Is micro 4/3 dead?

Some interesting thoughts about the future of micro 4/3 by Tony Northrup on https://www.youtube.com/watch?v=qjXSnNMZ0PU

Some of Tony’s main points in the first video are that:
(1) micro 4.3rds cameras were introduced because FF sensors were very expensive and smaller sensors made cameras much cheaper. But now sensors are so much cheaper that using an FF sensor no longer makes a camera significantly more expensive. So why bother, particularly if you have limited R&D funds and you’ve just introduced a new FF format.

(2) smaller sensors were said to allow smaller lenses to be made, but that is only true if you keep the numerical aperture the same (F no.). If you insist on the same light-collecting and thin-DOF qualities as the FF camera, then the lens’s optical aperture (entrance pupil) must stay the same, which means the lens diameter at least must stay the same, and in order to fit enough pieces of glass inside, probably the length does too. (I exempt here the older long teles with acres of empty space inside).

(3) If you keep the angle of view the same and also the optical aperture then the sensor collects the same amount of light whatever the sensor size you use, provided that the FL and F no. is altered to suit. FF has no more or worse dynamic range nor noise level than APS-C or micro 4/3 over the whole frame.
Of course the sensor size does affect the relationship between pixel density and Mp size, and for the highest Mp it may well be easier and more effective to get this in a larger sensor. But if we stick with (say) 12Mp for low light use then the sensor size makes no difference provided we can get lenses designed to suit.

This is where the theoretical arguments fall down, of course, since as consumers we need the lenses to be available.

(4) Tony also claims that claims of smaller-sensor compact cameras (including APS-C) are falling through the floor (my phrase), though I’m not convinced they are heading to zero.

So what’s my conclusion?
Tony is essentially right, with reservations. The advantage in small sensor sizes disappears when you demand equal low-light performance and DOF (which the lower end of the enthusiast market never did), and when you want to combine this with high pixel count then bigger sensors (currently) have a practical advantage. So in the current market, the demand for high Mp numbers and the availability of FF lens designs make more sense to make FF cameras with an APS-C option than micro 4/3.

1. MFT is not a sensor size, tony is simply wrong with what he says about sensors being cheaper originally. Four Thirds was introduced specifically with digital photography by Olympus with the E-1! Four Thirds is the sensor size not MFT which is the mount.

It looks like Tony and a few others have decided that MFT is the new whipping boy for click bait champions like him, be interesting to see what Olympus do on the 100th Centenary next year.

Olympus have always taken their own way with much of the industry catching up later!
Live View, mirrorless and tilting/swivelling screens spring to mind!

https://www.digitalcameraworld.com/features/olympus-om-d-e-m1x-what-we-want-to-see-and-what-were-likely-to-get

Miranda F wrote: (1) micro 4.3rds cameras were introduced because FF sensors were very expensive and smaller sensors made cameras much cheaper. But now sensors are so much cheaper that using an FF sensor no longer makes a camera significantly more expensive.

A full-frame sensor will always cost considerably more than 2.34x an APS-C one or 4x and Four-Thirds one. What those costs are currently, I don't know.

The Micro Four-Thirds lens system seems more expensive than FE to me, which is enough reason for it be discontinued. It is too bad. I really like the Panasonic bodies. Olympus seems to have some of the best sensors available today.

Edited by QuietOC - 01 November 2018 at 15:38

Jonas A-R wrote: Four thirds is the aspect ratio.

It is more the sensor size then - which includes the aspect ratio, but not every four thirds aspect ratio is four thirds.

Anyway, here is a curious observation by Thom Hogan. I think he is spot on, m43 needs to find its USP and that is size.

Jonas A-R wrote: Miranda F wrote: (3) If you keep the angle of view the same and also the optical aperture then the sensor collects the same amount of light whatever the sensor size you use, provided that the FL and F no. is altered to suit. FF has no more or worse dynamic range nor noise level than APS-C or micro 4/3 over the whole frame. Of course the sensor size does affect the relationship between pixel density and Mp size, and for the highest Mp it may well be easier and more effective to get this in a larger sensor. But if we stick with (say) 12Mp for low light use then the sensor size makes no difference provided we can get lenses designed to suit. This is plain wrong. If you expose for the same ISO, FF will provide 4 times as much light energy for the same AoV and aperture and will thus provide a 2 times better signal-to-noise ratio. That is, if you can use longer exposure times in natural light or can increase the light with strobes, larger formats have inherently better noise performance if everything else is equal.

Sorry, but your analysis is incorrect, and mine is right!
For a given angle of view and a given scene, the amount of light entering the lens (and heading toward the sensor) is determined only by the size of the lens opening - the entrance pupil = optical aperture = Focal length/numerical aperture.

Apart from the fact that the lens opening is circular and the sensor is rectangular, so some light is always lost in the corners, all the light reaches the sensor whatever size the sensor is. If you have a big sensor the light is spread out, if it is a small sensor it is concentrated. But it's the same amount of light.

I agree that if you use the exact same lens for both sensors then clearly a small sensor intercepts less light, but that is not the issue. Tony and I are both discussing a system where the lenses are designed to suit the sensor, and he makes the good point that camera systems with small sensors suffer because they don't have fast enough glass to suit them, not because of the sensor per se.

As Tony points out, much of the confusion occurs because people who should know better keep referring to 'FF' equivalent focal lengths without using 'FF equivalent' apertures.

Miranda F wrote: Jonas A-R wrote: Miranda F wrote: (3) If you keep the angle of view the same and also the optical aperture then the sensor collects the same amount of light whatever the sensor size you use, provided that the FL and F no. is altered to suit. FF has no more or worse dynamic range nor noise level than APS-C or micro 4/3 over the whole frame. Of course the sensor size does affect the relationship between pixel density and Mp size, and for the highest Mp it may well be easier and more effective to get this in a larger sensor. But if we stick with (say) 12Mp for low light use then the sensor size makes no difference provided we can get lenses designed to suit. This is plain wrong. If you expose for the same ISO, FF will provide 4 times as much light energy for the same AoV and aperture and will thus provide a 2 times better signal-to-noise ratio. That is, if you can use longer exposure times in natural light or can increase the light with strobes, larger formats have inherently better noise performance if everything else is equal. Sorry, but your analysis is incorrect, and mine is right! For a given angle of view and a given scene, the amount of light entering the lens (and heading toward the sensor) is determined only by the size of the lens opening - the entrance pupil = optical aperture = Focal length/numerical aperture. Apart from the fact that the lens opening is circular and the sensor is rectangular, so some light is always lost in the corners, all the light reaches the sensor whatever size the sensor is. If you have a big sensor the light is spread out, if it is a small sensor it is concentrated. But it's the same amount of light. I agree that if you use the exact same lens for both sensors then clearly a small sensor intercepts less light, but that is not the issue. Tony and I are both discussing a system where the lenses are designed to suit the sensor, and he makes the good point that camera systems with small sensors suffer because they don't have fast enough glass to suit them, not because of the sensor per se. As Tony points out, much of the confusion occurs because people who should know better keep referring to 'FF' equivalent focal lengths without using 'FF equivalent' apertures.

This is a strawman argument. You are talking about different situations, you are both right. But please, do not talk equivalence, as it always seems to turn out nasty and it is completely irrelevant. I do not take pictures to try to emulate a picture I could have made with a different camera.

Jonas A-R wrote: A m4/3 set to f/2, 1/400 and a FF set to f/4 1/100 does not have the same SNR although they have identical apertures. The FF system will have captured 4x the light energy, precisely as I stated above.

That's true if you expose for the same ISO, as you said earlier.

If you instead keep shutter speeds the same and let ISO float as needed, then everything - including motion blur and camera shake - remains equal, right?