Size advantage of E-M1 vs a compact DSLR

[cocoanud]

cocoanud wrote

In my backyard are an empty hot tub and an empty shot glass. It has just started raining. It is raining with the same intensity on both the glass and the tub. Which one do you think will collect more water?

Interesting thought there... i get the point that a larger sensor will gather a total of more light.

But in your example above, how long do you suppose it will(should) rain ?

Keeping this sub-thread on track, an f/1.4 lens on MFT (let's say focal length F) will need the same exposure time as f/1.4 2xF lens on full frame ... No ?

Assuming sensitivity is the same for all pixel sizes, the resulting images will be equally bright ? The one with MFT will have greater DoF ?

Now, can equivalence arithmetic say that an f/2.8 FF lens will give similarly bright image with the same exposure time as with our f/1.4 MFT lens ?

Sorry for dragging this on... But it's kinda intriguing.

Regards,

--

C

 

[jeffharris]

First, let's forget about FF cameras, shall we? The OP was asking about "compact DSLR cameras".

So, that means we should ONLY talk about Sony A7-series cameras, correct?

 

[ijm5012]

You hear a lot of people talk about "total light" when it comes to equivalence, and that an f/2.8 lens on m43 gathers less "total light" than an f/2.8 lens on a full frame camera. People love talking about this, but I can't for the life of me figure out what effect "total light" has on my final image.

I fully understand what effect FoV has on my image. If I like a 50mm perspective on a FF camera, I shoot with a 25mm lens on m43.

I also know what effect shutter speed has on a final image, where a shot with a SS of 1/500 on FF will have the same amount of motion blur or sharpness as an image taken on a m43 camera with a SS of 1/500.

The same goes with aperture. Ignoring any differences in actual light transmission between lenses, a shot on m43 @ ISO 200, f/2.8, 0 EC would yield the same shutter speed as a FF image shot @ ISO 200, f/2.8, 0 EC. Yes, DoF will be different, but that's just an inherent difference between formats (large format, medium formal, 35mm, APS-H, APS-C, M43, 1", etc.).

When I look at EXIF data for an image, regardless of what camera it was shot on, I see things like "focal length", "aperture", "shutter speed", exposure compensation", ISO", etc., but I never seem to find the "total light" output in the EXIF info. Strange...

When I compose a shot, I'm looking at things like FoV, exposure (histogram), shutter speed (to freeze or blur action). All of this is the same irrespective of the camera you use. Never when I'm setting up to take a shot do I think to myself, "OK now, what do I need to do to make sure I capture the right amount of total light?".

 

[texinwien]

f/5.6, ISO 1600 on FF will give you the same image brightness, shot noise and DoF as will f/2.8, ISO 400 on MFT.

Re 'shot noise': While this is true in theory, it doesn't tend to hold up in practice, at least not according to DXOMark. Last time I checked, there were no full frame cameras that were 2 full stops ahead of the best m43 sensors across their range of shared ISOs.

See this link for a comparison of the D750 and A7II to the E-M5, for instance.

So f/5.6 on FF will get you the same exposure and DOF as f/2.8 on m43, but it will also often result in poorer image quality in terms of shot noise and dynamic range at equivalent ISOs.

 

[Bobby J]

That both you and the reference poster were talking about image quality on a therotical 30x40 print. What I said was meant to spoof that. I have only rarely seen such a large print, although there are probably some who do. For about 98% of what most of us do, there is little difference readily visible between the latest m4:3 cameras and even FF. It's there to be sure, but only on close inspection. Perhaps a pro who is trying to produce the very best image possible for a client would be unwilling to use m4:3 and with good reason. For the vast majority of amateurs the convenience of the m4:3 system is likely more valuable than the ultimate IQ of the big, heavy systems. Even a growing body of pros are starting to use m4:3 as a "2nd" system. Kind of like the old days when I had a 35mm film camera and an RB 6x7 and even a Mamiya 645. The big cameras made the best images of course, but most of what I did was done with a OM-2 because it was so easy to carry and it was so versatile. And it was good enough for about 70% of what I did, even for clients.

Most of my prints are 18x24 or 16x20 being used primarily in camera club competitions. You won't see much difference in them when compared to APS or FF camera images. For practical purposes the difference is negligable. The largest print I've personally seen made from a m4:3 camera was 24x32 and it looked great. A local dealer even had a huge print made from an E-1, that's a 5 MP camera! It was beautiful and I suspect the print operater had as much to do with that as the actual camera, although it was a very good camera in it's day. As near as I could tell that print was about 40x60 and I have no other information on it. It was a portrait.

Even ISO 3200 is reasonably good with the latest m4:3 camera and looking at my files, I don't think I'd have a problem printing them at 16x20. Just how big do you need? Maybe 30x40? I think the OP probably is more interested in Pixel Peeping articles than actually making a 30x40.

 

[gear1box]

I've taken a look at the camerasize.com comparison between the E-M1 and the Canon SL1, and it seems that the size advantage is not as great as I'd have expected (although I can still see it in thickness). But I was wondering if the advantage is magnified once lenses are put on since m4/3 lenses are smaller than DSLR ones?

 

[cocoanud]

You hear a lot of people talk about "total light" when it comes to equivalence, and that an f/2.8 lens on m43 gathers less "total light" than an f/2.8 lens on a full frame camera. People love talking about this, but I can't for the life of me figure out what effect "total light" has on my final image.

I fully understand what effect FoV has on my image. If I like a 50mm perspective on a FF camera, I shoot with a 25mm lens on m43.

I also know what effect shutter speed has on a final image, where a shot with a SS of 1/500 on FF will have the same amount of motion blur or sharpness as an image taken on a m43 camera with a SS of 1/500.

The same goes with aperture. Ignoring any differences in actual light transmission between lenses, a shot on m43 @ ISO 200, f/2.8, 0 EC would yield the same shutter speed as a FF image shot @ ISO 200, f/2.8, 0 EC. Yes, DoF will be different, but that's just an inherent difference between formats (large format, medium formal, 35mm, APS-H, APS-C, M43, 1", etc.).

When I look at EXIF data for an image, regardless of what camera it was shot on, I see things like "focal length", "aperture", "shutter speed", exposure compensation", ISO", etc., but I never seem to find the "total light" output in the EXIF info. Strange...

When I compose a shot, I'm looking at things like FoV, exposure (histogram), shutter speed (to freeze or blur action). All of this is the same irrespective of the camera you use. Never when I'm setting up to take a shot do I think to myself, "OK now, what do I need to do to make sure I capture the right amount of total light?".

I agree, and that's what I am trying to ask in my previous post. But also looking for some objective answers since equivalence has been devious so far.

My belief was that with M43, with newer sensors and fast primes, I don't loose much up till ISO1600 when compared with FF. APS-C is a confused/confusing format right now given how much better the M43 generation has become.

(And I know... since I have shot with Oly E-500 14-54mm combo for about 3 years, then Nikon D90 and 17-50 f/2.8 Tamron for about 5 years and now with M43 with the lovely primes)

But then I see some comments which seem a bit out of reason and I chime in...

@Peace

--

C

 

[Jon555]

Most of the noise in an image isn't due to the camera but comes in with the light. Atoms don't chat to each other about when to release photons, they just do it at an average rate. The noise due to this averages at the square root of the light's photon count (so if you have 10,000 photons the average noise is 100). That's just Physics. Capture 4x the light and you have 1/2 the noise (assuming you make white the same so the signal from both is adjusted accordingly, strictly you have 4x the signal and 2x the noise).

Secondly do a thought experiment with a 20MP FF camera (6D say) vs a 20MP m43 (GX8). At the same f-stop you get the same light per unit area. Each FF pixel is 4x the area of the m43 one so gets 4x the Photons. Hence you can stop the FF lens down 2 stops to give the same DoF as m43 and you will get 1/4 the Photons in each pixel, which is exactly the same as the m43 camera and so the same noise level.

See the dpreview article linked previously for the long version.

 

[FingerPainter]

cocoanud wrote

In my backyard are an empty hot tub and an empty shot glass. It has just started raining. It is raining with the same intensity on both the glass and the tub. Which one do you think will collect more water?

Interesting thought there... i get the point that a larger sensor will gather a total of more light.

But in your example above, how long do you suppose it will(should) rain ?

Same amount of time for both containers will provide the same depth of water (but still, different volumes) . Same shutter speed for two formats will provide the same motion blur, but different amounts of total light.

Keeping this sub-thread on track, an f/1.4 lens on MFT (let's say focal length F) will need the same exposure time as f/1.4 2xF lens on full frame ... No ?

If you shoot f/1.4 on both formats, you will get the same exposure if you also use the same shutter speeds. Using the same shutter and aperture number on two different formats will yield different DoF, shot noise and diffraction blur. The smaller format will have more of all three.

Assuming sensitivity is the same for all pixel sizes, the resulting images will be equally bright ?

Depends what you mean by "sensitivity". Same ISO and same quantum efficiency will yield the same image brightness for a given set of scene luminance, shutter speed and aperture number.

The one with MFT will have greater DoF ?

Yes.

Now, can equivalence arithmetic say that an f/2.8 FF lens will give similarly bright image with the same exposure time as with our f/1.4 MFT lens ?

Sorry for dragging this on... But it's kinda intriguing.

Not at a problem at all. You are exploring knowledge. That's great. I'll answerr this last question over the following three paragraphs.

Most people already know that to get the same FoV you need to use an equivalent focal length. That's why everybody and their dog refers to a 25mm MFT lens as a 50mm equivalent. To determine a lens' FF equivalent focal length, multiply by the format's crop factor.

To get the same DoF and diffraction blur for a given FoV, you have to use the equivalent aperture number as well. To calculate the equivalent aperture number, multiply the aperture number you use on your format by the format's crop factor. The equivalent aperture number gives you equal aperture diameter. If you use the same aperture diameter and the same shutter speed on sensors of equal efficiency, you will get the same shot noise.

To get the same image brightness for a given combination of shutter speed and aperture diameter, you have to use the equivalent ISO. To calculate the equivalent ISO, multiply the ISO you use on the smaller format by the square of the crop factor. So ISO 200 on MFT is equivalent to ISO 800 on FF.

Some people get a little wary about this last point because they have learned to associate higher ISO with higher noise. But higher ISO doesn't cause higher noise. Less captured light is what causes higher noise. You get less captured light when either you have a lower exposure or a smaller sensor. The amount of light captured is the product of the exposure, the surface area of the sensor and the sensor quantum efficiency. (Sensor quantum efficiencies vary by much less than do sensor sizes or exposure values, so they are insignificant for purposes of this discussion.)

Exposure is the amount of light falling on the sensor per unit area during the time the shutter is open. For a given scene luminance, exposure is controlled by the product of shutter speed and aperture diameter (not aperture number) alone. Changing ISO on its own doesn't change exposure. The association of more noise with higher ISO exists because with a lower exposure you generally want to add more brightening. That's what increasing ISO does.

When shutter speed and aperture are both constrained, - i.e. you need a specific amount of motion blur and a specific DoF - a larger format provides no benefit regarding shot noise. It still retains any resolution advantage it might have. The benefit of FF wrt shot noise comes from two things.

1. In many shooting situations, shutter speed and aperture diameter are not both completely constrained.
2. In some cases, it can be appropriate to stop up to the wider aperture diameter most FF lenses can achieve.

Finally I'll mention that smaller formats provide no advantage regarding deeper DoF. Smaller formats have a deeper minimum DoF, but they have the same effective maximum DoF. That's because by the time you stop a smaller format lens down to an aperture diameter smaller than what the FF lens can achieve, any gains to sharpness due to deeper DoF have been cancelled out by additional unsharpness from diffraction. Thus there is no net increase in DoF from stopping down to such apertures, and indeed there may actually be a reduction of DoF by stopping down so far.

 

[FingerPainter]

First, let's forget about FF cameras, shall we? The OP was asking about "compact DSLR cameras".

So, that means we should ONLY talk about Sony A7-series cameras, correct?

Incorrect. A7 series are not DSLRs.

It might mena we should only talk about SL1, Digital Rebels, D3x00 series, D5x00 series, and most Pentax DSLRs, or it may men we should only talk about SL1.

 

[abe4652]

I'm not sure why I can't ask for a 2.8 lens in the FF world that gives me the equivalent of 80-300.

Say I shoot roving unicorns at dawn. I need 80-300, and I need 2.8, because of low light levels. I also want the increased IQ of a FF camera compared to a m43. In fact, I appreciate the increased DOF of the m43 because sometimes the unicorns get close, and face me rather than stand across the field of view.

You seem to think you need the same aperture number on FF to get the same image quality as on MFT. You don't.

No, I need a certain aperture number in order to be able to have a certain minimum shutter speed.

If you need 80-300 f/2.8 on FF then you need 40-150 f/1.4 on MFT. If you can get by with 40-150 f/2.8 on MFT then you can get by just as well with 80-300 f/5.6 on FF.

No, I can't, if I want a certain minimum shutter speed for a given level of illumination.

f/5.6, ISO 1600 on FF will give you the same image brightness, shot noise and DoF as will f/2.8, ISO 400 on MFT.

I don't know what "image brightness" is. I know that my camera maintains a given level of exposure based upon my decision to select a certain aperture by itself selecting a shutter speed. If my aperture is large enough, my shutter speed will be high enough.

OK, you want the improved IQ of the FF camera. Well that improvement, as far as shot noise goes, comes from capturing more light.

Make sense to me.

And to do that at the same shutter speed, you need a larger aperture diameter (not aperture number). That larger aperture diameter requires a larger lens, and produces a shallower DoF.

Makes sense to me.

 

[abe4652]

In fact, if I don't care about having shallower DOF, and in fact prefer having more DOF, the fact that the FF lenses gives me shallower DOF is not a plus.

FF will give you as deep DoF as MFT will. It will also give you shallower DoF when you need it. You don't need it, so as far as DoF goes, there is no difference to you between the systems. Many people would consider more at one end without less at the other to be an advantage.

Ok, but my point is that people keep throwing narrower DOF at me all the time whenever the issue arises between FF and m43. From my perspective, it's not important, so why can't we simply look at the other variables?

In this case, the FF lenses are much heavier and bigger when they give me the same maximum f stop and same effective reach.

Why does same f-number mean anything to you across formats? F-number is just a shortcut to make exposure calculations easier. It is exposure diameter that matters. That is what controls DoF, shot noise and diffraction.

I understand.

But if I shoot a bird at f/4 and 1/1000 on a FF, and get a proper exposure of the bird, I will get a proper exposure at f/4 and 1/1000 on m43.

No?

 

[FingerPainter]

f/5.6, ISO 1600 on FF will give you the same image brightness, shot noise and DoF as will f/2.8, ISO 400 on MFT.

Re 'shot noise': While this is true in theory, it doesn't tend to hold up in practice, at least not according to DXOMark. Last time I checked, there were no full frame cameras that were 2 full stops ahead of the best m43 sensors across their range of shared ISOs.

See this link for a comparison of the D750 and A7II to the E-M5, for instance.

Well, how fine do you want to slice this?

The theory doesn't say every FF will be precisely 2 stops better at everything when compared to every MFT. It gives a reasonable approximation of the difference of performance one can expect. There will be differences of small fractions of a stop due to engineering approach differences between models. If you want a more accurate approximation, you need to use more accutate comparison factors. If you want a precise and accurate comparison then you have to account for all engineering difference Since such differences are not generally universal by format. no one comparison factor will be perfectly precise and accurate for all cross-format model pair comparisons.

While the diagonal of an MFT sensor is almost exactly 1/2 of the diagonal of a FF sensor, its 4:3 aspect ratio, as opposed to FF's 3:2 aspect ratio, means the FF sensor has only 3.8 times the surface area of MFT, not 4x. When I compare the SNR 18% graph of the D750 to that of the E-M1 in the above link, I see a difference in SNR 18% performance at equal measured ISO that is essentially consistent with this difference in surface area. The difference ranges from about 1.5 stops at the EM-1's base ISO to 2.25 stops at its top ISO. That's an average difference of 15/8 stops, which is well within the tolerance of camera control increments (1/3 stop) of the amount you'd expect given the differences in sensor area.

So f/5.6 on FF will get you the same exposure and DOF as f/2.8 on m43, but it will also often result in poorer image quality w/r/t shot noise and dynamic range at equivalent ISOs.

Well that's partly because equivalent ISOs are not actually quite equivalent given the difference between the ratio of sensor areas and the ratio of squares of sensor diagonals. Most of the rest is due to differences in engineering design that result in similar differences between performance of models of the same format.

I happen to think that the best MFT cameras are typically slightly more efficient at low ISO's (where it matters least) compared to FF cameras at the same nominal ISO. They may get on the order of 1/3 stop better performance relative to what you'd otherwise expect from FF cameras at ISO 200. However, when you compare MFT to FF at the FF's equivalent ISO to MFT base, i.e. ISO 800, the difference has mostly disappeared. And as we see in the graph you linked to, at higher ISO,'s the MFT does worse than what you'd expect. MFTs pretty consistently have a higher slope to their graph than do FFs. IDK what causes this. Perhaps it is related to the different tone curve they use, which also results in a greater difference between reported and measured ISO.

The use of the sensor diagonal crop factor is a shortcut to compare performance across formats. It is not the most accurate for all forms of comparison. If you want to improve the accuracy of the comparison, you should actually use different conversion factors to compare different attributes. As long as FOV is measured relative to sensor diagonal, using crop factor for equivalence of FoV, DoF and diffraction is fine. When comparing noise, DR or any other factors based on total light, across formats with different aspect ratios, you will get more accurate comparisons if you use the ratio of sensor areas rather than square of crop factor.

The amount by which actual measured performance varies from the properly applied theory is usually less than the finest adjustment you can make on a typical camera control: 1/3 stop. The difference in performance between formats runs to multiple stops. So the use of the theory generally gives results that are significant as well as accurate within operational tolerances.

 

[FingerPainter]

I'm not sure why I can't ask for a 2.8 lens in the FF world that gives me the equivalent of 80-300.

Say I shoot roving unicorns at dawn. I need 80-300, and I need 2.8, because of low light levels. I also want the increased IQ of a FF camera compared to a m43. In fact, I appreciate the increased DOF of the m43 because sometimes the unicorns get close, and face me rather than stand across the field of view.

You seem to think you need the same aperture number on FF to get the same image quality as on MFT. You don't.

No, I need a certain aperture number in order to be able to have a certain minimum shutter speed.

No you don't. Increase your ISO. You don't need as low ISO on a larger sensor to get the same image quality.

If you need 80-300 f/2.8 on FF then you need 40-150 f/1.4 on MFT. If you can get by with 40-150 f/2.8 on MFT then you can get by just as well with 80-300 f/5.6 on FF.

No, I can't, if I want a certain minimum shutter speed for a given level of illumination.

So let's say you need to shoot at 1/500 to freeze unicorn movement. It's dawn, so EV 11. With your f/2.8 MFT lens, you have just enough light shooting wide open to stay at ISO 200.

Standing right beside you with my FF camera and f/5.6 lens (which is only slightly bigger than your lens), I can't get 1/500 at ISO 200. So I increase ISO to 800. Bingo. 1/500 is now possible. I take the shot and it turns out to have essentially the same motion blur, perspective, FoV, DoF, shot noise, and diffraction blur as your shot. Mine's a bit sharper though because I have more pixels. (I can't afford a D4s and I don't want a Df.)

f/5.6, ISO 1600 on FF will give you the same image brightness, shot noise and DoF as will f/2.8, ISO 400 on MFT.

I don't know what "image brightness" is.

Really? Well one way to think of image brightness is how close a middle tone is to appearing black or white. An image in which an object of middle gray appears close to white is brighter than an image where the same object appears close to black. Some people mistakenly call this "exposure." Exposure is actually the amount of light falling on the sensor per unit area. How bright an image is depends on both how high the exposure is and how high the ISO is set.

I know that my camera maintains a given level of exposure based upon my decision to select a certain aperture by itself selecting a shutter speed. If my aperture is large enough, my shutter speed will be high enough.

So you are in aperture priority mode. If you decrease your aperture number, the camera increases the shutter speed. What happens to the shutter speed if you increase the ISO instead and leave aperture unchanged?

OK, you want the improved IQ of the FF camera. Well that improvement, as far as shot noise goes, comes from capturing more light.

Make sense to me.

And to do that at the same shutter speed, you need a larger aperture diameter (not aperture number). That larger aperture diameter requires a larger lens, and produces a shallower DoF.

Makes sense to me.

 

[Martin.au]

I'm not sure why I can't ask for a 2.8 lens in the FF world that gives me the equivalent of 80-300.

Say I shoot roving unicorns at dawn. I need 80-300, and I need 2.8, because of low light levels. I also want the increased IQ of a FF camera compared to a m43. In fact, I appreciate the increased DOF of the m43 because sometimes the unicorns get close, and face me rather than stand across the field of view.

You seem to think you need the same aperture number on FF to get the same image quality as on MFT. You don't.

No, I need a certain aperture number in order to be able to have a certain minimum shutter speed.

No you don't. Increase your ISO. You don't need as low ISO on a larger sensor to get the same image quality.

If you need 80-300 f/2.8 on FF then you need 40-150 f/1.4 on MFT. If you can get by with 40-150 f/2.8 on MFT then you can get by just as well with 80-300 f/5.6 on FF.

No, I can't, if I want a certain minimum shutter speed for a given level of illumination.

So let's say you need to shoot at 1/500 to freeze unicorn movement. It's dawn, so EV 11. With your f/2.8 MFT lens, you have just enough light shooting wide open to stay at ISO 200.

Standing right beside you with my FF camera and f/5.6 lens (which is only slightly bigger than your lens), I can't get 1/500 at ISO 200. So I increase ISO to 800. Bingo. 1/500 is now possible. I take the shot and it turns out to have essentially the same motion blur, perspective, FoV, DoF, shot noise, and diffraction blur as your shot. Mine's a bit sharper though because I have more pixels. (I can't afford a D4s and I don't want a Df.)

I think you'll find, if you actually compared them side by side, that the sharpness from extra pixels is quite effectively offset by shooting at the higher ISO. That's been my observation comparing 36mp against m4/3. You'll also see a little more noise typically in the ff shot, as very few cameras achieve the two stops advantage expected from sensor area. The 16mp ff cameras are close.

f/5.6, ISO 1600 on FF will give you the same image brightness, shot noise and DoF as will f/2.8, ISO 400 on MFT.

I don't know what "image brightness" is.

Really? Well one way to think of image brightness is how close a middle tone is to appearing black or white. An image in which an object of middle gray appears close to white is brighter than an image where the same object appears close to black. Some people mistakenly call this "exposure." Exposure is actually the amount of light falling on the sensor per unit area. How bright an image is depends on both how high the exposure is and how high the ISO is set.

I know that my camera maintains a given level of exposure based upon my decision to select a certain aperture by itself selecting a shutter speed. If my aperture is large enough, my shutter speed will be high enough.

So you are in aperture priority mode. If you decrease your aperture number, the camera increases the shutter speed. What happens to the shutter speed if you increase the ISO instead and leave aperture unchanged?

OK, you want the improved IQ of the FF camera. Well that improvement, as far as shot noise goes, comes from capturing more light.

Make sense to me.

And to do that at the same shutter speed, you need a larger aperture diameter (not aperture number). That larger aperture diameter requires a larger lens, and produces a shallower DoF.

Makes sense to me.

 

[FingerPainter]

I'm not sure why I can't ask for a 2.8 lens in the FF world that gives me the equivalent of 80-300.

Say I shoot roving unicorns at dawn. I need 80-300, and I need 2.8, because of low light levels. I also want the increased IQ of a FF camera compared to a m43. In fact, I appreciate the increased DOF of the m43 because sometimes the unicorns get close, and face me rather than stand across the field of view.

You seem to think you need the same aperture number on FF to get the same image quality as on MFT. You don't.

No, I need a certain aperture number in order to be able to have a certain minimum shutter speed.

No you don't. Increase your ISO. You don't need as low ISO on a larger sensor to get the same image quality.

If you need 80-300 f/2.8 on FF then you need 40-150 f/1.4 on MFT. If you can get by with 40-150 f/2.8 on MFT then you can get by just as well with 80-300 f/5.6 on FF.

No, I can't, if I want a certain minimum shutter speed for a given level of illumination.

So let's say you need to shoot at 1/500 to freeze unicorn movement. It's dawn, so EV 11. With your f/2.8 MFT lens, you have just enough light shooting wide open to stay at ISO 200.

Standing right beside you with my FF camera and f/5.6 lens (which is only slightly bigger than your lens), I can't get 1/500 at ISO 200. So I increase ISO to 800. Bingo. 1/500 is now possible. I take the shot and it turns out to have essentially the same motion blur, perspective, FoV, DoF, shot noise, and diffraction blur as your shot. Mine's a bit sharper though because I have more pixels. (I can't afford a D4s and I don't want a Df.)

I think you'll find, if you actually compared them side by side, that the sharpness from extra pixels is quite effectively offset by shooting at the higher ISO.

Well no, having done so*, I don't find that at all.

Since shooting at equivalent ISO makes no difference to noise, there is no factor to negate the sharpness gained from more pixels on a large sensor.

*Not shooting mythical creatures with mythical lenses, but rather smaller low picel-count cameras and larger, higher pixel count camersa t equivalent settings.

That's been my observation comparing 36mp against m4/3.

Perhaps you were examining different portions of each image.

You'll also see a little more noise typically in the ff shot, as very few cameras achieve the two stops advantage expected from sensor area. The 16mp ff cameras are close.

Well, as I said in my recent reply to texinwien, the full two stops is a slight overexpectation because the FF sensor area isn't really twice as great, duie to teh differenc ein aspect ratio. Overall, 11/6 stops is a better approximation, but the relationship isn't linear with ISO. At low (for MFT) ISO, it may be 1.4 - 1.7 stops and at high (for MFT) ISO it might be 2-2.5 stops.

 

[dulynoted]

I've taken a look at the camerasize.com comparison between the E-M1 and the Canon SL1, and it seems that the size advantage is not as great as I'd have expected (although I can still see it in thickness). But I was wondering if the advantage is magnified once lenses are put on since m4/3 lenses are smaller than DSLR ones?

However, if you care about weight, then why E-M1? The E-M10 and E-M5 II will beat the weight and size with identical image quality.

You need to research your claims. E-M5 II is 1 gram lighter than E-M1 (496 vs. 497). In fact, because of better ergonomics, E-M1 will feel lighter in your hand than E-M5 II. The latter is still a great camera, though.

Not necessarily true, ergonomics are subjective and the em5 mII certainly feels much smaller than the em1

The EM5, EM5 II and EM10 all have rather poor ergonomics, pretty much demanding the use of an expensive add-on grip. Sure, they're sort of nice looking in a 1970s Gremlin kind of way, but that's a different issue entirely. The EM1 is a different class entirely, as are any of the GH-series cameras.

We will have to agree to disagree on that one Jeff. I can't stand the em1 grip personally and much prefer the feel of the em5 II (never held the first one). So far though my g6 still feels best

 

[FingerPainter]

In fact, if I don't care about having shallower DOF, and in fact prefer having more DOF, the fact that the FF lenses gives me shallower DOF is not a plus.

FF will give you as deep DoF as MFT will. It will also give you shallower DoF when you need it. You don't need it, so as far as DoF goes, there is no difference to you between the systems. Many people would consider more at one end without less at the other to be an advantage.

Ok, but my point is that people keep throwing narrower DOF at me all the time whenever the issue arises between FF and m43. From my perspective, it's not important, so why can't we simply look at the other variables?

Along with the shorter DoF, which may not be something you need, come less noise and less diffraction blur.

In this case, the FF lenses are much heavier and bigger when they give me the same maximum f stop and same effective reach.

Why does same f-number mean anything to you across formats? F-number is just a shortcut to make exposure calculations easier. It is exposure diameter that matters. That is what controls DoF, shot noise and diffraction.

I understand.

But if I shoot a bird at f/4 and 1/1000 on a FF, and get a proper exposure of the bird, I will get a proper exposure at f/4 and 1/1000 on m43.

No?

Yes you will, but if the bird is sitting on a fence or branch a couple of metres in front of a distracting background, the background will be a lot more distinct and distracting on m43. The dark parts of the frame will have more visible noise. You'll have more blurring from diffraction. To avoid all that, you'd need to shoot at f/2 and at 1/4 the ISO you used on FF. Often m43 just can't do that.

The important point is you shouldn't assume that you should always shoot at the same aperture, shutter and ISO settings when you use a larger format as you would when you use a smaller format. Doing so will give you a shorter DoF (which you might not want) as well as less noise and diffraction blur, (which are improvements you probably do want, otherwise why don't you just use a phone camera or a compact zoom camera?). Some of the time you will get better pictures if they are different. You certainly don't need to in order to get the same image brightness.

 

[Sergey_Green]

f/5.6, ISO 1600 on FF will give you the same image brightness, shot noise and DoF as will f/2.8, ISO 400 on MFT.

Re 'shot noise': While this is true in theory, it doesn't tend to hold up in practice, at least not according to DXOMark. Last time I checked, there were no full frame cameras that were 2 full stops ahead of the best m43 sensors across their range of shared ISOs.

If you are not comparing 16MP to 16MP,

http://www.dpreview.com/previews/nikon-d4s/7

Or take Nikon Df, and why is it so hard to find the comparison that is valid?

See this link for a comparison of the D750 and A7II to the E-M5, for instance.

Sure. While Sony may have slightly grainier noise, although it is about the same with 2 stops difference, it will also retain more detail in the images. It is 24MP camera.

http://www.dpreview.com/reviews/sony-alpha-7-ii/11

So f/5.6 on FF will get you the same exposure and DOF as f/2.8 on m43, but it will also often result in poorer image quality in terms of shot noise and dynamic range at equivalent ISOs.

As above, you may see more noise (you may?) if you take the higher MP camera for comparison. In practice you will end up with easier post-processing route on 24-36MP images, than dealing with noise from the 16MP camera. As an example. Besides the fact that the most shooters prefer shallower DoF over noise anyway, and there is only so much you need to stop the lens down when shooting from afar, as landscapes for example.

--
- sergey

 

[rovingtim]

You make a valid point. My point was that if you want a zoom with f/2.8 speed in the range of a 70-200mm FF equivalent for a FF or APS-C sensor camera you are looking at big bucks and big weight.

Dude, why do you always adjust for focal length (one is 200mm while the other is 100mm) but don't adjust for aperture (both f2.8)?

Loading…

camerasize.com