
better detail in your subjects if you use a 1.6 crop factor camera
because with higher pixel density the 40D capture more detail from
the same distance to subject.
The higher pixel density does not always translate into greater details. Greater pixel density means the sensor can resolve more line pairs per mm. But that also puts more demand on the lens, which must also resolve enough to take advantage of the sensor's resolution. Bob Atkins calculated that the 8mp 1.6x Canon sensor, IIRC, has a resolution equal to about 70 line pairs per mm. The 40D would be slightly higher in resolution than 70 lp/mm So, a telephoto lens must be capable of at least 70 line pairs per mm to take full advantage of the 1.6x crop sensor. The 5D, on the other hand, is not as demanding on the lens because its sensor has lower resolution and because images from it do not have to be enlarged as much as an image from the 40D sensor. The 5D is like a mini-medium format camera, and it is well known that medium format camera lenses need not be as sharp as 35mm lenses to produce a sharp image because of the larger negative.
that is why I stick with 1.6x crop factor and hoping that they will
eventualy improve the 40D line to match up the 1D line so that I
don't have to go with a 1.3x crop factor and lose that advantage.
It is no doubt that sensor technology will improve. But keep in mind that these improvements can also be implemented on the full frame DSLR cameras. That means the gap between full frame and 1.6x crop cameras of the same generation will likely remain the same. The 40D is one generation ahead of the 5D, but the 5D MKII will have the same technological advances found in the 40D. The 40D has closed the gap somewhat on the 5D in terms of image quality, but the 5D MKII should widen the gap again.

Full frame bodies may well have longer product cycles, and they may lag slightly behind the APS-C models, at least for the time being, but when they are updated, they will pull ahead of 1.6x crop cameras once again. The full frame can always "reel in" any temporary edge than the crop sensor cameras have because of technical advances. Besides, as sensor resolution continue to increase, adding more pixels to the crop sensor will simply exceed the ability of the lenses to resolve enough to take advantage of them. It is a game the crop sensor cameras cannot win.
 
It's kinda like the "chicken and the egg" scenario.
how can you compare this scenario to the chicken and the egg? with
the chicking and the egg, there is a paradox..which one came first?
the chicken or the egg?

it is a paradox because you can't have chicken without the egg and
you can't have eggs without chickens.

so..which one came first? :)
Well, I did say "kinda like" -- at least it was better than saying
"it's kinda like why the chicken crossed the road". : )
My only logical explanation for this is
that the chickens were made first by God or similar powerfull and
intelligent being who can program the DNA to produce the
chickens..then those chickens can reproduce by making eggs.

without God, this is just a plain paradox..
OK, let's presume there's a God for a moment that designed the entire
Universe (well, now the Multiverse, as we're coming to understand),
including chickens. Let's also presume that God had some sort of
purpose with this design, rather than "Hey, let's see what happens!"

So, given all that, I have one question: where do mosquitos and
flies fit in?
they help reduce the amount of humans on Earth..amount is still growing so fast that it is obvious moskitos are losing the battle.

I don't give a flying frak what part they serve in an
ecological cycle, I want to know why this omniscient God couldn't
find a way around that. : )
Maybe there was a good reason for them...as well as virus :)
as for the FF sensor needing a longer lens, this is another
scenario...a scenario of compensation, but I understand that putting
a longer lens to get the same framing will give less DOF.
For the same perspective and f-ratio, yes. Well, there! You got it!
'Never argue with an idiot, they drag you down to their level and
beat you with experience'
Yeah, I hear you. I argue with myself all the time with the same
results. : )

--
--joe

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



http://www.pbase.com/zylen

'Never argue with an idiot, they drag you down to their level and beat you with experience'
 
By now you've been totally confused by the various opinions. I own an XT, an XTi, and a 5D. The only way you are going to get an answer to your question is to rent a FF camera and give it a try.
Hello,

I'm thinking of upgrading my 20D sometime soon, and am thinking of
going with the 40D, or perhaps waiting a few months to see what Canon
shakes out of its camera tree. A successor to the 40D may be amongst
the offerings.

Something I don't understand is the frame size of higher priced DSLRs
like the 5D which offer full frame.

The 20D doesn't. Can somebody explain what the difference is, and
what I actually see or don't see after I take a shot ?

The 20D and 40D are amazing cameras. Quality is there, and so I
wonder, is a full-frame camera so different in terms of quality ? or
is it that what it takes when the shutter is the only thing that is
different ( more area )

Thanks for any explaination gentlemen,

Dave.
--
Galleries: http://www.pbase.com/jon_b
 
OK, let's presume there's a God for a moment that designed the entire
Universe (well, now the Multiverse, as we're coming to understand),
including chickens. Let's also presume that God had some sort of
purpose with this design, rather than "Hey, let's see what happens!"

So, given all that, I have one question: where do mosquitos and
flies fit in?
they help reduce the amount of humans on Earth..amount is still
growing so fast that it is obvious moskitos are losing the battle.
Well, there you go. He should have kept the dinosaurs and skipped the mosquitos, then. : )
I don't give a flying frak what part they serve in an
ecological cycle, I want to know why this omniscient God couldn't
find a way around that. : )
Maybe there was a good reason for them...as well as virus :)
Well, maybe. OK, explain Mullets and bell bottoms, then. What kind of God would allow that? : )

--
--joe

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

Keyboard shortcuts

Back
Top