E5 vs D700--tests

[nugat]

Do current lenses outresolve sensors

I still doubt camera FF Nikkors can reach 100lp/mm ACROSS the image circle.
(Well, there is one nikkor that sure can do even more--the printing Nikkor).
The parameters of the mount diameter and distance to the sensor and its size are not optimal from the digital point of view. They are on 4/3: big mount diameter, almost the same flangeback distance as FF and small sensor. Therefore Zuikos can fly across the 4/3 frame with resolution, vignetting, CA etc. But the price is the SHG size.
For FF Nikkors to try match that behavior they'd need to be 2-3 times bigger than now to start with, IMHO.
It would be interesting to mount the new cine Leica primes on a D3x, supposedly they cover the FF (43mm) image circle.
I am afraid that would require alterations to the mount and mirror though (flange back distance). And those babies weigh 2 kg and cost 15,000$ a pop.
When I get the ordered SHG glass I'll also borrow D3x and do some tests with Nikkors.
As for diffraction, I believe it is fair to assume that Zuikos (similar to Leica) are diffraction limited ie. only physics limits them, not construction. In 4/3 and with the 12Mpixel sensor the Airy disk becomes bigger than the assumed CoC (circle of confussion, for 4/3 CoC= 0,015mm) at f13.
But frankly, before we get to diffraction at f13, there will be focussing errors, camera shake, mirror slap, visibility ....So even on 4/3, with Zuikos, diffraction is a theoretical problem, I think.
(nice diffraction calculator here:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm)

PS. Even if current nikkors seriously outresolve the 24Mpix sensor the gain in resolution from going to a 37mpix sensor will be 25% in lw/ph.

 

[jonoslack]

the d3x/a900/5dII do show that with good glass higher MP than the d700´s 12MP does not hurt IQ at low and medium ISO.
Upscaling images makes me allways feel like adding information which has not been there in reality.

HI Tom
I quite agree - but downsizing is a perfectly good thing.
I think that all the maths about this rather beggars the question of the reality.
The reality I see when comparing my D700 images to my A900 images (I did this in some detail recently when doing two wedding books, 1 with each camera).

1. pixel peeping at 100% shows how much better the D700 pixels are - especially in low light
2. this relates to nothing in the real world because you are looking at a much larger portion of the image.
3. downsizing the A900 pictures produced high ISO values very little different from the D700
4. cropping the A900 pictures left lots of resolution, cropping the D700 shots was actually a real problem.

When you consider the pixel density of relative cameras and sensors, and look at the quality of the new 16mp sensor it seems to me that either:
1. a 40mp FF camera will be just fine
or
2. the pixel density on the E5 is too high :salute:

I think 1 is the case

pixel density:
D700: 1.4 MP/cm²
D3X: 2.9 MP/cm²
D7000: 4.4 MP/cm²
E5: 5.1 MP/cm²

just as a matter of interest, the pixel density of the new Panasonic LX5 is
24 MP/cm² . . . . imagine that on a full frame sensor!

The E5 proves that 5.1 MP/cm² does a fine job - so lets see a new Sony/Nikon full frame sensor with that kind of sensor density.

I know I'd rather downsize images to remove noise than upsize them to increase detail!

. . . . and I don't care about Nyquist and MTF charts proving that each pixel is less good - experience tells me that in the real world more of them (within reason) is better!

all the best

 

[ptomsu]

Re: Do current lenses outresolve sensors

I still doubt camera FF Nikkors can reach 100lp/mm ACROSS the image circle.
(Well, there is one nikkor that sure can do even more--the printing Nikkor).
The parameters of the mount diameter and distance to the sensor and its size are not optimal from the digital point of view. They are on 4/3: big mount diameter, almost the same flangeback distance as FF and small sensor. Therefore Zuikos can fly across the 4/3 frame with resolution, vignetting, CA etc. But the price is the SHG size.
For FF Nikkors to try match that behavior they'd need to be 2-3 times bigger than now to start with, IMHO.
It would be interesting to mount the new cine Leica primes on a D3x, supposedly they cover the FF (43mm) image circle.
I am afraid that would require alterations to the mount and mirror though (flange back distance). And those babies weigh 2 kg and cost 15,000$ a pop.
When I get the ordered SHG glass I'll also borrow D3x and do some tests with Nikkors.
As for diffraction, I believe it is fair to assume that Zuikos (similar to Leica) are diffraction limited ie. only physics limits them, not construction. In 4/3 and with the 12Mpixel sensor the Airy disk becomes bigger than the assumed CoC (circle of confussion, for 4/3 CoC= 0,015mm) at f13.
But frankly, before we get to diffraction at f13, there will be focussing errors, camera shake, mirror slap, visibility ....So even on 4/3, with Zuikos, diffraction is a theoretical problem, I think.
(nice diffraction calculator here:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm)

PS. Even if current nikkors seriously outresolve the 24Mpix sensor the gain in resolution from going to a 37mpix sensor will be 25% in lw/ph.

Well, you took the example of MFT charts for Olympus. If we believe in what Oly publishes here, we also need to believe in what Nikon or others publish - right?

I agree that all vendors try to draw their MFT charts differently, so that most of users cannot follow (or simply give up) and just believe the marketing message. Shame on the industry that there is nothing like a unified measurement here for MFTs.

In fact I see NO reason why top Nikkor glass should not be able to fulfill what they publish in MFT. And this is on par with top Oly glass - right? And I am pretty sure in you take the latest development of Canon glass, it will be the same.

Leica was always trying to do better especially when it came to optics and for sure they did so for the S system. Bur to think that any R glass would have done some job on the S2 sensor size is simply comparing apples with oranges - R glass NEVER was designed for that image circle! So Leica had to go a totally different path for their S system and develop glass from the scratch - which the again did with the expected quality and finally success!

But to be honest, if we take MFT charts from Olympus and start deriving conclusions from them, then it is only fair to do the same with other vendors.

WRT size of lenses - the Oly lenses I am talking about are all pro lenses with high apertures like 2.0/14-35. Which is as big as the 2.8/24-70 from Nikon. And here you are right, you gain 1 f stop while haveing similar size if you go 43. But that was always true and actually nothing new.

 

[ptomsu]

HI Tom
I quite agree - but downsizing is a perfectly good thing.
I think that all the maths about this rather beggars the question of the reality.
The reality I see when comparing my D700 images to my A900 images (I did this in some detail recently when doing two wedding books, 1 with each camera).

1. pixel peeping at 100% shows how much better the D700 pixels are - especially in low light
2. this relates to nothing in the real world because you are looking at a much larger portion of the image.
3. downsizing the A900 pictures produced high ISO values very little different from the D700
4. cropping the A900 pictures left lots of resolution, cropping the D700 shots was actually a real problem.

When you consider the pixel density of relative cameras and sensors, and look at the quality of the new 16mp sensor it seems to me that either:
1. a 40mp FF camera will be just fine
or
2. the pixel density on the E5 is too high :salute:

I think 1 is the case

pixel density:
D700: 1.4 MP/cm²
D3X: 2.9 MP/cm²
D7000: 4.4 MP/cm²
E5: 5.1 MP/cm²

just as a matter of interest, the pixel density of the new Panasonic LX5 is
24 MP/cm² . . . . imagine that on a full frame sensor!

The E5 proves that 5.1 MP/cm² does a fine job - so lets see a new Sony/Nikon full frame sensor with that kind of sensor density.

I know I'd rather downsize images to remove noise than upsize them to increase detail!

. . . . and I don't care about Nyquist and MTF charts proving that each pixel is less good - experience tells me that in the real world more of them (within reason) is better!

all the best

Jono,

I see here the discussion about Sony (A900) versus Nikon (D3X and D700) coming up again. I must iterate here, I am maybe the only one, but I DID NOT LIKE SONY COLORS OUT OF THE BOX, so I went back to Nikon.

I agree that more pixels is always better, if the design is done right. This is also the reason I am waiting for the next incarnation of high resolution Nikon DSLR (as maybe you are waiting in the Sony camp) to be able to push the optics to the limits and more importantly maybe find a "cheap" replacement for my MFD gear.

And sorry - I cannot resist - I only hope for you that Sony does not stop the development of their FF DSLRs as was already rumored a lot

 

[nugat]

@ nugat

If I go to the Oly website, I only can find MTF for e.g. SWD 14-35 which shows a highest contrast value of approx. 95 with 10 lp/mm or 75 with 60 lp/mm at Wide end. From this I cannot see how you come to 100 lp/mm at 50 percent contrast - is this value true???? If so then your calculation is right that 100 lp/mm x 13mm results in 1300 lp/ph.

Now if I do the same for D3X FF ....

D3X with a sensor size of 24mm ph x 36 mm pw with 24.5 MP results in 864 mm2 and this means 28356 pixels/mm2 which results in approximately 168 pixels/mm (which BTW is only 70% of the pixel density of the E5, so there is for sure room for improvement, as long as Nikkor lenses add to this). Given a ph of 24mm in FF you end up in 4032 pixels. With Nyquist you end up in 2016 lp/ph.

Looking at the Nikon website, for the 2.8/24-70 for MTF you find a highest contrast value of approx. 98 with 10 lp/mm or 88 with 30 lp/mm. If I do a similar extrapolation as you assumed for Zuiko at 50% contrast value, this should at least also result in 100 lp/mm. This gives you then 2400 lp/ph - slightly above the Nyquist limit of this 24.5 MP sensor.

I never said I extrapolated the Zuiko 100lp/mm from the Oly MTF graphs alone. I would not know how to extrapolate from 20/60 lp at 95% and 75% contrast, the number of lines for 50% contrast. If there is a formula I'd like to learn it. Similarily if the nikkor shows for 10/30lp contrast of 98% and 88% respectively I do not know from that what the lp figure would be for 50%.
But if you put these two graphs together, the correlation that emerges is that Oly for similar contrast gets almost double the resolution. Eg, in the center Oly 75%/60lp, Nikkor 88%/30lp. (88/75=17% deviation from 2x). In simple words the function looks very similar, but the units on the Oly resolution axis are double. From some more trusty (for me) sources (Puts/imx.nl/photo, lenstip.com, even dpreview.com after recent corrections) we might find that good FF glass tops out at +50lp/mm MTF50, never reaching , or physics forbid, exceeding the Nyquist limit. Therefore Oly glass might just offer 100lp/mm. (Excellent Leica glass delivers 80lp/mm).
Some info herebut it's good to read more there)
http://www.imx.nl/photo/leica/leica/leica/page105.html
http://www.imx.nl/photo/leica/camera/page163/page163.html
PS.
The very conservative lenstip.com finds in recent tests the resolution of the new panny pancake 28/2.5 at 70lp/mm. The conclusion is the glass is NOT very good overall.
PSPS.
I know, I said I do not trust manufacturers and reviewers. OK , my own tests above show E5+12-60mm at 2400lw/ph (1200lp/ph). Waiting for that 14-35/2 next week.

 

[ptomsu]

I never said I extrapolated the Zuiko 100lp/mm from the Oly MTF graphs alone. I would not know how to extrapolate from 20/60 lp at 95% and 75% contrast, the number of lines for 50% contrast. If there is a formula I'd like to learn it. Similarily if the nikkor shows for 10/30lp contrast of 98% and 88% respectively I do not know from that what the lp figure would be for 50%.
But if you put these two graphs together, the correlation that emerges is that Oly for similar contrast gets almost double the resolution. Eg, in the center Oly 75%/60lp, Nikkor 88%/30lp. (88/75=17% deviation from 2x). In simple words the function looks very similar, but the units on the Oly resolution axis are double. From some more trusty (for me) sources (Puts/imx.nl/photo, lenstip.com, even dpreview.com after recent corrections) we might find that good FF glass tops out at +50lp/mm MTF50, never reaching , or physics forbid, exceeding the Nyquist limit. Therefore Oly glass might just offer 100lp/mm. (Excellent Leica glass delivers 80lp/mm).
Some info herebut it's good to read more there)
http://www.imx.nl/photo/leica/leica/leica/page105.html
http://www.imx.nl/photo/leica/camera/page163/page163.html
PS.
The very conservative lenstip.com finds in recent tests the resolution of the new panny pancake 28/2.5 at 70lp/mm. The conclusion is the glass is NOT very good overall.
PSPS.
I know, I said I do not trust manufacturers and reviewers. OK , my own tests above show E5+12-60mm at 2400lw/ph (1200lp/ph). Waiting for that 14-35/2 next week.

If you have 20/60 lp at 95% and 75% contrast respectively it is pretty fair to assume that you would have 100lp at 50% contrast - right=? Where else did you get the 100lp at 50% contrast then? (just want to understand ....)

For me the MTF of the Nikkor shows similar values at least in the center of the image. Sure, if you come out of the center the Oly glass wins. And if you compare that the Zuiko values are at f2.0 and the Nikkoe values at f2.8, then Oly is the clear winner from the lens side.

I think that this reflects the real state, as building a perfect lens for 43 size is easier (or gives you more constructive freedom) compared to FF (4x the image size).

Having said that I am still not sure if 43 is the sweetspot in image size - but I agree this is a more theoretical question to answer. But one other thing comes to my mind here - APSC would be the next size which established over the last decade. Why we do not find more perfect glass for APSC might be the reason because this is always considered kind of the cheap product lines of a company and not the top or pro lines - think that could be a reason.

I think our calculations are pretty ok still!

PS: Anyway I am interested in how the 2/14-35 performs with the E5.

PS1: I am going to order the GH2 as I really want to see how this beast performs with its 16MP resolution a M43 glass.

PS2: Not quite sure if I really want to abandon my Nikon FF line after all I found and discovered so far.

 

[Godfrey]

...
PS: Anyway I am interested in how the 2/14-35 performs with the E5.
...

According to several folks I chat with on another forum who have the 14-35/2, the E-5 has now made this their absolute favorite lens, rarely off the camera. They all complained, intermittently, of focusing issues with the 14-35/2 on the E-3 and other bodies although not a one ever complained about its performance optically.

I suspect that for what I like to shoot, I could buy one and super-glue it to an E-5, and be completely happy. ;-)

 

[nugat]

If you have 20/60 lp at 95% and 75% contrast respectively it is pretty fair to assume that you would have 100lp at 50% contrast - right=? Where else did you get the 100lp at 50% contrast then? (just want to understand ....)
.

It is not obvious to me. There shoud be a function formula to assume that, but supposedly this graph is a physical representation, not a simple function. I can be wrong of course.
I said I regret getting into this resolution stuff. The zuiko 100lp/mm is based on circumstantial evidence not a fact I know for sure. Leica best glass reaches 100lp/mm and Zuiko is not worse supposeedly. Etc etc. Let's just leave it and wait for the D4x to check.

 

[ptomsu]

It is not obvious to me. There shoud be a function formula to assume that, but supposedly this graph is a physical representation, not a simple function. I can be wrong of course.
I said I regret getting into this resolution stuff. The zuiko 100lp/mm is based on circumstantial evidence not a fact I know for sure. Leica best glass reaches 100lp/mm and Zuiko is not worse supposeedly. Etc etc. Let's just leave it and wait for the D4x to check.

Ok, got it

I am grateful that you brought up this discussion, good stuff and a typical example why I LIKE this forum

 

[Riley]

I think it's the same sensor isn't it?

What gets me is everyone going on about how small 4/3 sensors are in comparison, but actually, in terms of height, there really is very little difference, more in width, but of course thats 3:4 rather than 2:3 aspect ratio.

There isn't any reason why 4/3 sensors shouldn't be within a gnat's crotchet of APSc in terms of high ISO and quality . . . just that they haven't been in the past.

all the best

and all true
one of the things that happens in comparison becomes a duel of percentages, this sensor is 50% bigger than that, with the underlying intuition that the IQ must be 50% better

Knight Palm woke me up to this and did comparisons in the following table to e/v, this alleviates the perceptual error created by using percentages, which actually squares the differences in sensor size. In this way 4/3rds is (1.94 - 1.38) 0.56 stop behind Canon APSC and (1.94 - 1.21) 0.73 stop behind 1.5x. This is in my view a far better way of handling this, as it gives you the F stops apart (which is what really matters) and accounts for format differences unique to 4/3rds.

and just an observation but,
this method seems to be so far been avoided by the 'equivalence' conspirators,
and noting that 5D is 1 point off being real FF

 

[ptomsu]

and all true
one of the things that happens in comparison becomes a duel of percentages, this sensor is 50% bigger than that, with the underlying intuition that the IQ must be 50% better

Knight Palm woke me up to this and did comparisons in the following table to e/v, this alleviates the perceptual error created by using percentages, which actually squares the differences in sensor size. In this way 4/3rds is (1.94 - 1.38) 0.56 stop behind Canon APSC and (1.94 - 1.21) 0.73 stop behind 1.5x. This is in my view a far better way of handling this, as it gives you the F stops apart (which is what really matters) and accounts for format differences unique to 4/3rds.

and just an observation but,
this method seems to be so far been avoided by the 'equivalence' conspirators,
and noting that 5D is 1 point off being real FF

What does he mean with EV ?????

 

[Riley]

What does he mean with EV ?????

ev is exposure value
in this case the chart identifies the exposure value from FF for all formats. In order to depict the difference in ev between other formats just subtract one form the other. A difference in value of 1.00, is one stop

 

[ptomsu]

ev is exposure value
in this case the chart identifies the exposure value from FF for all formats. In order to depict the difference in ev between other formats just subtract one form the other. A difference in value of 1.00, is one stop

Ok, if this is Exposure Value - I thought EV is calculated of f-stop and shutter speed ...

How does that relate to sensor size?

 

[Riley]

Ok, if this is Exposure Value - I thought EV is calculated of f-stop and shutter speed ...

How does that relate to sensor size?

we say (as an approximation) that FF is 2 stops from 4/3rds right?
the actual difference in light is 1.94 stops, and so it goes for other formats there which are listed in their relationship to FF

 

[ptomsu]

we say (as an approximation) that FF is 2 stops from 4/3rds right?
the actual difference in light is 1.94 stops, and so it goes for other formats there which are listed in their relationship to FF

Issue is, if this tells really something about IQ etc. While FF is 2 stops from FF in 43, this means that you will get same DOF at a common focal length using e.g. 2.0 in 43 instead of 4.0 in FF. That is true, but it nowhere reflects the IQ achievable.

I think that one never can argue against 4x the sensor size between FF and 43, which makes it MUCH easier to achieve good higher ISO because of the size of the pixels at same resolution. Now today vendors obviously manage to overcome a lot of these restrictions with some advanced sensor technologies and advanced signal processing. Having said that - if one would make use of those advanced technologies also for FF it would give another quantum leap over smaller size sensors like 43 again. These are simply physical restrictions.

What worries me more is that all DSLR vendors still deal with AA filters - although Leica and MF vendors have proven that this is not necessary! If they would work without AA filters (or at least very weak ones as the E5 does) the IQ would be much higher. But it would also need much more advanced processing. Which in turn can no longer be the real issue today!

I would highly prefer say a Nikon D4X FF with maybe 20MP and WITHOUT AA filter, which in turn would be much slower (maybe only 3 FPS, as I need no image firing machine) but does real advanced processing of the image. In turn all the vendors are keen to show highest FPS rates and thus they need to filter heavily, in order to take details out of the taken images in order to make this processing work for so many frames per second - really weird.

What 43 really brings is the optimized sensor size for the lenses (or vice versa) which is really unique on the market and only a second vendor has done a complete reinvention of a system for digital which is Leica with their S System. All the others try to tweak their existing lens mounts and distances for digital, which finally never can deliver the same results as a system developed from the scratch.

I think that 43 has really done one of the best things when defining the sensor size as it is and the lenses resulting from that. Had they taken an APSC sensor size, that would have resulted in significantly larger lenses for the same IQ level as current 43 glass offers. Another major quantum leap will be possible with M43, as lenses can still get smaller than 43 glass. But so far real Pro lenses do not exist for M43 as we all know, maybe this is the next step.

Had a vendor taken FF and designed a complete new digital system around that, then everything (mount and lenses) would be significantly larger compared to today's FF. Somehow you can sense this by looking at the S system, which has a larger than FF sensor, but not as much, but think of a pretty similar lens size for FF digital if it would have been designed from scratch for digital. Or take the Hasselblad H system, which was designed with digital in mind and see how large the lenses are there - of course for medium format. But they are much larger compared to Phase, which is still based on the "old" analog system parameters.

Etc, etc.

 

[Pesitalia]

we say (as an approximation) that FF is 2 stops from 4/3rds right?
the actual difference in light is 1.94 stops, and so it goes for other formats there which are listed in their relationship to FF

I am a little bit confused here. I tought that EV values are independent from sensor format; in fact if I take a picture in a sunny day with a D700 and an E5 both set at 100 ISO, I will get 1/125sec at f/8 for both as long as the lenses have the same equivalent focal lenght or angle of view (let's say a Nikkor 50mm and a PL 25mm). The only difference would be in the DOF which will be about twice for the E3+25mm. In terms of light we will get same EV for both (likely EV=13).

Form another perspective the amount of light hitting 4/3 sensor is going to be about 1/4 of the light hitting FF because 4/3 lens diaphram is set at 3,125mm diam. while FF diaphram diameter will be 6,25mm:

FF exposed area thru the lens = 30,7 mm2
4/3 exposed area thru the lens = 7,7 mm2.

On the other end the 4/3 sensor has just about 1/4 the area of a FF sensor, thus the amount of light per unit area is the same, which justifies the equal settings for shooting under same conditions.

Really, the only difference remains DOF.

carlo

 

[ptomsu]

I am a little bit confused here. I tought that EV values are independent from sensor format; in fact if I take a picture in a sunny day with a D700 and an E5 both set at 100 ISO, I will get 1/125sec at f/8 for both as long as the lenses have the same equivalent focal lenght or angle of view (let's say a Nikkor 50mm and a PL 25mm). The only difference would be in the DOF which will be about twice for the E3+25mm. In terms of light we will get same EV for both (likely EV=13).

Form another perspective the amount of light hitting 4/3 sensor is going to be about 1/4 of the light hitting FF because 4/3 lens diaphram is set at 3,125mm diam. while FF diaphram diameter will be 6,25mm:

FF exposed area thru the lens = 30,7 mm2
4/3 exposed area thru the lens = 7,7 mm2.

On the other end the 4/3 sensor has just about 1/4 the area of a FF sensor, thus the amount of light per unit area is the same, which justifies the equal settings for shooting under same conditions.

Really, the only difference remains DOF.

carlo

This was precisely my point. I do not understand that table either.

 

[nugat]

f# is a dimensionless figure telling us about ratios of light falling per AREA UNIT in the focus plane. f4 is always 2x more light than f2.8, and 2x less than f5.6 , for a given scene ( framing) with the same lighting. Hand held exposure (light ) meters gave the same EI pairs of shutter/aperture regardless of format (for the same ISO). DOF on the other hand is a subjective and conventional measure of sharpness based solely on the assumed circle of confusion (CoC). We are still using the 19th century notion of sharpness, whereas a 8x10 inch contact print looked upon from one foot was deemed sharp if the CoC was not bigger than 1/100inch.

 

[robertro]

Beyond the limitations of maintaining support for non-digital lenses, I have to imagine that theoretical limits are largely overcome by financial concerns.

I can just imagine the conversation at the Nikon (and likely Canon) lens engineering R&D meetings; "you want $2 million to develop a new high end FF lens that'll sell 2000 copies - he wants the same amount to develop a slow crop zoom that'll sell 50,000 copies and help achieve margins for our new entry-level body - OK you full-frame guy get $100k, low-cost guy $1.9M. Do your best".

 

[nugat]

new pics from E5 vs D700

I went for a walk today with both combos: D700+24-120VR and E5+12-60.
You can download the direct comparisons here for private use only (the "D700 vs E5" album):

http://gallery.me.com/nugat#gallery

These are big jpegs: 6-11MB. At around 2:1 compression they are visually lossless, full size, no crops. Only the "Palace" I bothered to compare here and include Oly raw (ORF) to jpeg developed in Oly Viewer, as well as the out of camera (OOC) E5 jpeg. I could not see any difference ( I did couple more comparisons for myself ), so then I used only LR3 Oly OOC jpegs and Nikon NEF to jpeg.
Olympus is here exactly what E5 offered in Superfine JPG out of camera (I also recorded orfs). To level the ground I processed Nikon raws (14 bit NEFs/d-lighting normal) in LR3 Adobe Preset , as the Nikon OOC JPGs are immediately worse than Olympus in resolution. Fortunately LR3 developer is pretty good and no work on color had to be done. But most of these NEFs required exposition corrections (highlights, black level). Olympus on the other hand nailed the expossition almost perfect all the time. Both E5 and D700 were in "P" and exposure was full frame/center weighed, focus with one central spot. VR/IS on. Both camera chose very similar EI values and ISO at/near the native 200 (full EXIF included, or click "i" for info).

I will save you the trouble and give my appraisal immediately this time.
Both combos cost near the same money, are very similar in many aspects (Oly a bit smaller) and deliver similar pictures near native ISO.
Except Oly does it out of camera as JPEG, and on Nikon NEFs one needs to do work to get close. Close, because Oly OOC JPG resolution is a tad better than NEFs developed at the preset sharpening of 25. Again, I had the weird phenomenon of some Oly pics' sharp lines aliasing on the one monitor with the bigger pixels (EIZO), to smooth out at 100% on iMac 27 inch (smaller display pixels). Sharp people make sharp pictures. Sorry, Olympus people do so.
PS.
I could get Olympus jpegs sharpness working on NEFs with NIK Output Sharpener. But NIK on Oly raw/tiff again increased the distance.
In one of the pictures NEF seemed cleaner in shadows, but more comparisons should be done on that.
PSPS
When I get the SHG glass will do the lest leg of comparisons.