'Can Micro 4/3 lenses actually resolve the additional resolution of high res mode?'

The short answer is yes. Even the kit lens.

 

Link: Can m43 lenses actually resolve the additional resolution of high res mode?

Per convention with Imatest results, the resolution is expressed as "lines per image height." In high resolution mode, the height at 50 MP is 5770 pixels. By the test, resolution is a little over 50% of the theoretical maximum. The same test applied to a FF Sony (42 mp) with high performance lenses is as much as 4500 lph, or 85% of theoretical. In any case, the indicated resolution depends on a composite the resolution of the sensor and the lens, approximated as the root sum of variances.

 

Resolution of the sensor is further degraded by the use of an anti-aliasing filter. Most 24 MP sensors (or less) use AA filters, and switching to a high resolution mode would not change that. Roughly speaking, the AA filter reduces resolution by about 30%. The following test was performed using a Sony A7Riii, which has no AA filter. The peak resolution of a Batis 40/2 lens is shown as 5000 lph, which is 95% of theoretical.

 

Review Zeiss Batis 2/40 CF | Zeiss lens review | Lens | Reviews

 

Sharpness isn't everything, but it's a start.

Edited May 9, 2019 by Ed_Ingold

Sharpness may not be everything, but it's the one thing you can't add in post. I'll take sharp any day since you can always soften in post.

An additional consideration, the high resolution mode also cancels out a lot of noise you would normally see when shooting these cameras in normal one shot mode and higher ISO settings. I have some images, like the one below, shot at ISO 3200 and processed in Adobe Camera raw with no luminance noise filtering used, and the results are very nice compared to had I used the regular shot mode at the same ISO setting. The exposure are captured extremely fast so easily done handheld but there are severe limitations to the use of this mode. Any amount of substantive movement, limbs blowing in the wind, birds, etc will not be captured like in that single moment of time if shot using a higher shutter speed in the normal one-shot mode, so you need to limit the use to fairly static subjects, unless you want to show that affect for some reason. I would have loved using this when in Paris and shooting in museums and cathedrals.

 

Here is a link to the full-sized image, captured with the E-M1X and 12-100mm f4.

 

Olympus E-M1X - Greg

 

 

 

Here is a link to the full-sized image, captured with the E-M1X and 12-100mm f4.

 

Olympus E-M1X - Greg

Thanks for posting that. What a huge file! At a pixel level, yes, you can see the noise. If you downsample to 4042x3032 you get a very nice 12Mpx file. Of course in practice you might to just shoot a single exposure at 20Mpx, at a lower ISO, using IBIS. That would arguably get you a much better image.

Yeah, that's what I do most of the time. That was me just playing with the handheld high-rez mode to see what was possible. Outdoors most all the time and indoors much of the time I would never use it as rarely is everything that still.

Okay, got it.

 

A little bloke on steroids is just as good as a big bloke.

 

Really?

I'm disappointed to learn that ISO 3200 is a high bar for M43.

A little bloke on steroids is just as good as a big bloke.

 

Really?

I'd rather a Ferrari F355 than a Chevrolet Corvette from that era.

 

I'm disappointed to learn that ISO 3200 is a high bar for M43.

We're talking the equivalent of an 80Mpx 36mm sensor. It's not disappointing at all - it's just the state of modern sensor design. :-)

 

Going the other way, imagine an A7rIII sensor divided by 4, which would make a 10.5Mpx Micro 4/3 sensor. Good enough? Well, maybe!

I'd rather a Ferrari F355 than a Chevrolet Corvette from that era.

 

 

We're talking the equivalent of an 80Mpx 36mm sensor. It's not disappointing at all - it's just the state of modern sensor design. :)

 

Going the other way, imagine an A7rIII sensor divided by 4, which would make a 10.5Mpx Micro 4/3 sensor. Good enough? Well, maybe!

Honestly I think that these days MP is one of the least interesting things to judge a camera by. Stick a new model 10.5mp sensor in something like that Pen F from a couple years ago, with its design and excellent color processing, add a set of excellent primes and to me you’ve got a very compelling package. (My favorite kit is an X-Pro2 with 18/35/60 but if I weren’t knee deep in the Fuji system an Olympus like that would draw me in.)

We're talking the equivalent of an 80Mpx 36mm sensor. It's not disappointing at all - it's just the state of modern sensor design. :)

Noise matters! ISO 3200 is fairly typical for well-lighted interiors.

 

Higher resolution tends to mask noise, simply because noisy pixels are smaller and evenly distributed. That said, ISO 3200 is clean in my 10 year old Nikon D3 (12 MP), and ISO 12800 still clean in a Sony A9 (24 MP) and A7Riii (42 MP). Size (pixels) matters!

I think a life spent pixel peeping can cause you to lose sight of the ability of photographs to inform and excite. However, we're all in this craft for our own purposes and reasons, and we should respect the choices of others whether in subject matter or equipment.

 

What I take away from this thread is that the second poster is fairly contemptuous of M4/3 and compares it unfavourably to Sony A7 full frame equipment, despite the fact that the OP was making no inter-system comparisons or claims.

 

It reminds me of those kids at school who always claimed to have something back home better than you did.

 

On the other hand, I imagine there are guys out there with medium format cameras and high-end lenses who could beat the pants off the Sony A7 image quality, but these guys don't hang around the forums dissing other people's cameras.

What I take away from this thread is that the second poster is fairly contemptuous of M4/3 and compares it unfavourably to Sony A7 full frame equipment, despite the fact that the OP was making no inter-system comparisons or claims.

The OP's question was whether good M43 lenses can take full advantage of the sensor resolution. If the resolution test results are only 50% of theoretical, the answer is "no." There are cameras and lenses which render 90% of that theoretical value, and the Sony/Zeiss results were merely an example to illustrate that point.

 

When discussing resolution testing, it's hard to avoid pixel peeping. Comparing photos of grandchildren seldom involves technical issues, but that's a different forum.