An E-M5 II User’s Test of the Nikon Z6

Terrible Photographer said:

Kamukix said:

Remind me to stay far far FAR away from that 24-70 f/2.8 then haha. I've used some great glass over the years, but the numbers that I keep seeing from the Z glass make my mind hurt a bit haha. Test charts are of course one thing, but I know the reality to be just as insane. I absolutely LOOOOOOOOOOOOVED my 25/45 1.2 Pro lenses...but I know there is a limit to what you can do with glass and a sensor that small. I can't imagine using a 70-200 though on my Z6, it'd be okay with good handling discipline but I still have zero desire to go down that road these days. If I did however I'd definitely prefer to do it with a DSLR again most likely due to the extra weight in my hand I feel like I'm fighting with it less when there is that much weight. I'm well past strong enough to still lug that kinda gear around all day....but I just don't want to anymore haha.

Click to expand...

It took me almost an entire year of testing multiple copies of lenses to figure out that YES, THERE ARE LIMITATIONS.

The only m4/3rds lenses I have that actually have sharp corners by f/5.6 are the 300, 40-150, and 75. The 25 1.2, 45 1.2, 17 1.8, 12-40, 7-14 never actually get the last 1/4 peripheral of the frame perfectly sharp. The 45 almost gets there if I'm focusing somewhat close, like 4-7 feet.

My Nikon 85 1.8 at f/5.6 is much more uniform and sharper across the field vs the 45 1.2 Pro at f/2.8-5.6. Kind of a bummer, but it's reality.

Click to expand...

James Stirling said:

s_grins said:

I want letter size paper at min. distance that lens allows. Focus point in the center and normal to the surface. Also, RAW without optical DxO correction ( you can use FastStone software) Can you do it?

Thanks

PS zoom it in for shortest FL

Click to expand...

Oddly enough if I was photographing sheets of letter paper at minimum focus distances which is 28cm on the Nikon and 20cm I think on the Olympus I would use an appropriate lens and it would not be a a m43 lens at 7mm or a FF lens at 14mm . So for me it would be a rather daft way to look at it. But if that is how you shoot UWA lenses knock yourself out I am looking forward to your results .I assume by the precise suggestions you will already have the results at hand , I wouldn't want you wasting your time because of a daft suggestion online

Click to expand...

James Stirling said:

Stan Powers said:

Curious how we are all so different.

I found my E-M1II better (vs Z6) in low light as I able to handhold at a lower ISO. Faster AF acquisition, less noise and sharper images.

Choice is good!

Click to expand...

Stan ,assuming a two stop difference in IBIS between these cameras , there are CIPA standards for this . But hey some here think their random opinion trumps scientific testing recognised and quoted by Olympus and all the other makers with IBIS.

This two stop advantage would be eaten away with the two stop noise advantage . You also better be photographing rocks as for any movement the IBIS benefits blow away. I can shoot at 4 stops below the reciprocal shutter speed with my Z7 I assume the Z6 will be about the same.

So using say a 50mm lets go with common shutters speed so I will get down to 1/4 sec at the same ISO as { the red dot on the chart} . Now add your two stops and you are now at 1 sec { just how low light do you mean} .



This is a z7 shot handheld at 1/4th at 64ISO F/5.6 50mm in a room where the only light is a desk lamp with one of those energy saving bulbs .



This is a 100% comparison of the E-M1 II RAW low light 800 ISO VS the Z6 RAW low light at 3200 ISO showing what one would expect , a clear two stop advantage to the Nikon



Browsing through your linked flickr account the first z6 comparison with the E-M1II I came across . Is done in exactly the sort of comical way one expects here. Namely taking the Nikon at F/4 1/50th at 44mm and 5600 ISO and the Olympus at 0.6 at f/4 and 200ISO So you are going with zero stops ability for IBIS with the z6 .

It is always amusing here that no matter the photographic challenges compared to other formats. There are only ever two answers namely take a 16 shot { or however many } pixel shift image or they can shoot handheld IBIS for wildly exaggerated times and I am still waiting to see what I consider even remotely sharp handheld multi-second shot.

--
Jim Stirling:
It is not reason which is the guide of life, but custom. David Hume

Click to expand...

Tommi K1 said:

kenw said:

Tommi K1 said:

Bridgehouse said:

• Choosing to buy into a FF system is very much an individual thing. I had distinct reasons for doing so. I wanted FF for digitising film - I still shoot a lot of 35mm and it's very easy when digitising 1:1. That's why I also picked the Z7 - 45mp means *I'm close* to out resolving the grain on even the most finely grained B&W film stocks.

Click to expand...

You "out resolve" 35mm film with 16Mpix 4/3" sensor really. You need to step up to 4x5" film to get more.

https://theonlinephotographer.typep...ormat-film-we-have-the-definitive-answer.html

"Image quality is a multidimensional thing, some of which can be quantified and some not. Still, by no measure of image quality does a good Micro 4/3 camera and lens perform more poorly than a good medium format film rig, and by some measures it performs considerably better. My overall subjective evaluation is that the aggregate image quality of Micro 4/3 today, in film terms, falls midway between 6x7 medium format and 4x5-inch large format."

Click to expand...

The disconnect here is that the table and linked article relate to comparing direct capture on film vs direct capture on a digital sensor. That is actually a different question than the resolution required to sample the film and not degrade the resolution any further than the original direct capture to film achieved.

Click to expand...

If you would read the article of the image, he stated that is his conclusion of all the the tests between film and digital in Mpix count.

Member said:

I don’t have that much direct experience in image processing but from the little that I do I believe the rule of thumb is usually a factor of two or so in linear dimension.

Click to expand...

Both of them are as good as one can get when it comes to film, digital sensors and printing. Far more experience and knowledge than anyone in DPR or in its community, and ever will get (because many of the technologies are already gone or never available for them):

Member said:

"Almost everyone you can find who is still arguing that Micro 4/3 can't match up to professional film has not done substantial amounts of serious work in both media. I believe the technical term is 'talking through one's hat.'"

Click to expand...

Click to expand...

kenw said:

kenw said:

Tommi K1 wrote

Click to expand...

So the vaunted Ctein has clearly stated that a 24MP camera is not good enough for him to digitize 35mm film.

Click to expand...

Tommi K1 said:

kenw said:

Member said:

Tommi K1 wrote

Click to expand...

So the vaunted Ctein has clearly stated that a 24MP camera is not good enough for him to digitize 35mm film.

Click to expand...

Looks like you do not read what was written.

Digitalization is different thing than is printing. And making a FINAL OUTPUT, is not at all same thing as doing FIRST INPUT.

So please read the articles, comprehend that what a various different parts they go around and what they actually both write about. And you will find that the graph that I quoted is the overall conclusion of the high DPI scanning of even large format films etc, but put the general context for the 35mm film.

Click to expand...

kenw said:

Tommi K1 said:

Bridgehouse said:

- Choosing to buy into a FF system is very much an individual thing. I had distinct reasons for doing so. I wanted FF for digitising film - I still shoot a lot of 35mm and it's very easy when digitising 1:1. That's why I also picked the Z7 - 45mp means I'm close to out resolving the grain on even the most finely grained B&W film stocks.

Click to expand...

You "out resolve" 35mm film with 16Mpix 4/3" sensor really. You need to step up to 4x5" film to get more.

https://theonlinephotographer.typep...ormat-film-we-have-the-definitive-answer.html

"Image quality is a multidimensional thing, some of which can be quantified and some not. Still, by no measure of image quality does a good Micro 4/3 camera and lens perform more poorly than a good medium format film rig, and by some measures it performs considerably better. My overall subjective evaluation is that the aggregate image quality of Micro 4/3 today, in film terms, falls midway between 6x7 medium format and 4x5-inch large format."

Click to expand...

The disconnect here is that the table and linked article relate to comparing direct capture on film vs direct capture on a digital sensor. That is actually a different question than the resolution required to sample the film and not degrade the resolution any further than the original direct capture to film achieved.

In general in most any resampling process you want the resampling operation to occur at higher resolution, finer quantization and higher SNR than the original capture. Depending on what you are doing that might range from a factor of two up to a factor of ten. I don’t have that much direct experience in image processing but from the little that I do I believe the rule of thumb is usually a factor of two or so in linear dimension. So if Tech Pan captures about 16MP you’d want to digitize that (either camera or scanner) at about 64MP.

This is essentially the repeated photocopying problem. Copies get worse and worse if repeatedly resampled at the same quality. If you want to preserve the original quality of any sort of capture (image, sound, RF) you do subsequent processing at higher quality than the original.

Click to expand...

Bridgehouse said:

I don’t want to start a debate about film resolution, as it’s the subject of a lot of debate and argument already...

But in essence, and from my experience of digitising thousands of shots with a variety of I tend to subscribe to the formula of 4 pixels per line pair to fully out resolve film.

An average black and white film stock is around 50 lppmm resolving power so:

24*4*50 = 4800 36*4*50 = 7200

4800*7200 = 34.5 mpix

The more conservative 2 pixels per line gives 2400*3600 = 8.6 mpix - but I personally don’t think this works with colour sensors - it’s more appropriate for monochrome sensors

Velvia is quite different. Fuji state:



And at 160 lpmm:

24*4*160 = 15,360

36*4*160 = 23,040

23040*15360 = 354mpix

At the conservative 2 pixels per line you get:

24*2*160 = 7680

36*2*160 = 11520

11520*7680 = 88.5mpix

This is in ideal circumstances of course, but in my experience even 45mpix isn’t truly revealing all captured detail on Velvia. Sure, with black and white 16mpix is more than adequate, but when dealing with slide film more mpix is always a good thing, and even the 45 of the Z7 in some circumstances isn’t really enough to capture everything..

Click to expand...

Rohith Thumati said:

E-M5 II - Tree - Trunk color needed a good amount of work to get to look like this

Click to expand...

dpreviewreader said:

Bridgehouse said:

I don’t want to start a debate about film resolution, as it’s the subject of a lot of debate and argument already...

But in essence, and from my experience of digitising thousands of shots with a variety of I tend to subscribe to the formula of 4 pixels per line pair to fully out resolve film.

An average black and white film stock is around 50 lppmm resolving power so:

24*4*50 = 4800 36*4*50 = 7200

4800*7200 = 34.5 mpix

The more conservative 2 pixels per line gives 2400*3600 = 8.6 mpix - but I personally don’t think this works with colour sensors - it’s more appropriate for monochrome sensors

Velvia is quite different. Fuji state:

And at 160 lpmm:

24*4*160 = 15,360

36*4*160 = 23,040

23040*15360 = 354mpix

At the conservative 2 pixels per line you get:

24*2*160 = 7680

36*2*160 = 11520

11520*7680 = 88.5mpix

This is in ideal circumstances of course, but in my experience even 45mpix isn’t truly revealing all captured detail on Velvia. Sure, with black and white 16mpix is more than adequate, but when dealing with slide film more mpix is always a good thing, and even the 45 of the Z7 in some circumstances isn’t really enough to capture everything..

Click to expand...

Roger Clark on his website writes

"Fujichrome Velvia (both the ISO 50 and ISO 100 velvia) have an lpm1.6 = 80 lpm. Equation 1 gives 10 megapixels for intensity detail, but color detail would require 16 megapixels."

Some years ago when digital was becoming popular, digital vs film debates were common. I saw comparisons and 16MP digital did match 35mm Velvia images. Your calculations above are not validated by actual results.

Click to expand...

medon78 said:

Thanks for this interesting read! Actually I am lusting after the Z6 as well, so your insight is quite useful for me.

Regarding the color issues on the tree trunk:

Lightroom profiles are definitely not the same quality across all cameras.

I remember that on most of my older Canon APS-C bodies, LR would produce some heavy color shift in pushed shadow areas, where Canon's own DPP raw converter would do just fine.

Same goes for my Olympus bodies -- while I sill appreciate LR for its sheer power, I stopped using it for the colors being not as nice as with the camera makers' own raw tools. Unfortunately of course, neither Olympus Work Space nor DPP allow for local adjustments, AFAIK.

BR Medon
...

medon78 said:

E-M5 II - Tree - Trunk color needed a good amount of work to get to look like this

Click to expand...

Click to expand...

M_digicapt said:

medon78 said:

Thanks for this interesting read! Actually I am lusting after the Z6 as well, so your insight is quite useful for me.

Regarding the color issues on the tree trunk:

Lightroom profiles are definitely not the same quality across all cameras.

I remember that on most of my older Canon APS-C bodies, LR would produce some heavy color shift in pushed shadow areas, where Canon's own DPP raw converter would do just fine.

Same goes for my Olympus bodies -- while I sill appreciate LR for its sheer power, I stopped using it for the colors being not as nice as with the camera makers' own raw tools. Unfortunately of course, neither Olympus Work Space nor DPP allow for local adjustments, AFAIK.

BR Medon
...

M_digicapt said:

E-M5 II - Tree - Trunk color needed a good amount of work to get to look like this

Click to expand...

Click to expand...

Which program do you use when you need to llift shadows?

Thanks

Click to expand...

M_digicapt said:

Which program do you use when you need to llift shadows?

Thanks

Click to expand...

dpreviewreader said:

Bridgehouse said:

I don’t want to start a debate about film resolution, as it’s the subject of a lot of debate and argument already...

But in essence, and from my experience of digitising thousands of shots with a variety of I tend to subscribe to the formula of 4 pixels per line pair to fully out resolve film.

An average black and white film stock is around 50 lppmm resolving power so:

24*4*50 = 4800 36*4*50 = 7200

4800*7200 = 34.5 mpix

The more conservative 2 pixels per line gives 2400*3600 = 8.6 mpix - but I personally don’t think this works with colour sensors - it’s more appropriate for monochrome sensors

Velvia is quite different. Fuji state:



And at 160 lpmm:

24*4*160 = 15,360

36*4*160 = 23,040

23040*15360 = 354mpix

At the conservative 2 pixels per line you get:

24*2*160 = 7680

36*2*160 = 11520

11520*7680 = 88.5mpix

This is in ideal circumstances of course, but in my experience even 45mpix isn’t truly revealing all captured detail on Velvia. Sure, with black and white 16mpix is more than adequate, but when dealing with slide film more mpix is always a good thing, and even the 45 of the Z7 in some circumstances isn’t really enough to capture everything..

Click to expand...

Roger Clark on his website writes

"Fujichrome Velvia (both the ISO 50 and ISO 100 velvia) have an lpm1.6 = 80 lpm. Equation 1 gives 10 megapixels for intensity detail, but color detail would require 16 megapixels."

Some years ago when digital was becoming popular, digital vs film debates were common. I saw comparisons and 16MP digital did match 35mm Velvia images. Your calculations above are not validated by actual results.

Click to expand...

medon78 said:

M_digicapt said:

Which program do you use when you need to llift shadows?

Thanks

Click to expand...

Both Olympus Work Space and Canon Digital Photo Professional (DPP) feature shadow lifting capabilities, see below.

They are locally adjusting shadows, and they both replicate what the respective cameras can do in their built-in JPEG engines.

Both features are not as powerful as Adobe's shadow recovery though.


Olympus Work Space -- Auto Gradation


Canon Digital Photo Professional -- Auto Lighting Optimizer

Click to expand...

kenw said:

dpreviewreader said:

Bridgehouse said:

I don’t want to start a debate about film resolution, as it’s the subject of a lot of debate and argument already...

But in essence, and from my experience of digitising thousands of shots with a variety of I tend to subscribe to the formula of 4 pixels per line pair to fully out resolve film.

An average black and white film stock is around 50 lppmm resolving power so:

24*4*50 = 4800 36*4*50 = 7200

4800*7200 = 34.5 mpix

The more conservative 2 pixels per line gives 2400*3600 = 8.6 mpix - but I personally don’t think this works with colour sensors - it’s more appropriate for monochrome sensors

Velvia is quite different. Fuji state:

And at 160 lpmm:

24*4*160 = 15,360

36*4*160 = 23,040

23040*15360 = 354mpix

At the conservative 2 pixels per line you get:

24*2*160 = 7680

36*2*160 = 11520

11520*7680 = 88.5mpix

This is in ideal circumstances of course, but in my experience even 45mpix isn’t truly revealing all captured detail on Velvia. Sure, with black and white 16mpix is more than adequate, but when dealing with slide film more mpix is always a good thing, and even the 45 of the Z7 in some circumstances isn’t really enough to capture everything..

Click to expand...

Roger Clark on his website writes

"Fujichrome Velvia (both the ISO 50 and ISO 100 velvia) have an lpm1.6 = 80 lpm. Equation 1 gives 10 megapixels for intensity detail, but color detail would require 16 megapixels."

Some years ago when digital was becoming popular, digital vs film debates were common. I saw comparisons and 16MP digital did match 35mm Velvia images. Your calculations above are not validated by actual results.

Click to expand...

You are making the same mistake that Tommi K1 is making.

Click to expand...

kenw said:

What Roger Clark was referring to is what resolution sensor taking a picture of a scene through a lens will capture the same detail as 35mm film taking a picture of a scene through a lens. That is the many times mentioned around 16MP.

But what is being discussed here is how much resolution must you scan or digitally photograph the film at to not further degrade the resolution the film captured. Naively the answer seems 16MP but that is in fact wrong. You will end up with less than 16MP of resolution in your scan/photo if you do this.

And in fact just as Ctein clearly stated that even 24MP is insufficient to digitize 35mm film at so does Roger Clark:

https://clarkvision.com/articles/scandetail/index.html

The most relevant quote from there:

From the above scans, it shows that 35mm shows increasing detail at least to 6000 dpi in the drum scans (5650 x 8500 pixel image = 144 MBytes), and similar to the best consumer film scanners available today (4000 dpi). The more common film scanners (2400 to 2800 dpi) do not record all the detail on images such as this one (they probably will record all detail on grainier film, or hand-held snapshots). Note, 6000 DPI on 35mm is 48 megapixels (RGB pixels)

So Ctein and Roger Clark both adamantly agree that if you were going to try to “scan” 35mm film with a camera that the camera needs to be much more than 24MP. They also clearly state that if you want to directly take a picture of a scene with a digital camera that 16MP is plenty to exceed the results of 35mm film.

This is not a contradiction and is in fact completely logically consistent if you take the time to understand how MTFs of optical systems, sensors, film and scanners work. If you want to preserve the detail of the previous stage in the imaging chain you must sample at a higher resolution than the previous stage outputs. If you instead sample at the same resolution you will in fact degrade the image and lose resolution.

None of this is new or rocket science. Ctein, Clark and many others have been clear on this for ages. It is pretty fundamental to film scanning.

Click to expand...