Olympus EM1-III vs Sony a7S-III (another m43 vs FF)

A 12mp FF sensor will show equal to more detail than a 20mp m43 sensor. These are my findings. Some people will refuse to believe but I tried both side by side and it was clear.
could you pls provide RAW files (daylight and high ISO)?
If people want to run topaz denoise or super resolution dxo depp prime or whatever, they can. But I find this workflow highly unrealistic and incompatible with everyday use.
Why? My Deep Prime files on an M1 Mac export in ~45 seconds. I suspect the 12 MP A7S files would go substantially faster. Provide a link to the RAWs and I'll run them both through.
 
...Another big difference I noticed: noise in the sky at base ISO

In a situation where there's plenty of light, the base ISO of 80 on the Sony proves itself very useful, vs the base ISO of 200 on the Olympus.

Here some example of sky shots on the em1.2 and sony a7sIII at their base ISO, viewed at 400% magnification. EXIF included.

in this situation the lower base ISO and 14 bit color depth makes a difference even if there's plenty of light available.
...Moreover, ISO 100 is available as a camera setting on the Oly.
ISO100 in the E-M1II is pretty stupid as it's overexposed ISO320.
I'm not sure why Olympus added it but I actively try to avoid it as it's best way to clip the highlights.
In comparison ISO64 is overexposed ISO200 and it's fine.
OK. I mistakenly thought that the minimum ISO on the E-M1.2 and E-M1.3 was ISO 100 as on many other Oly cams. But I now see it goes all the way down to ISO 64, which is preferable to ISO 100 for the reason you mention. Weird that Oly chose to increase the gain when going from ISO 64 to ISO 100. It should of course have been kept unchanged and the values simply scaled 2/3 stop upwards when producing the OOC jpeg.
 
Last edited:
Where is the surprise? I have owned the S5 and the quality of the 24 megapixels file was impressive you could push them way beyond any MFT

There are two stops benefit of DR and is superior in all aspects EXCEPT I had to lug a 2.3 Kg lens shooting wildlife and that just did not cut the cake

But for prime work the new Lumix S are 320 grams so actually the camera weight less than my EM1+17mm 1.2 pro

For shooters of landscape, portraits, night photography and events not requiring long lens or high frame rate there is ZERO reason to get an OMD EM1MKIII nor a G9 or an APSC camera. The S5 costs the same of the OMD EM1MKIII nobody sane of mind would buy the Olympus if they do not need lighter long lenses and some form of agility that the full frame platform does not have due to the lens size at tele end

I guess OMD knows that hence are focussing on wildlife and macro and Panasonic on video capabilities
It's important to also know your use case and habits. For the longest time I would never even give full frame a nano second of a thought because every prime brighter than f/1.8 and every zoom faster than f/4 was just more than I was willing to carry. That's changing now with Tamron releasing ultra compact and good f/2.8 zooms and Samyang releasing good, compact and affordable f/1.8 and f/1.4. primes. I suspect that is the case for a lot of other people.
I have used the Tamron 28-75/2.8 for some time. It’s right on the edge of what I want to carry without a sling. I just bought the Samyang 35/1.8 and it’s OK as a small, cheap and light lens you can shoot at f2.2. For more typical prime IQ, the Sigma 35/1.8 would be the pick. The Tamron 17-28/2.8 is lighter than the 28-75, but not as good optically. There are lots of UWA FE primes.

I find the A7R4 also on my personal limit - smaller than a G9 but 7g heavier.

An A7R with a small prime is pretty handy, if you can accept its functional limits and bad habits.

FF for me is really about landscape shooting with a tripod, filters, backpack etc. It is getting to the point where it suits me on a ramble as a handheld camera.

The EM1.2 with 12-40 Pro is still the go to camera for general use. It would be hard choosing what to take on an extended holiday.

The killer light FE zoom is the new Sony 28-50/4-5.6 lens for the A7C. Sony are also producing lighter and slower primes to match.

I general, quality equipment is still cheaper in MFT but not so much that there is a clear winner for every use case, at least for me.

Andrew
 
For these to be valid comparisons, the shutter speed must be the same, and the aperture must be equivalent. Using constant ISO and different shutter speeds is going to bias the results in favor of the longer shutter speed which is being set on the Sony.

For the most part, actual ISO values are irrelevant as long as the values are in the ISO invariant range on each camera. But using "equivalent" ISO settings would be more appropriate. Since true ISO on these cameras is likely different, that means setting ISO to simply avoid saturation.

And the bottom line is: so? Used properly, I would expect the Sony to produce better images. At about twice the cost of the Olympus. You will get what you pay for, and pay for what you get.

And I still prefer shooting with my Olympus over my Nikon D750 or Z7. It's only my D500 that I prefer when shooting wildlife, but not if I'm going to be hefting the lens around all day.
 
For these to be valid comparisons, the shutter speed must be the same, and the aperture must be equivalent. Using constant ISO and different shutter speeds is going to bias the results in favor of the longer shutter speed which is being set on the Sony.
It really depends on what you are after - for example, if you want to know how much low light performance can be improved, the comparison should be between the affordable 4/3rd with the fastest lens and an affordable FF with the fastest lens.

There is no a priori constraint that if you move to the FF, you must maintain the same DOF.
For the most part, actual ISO values are irrelevant as long as the values are in the ISO invariant range on each camera. But using "equivalent" ISO settings would be more appropriate. Since true ISO on these cameras is likely different, that means setting ISO to simply avoid saturation.

And the bottom line is: so? Used properly, I would expect the Sony to produce better images. At about twice the cost of the Olympus. You will get what you pay for, and pay for what you get.

And I still prefer shooting with my Olympus over my Nikon D750 or Z7. It's only my D500 that I prefer when shooting wildlife, but not if I'm going to be hefting the lens around all day.
 
Where is the surprise? I have owned the S5 and the quality of the 24 megapixels file was impressive you could push them way beyond any MFT
Well the S5 is 24 mp and the a7SIII only 12 mp.

The surprise for me was the resolving power of a full frame 12 mp being equal to slightly superior to a 20 mp m43.
There are two stops benefit of DR and is superior in all aspects EXCEPT I had to lug a 2.3 Kg lens shooting wildlife and that just did not cut the cake

But for prime work the new Lumix S are 320 grams so actually the camera weight less than my EM1+17mm 1.2 pro
Yes a full frame prime setup is lighter than a m43 equivalent.
Interesting, what FF kit would give me the same angle of view as a 90mm FF, aperture opens to f1.8 and, when mounted on my camera (Olympus E-M5 III), weighs 530 grams (414g body+battery & 116g lens)?
Canon RP w/RF 50mm F1.8 1.07 lb / 485 g (Body with Battery and Memory) + 160 g (0.35 lb) = 645g

Using the "equivalent", I can argue that Canon RP is actually lighter per amount of light collected at f/1.8. ;-)
Absolutely. More light, but then you have to deal with the RF 50mm f/1.8's gag inducing bokeh. Of course, if we want to get truly ridiculous for mass per amount of light collected, we could always toss something like the GM5 + 20 f/1.7 into the mix.
Surely, Canon is not "lighter" but if 4/3rd is counting on "light system", I would be worried.
Why? That's the driving reason behind why the GM5 and GR III are my most used cameras. They see more use than my E-M1 III and Z7. I can chuck the GM5 and 2 or 3 lenses in a bag and have less weight and size to carry around than the EOS RP body alone.
I think there is a reason why GM5 is no longer offered and unless you know your use case is some universal practice, your personal anecdote is just that.
 
For these to be valid comparisons, the shutter speed must be the same, and the aperture must be equivalent. Using constant ISO and different shutter speeds is going to bias the results in favor of the longer shutter speed which is being set on the Sony.
It really depends on what you are after - for example, if you want to know how much low light performance can be improved, the comparison should be between the affordable 4/3rd with the fastest lens and an affordable FF with the fastest lens.

There is no a priori constraint that if you move to the FF, you must maintain the same DOF.
And no a priori constraint that if you move to MFT to maintain the same noise level or range of AoV. Right?

BTW: You accused me in an expired parallel thread of being obtuse when I said I preferred MFT on the basis of image quality. I wasn’t.

For the most part, actual ISO values are irrelevant as long as the values are in the ISO invariant range on each camera. But using "equivalent" ISO settings would be more appropriate. Since true ISO on these cameras is likely different, that means setting ISO to simply avoid saturation.

And the bottom line is: so? Used properly, I would expect the Sony to produce better images. At about twice the cost of the Olympus. You will get what you pay for, and pay for what you get.

And I still prefer shooting with my Olympus over my Nikon D750 or Z7. It's only my D500 that I prefer when shooting wildlife, but not if I'm going to be hefting the lens around all day.
 
Where is the surprise? I have owned the S5 and the quality of the 24 megapixels file was impressive you could push them way beyond any MFT
Well the S5 is 24 mp and the a7SIII only 12 mp.

The surprise for me was the resolving power of a full frame 12 mp being equal to slightly superior to a 20 mp m43.
There are two stops benefit of DR and is superior in all aspects EXCEPT I had to lug a 2.3 Kg lens shooting wildlife and that just did not cut the cake

But for prime work the new Lumix S are 320 grams so actually the camera weight less than my EM1+17mm 1.2 pro
Yes a full frame prime setup is lighter than a m43 equivalent.
Interesting, what FF kit would give me the same angle of view as a 90mm FF, aperture opens to f1.8 and, when mounted on my camera (Olympus E-M5 III), weighs 530 grams (414g body+battery & 116g lens)?
Canon RP w/RF 50mm F1.8 1.07 lb / 485 g (Body with Battery and Memory) + 160 g (0.35 lb) = 645g

Using the "equivalent", I can argue that Canon RP is actually lighter per amount of light collected at f/1.8. ;-)
Absolutely. More light, but then you have to deal with the RF 50mm f/1.8's gag inducing bokeh. Of course, if we want to get truly ridiculous for mass per amount of light collected, we could always toss something like the GM5 + 20 f/1.7 into the mix.
Surely, Canon is not "lighter" but if 4/3rd is counting on "light system", I would be worried.
Why? That's the driving reason behind why the GM5 and GR III are my most used cameras. They see more use than my E-M1 III and Z7. I can chuck the GM5 and 2 or 3 lenses in a bag and have less weight and size to carry around than the EOS RP body alone.
I think there is a reason why GM5 is no longer offered and unless you know your use case is some universal practice, your personal anecdote is just that.
I am sorry, but size is clearly a concern for the market. There is a reason why Canon saw fit to make the RP small, why the A7C is doing so well, why Samyang’s “Tiny” line sells well, and Tamron is doing well with zooms.



The GM5 may be dead, but compact bodies are doing very well in quite a few markets. That is a strength that m4/3 can continue to play to.
 
For these to be valid comparisons, the shutter speed must be the same, and the aperture must be equivalent. Using constant ISO and different shutter speeds is going to bias the results in favor of the longer shutter speed which is being set on the Sony.
It really depends on what you are after - for example, if you want to know how much low light performance can be improved, the comparison should be between the affordable 4/3rd with the fastest lens and an affordable FF with the fastest lens.

There is no a priori constraint that if you move to the FF, you must maintain the same DOF.
And no a priori constraint that if you move to MFT to maintain the same noise level or range of AoV. Right?

BTW: You accused me in an expired parallel thread of being obtuse when I said I preferred MFT on the basis of image quality. I wasn’t.
For the most part, actual ISO values are irrelevant as long as the values are in the ISO invariant range on each camera. But using "equivalent" ISO settings would be more appropriate. Since true ISO on these cameras is likely different, that means setting ISO to simply avoid saturation.

And the bottom line is: so? Used properly, I would expect the Sony to produce better images. At about twice the cost of the Olympus. You will get what you pay for, and pay for what you get.

And I still prefer shooting with my Olympus over my Nikon D750 or Z7. It's only my D500 that I prefer when shooting wildlife, but not if I'm going to be hefting the lens around all day.
Do you have a FF camera?
 
For these to be valid comparisons, the shutter speed must be the same, and the aperture must be equivalent. Using constant ISO and different shutter speeds is going to bias the results in favor of the longer shutter speed which is being set on the Sony.
It really depends on what you are after - for example, if you want to know how much low light performance can be improved, the comparison should be between the affordable 4/3rd with the fastest lens and an affordable FF with the fastest lens.

There is no a priori constraint that if you move to the FF, you must maintain the same DOF.
And no a priori constraint that if you move to MFT to maintain the same noise level or range of AoV. Right?

BTW: You accused me in an expired parallel thread of being obtuse when I said I preferred MFT on the basis of image quality. I wasn’t.
For the most part, actual ISO values are irrelevant as long as the values are in the ISO invariant range on each camera. But using "equivalent" ISO settings would be more appropriate. Since true ISO on these cameras is likely different, that means setting ISO to simply avoid saturation.

And the bottom line is: so? Used properly, I would expect the Sony to produce better images. At about twice the cost of the Olympus. You will get what you pay for, and pay for what you get.

And I still prefer shooting with my Olympus over my Nikon D750 or Z7. It's only my D500 that I prefer when shooting wildlife, but not if I'm going to be hefting the lens around all day.
Do you have a FF camera?
In what way is that relevant?

And to satisfy your curiosity, please have a look at my gear list here on DPR. It includes all my gear, current as well as past.

P.S. Perhaps I should also mention that I have used medium format although I never personally owned such a camera.
 
Last edited:
some new recent findings:

I did some landscape shots today and realised how little megapixels matter for landscape. The atmospheric distortions does reduce resolution by a-lot even on wide angle lenses.

Not seeing a difference between Olympus files and Sony files in terms of details tbh.

But again the a7sIII files support dehazing and sharpening a lot better than the em1.3.

Atmospheric distortion also reduce the advantage of HHHRS. During winter time HHHRS works better though.

So far the only situation I know of that requires high megapixels is group shots of people, where seeing individual faces is a real asset. But then again, for that kind of use, even 20mp is low.
 
some new recent findings:

I did some landscape shots today and realised how little megapixels matter for landscape. The atmospheric distortions does reduce resolution by a-lot even on wide angle lenses.

Not seeing a difference between Olympus files and Sony files in terms of details tbh.

But again the a7sIII files support dehazing and sharpening a lot better than the em1.3.

Atmospheric distortion also reduce the advantage of HHHRS. During winter time HHHRS works better though.

So far the only situation I know of that requires high megapixels is group shots of people, where seeing individual faces is a real asset. But then again, for that kind of use, even 20mp is low.
Sorry to hear that you live in an area with so poor atmospheric conditions that you can no longer see the difference between 12 and 20 MP. Happy to be in a region where the difference is readily visible. :-)
 
some new recent findings:

I did some landscape shots today and realised how little megapixels matter for landscape. The atmospheric distortions does reduce resolution by a-lot even on wide angle lenses.
The dedicated landscape photographers I've met and/or worked with in the past are rather big proponents of "more megapixels" in their work (some might say, obsessively so!).

Try A7RIV with 21 Loxia then on A7SIII and see if you reach the same conclusion.
Not seeing a difference between Olympus files and Sony files in terms of details tbh.

But again the a7sIII files support dehazing and sharpening a lot better than the em1.3.

Atmospheric distortion also reduce the advantage of HHHRS. During winter time HHHRS works better though.
My experience is rather the opposite. HHHR provides noticeable improvement when atmospheric conditions deteriorate. Here are some b-roll shots from a gig with local fire department fending off wild fires last summer. HHHR worked very well cutting through the heat and smoke haze.





f06ac2d83f9247779c5e0130bdcec677.jpg



1ff2c47153c64fecbbece68d37eb8b16.jpg
 
some new recent findings:

I did some landscape shots today and realised how little megapixels matter for landscape. The atmospheric distortions does reduce resolution by a-lot even on wide angle lenses.
The dedicated landscape photographers I've met and/or worked with in the past are rather big proponents of "more megapixels" in their work (some might say, obsessively so!).

Try A7RIV with 21 Loxia then on A7SIII and see if you reach the same conclusion.
Not seeing a difference between Olympus files and Sony files in terms of details tbh.

But again the a7sIII files support dehazing and sharpening a lot better than the em1.3.

Atmospheric distortion also reduce the advantage of HHHRS. During winter time HHHRS works better though.
My experience is rather the opposite. HHHR provides noticeable improvement when atmospheric conditions deteriorate. Here are some b-roll shots from a gig with local fire department fending off wild fires last summer. HHHR worked very well cutting through the heat and smoke haze.

f06ac2d83f9247779c5e0130bdcec677.jpg

1ff2c47153c64fecbbece68d37eb8b16.jpg
I think you're interpreting the lack of luminance noise with extra detail.

The extra detail you're seeing is because there's less noise pattern to hide detail. The resolving power is remaining constant.

--
...
 
Where is the surprise? I have owned the S5 and the quality of the 24 megapixels file was impressive you could push them way beyond any MFT
Well the S5 is 24 mp and the a7SIII only 12 mp.

The surprise for me was the resolving power of a full frame 12 mp being equal to slightly superior to a 20 mp m43.
There are two stops benefit of DR and is superior in all aspects EXCEPT I had to lug a 2.3 Kg lens shooting wildlife and that just did not cut the cake

But for prime work the new Lumix S are 320 grams so actually the camera weight less than my EM1+17mm 1.2 pro
Yes a full frame prime setup is lighter than a m43 equivalent.
Interesting, what FF kit would give me the same angle of view as a 90mm FF, aperture opens to f1.8 and, when mounted on my camera (Olympus E-M5 III), weighs 530 grams (414g body+battery & 116g lens)?
Canon RP w/RF 50mm F1.8 1.07 lb / 485 g (Body with Battery and Memory) + 160 g (0.35 lb) = 645g

Using the "equivalent", I can argue that Canon RP is actually lighter per amount of light collected at f/1.8. ;-)

Surely, Canon is not "lighter" but if 4/3rd is counting on "light system", I would be worried.
So, a lens that does not nearly "give me the same angle of view as a 90mm FF" and on-body weighs 231g more isn't what I asked about.

This is what I replied to: "Yes a full frame prime setup is lighter than a m43 equivalent." I suppose "equivalent" is open to interpretation but angle of view is what I specifically asked about and, to me, aperture is primarily for exposure and only secondarily for bokeh.

Please don't infer that I was talking about what "[micro] 4/3rd is counting on". I did not say why I was interested.
The point you refuse to acknowledge is that FF w/50mm f/1.8 lens can weigh less than 1.5 lb. As for FOV, you could get closer for 90mm FOV. But what is it about 90mm FF?
I fully acknowledge and understand everything you have stated. However, even the far wider AOV FF 50mm at 1.5 lb is far heftier than 116g, and not "lighter" as Funny Valentine inferred. I like 90mm FF FL for portraits (why are my reasons pertinent?). Getting closer @50 FF doesn't have the same look and makes proper lighting more difficult. I don't want to further digress into the differences but if you are truly interested in answering your question, simply google "wide angle vs telephoto portrait".

I asked a simple question and thought I was unambiguous ("same AOV as a 90mm FF, aperture opens to f1.8"). Super shallow DOF is not a factor. I will probably rent such a FF combination for a poorly lit private "party" that I have been hired to photograph in October. The attendees (celebrities) do not want a 50 poked a few inches from their cheeks and would not be conducive for semi-candids.

If I can't find a suitable FF solution, I could rent a 45 1.2 or pray that AI noise reduction works as well as advertised :) but I prefer to get the higher resolution that FF can provide.

So, thank you for your effort.
For shooters of landscape, portraits, night photography and events not requiring long lens or high frame rate there is ZERO reason to get an OMD EM1MKIII nor a G9 or an APSC camera. The S5 costs the same of the OMD EM1MKIII nobody sane of mind would buy the Olympus if they do not need lighter long lenses and some form of agility that the full frame platform does not have due to the lens size at tele end

I guess OMD knows that hence are focussing on wildlife and macro and Panasonic on video capabilities
Wise decisions
I tired of waiting for a genuine answer to my want of a 90ish 1.8ish FF mirrorless kit, comparably sized to my E-M5 III + 45 1.8, based on Funny Valentine's assertion that "a full frame prime setup is lighter than a m43 equivalent". So I posed the question in the Open Talk forum. Moderator Labe suggested a Sony A7C with a Sony FE 85 1.8. At 880g it is 1.7X heavier than 530g, not what I would consider "lighter than a m43 equivalent" - and for only 4 more MP.
 
some new recent findings:

I did some landscape shots today and realised how little megapixels matter for landscape. The atmospheric distortions does reduce resolution by a-lot even on wide angle lenses.

Not seeing a difference between Olympus files and Sony files in terms of details tbh.

But again the a7sIII files support dehazing and sharpening a lot better than the em1.3.

Atmospheric distortion also reduce the advantage of HHHRS. During winter time HHHRS works better though.

So far the only situation I know of that requires high megapixels is group shots of people, where seeing individual faces is a real asset. But then again, for that kind of use, even 20mp is low.
Sorry to hear that you live in an area with so poor atmospheric conditions that you can no longer see the difference between 12 and 20 MP. Happy to be in a region where the difference is readily visible. :-)
You would need a big group for 20Mpix to be insufficient. Must have big leaves if landscape needs less than faces.

Andrew
 
Anyway if you're photographing something that has more than 3 or 4 KM layer of atmosphere between your lens and the subject, the effective resolution will be below 5 or 6 megapixels. The limiting factors will be the atmosphere and the lens (diffraction effects are amplified when focused close to infinity).

I've photographed mountains 50-70 km away and they look like paintings, even with HHHRS.

Effective resolution must be lower than 2 mp here. Shot on Olympus em1.2.

View attachment dc5144629ecd42a4ba258803b8f8d136.jpg

--
...
 
Last edited:
Last edited:
I've photographed mountains 50-70 km away and they look like paintings, even with HHHRS.

Effective resolution must be lower than 2 mp here. Shot on Olympus em1.2.

View attachment dc5144629ecd42a4ba258803b8f8d136.jpg
How you do 50MP HHHRS on Em1.2? I think you are telling stories as evidence lack of EXIF. Unless you try to handhold pixel shift on your camera - no wonder soft image ;-)
that's tripod HRS at ISO 1600 on the em1.2.



--
...
 
Anyway if you're photographing something that has more than 3 or 4 KM layer of atmosphere between your lens and the subject, the effective resolution will be below 5 or 6 megapixels. The limiting factors will be the atmosphere and the lens (diffraction effects are amplified when focused close to infinity).
Could you please link to the reliable source(s) I am sure you have ready to support these contentions? ;-)
I've photographed mountains 50-70 km away and they look like paintings, even with HHHRS.

Effective resolution must be lower than 2 mp here. Shot on Olympus em1.2.

View attachment dc5144629ecd42a4ba258803b8f8d136.jpg
 
Last edited:

Keyboard shortcuts

Back
Top