I Had Been Wondering What Actual Difference In Fine Detail I Would See With Various Sensor Sizes So I Conducted A Quick Test

My four digital cameras tested for rendering of the finest detail.  From the left: the 20mp OM Systems OM-1 with the 12-40mm f/2.8 PRO II lens, the 26mp Fujifilm X-H2S with the 16-55mm f/2.8 Red Badge lens, the 40mp Fujifilm X-T5with the 16-55mm f/2.8 Red Badge lens and the 45mp Nikon Z7II with the 24-120mm f/4 S lens. (click to enlarge)

I now own and use four different digital cameras.  All are mirrorless interchangeable lens cameras.  All have different format sensors with differing numbers of pixels.  So, I started wondering, in a practical, day-to-day sense, just what differences in fine detail would I see if I made identical images with all four cameras then examined the image files.  That is what I did.

My four current digital cameras are the Nikon Z7II with a 45.7mp sensor, a Fujifilm X-T5 with a 40.2mp sensor, a Fujifilm X-H2S with a 26.1mp sensor and an OM-1 with a 20.4mp sensor.  The Nikon has more than twice the pixels and the X-T5 has about twice the pixels of the OM-1. 

Add to this the OM-1 has a 50mp handheld high resolution mode as well as an 80mp tripod mounted high resolution mode.  The X-T5 has a high resolution mode as well but I did not test that as my previous tests using that feature showed highly flawed fine detail (a crosshatched pattern).  See my previous post, here, if you want to look at the flaws I found.  All that said, would I be able to see significant differences in the rendering of fine detail when comparing the image files from these cameras?  Let's test them and find out.

First, the test I conducted most likely wouldn't qualify for a highly scientific effort.  I'll call it a practical test.  It is repeatable, however.  I did my best to make all aspects of the test equal in every way and minimize any differences between the cameras and resulting image files.  Here is what I did:

-I first placed premium lenses on each camera to eliminate comparing excellent lenses to not so great lenses.

-OM-1 had the OM Systems 12-40mm f/2.8 II PRO lens 

-Fujifilm X-H2S and X-T5 had the Fujifilm 16-55mm f/2.8 Red Badge lens 

-Nikon Z7II had the Nikon 24-120mm f/4 S lens

-All images were made at the 50mm 35mm equivalent focal length as that is in the middle of each lens' zoom range and using the middle eliminates any lack of sharpness that is sometimes found at either the short end or long end of zoom lenses.

-Only used the sharpest portion of the lens (center) for evaluation.

-All cameras set to aperture priority and lenses at their approximate sharpest aperture using the the same "effective" aperture so depth of field was the same (OM-1 @ f/4; Fujiflm @ f/5.6; Nikon @ f/8).

-All cameras mounted on a sturdy tripod.

-IBIS was turned off.

-2 second shutter delay was implemented.

-All cameras had the first curtain electronic shutter enabled to eliminate potential 'shutter shock.'

-Raw format was used with the exception of the high resolution OM files which were set to raw + JPEG (JPEG can't be disabled in the 80mp mode).

-Lighting was constant during all the tests.

-Base ISO was used for each camera.

-Manual focus with about 10X magnification and focus peaking was used to accurately focus each exposure.  This test was NOT to test the AF but that is what it would have been if I had used AF-S or AF-C.

-Since Lightroom doesn't do the best job in converting X-Trans sensor files, I sent all the images to DXO PhotoLab 5 for raw conversion to DNG and the application of DXO's Deep Prime noise reducing and sharpening algorithms.  DXO also applied their specifically engineered camera/lens combination correction feature which eliminates vignetting, aberrations, distortion and also improves detail and lens sharpness.  By using DXO instead of Lightroom, I felt that the test would be more fair and consistent between brands and sensor types.  

Results 

Sort of a disclaimer.  There may be more scientific tests that have been conducted looking at this same issue.  Those test results may differ or be more quantitative than mine.  However, I wanted to run a practical test using my cameras with my copies of my lenses.  You may get different results with your gear.

Also, remember that to double resolution in an image file you need four times the number of pixels.  In other words, to double the resolution of a 20mp sensor, you need a sensor with 80mp.  So, I don't think we can expect large differences in detail in these image files.

In the images I'm posting below, you won't be able to see what I see on my monitor so I will attempt to describe what differences I see in the various files.  For these images, I cropped each image down to only 1.8% of its original size so I could see the finest detail rendered as well as when detail fell apart.  I then looked at each cropped image at 100%, 200% and 300%.

At 100% magnification all of the images look pretty much the same.  

I really don't see any meaningful difference in the details from the 20mp OM-1 and the 26mp X-H2S at any magnification.  I expected that.

I can see a tiny difference in fine detail between the OM-1/X-H2S and 40mp X-T5 files.  Again, I kind of expected to see some difference.  In my judgment, however, not enough to make a meaningful difference that I would want to sell the 20/26mp cameras and run out and buy the 40mp camera, but nonetheless a bit of a difference.  For example, I can just barely read the word "Splenda®" in the X-T5 file whereas it is a bit broken up in the OM-1 and X-H2S files.

I don't see any meaningful differences between the 40mp X-T5 and 45mp Z7II image files.  Again, expected.  However, they render just a little different but detail is about the same.  What do I mean by that?  In some areas the letters render a tiny bit better in the Nikon file.  I think they look different because the X-T5 has an X-Trans file and the Nikon has a Bayer Array file.  I think the difference is the way they capture original data in the photosites in the sensor.  This comment is up for further review.  To be honest, I'm not sure if I'm seeing different or not.  I'll possibly do some additional testing in the future.

The 80mp high resolution file from the OM-1 has better detail than the 40mp X-T5 file and the 45.7mp Nikon file.  Not much, but better fine detail.  It is the best file of the bunch.  I can read some of the fine print with this file that I can't with the Nikon or X-T5 files.  Just barely, however.  Again, the differences are really subtle.

The 50mp handheld high resolution file from the OM-1 is not quite as good as the 40mp X-T5 file or the 45.7mp Nikon file.  However...The 50mp high resolution feature is supposed to be handheld as it uses the subtle camera movement between frames to be combined and made into the 50mp file.  I had the camera on a tripod so this comparison might not be entirely valid.  OM would know but I don't know anyone there to ask.  The reason I didn't handhold the camera is that the light was relatively low and the shutter speed would have been relatively long or the ISO higher which would have skewed the results.  I might try this again outdoors on a sunny day in the future just to ensure the results are accurate.

This was unexpected.  The OM-1's high resolution JPEGs look as good or even a bit better than the commensurate 50mp and 80mp raw files.  Now that's interesting.  However, this reinforces my earlier tests when first trying out this feature on the camera.  In other informal tests, the detail was better in the high resolution JPEG files than the raw files.  

Observations 

-The OM was by far the easiest to manually focus with the nice sliding focusing clutch.

-The Z7II was the most difficult to focus, not because of the camera but because of how the lens' manual focusing is engineered.  The lens isn't as responsive to minute focusing ring movements as much as are the OM and Fujifilm lenses.  I had to rock it back and forth numerous times just to ensure I had the absolute best focus.

-The OM-1 had the best focus peaking in my opinion.  It seemed to 'snap' into focus more visibly and the (red) color move forward and backward showing what was in focus a bit better.

-I think, considering having the right circumstances, the high resolution image file feature of the OM-1 is a worthwhile option if you need larger and more detailed files.

Here are the images starting with the least number of pixels to the most.

This as well as the five following images represent the center 1.8% of the overall image.
20mp OM-1; 12-40mm f/2.8 lens @ 25mm; f/4; ISO 200

26mp Fujifilm X-H2S; 16-55mm f/2.8 @ 33mm; f/5.6; ISO 160. (click to enlarge)

40mp Fujifilm X-T5; 16-55mm f/2.8 lens @ 33mm; f/5.6; ISO 125 (click to enlarge)

45mp Nikon Z7II; 24-120mm f/4 S lens @ 50mm; f/8; ISO 64 (click to enlarge)

50mp OM-1 high resolution mode; 12-40mm f/2.8 II PRO lens @ 25mm; f/4; ISO 200

80mp OM-1 high resolution mode; 12-40mm f/2.8 II PRO lens @ 25mm; f/4; ISO 200

Conclusion 

My general conclusion is that there is a small bit of difference in the rendering of very fine detail under ideal conditions (tripod mounted, no wind, exposure delay, optimum aperture, perfectly focused, base ISO, etc.) between the 20/26mp sensors when compared to the 40/45mp sensors.  But unless you need very large files for very large prints (4 ft. wide or larger), I don't see enough fine detail differences in any of these files that it will make a difference with what I would call mainstream sized prints.  Also, if you need to crop in more than 50-75% of a file, again, I don't think any differences in fine detail will make a practical difference in viewing these different sized files.  If you are photographing under less than ideal conditions, you may lose any differences in detail.  

I think the quality of the lens, ISO, aperture and your shooting technique will impact the fine detail more than what the sensor can ultimately record.  If you aren't meticulous in your photography, you may not see any differences.

If you are a fan of cameras with 35mm sensors, that's great!  You have all the resolution and fine detail you will ever need.  If you are a fan of Fujifilm cameras, I don't think you are missing anything from those who have 35mm sensor sized cameras, especially if you are using an X-T5 or X-H2.  If you are a fan of Micro 4/3, unless you need a really big print or need to crop in significantly and then make a big print, you aren't missing much.  A little but not much.  I don't think anyone will tell the difference in detail in your images versus those with a sensor with more pixels under normal circumstances.

I've written before that I enjoy punching in to a 45mp file and seeing the fine detail.  However, almost all of that fine detail is there in my 20/26mp files, its just that they don't enlarge as much when viewed at 100%.  I've also written that I don't need 45mp files.  That opinion has been reinforced to me after conducting this test.

One more thought.  There are other factors that have to do with the image quality you get from different format sensors.  Even though dynamic range between formats has fewer differences than before, if you are a photographer who routinely photographs high dynamic range scenes, that may be a factor.  But for the vast majority of us enthusiasts, I no longer think sensor size and an extreme number of pixels really makes a difference in our day-to-day photographic endeavors.

That will do it for now.  I hope you have found this interesting.  If you have questions please leave them in the comments below

Join me over at my website, https://www.dennismook.com.  

Thanks for looking. Enjoy!  

Dennis A. Mook  

All content on this blog is © 2013-2023 Dennis A. Mook. All Rights Reserved. Feel free to point to this blog from your website with full attribution. Permission may be granted for commercial use. Please contact Mr. Mook to discuss permission to reproduce the blog posts and/or images.