EF-M 22mm is sharp enough for 32 megapixel?

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,534
My 11-22 EF-M is a great lens for my M and M2. When I Switch to M5, the corners are not as sharp due to the higher resolution, especially with the IS being turned on. May be I have a bad copy.
My first copy was very soft at the edges on the M5. The second very good, close to that of the 22mm prime.
 
Upvote 0
Jan 29, 2011
10,675
6,121
Positing the question in the first place illustrates a misunderstanding of system resolution.

No sensor or lens 'out resolves' another, as a system a different combination might give you more or less resolution but that is irrelevant. Put the bottom of a bottle in front of two sensors one with more resolution and you will get more resolution from the one with more MP, you will never get the stated increase of the difference between sensors however well the lens resolves, but you will get an increase irrespective of the lens you put in front of it.

 
  • Like
Reactions: 1 user
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,534
Positing the question in the first place illustrates a misunderstanding of system resolution.

No sensor or lens 'out resolves' another, as a system a different combination might give you more or less resolution but that is irrelevant. Put the bottom of a bottle in front of two sensors one with more resolution and you will get more resolution from the one with more MP, you will never get the stated increase of the difference between sensors however well the lens resolves, but you will get an increase irrespective of the lens you put in front of it.

What you say is technically correct, and is often quoted. But, it is not unreasonable to talk about the sensor or the lens outresolving the other in some circumstances. There are many processes in physics, chemistry, engineering, biology etc where the overall result depends on the contributions from all of the components. However, one of the components may be a rate limiting process and to all intents and purposes it is the major factor that determines the overall properties of the system. If the the resolutions or MTFs of the lens and the sensor are not too different, then both have crucially to be taken into account for calculating the overall MTF of the system. If they are very different, then the magnitude of the weaker component dominates. Take one extreme where the sensor has just one pixel or sensel. The overall resolution will be independent of the MTF of the lens, which truly outresolves the sensor. This can be expressed mathematically as follows. To a first approximation, the overall resolution of the system, r , depends on the resolution of the lens r1 and that of the sensor r2 according to the equation: 1/r = 1/r1 + 1/r2. If r1 >> r2, then 1/r ~ 1/r2, and the lens is outresolving the sensor. And conversely if If r1 << r2, then 1/r ~ 1/r1, and the sensor is outresolving the lens.
 
  • Like
Reactions: 2 users
Upvote 0
Jan 29, 2011
10,675
6,121
What you say is technically correct, and is often quoted. But, it is not unreasonable to talk about the sensor or the lens outresolving the other in some circumstances. There are many processes in physics, chemistry, engineering, biology etc where the overall result depends on the contributions from all of the components. However, one of the components may be a rate limiting process and to all intents and purposes it is the major factor that determines the overall properties of the system. If the the resolutions or MTFs of the lens and the sensor are not too different, then both have crucially to be taken into account for calculating the overall MTF of the system. If they are very different, then the magnitude of the weaker component dominates. Take one extreme where the sensor has just one pixel or sensel. The overall resolution will be independent of the MTF of the lens, which truly outresolves the sensor. This can be expressed mathematically as follows. To a first approximation, the overall resolution of the system, r , depends on the resolution of the lens r1 and that of the sensor r2 according to the equation: 1/r = 1/r1 + 1/r2. If r1 >> r2, then 1/r ~ 1/r2, and the lens is outresolving the sensor. And conversely if If r1 << r2, then 1/r ~ 1/r1, and the sensor is outresolving the lens.
In some circumstances I'd agree, in almost any practical application of currently available cameras and lenses I wouldn't, not least of which because in general we actually have very little choice and this is especially true of the M series. The two exceptions I can think of to hand would be if you found yourself focal length limited regularly or if you regularly print large and require narrow depth of field.

For instance if I want a 22mm and IQ from the M series people would argue the 11-22 is great and the 22 prime is not sharp enough for the M6 II, I'd say compare the 22 prime at f5.6 (which is the fastest aperture of the 11-22 at 22), do that and there's nothing in it.



Take another example, the 35 f1.4L II and the 35mm f2 IS, put them both on a 5DSr and shoot at f2, the differences are minimal though the L lens is, unsurprisingly, slightly better. But it cost three times the money and didn't affect the compelling nature of the image one iota. Stop them down to f5.6 and there really is no difference in IQ.



You, Alan, look at these things from a very specific and atypical point of view, you are almost always focal length limited and pushing the limits of your resolution, but as I say, that is not typical and in general for people that are focal length limited they are cropping from close to the center of their image which is generally the highest resolving area of the lens. I'd venture nobody is shooting birds with an M6 II and a 22mm lens.

If anybody asked me if they should buy a 35mm f1.4L or a 35mm f2 lens because their camera can out resolve the more modest lens I'd say get the cheap prime and spend the $1,000 you saved on photography education, lighting, or a trip somewhere you'd find compelling content.
 
  • Like
Reactions: 1 user
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,534
I wasn't replying to any particular lens or sensor or what lenses I use but the general principles pertinent to:

Positing the question in the first place illustrates a misunderstanding of system resolution.

No sensor or lens 'out resolves' another, as a system a different combination might give you more or less resolution but that is irrelevant. Put the bottom of a bottle in front of two sensors one with more resolution and you will get more resolution from the one with more MP, you will never get the stated increase of the difference between sensors however well the lens resolves, but you will get an increase irrespective of the lens you put in front of it.

 
Upvote 0
Jan 29, 2011
10,675
6,121
I wasn't replying to any particular lens or sensor or what lenses I use but the general principles pertinent to:
I understand that, and within the real world situations most of us shoot in most of the time with the gear actually purchasable I say the question displays a complete lack of understanding of the concept of system resolution, which was my original point. The maths can be debated and we can theorize down to one sensel sensors but why when it really makes so little (virtually no) difference to actual real world images and output almost all the time?

Now let's get back to the practical uses of 4k 8k and DR ;)
 
Upvote 0