September 02, 2014, 06:30:18 PM

Author Topic: Does Canon really have a Flagship Stills Camera?  (Read 28147 times)

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 14041
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #90 on: April 21, 2013, 04:59:23 PM »
As far as I understood it so far, the fact that in general FF cameras have lower noise is mainly due to the fact that the sensors are the most recent types that have been developed. I have no experience with most models that are out there, but I would think that newer crop sensors will give lower noise than older FF sensors - simply because they are newer. Any attempt to quantify high-ISO performance simply on size and MP-count alone don't work because no two sensors of different size are the same otherwise.

I am open to hearing that all I have just written is nonsense... But what I would really like to hear is what a real expert has to say. Someone who builds the sensors, or who does research in that field...

Check out the DxOMark ISO 'scores' (the highest ISO than meets a specific noise criterion, so the higher the number, the lower the camera noise).  For the newest Canon APS-C camera which they've tested, the T4i/650D, the score is ISO 722.   Compare that to the ISO score for the 5D (the original one): ISO 1368.  So the 8 year old FF camera with sensor tech that's also 8 years older has nearly a full stop better ISO performance than a very recent APS-C sensor.  By comparison, a modern FF sensor (the 5DIII) scores an ISO 2293, close to 2 stops better than its contemporary APS-C sensor.

Here's what one academic expert, Emil Martinec, has to say:

For comparing pixels on a patch of sensor of a given area, the normalizations given above are the appropriate ones. However, one is often interested in comparing the overall noise of the image coming from different sizes of sensors, so noise per area is less relevant that noise normalized to the image frame. Thus another metric for comparison is to consider the entire frame, and if comparing two different sensor formats (such as the 1D3 and 40D considered above), shoot with different focal lengths (say 130mm on the 40D and 160mm on the 1D3) so that the field of view is kept fixed. In the case of fixed framing, we should refer the noise characteristics not to a fixed spatial scale in microns, but rather to a fixed percentage of the frame height. An appropriate figure of merit is to divide the per pixel noise values by the frame height in pixels, again because the noise combines linearly in the number of pixels combined. The clear rule of thumb that emerges from such an exercise (not surprisingly) is that larger size sensor formats are less noisy than smaller size sensor formats. A simple example is the 40D relative to the 1D3; both are 10MP cameras, with 2592 pixels in the frame height, and so in this particular example the frame-referred noise levels are the same as the pixel-level noises, and these are lower (as measured in ADU) at all exposure levels and ISO's for the 1D3. The read noises are similar in photo-electron units, but the 1D3's bigger photosites capture more photons, hence have higher S/N ratio. Bigger sensors have higher S/N ratios, because bigger sensors collect more photons.
« Last Edit: April 21, 2013, 05:18:53 PM by neuroanatomist »
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: Does Canon really have a Flagship Stills Camera?
« Reply #90 on: April 21, 2013, 04:59:23 PM »

kyamon

  • PowerShot G1 X II
  • ***
  • Posts: 62
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #91 on: April 21, 2013, 05:11:48 PM »
I am sorry to say this but honestly, I think bzzzz'ing each other is below the general level of maturity of this forum.

Now if you go back and read my post you will see that we probably agree regarding the framing. What I am saying is that the only difference between a crop sensor and a FF in most dSLR bodies is that the crop only collects a part of the image that makes it through the lens. Of course it collects less light than a FF sensor, but the additional light on the FF hits areas that simply don't exist in the crop sensor. The amount of light per unit surface is the same in both cases, because the image that is produced in the plane of the sensor depends on the lens and not on the sensor. Just that in one case you hold a large sensor in the beam path while in the other case you hold a small sensor in the same beam path.
It is as if you would take a movie projector and put a small projection wall at a given place. You don't get the full image, but what you get does not get any darker.

First: an APS-C sensor is smaller than a FF sensor, and thus a lower amount of light hits the total sensor surface. However, the framing is also different, and the image from the APS-C sensor has exactly the same size on the FF sensor because that is defined by the lens, not the sensor. Consequently, the amount of light per unit surface is the same on both sensors, and the noise performance can not depend on the sensor size.

BZZZZZT!

Worng.

The framing is the same only if you multiply the focal length by the ratio of the diagonals of the two lenses.

And the amount of light collected by the sensors is only the same if you perform a similar modification of the apertures.

To make the math simpler, let's start with a 100mm f/2 lens mounted to a 135 format ("full frame") camera. And we'll try to figure out what we need to get an equivalent image with a 4/3 camera, which has close enough to a 2x "crop factor" as makes no difference.

If you just mount the 100mm lens to the 4/3 camera, you'll only get the inner quarter of the image that you would have gotten on the 135 camera. On the 135 camera, you can get the exact same image by cropping out all but the inner quarter.

To match the field of view, we need to use a lens of half the focal length -- a 50mm lens. But what aperture?

The 100mm f/2 lens has a 100 / 2 = 50mm aperture. All the light headed to the lens that falls within that 50mm diameter circle makes its way to the sensor. To gather the same amount of light, we need a 50mm lens with a 50mm aperture. That means that our 4/3 camera needs a (50 / 50 = 1) 50mm f/1 lens to gather as much light as a 135 format camera with a 100mm f/2 lens. In both cases, both cameras are capturing all the light that falls onto a 50mm circle.

Because this is all simple geometry, it so happens that the depth of field, background blur, and the rest are also comparable.

But...even if everything else is equal, the larger format still retains a number of advantages, mostly because the image on the sensor doesn't need to be magnified as many times in absolute terms to print size. Imagine making a contact print with 8" x 10" film and comparing it with an 8" x 10" enlargement even from 4" x 5" film to understand why.

Cheers,

b&

kyamon

  • PowerShot G1 X II
  • ***
  • Posts: 62
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #92 on: April 21, 2013, 05:25:24 PM »
As far as I understood it so far, the fact that in general FF cameras have lower noise is mainly due to the fact that the sensors are the most recent types that have been developed. I have no experience with most models that are out there, but I would think that newer crop sensors will give lower noise than older FF sensors - simply because they are newer. Any attempt to quantify high-ISO performance simply on size and MP-count alone don't work because no two sensors of different size are the same otherwise.

I am open to hearing that all I have just written is nonsense... But what I would really like to hear is what a real expert has to say. Someone who builds the sensors, or who does research in that field...

Check out the DxOMark ISO 'scores' (the highest ISO than meets a specific noise criterion, so the higher the number, the lower the camera noise).  For the newest Canon APS-C camera which they've tested, the T4i/650D, the score is ISO 722.   Compare that to the ISO score for the 5D (the original one): ISO 1368.  So the 8 year old FF camera with sensor tech that's also 8 years older has nearly a full stop better ISO performance than a very recent APS-C sensor.  By comparison, a modern FF sensor (the 5DIII) scores an ISO 2293, close to 2 stops better than its contemporary APS-C sensor.

Here's what one academic expert, Emil Martinec, has to say:

For comparing pixels on a patch of sensor of a given area, the normalizations given above are the appropriate ones. However, one is often interested in comparing the overall noise of the image coming from different sizes of sensors, so noise per area is less relevant that noise normalized to the image frame. Thus another metric for comparison is to consider the entire frame, and if comparing two different sensor formats (such as the 1D3 and 40D considered above), shoot with different focal lengths (say 130mm on the 40D and 160mm on the 1D3) so that the field of view is kept fixed. In the case of fixed framing, we should refer the noise characteristics not to a fixed spatial scale in microns, but rather to a fixed percentage of the frame height. An appropriate figure of merit is to divide the per pixel noise values by the frame height in pixels, again because the noise combines linearly in the number of pixels combined. The clear rule of thumb that emerges from such an exercise (not surprisingly) is that larger size sensor formats are less noisy than smaller size sensor formats. A simple example is the 40D relative to the 1D3; both are 10MP cameras, with 2592 pixels in the frame height, and so in this particular example the frame-referred noise levels are the same as the pixel-level noises, and these are lower (as measured in ADU) at all exposure levels and ISO's for the 1D3. The read noises are similar in photo-electron units, but the 1D3's bigger photosites capture more photons, hence have higher S/N ratio. Bigger sensors have higher S/N ratios, because bigger sensors collect more photons.

As far as I understand this it is quite similar to what I wrote/asked, except that he specifically proposes a way to standardise the measurement. With the last sentence of what you quote he basically says that the noise generated in a single pixel is independent of its size (this was one of the questions I had asked), and thus the S/N gets better if the pixel collects more photons.
Comparing particular frames is one way to do it, and possibly the most reasonable one from a photography point of view. But it is not the only one, and depending on the type of photography, the settings you have available, etc. it is also not necessarily what you need to know, is it?

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 14041
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #93 on: April 21, 2013, 05:35:29 PM »

Comparing particular frames is one way to do it, and possibly the most reasonable one from a photography point of view. But it is not the only one, and depending on the type of photography, the settings you have available, etc. it is also not necessarily what you need to know, is it?

No, it's not the only one...but I'd argue it's the most relevant and important one, for a photographer.  Wearing my microscopist hat, there are times (such as when imaging fluorescently-stained samples) where the pixel characteristics of an imaging sensor are more important (because the relevant portion of the sample covers a tiny fraction of the sensor, a few pixels wide in some cases).  But more often than not, when that's the case I don't use a planar imaging sensor, but rather a photomultiplier tube, with spatial information derived from rastered laser illumination.

Photographers generally acquire images...so comparing images is most appropriate. 

One could make a similar argument about the effect of sensor size on DoF.  We generally say that FF gives shallower DoF, and of course that's true when talking about identical framing where you're either using a longer lens or are closer to the subject.  But if you hold focal length and distance constant, you actually get shallower DoF with APS-C.  Of course, your head shot has now cut off the subject's hair and mouth to give you a nose-and-eyes portrait instead of a head shot.  Not a very useful point of view for most...
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

Don Haines

  • Canon EF 300mm f/2.8L IS II
  • *******
  • Posts: 3076
  • Posting cat pictures on the internet since 1986
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #94 on: April 21, 2013, 05:43:42 PM »
There is a lot more to a camera than a sensor. There is autofocus system, frame rate, physical controls, user interface, physical size, battery life, sealing, build quality, and a whole lot more.

You can not make a blanket statement like all FF cameras are better than all crop cameras because all FF sensors are better than all crop sensors. Other things come into play, particularly AF systems. Any in focus crop picture beats any out of focus FF picture..... we are dealing with systems that are balanced towards particular needs. There is no absolute winner for all situations..... there will always be cases where the unlikely candidate shines. Heck, an iPhone beats a 1DX as an inspection camera, but you don't see me disparaging the 1DX because of that.
The best camera is the one in your hands

kyamon

  • PowerShot G1 X II
  • ***
  • Posts: 62
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #95 on: April 21, 2013, 06:02:26 PM »

Comparing particular frames is one way to do it, and possibly the most reasonable one from a photography point of view. But it is not the only one, and depending on the type of photography, the settings you have available, etc. it is also not necessarily what you need to know, is it?

No, it's not the only one...but I'd argue it's the most relevant and important one, for a photographer.  Wearing my microscopist hat, there are times (such as when imaging fluorescently-stained samples) where the pixel characteristics of an imaging sensor are more important (because the relevant portion of the sample covers a tiny fraction of the sensor, a few pixels wide in some cases).  But more often than not, when that's the case I don't use a planar imaging sensor, but rather a photomultiplier tube, with spatial information derived from rastered laser illumination.

Photographers generally acquire images...so comparing images is most appropriate. 

One could make a similar argument about the effect of sensor size on DoF.  We generally say that FF gives shallower DoF, and of course that's true when talking about identical framing where you're either using a longer lens or are closer to the subject.  But if you hold focal length and distance constant, you actually get shallower DoF with APS-C.  Of course, your head shot has now cut off the subject's hair and mouth to give you a nose-and-eyes portrait instead of a head shot.  Not a very useful point of view for most...

:) True (the last statement).

I have done my share of scientific imaging applications, and of course there the criteria are usually different, as are the technical requirements and/or constraints.

But I think this discussion is now really getting too far off-topic (among other things because I agree that there are many more things to a camera than the sensor, and that a good sensor far from makes a good photographer!). If we continue we should probably start a new thread and leave this one to figure out if the flagship actually is the ship carrying the flag, or if Canon was mistaken :)

Sella174

  • 5D Mark III
  • ******
  • Posts: 672
  • So there!
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #96 on: April 22, 2013, 06:13:35 AM »
The 5DII and 50D are of a similar tech (and the 50D has the advantage of gapless microlenses) ...

If the 50D has an "advantage" then they are not completely of a "similar tech."

... but even if the 50D were only 8 MP, the 5DII would still blow it out of the water for ISO performance.

Conjecture, i.e. not empirically proven.

Is the 300D's sensor exposed to the same amount of that light as the 1D X's sensor?

Does the light that the APS-C sensor "crops out" affect the same (APS-C sized) centre area of the "full-frame" sensor?

The lens may well gather the exact same amount of light regardless of what, if anything, is attached to the rear mount. But that's completely irrelevant when what actually is attached to the rear amount throws away half of the light that the lens gathers.

No, because the light that is "thrown away" does not affect the light that is "gathered."

You do know that the "crop" part of "crop sensor camera" means that the borders get cropped away and the light that the lens would have projected onto them gets absorbed by the black interior of the camera, don't you?

Yes, but do you?

And the amount of light collected by the sensors is only the same if you perform a similar modification of the apertures.

Here's what one academic expert, Emil Martinec, has to say:

Thus another metric for comparison is to consider the entire frame, and if comparing two different sensor formats (such as the 1D3 and 40D considered above), shoot with different focal lengths (say 130mm on the 40D and 160mm on the 1D3) so that the field of view is kept fixed.

By trying to keep the framing and/or DoF the same, you are changing lenses. But the objective is not to compare lenses, but sensors. Simple scientific procedure tells you that is a big no-no. You cannot change parameters and still expect the results to make sense ... unless you are forcing a conclusion.

But if you hold focal length and distance constant, you actually get shallower DoF with APS-C.

Huh? Using the exact same lens? Are you sure?
Happily ignoring the laws of physics and the rules of photography to create better pictures.

canon rumors FORUM

Re: Does Canon really have a Flagship Stills Camera?
« Reply #96 on: April 22, 2013, 06:13:35 AM »

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 14041
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #97 on: April 22, 2013, 08:23:09 AM »
The 5DII and 50D are of a similar tech (and the 50D has the advantage of gapless microlenses) ...

If the 50D has an "advantage" then they are not completely of a "similar tech."

True...but that tech advantage favors the APS-C sensor, and the FF sensor still wins by a big margin.

In your very detailed responses, I notice you conveniently ignored the fact that DxO's measurements show that a current Canon APS-C sensor has ~1 stop more ISO noise than an 8-year old Canon FF sensor, and close to 2-stops more noise than a current Canon FF sensor.  I wonder why you didn't respond to that?  ::)

But if you hold focal length and distance constant, you actually get shallower DoF with APS-C.

Huh? Using the exact same lens? Are you sure?

Yes, I'm sure.  Go visit a DoF calculator (e.g., dofmaster.com), or verify it empirically as I've done.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

kyamon

  • PowerShot G1 X II
  • ***
  • Posts: 62
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #98 on: April 22, 2013, 08:43:50 AM »
Yes, I'm sure.  Go visit a DoF calculator (e.g., dofmaster.com), or verify it empirically as I've done.

But did you also look how the DoF calculator works? The only difference in the mathematical expressions that give you the DOF is the circle of confusion, which is a very ill-defined parameter anyways. They use a particular value for any sensor size, so by definition they have to get different DoF values for every sensor size.
But this only makes sense if you somehow take an image from the different sensors and look at them at the same size, and there are also certain requirements on pixel density, I would say. I don't think that the DoF calculator can be taken as an argument in this discussion - it just serves to give you an order of magnitude for some concept that is very hard to quantify.
Since you say you tested it empirically, however, I would be interested in seeing your data... Because I never quite understood why people say that FF and APS-C have different DoF (provided the same optics, aperture, and distance to the object are used).

Sella174

  • 5D Mark III
  • ******
  • Posts: 672
  • So there!
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #99 on: April 22, 2013, 09:03:34 AM »
... show that a current Canon APS-C sensor has ~1 stop more ISO noise than an 8-year old Canon FF sensor, and close to 2-stops more noise than a current Canon FF sensor.

Current 20MP "full-frame" sensors equal an 8MP APS-C sensor ... larger photo-sites (generally) equal less noise ... make an APS-C sensor using current "full-frame" sensor "technology" and ISO, noise, everything will be the same.

Or the inverse ... make a 46.5MP "full-frame" sensor using current APS-C sensor "technology" and ISO, noise, everything will be the same.

Go visit a DoF calculator (e.g., dofmaster.com) ...

dofmaster.com changes the CoC when it changes the sensor size, so it's of not much use because if the CoC stays the same, then the DoF stays the same.

... or verify it empirically as I've done.

I have ... seems we get different results.
Happily ignoring the laws of physics and the rules of photography to create better pictures.

TrumpetPower!

  • 1D Mark IV
  • ******
  • Posts: 951
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #100 on: April 22, 2013, 09:19:02 AM »
By trying to keep the framing and/or DoF the same, you are changing lenses. But the objective is not to compare lenses, but sensors.

Strange. I thought we were photographers trying to compare camera systems.

Or do your clients consider it reasonable for you to hand them portraits with the head chopped off at the mouth and the body chopped off at the navel, because that's what happens when you switch to a crop format camera?

Cheers,

b&

Sella174

  • 5D Mark III
  • ******
  • Posts: 672
  • So there!
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #101 on: April 22, 2013, 09:28:50 AM »
... I never quite understood why people say that FF and APS-C have different DoF (provided the same optics, aperture, and distance to the object are used).

Marketing through misinformation, obfuscation, leading and derogation.
Happily ignoring the laws of physics and the rules of photography to create better pictures.

privatebydesign

  • Canon EF 300mm f/2.8L IS II
  • *******
  • Posts: 2267
  • Ermintrude says "moo"
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #102 on: April 22, 2013, 09:32:00 AM »
To come out with a meaningful figure across sensor sizes you have to standardise output. The "standard" DOF calculator assumes an 8"x10" print viewed at 12" and average eyesight. If you magnify a sensors output more to get to that 8x10 then you have to use a smaller CoC value, because you are enlarging it more.

Crop cameras, as always, can cause confusion. As Nero says, for the same lens from the same place a crop sensor image will return less dof than a FF camera if both images are uncropped (this would have the drawback of completely different framing though, so isn't an accurate comparison). If you crop both images to the same framing the dof is the same (this would emulate focal length limited situations, birding etc). If you use two different lenses (or a zoom) from the same spot to get the same framing (this would be the more normal scenariofor most people) the crop camera image has less dof. You see, depending on how you work the figures a crop camera can be said to have more dof, less dof, or the same dof as a FF sensor.

By the most pragmatic view, two people standing in the same place and taking the same framed image with different sensor sizes (which necessitates different focal lengths), a smaller sensor output has more dof because it uses a shorter focal length to achieve the same framing. Ever wonder why f8 on a P&S renders everything in focus? It is because you are using a 6mm lens.

One of the great mysteries to me is why people debate aperture (particularly very fast lenses) and use cropped images to do it; if you crop an image you decrease it's dof. Want an f1.2 image but only have an f1.8 lens? Stand back and crop.

« Last Edit: April 22, 2013, 09:37:24 AM by privatebydesign »
The best time to plant a tree is twenty-five years ago. The second best time is today.

canon rumors FORUM

Re: Does Canon really have a Flagship Stills Camera?
« Reply #102 on: April 22, 2013, 09:32:00 AM »

Sella174

  • 5D Mark III
  • ******
  • Posts: 672
  • So there!
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #103 on: April 22, 2013, 09:33:15 AM »
...

Out of context quote.

Let's not be silly!
Happily ignoring the laws of physics and the rules of photography to create better pictures.

TrumpetPower!

  • 1D Mark IV
  • ******
  • Posts: 951
    • View Profile
Re: Does Canon really have a Flagship Stills Camera?
« Reply #104 on: April 22, 2013, 09:36:48 AM »
...

Out of context quote.

Let's not be silly!

So then enlighten us, oh not-silly one.

Why should a photographer care more about comparing sensors than comparing camera systems? In what situation is the type of comparison you insist on making ever relevant to a photographer using a camera to create photographs?

b&

canon rumors FORUM

Re: Does Canon really have a Flagship Stills Camera?
« Reply #104 on: April 22, 2013, 09:36:48 AM »