Aperture equivalency in M4/3's vs Full Frame

[Syringer]

I was reading this review: http://www.cameralabs.com/reviews/Olympus_M_Zuiko_Digital_45mm_f1-8/verdict.shtml

Which states: "So in full-frame terms, the Olympus 45mm f1.8 acts like a 90mm f3.6 lens."

Is that entirely accurate? Obviously the 45/90mm crop factor is, but do aperture sizes really get cut in half?

 

[blastingcap]

The short answer is yes.

The focal length of a lens never changes regardless of the sensor placed behind it. Ever.

However, if you use a smaller sensor with that lens, it will capture less light, and the angle of the light rays that do strike it, will be at a smaller angle, thus more DoF.

So if you were to take a 90mm f/3.6 lens and place a FF sensor behind it, it would like a lot like a 45mm f/1.8 lens with a MFT sensor behind it.

Imho DoF control is overrated. In real life, few non-newbies need paper-thin DoF, and such DoF can be a hindrance when you are trying to get more things in focus. It's cool at first but can be overused, like any other photographic effect. Also note that something like the Panaleica 25mm f/1.4 or Oly 45mm f/1.8 is pretty sharp in the center even wide open, whereas many FF lenses are not as sharp wide open, so if you have to stop down a 90mm f/1.8 lens to f/2.8 on FF to get it as sharp as the Oly 45mm f/1.8 at f/1.8, the FF advantage is much reduced.

 

[Sid59]

I was reading this review: http://www.cameralabs.com/reviews/Olympus_M_Zuiko_Digital_45mm_f1-8/verdict.shtml

Which states: "So in full-frame terms, the Olympus 45mm f1.8 acts like a 90mm f3.6 lens."

Is that entirely accurate? Obviously the 45/90mm crop factor is, but do aperture sizes really get cut in half?

I just assumed the aperture is constant and light gathering remains the same but what changes from FF,DX is the DOF and focal length. Obviously, we can't just say aperture changes since it's tied to FL and DOF.

DPreview forums should have some threads about it.

 

[AkumaX]

I was reading this review: http://www.cameralabs.com/reviews/Olympus_M_Zuiko_Digital_45mm_f1-8/verdict.shtml

Which states: "So in full-frame terms, the Olympus 45mm f1.8 acts like a 90mm f3.6 lens."

Is that entirely accurate? Obviously the 45/90mm crop factor is, but do aperture sizes really get cut in half?

Yup! If you got a 5D and stuck a 90mm f/3.6 lens, that's the equivalent of the m4/3 45mm @ f/1.8

The chart in the middle of this page says it best!

http://www.dpreview.com/reviews/sony-cybershot-dsc-rx100

 

[Throckmorton]

The DOF is the same as a 45mm f1.8 lens on a full frame or APS-C or whatever. If that could change by changing sensor size, then cropping would alter DOF, which obviously isn't true!


But I guess what matters here is that the DOF you get at 45mm f1.8 is the same as what you get with 90mm f3.6... and the comparison maters because the field of view is the same as 90mm on a full frame camera.

 

[runawayprisoner]

I think the basic conclusion is... since the focal length is effectively longer (because the image is cropped) but with the same DOF (from the photographer's point of view), scaling should apply to aperture as well.

Otherwise, 90mm f/1.8 is razor thin DOF in full-frame's term.

 

[AkumaX]

I think the basic conclusion is... since the focal length is effectively longer (because the image is cropped) but with the same DOF (from the photographer's point of view), scaling should apply to aperture as well.

Otherwise, 90mm f/1.8 is razor thin DOF in full-frame's term.

Yeah, I have an 85mm f/1.8 on a 5D2, and it's lots of fun!

 

[slashbinslashbash]

Just note, the equivalence here is ONLY in terms of DOF and focal length. In terms of exposure, of course the f/1.8 lens is letting in f/1.8 worth of light.

 

[iGas]

Background separation is a great quality with large aperture and FF, just like medium format or large format vs 35mm in the film world.

 

[blastingcap]

Just note, the equivalence here is ONLY in terms of DOF and focal length. In terms of exposure, of course the f/1.8 lens is letting in f/1.8 worth of light.

Yeah, for exposure times, but also keep in mind that the sensor is also smaller so the tolerance for high ISOs goes down as you go down in sensor size. I'm not dissing small-sensored cameras by the way (my primary camera is a RX100 with 1" sensor), just saying.

 

[Throckmorton]

Just note, the equivalence here is ONLY in terms of DOF and focal length. In terms of exposure, of course the f/1.8 lens is letting in f/1.8 worth of light.

No it's not. The total amount of light depends on the diameter of the front element and how much of incident light it manages to focus onto the sensor. A smaller lens/sensor captures less light, even at the same f number.

You could make a lens the size of a rice grain with f1.8. Doesn't mean it captures as much light as a full frame lens.

 

[slashbinslashbash]

No it's not. The total amount of light depends on the diameter of the front element and how much of incident light it manages to focus onto the sensor. A smaller lens/sensor captures less light, even at the same f number.

You could make a lens the size of a rice grain with f1.8. Doesn't mean it captures as much light as a full frame lens.

Yes and no. A 16mm f/2.8 and a 300mm f/2.8 admit the same amount of light, despite the front element of the latter typically being 2-3x the size of the front element of the former. Given an identical subject, and identical lighting, both lenses will give identical exposures when mounted on an identical camera.

Your hypothetical rice-grain-sized lens would likely not produce an image circle large enough to cover a full-frame sensor, but what image it does produce, would be of identical exposure to a standard full-frame lens at f/1.8.

Now you may say, "so it's obviously letting in less light". Yes, the number of photons able to pass through a rice-grain-sized lens is obviously less than the number of photons able to pass through a standard-sized lens. However, photographically speaking, they still give the same exposure values. ISO and shutter speed remain the same; the f/1.8 is "real" regardless of how much of the sensor the image covers.

 

[Throckmorton]

Yes and no. A 16mm f/2.8 and a 300mm f/2.8 admit the same amount of light, despite the front element of the latter typically being 2-3x the size of the front element of the former. Given an identical subject, and identical lighting, both lenses will give identical exposures when mounted on an identical camera.

Your hypothetical rice-grain-sized lens would likely not produce an image circle large enough to cover a full-frame sensor, but what image it does produce, would be of identical exposure to a standard full-frame lens at f/1.8.

Now you may say, "so it's obviously letting in less light". Yes, the number of photons able to pass through a rice-grain-sized lens is obviously less than the number of photons able to pass through a standard-sized lens. However, photographically speaking, they still give the same exposure values. ISO and shutter speed remain the same; the f/1.8 is "real" regardless of how much of the sensor the image covers.

Yes the brightness of the exposure is the same, but that's separate from the amount of light absorbed by each pixel. The rice grain lens would have a smaller sensor so with the same resolution the pixels would be a lot smaller

 

[slashbinslashbash]

Yes the brightness of the exposure is the same, but that's separate from the amount of light absorbed by each pixel. The rice grain lens would have a smaller sensor so with the same resolution the pixels would be a lot smaller

No, actually the light absorbed by each pixel is the same. That's what determines the exposure, the amount of light absorbed by each pixel. The total amount of light, summed across all pixels, will obviously be smaller with a smaller lens, but each individual pixel (in the middle of the image at least) will receive the same amount of light as it would from a bigger lens.

Assuming we're keeping the pixel size the same here (i.e. resolution goes down as the sensor gets smaller). Obviously smaller pixels = less light hitting each pixel.

 

[runawayprisoner]

I think you're taking some things backward.

Different focal lengths make light (photons) travel different distances. It's not like the inside of all lens is a complete vaccuum, so taking that into account, shorter focal lengths at the same aperture will always capture more incident light, and that means shorter focal lengths = higher exposure.

And then with cropped sensors, only a part of that incident light hits the sensor. On m4/3 cameras, only half of the light is captured. The other half is just gone. It's not like you can get the other half of the photo no matter how many more effective pixels you put on the sensor.

And effective exposure may remain the same, but shutter speed and actual exposure may still be different depending on ISO settings, and how the camera handles those ISO settings. I know there are m4/3 cameras with far better noise control at higher ISOs than APS-C or even full frame bodies, and their ISO settings seriously inflate their shutter speed and exposure.

 

[Throckmorton]

No, actually the light absorbed by each pixel is the same. That's what determines the exposure, the amount of light absorbed by each pixel. The total amount of light, summed across all pixels, will obviously be smaller with a smaller lens, but each individual pixel (in the middle of the image at least) will receive the same amount of light as it would from a bigger lens.

Assuming we're keeping the pixel size the same here (i.e. resolution goes down as the sensor gets smaller). Obviously smaller pixels = less light hitting each pixel.

The individual pixels ARE smaller on a smaller sensor! The micro 4/3 sensor is half the size of a full frame and the megapixels are about the same, so there's 1/4 the light absorption for each pixel at the same aperture and focal length.

 

[slashbinslashbash]

The individual pixels ARE smaller on a smaller sensor! The micro 4/3 sensor is half the size of a full frame and the megapixels are about the same, so there's 1/4 the light absorption for each pixel at the same aperture and focal length.

Yes, but this is not always the case, and you've got to keep all other things equal if you want to talk about this stuff in a meaningful way. Talking about ISO performance varying according to pixel size, sensor size, etc. is just confusing the issue. The exposure is calculated based on the amount of light hitting a given area of the sensor, period... no matter whether there's a lot of pixels in that area or not, or whether there's more pixels outside that area or not.

The exposure numbers have been the same since the early days of FILM photography, when pixels didn't enter the equation at all! You're too stuck in the current iteration of photographic technology to think about the problem in the right way. You've got to abstract things a bit, wipe away the technicalities of pixels and sensors and just think about light coming through the lens and hitting an object that is sensitive to light.

I have a 5D, and I had a 20D. They used almost the same sensor technology, the 20D had 8.2MP while the 5D had 12MP. Their pixel pitch wasn't exactly the same, but it was relatively close. I think 6.4µ for the 20D and 8.2µ on the 5D. You could stick the same lenses on both, set the ISO's and shutter speeds and apertures to the same values, and you'd get approximately the same image from both (one just cropped to the middle).

I could take a Sigma 30mm f/1.4 and it would work great with the 20D. Put it on the 5D, and the corners would be quite a bit darker, but the middle part of the image would be exactly the same as it was on the 20D.

Or take the opposite situation, take a full-frame lens like a Canon 70-200, put it on the 5D and you get the roughly the same photo as you do with the lens mounted on the 20D with the same exposure settings; but the 20D is cropped. The cropping throws away a good amount of the light that was transmitted through the lens (and that landed outside of the smaller sensor area of the 20D) but the exposure -- for the pixels that were exposed -- is the same.

ISO is ISO, aperture is aperture, and from the photographer's point of view it doesn't change at all from full-frame, to crop, to m4/3, to P&S (outside of DOF, of course). Yes the manufacturers might "juice" the numbers somehow, but overall they are trying to match the final result with the same film standards that photographers have used for 100 years. There is too much established already for a company to mess with anything too much.

There are plenty of photogs who use light meters, which tell you what your exposure needs to be based on the light available. These meters are the exact same as they were back in the film days. If a camera's exposure didn't match the numbers from the light meter, it would be found out pretty quickly. Granted I have seen some cases where one camera or lens might vary from another by 1/2 stop or so, but those are the rare cases.

 

[Throckmorton]

Yes I understand that "exposure" is the brightness of light hitting the imaging surface, but for a given image brightness with the same exposure, the smaller sensor with smaller pixels needs more voltage boost to get to the same ISO.

 

[slashbinslashbash]

Yes I understand that "exposure" is the brightness of light hitting the imaging surface, but for a given image brightness with the same exposure, the smaller sensor with smaller pixels needs more voltage boost to get to the same ISO.

Who cares about the voltage boost? Does that make the ISO any less "real"? Do you ever change the voltage boost directly in the camera settings?

This is like talking about the pigments used in film; yes, they might lead to different results, but it's not something you can change directly. If you don't like the results that a certain camera gets at certain ISO, you choose a different camera... same as if you don't like the results a certain film gives, you choose a different film. Unless you're some kind of mad scientist type, you don't modify the voltage regulators in the camera, or chemically alter the film. It's beyond your capacity to change... so why get so technical?

Of course I can understand a technical discussion of various cameras' capabilities, but to the photographer, all of that stuff is abstracted away. All that you have control over is your ISO, your shutter speed, and your aperture.

And that's what the OP was asking: is the f/1.8 in a m43 lens technically different from the f/1.8 in a full-frame lens? And the answer is, no. The aperture of a lens is measured the exact same way regardless of what kind of camera it is mounted to. If you take a photo with 1/100s, f/1.8, ISO 200 on a full-frame, you ought to be able to get the exact same photo (exposure-wise; DOF notwithstanding) with 1/100s, f/1.8, ISO 200 on a m43 camera. Full-stop. Maybe the m43 will have more noise due to the sensor. I'm sure you could find an old full-frame camera which is noisier. Of course there will always be differences from body to body, but when it comes to exposure, it is exactly the same (or at least it tries to be, within say a 1/3rd stop variance).

 

[blastingcap]

I think he's saying that the extra voltage also generates noise, and that's why people try to avoid high ISO whenever possible. Jacking up ISO is a last resort if you can't use a tripod, flash, image stabilization, faster lens, bigger sensor, etc.

 

[Throckmorton]

Who cares about the voltage boost? Does that make the ISO any less "real"? Do you ever change the voltage boost directly in the camera settings?

This is like talking about the pigments used in film; yes, they might lead to different results, but it's not something you can change directly. If you don't like the results that a certain camera gets at certain ISO, you choose a different camera... same as if you don't like the results a certain film gives, you choose a different film. Unless you're some kind of mad scientist type, you don't modify the voltage regulators in the camera, or chemically alter the film. It's beyond your capacity to change... so why get so technical?

Of course I can understand a technical discussion of various cameras' capabilities, but to the photographer, all of that stuff is abstracted away. All that you have control over is your ISO, your shutter speed, and your aperture.

And that's what the OP was asking: is the f/1.8 in a m43 lens technically different from the f/1.8 in a full-frame lens? And the answer is, no. The aperture of a lens is measured the exact same way regardless of what kind of camera it is mounted to. If you take a photo with 1/100s, f/1.8, ISO 200 on a full-frame, you ought to be able to get the exact same photo (exposure-wise; DOF notwithstanding) with 1/100s, f/1.8, ISO 200 on a m43 camera. Full-stop. Maybe the m43 will have more noise due to the sensor. I'm sure you could find an old full-frame camera which is noisier. Of course there will always be differences from body to body, but when it comes to exposure, it is exactly the same (or at least it tries to be, within say a 1/3rd stop variance).

What I'm saying is that 800 ISO on a tiny sensor like the ones we are talking about is going to be much noisier than the same ISO on an APS-C or full frame sensor. It's abstracted but that doesn't mean it doesn't matter

Otherwise why aren't we using tiny DSLRs that fit in your pocket? Obviously the amount of light hitting each pixel does matter. It's the opposite of the film world where a smaller piece of film capturing the same field of view just has a lower resolution... Unless the micro 4/3 sensors suddenly are reduced to 5 megapixel.

 

[slashbinslashbash]

What I'm saying is that 800 ISO on a tiny sensor like the ones we are talking about is going to be much noisier than the same ISO on an APS-C or full frame sensor. It's abstracted but that doesn't mean it doesn't matter

Otherwise why aren't we using tiny DSLRs that fit in your pocket? Obviously the amount of light hitting each pixel does matter. It's the opposite of the film world where a smaller piece of film capturing the same field of view just has a lower resolution... Unless the micro 4/3 sensors suddenly are reduced to 5 megapixel.

Ok, I have no problem with that. It is perfectly fine to say that ISO 400 on Camera X is as noisy as ISO 800 on Camera Y. Make those comparisons all you want. But don't try to say that ISO 400 on Camera X leads to a different exposure than ISO 400 on Camera Y.

Also, it is a fact that high ISO performance has generally been getting better even as the pixel density goes up. The sensor tech IS getting better. E.g. the 5D3 vs. the 5D2, the 5D3 has more pixels but better high ISO performance. Not the most extreme example, but an example. Could also compare, say, the 7D vs. 40D. Or even 7D vs. 10D. There has been a rising tide that has lifted all sensor size boats. So yeah, in any given generation of cameras, the bigger sensor will generally outperform the smaller sensor, but the smaller sensors of today are able to match the larger sensors of a few years ago.