Into the wild: Olympus E-M1 Mark III sample gallery

Welll I shoot mFT Em1.2. I compare with D800e which I also have and I do not witness it. How handheld Hires compares..I don't know. Never so a comparison. The very different tripod hires is somewhere between the best FF and the 50 MP medium format files. Its use is however restricted.

Yes I'd love to see a blind comparison one day. It's just too easy when you know what camera shot what, to say "oh yeah but the tonality is totally different"...just one example, this is the same with "look flat". Are you sure? I'm not saying you are wrong but I'd like to have a blind test. I don't think it would be too hard to organise for DPreview, smartphones review sites do it from time to time (at least they ask people what photo they prefer and at the end they disclose what phone took what).

I would like to see an honest comparison between the four sensors (MF, FF, APS-C and m4/3) in "real life" meaning taking the same photos 4 times and in different situations : low shutter speed, fast shutter speed, good light, bad light, very dark with and without subject movement, low iso, high iso, with Lightroom noise reduction, without LR noise reduction, etc. All that with the same 24mm f/5.6 (FF equivalent) for instance to get a good comparison. I've never seen that anywhere.

i feel the same way, wide angle looks like smarphone quality and tele shots like superzoom compact. I really think the MFT system has great cameras, but this sample gallery from a $4500 camera+lens makes me doubt they have a bright future..

@phouphou, m4/3 is stagnating image quality-wise for a few years, even my Fuji X100F images are more organic and plesant than my E-M1.2 + 12-40/2.8 Pro combo. I don't know if it's something special in focus/oof areas transition, or color transitions, but m4/3 images always look more "crunchy" than Fuji, and Fuji being "smoother" and pleasant.

It really depends on the user, skill and application. I don’t really rate these photographs and would never judge a system from a DP review gallery. The only way to see the real difference is to shoot across formats personally with your own personal workflow.

You’ll find that there is actually little difference between formats especially in good light. It is also highly dependent on the lens in front of the sensor.

For lowlight each system comes into its own. FF for moving subjects, MFT will give you a bigger shooting envelope with static subjects. I shoot across both FF and MFT. if the outcome is for print or screen you will not see any appreciable difference at lower iso unless pixel peeping a screen This is not how we view photographs.

Nah... don't need a blind test to know that FF sensors produce better images. But one needs to always keep equivalence in mind.

When considering the exposure triangle in general things can get pretty complicated. When you want to test how different formats perform. If you keep shutter speed and ISO the same then the DOF in the comparison would be different. However since a certain look is usually desired one needs to keep DOF the same between the two systems. And adjust ISO and shutter speed accordingly. Obviously if you have a required min shutter speed to avoid motion blur an image with a FF sensor would end up using an ISO value for the FF sensor that is 2 stops higher than that of the mFT.

This is all assuming that the MP count is the same between the two sensors. In the end the DR at whatever ISO setting is what comes into play. After all with the same DOF, and shutter speeds, and same tech, the resulting images should look about the same (before you start playing with DR)

For example, I find that a “crunchy” look is typically the result of oversharpening. Personally, I have noticed that I generally find my pictures more pleasant without any sharpening at all, even if it sometimes makes them a bit soft. I prefer that softness over the crunchiness.

Of course this is all general knowledge. But when I moved from crop to FF the gain was way more than I expected, yet way less at the same time. Essentially shooting wide open with FF compared to wide open on crop with the same lens resulted in a "better" image on FF. But of course DOF was different.

In the end the win that FF has over the smaller sensors is at the "edges" of photographic situation. At high ISOs, and DR.

But others have done far more extensive testing than me, and they come out to this sort of conclusion.
https://www.dpreview.com/forums/thread/4342784

As for the comment about iphones. Remember, it isn't exactly a fair comparison. One's iphone is essentially cheating. If you followed the same process as the iphone in night mode with a mFT, crop, or FF sensor you would end up with a better image than the iphone. After all the iphone is taking multiple images and stacking them.

Take the same number of images with a mFT sensor as an iphone and PP to stack and adjust exposure as desired and you will definitely get a better image than the iphone.

Now try to edit images coming out of a phone. Its a joke, they fall apart quick. Especially when it comes to shadow and highlight recovery.

It’s also relevant in how you shoot. I find that it’s always best to over expose MFT slightly, never underexpose. The actual difference in DR is not that significant in most scenarios. The current 20mp MFT actually have more dynamic range than a lot of Canon FF sensors. I’m doubt anyone could actually tell the difference if print if you compare a 20mp -24mp sensor vs 20mp MFT. Again MFT primes are very comparable. Where FF gives a real advantage is high resolution. 36-45mp. Even then software like alien skin blow up software mitigates much of this advatage and you can print very large and still struggle to find the difference. Unless you are consistently printing wall sized prints FF is really overkill for most users.

@danart

Sorry can't agree with you there. Even the hold 6D sensor will do better in terms of DR in general. After all DR isn't just at base ISO. From ISO 800 and beyond mFT doesn't stand a chance. And that is old tech without the onboard ADC. It takes them a while, but now that they are introducing the sensors with onboard ADC mFT definitely doesn't stand a chance.

EOSR, 5DIV, A7III, D810 etc etc... they are all several stops better at DR at all ISO settings. The 6D essentially matches the OM-D-E MkIII till ISO 300 or so (with the advantage towards the oly) and then the 6D pulls ahead beyond that.

And it isn't just about DR. Details are simply better retained in FF cameras vs crop or mFT.

@danart
Can't agree with you on few points:

- Crunchy look is there because there is still some noise present even at base ISO, so whole picture is like covered with small dust / sand particles. Fuji starts to introduce that crunch at much higher ISOs (starting at 800 and up, while Olympus at 200 and up). It is very visible in OOF areas, sky and water

- Also, highlights and tonal palette rollofs/transitions are harsher with m4/3 than with larger sensors, which adds "crunch" to overall picture

- DR at base ISO might be the same or even better with current mft sensors vs older FF, but it starts to sink down very fast. I've shot a concert with Olympus and Canon 6D, and no way Olympus beats it at middle and high ISO, both in terms of noise and DR.

Er...the m4/3 20 MP sensor has dynamic range that equals or surpasses APS-C sensors. Yes, they lose 3/4 of a stop of light compared to APS-C, but for static subjects you make that up with IBIS. As for flatness, everything looks flat compared to medium format.

As for lawny and SirPalomid's claim on dynamic range, the Olympus sensor is better up to ISO 800 and then the two are neck and neck onward. Sorry.

http://photonstophotos.net/Charts/PDR.htm#Canon%20EOS%206D%20Mark%20II,Olympus%20OM-D%20E-M1%20Mark%20II

@SirPalomid - with regards to the crunchy sandy look that you are describing, at base, this is not an issue for me. You understand that you need to shoot MFT different to as you would a FF? They are different and require a different shooting workflow. You will get noise around base if you are not exposing correctly - you always over expose with MFT. You should always be pulling the exposure slider left in PP. Any under exposure will give you a greater penalty vs FF. however the tonal roll off and highlights etc.. will not be distinguishable in print or screen which is my main application. The reason being is that print only gives you limited DR vs the actual camera out put. Print more and you will see. If you pixel peep you will note a difference and yes at higher ISO’s FF will come into play, but at base to low iso with the correct workflow and exposure MFT print can give you exceptional results indistinguishable from FF.

@unhappymeal
The PDR numbers firstly are not measuring what many think they are , and the way Olympus interprets their in camera ISO setting games that measurement. Take a look at DPreview DR test shots you will see that at their base ISO m43 cameras are well behind the best APS . The combination of smaller sensor and higher base ISO is the issue

Sorry, I don’t follow. If your respective lenses resolve 15MP of details on both sensors, then that’s what you get regardless of pixel density. If one image is less sharp, then that means that the lens for that image did not, in fact, deliver 15MP of details, doesn’t it? What am I missing?

Why does half the pixel pitch for half the focal length give more details? I have provided an explanation for why it would not. What would be a counter-explanation?

“FF sensors are nearly 4 times the size.”

They are, but so are the images projected onto them when using equivalent lenses. Why would that not make up for it?

“If we downsample a FF 45mp image to 20mp we get more detail than a native 20mp sensor and this is due to pixel density. ”

No, it’s due to oversampling. Which you can tie to pixel density within one format but not when comparing across them.

“A MFT image is essentially operating in a similar manner , it’s a roughly a 76mp file at the equivalent FF pixel level but with a maximum output of 20mp.”

If the sensor is 20MP and the final image is also 20MP, then there is no oversampling. It’s not a 76MP file at the equivalent FF pixel level because 56MP are missing around them, so the 20MP are the whole image. (Unless you were going to crop the FF image anyway because you don’t have long enough lenses.)

The second paragraph is true but does not have much to do with the discussion, does it?