Oly, Canon, Fuji. Though, I'm not at all confident in that pick.
The first one seems quite flat to me, skin-tone wise. The second is maybe overexposed but there's more going on. It also has the most DOF, which to me says it's a 50mm at 2.8 but from slightly farther back than the other lenses.
I'm going by size of catch light in the eyes. Medium, small and large in that order. So either go by format size or fstop? I think fstop so aps-c, full frame, and m4/3. But in any case hard to tell.
" when you go out of your way to make all the shots look the same, of course it's going to be hard to tell them apart,"
Huh....? The OP didn't go out of his way at all.
"but I don't really shoot that way,"
Don't shoot what way?
"It's most difficult for me to notice format differences on people pix."
So..format only matters in stuff like landscapes and such.
This supports a point that I have been telling others for awhile. Format does not matter except for extreme situations (large prints, very poor lighting,etc.) I shoot weddings with a 7d and a 5d MKII and NOBODY knows the difference between images.
As has been noted by the catchlights (which I didn't see, but it is indeed a way to tell apparently):
Image 1 is the Fuji X-M1 with 35/1.4 @ f/2
Image 2 is the Canon 1Ds Mark II with Sigma 50/1.4 @ f/2.8
Image 3 is the OM-D E-M5 with PL 25/1.4 @ f/1.4
Some fun with stats:
- Only 14% of voters correctly picked all three. This is slightly LOWER than random chance (16%), but given the sample size, you can say that the ability for people to identify all the images correctly is no better than eeny-meeny-miney-mo.
Here's where it gets interesting:
- 27% correctly identified the first image as APS-C - Lower than random chance of 33%
- 31% correctly identified the second image as Full Frame - again slightly lower than random chance.
- Only 20% correctly identified the final image as m4/3 - WELL below random chance.
Thus, the ability for people to identify format based on this image is WORSE than randomly picking.
The most interesting statistic: nearly HALF of the people participating in the poll (47%) thought the Full Frame image was the Micro 4/3 image.
For this test, it doesn't conclusively prove much of anything....this test doesn't really stress the limits of these cameras, but it is interesting to see that for a typical shot, people can't generally tell what format of camera took a picture. This is especially true in my experience when you are just looking at any random image...the image quality of modern cameras is really good at most levels...it's only at the extremes where quality differences really start to stand out. Of course, there are other benefits to larger formats...shallower depth of field, availability of specialty lenses like tilt/shift, etc....
This seems like fun, but to come to any reasonable conclusion shots should be taken of a controlled subject, from a tripod, and reproductions should feature "normal" viewing plus 100% details of sharp high contrast areas.
It's fun to say everything looks equivalently identical on a web screen reproduction, but it's something else to see enlarged files from various sized sensors and lenses.
Anyway, I hope someone knocked down all the leaded milk bottles at the amusement park.
Cool, I got it right.
The reasoning:
I recognised the PL 25mm in the third shot, so that just had to be the Oly.
The second shot has better fine detail than the first shot, so that second one had to be the Canon.
I thought 2 was oversharpened, which led me to guess it was m4/3. my experience has been that larger formats can take more sharpening, so I guessed on that rather than the amount of fine detail present.
This exercise really does show how close they are. However, its been my experience that the larger the format, the richer the detail of the skin is. Despite being old tech, the FF IHMO still has a slight edge on color transitions and a tonality of the skin.
Gunzorro - I disagree with all your points after the 'but...'
This was not a test of detail (which is highly lens dependent and has nothing to do with format), this was not a test of noise (mainly because the FF cam is older, and I can tell you right now who wins: for same shutter speed and aperture, it will go Fuji best, then the Canon and OM-D very close...maybe a slight edge to the Canon...for 'handholdability at same depth of field', the OM-D will be best because it can shoot at base ISO in these conditions, while the Canon would need ISO 1600 (though here the OM-D was at ISO 400, two stop slower than the Canon)....
A tripod misses the point, since this was about a test under normal shooting conditions - I'd never shoot a candid portrait like this with a tripod. All three files are sharp...so what's the point?
This test was about rendering, and the 'look' you get from the files. Trust me, all three cameras can resolve tons of detail and all three do a pretty good job in moderate ISO conditions. The Fuji and Canon are a bit smoother at base ISO, the Fuji is cleaner at really high ISOs...that's not the point...the point is simply if the average photographer can see a noticeable difference between images shot with different formats under typical indoor conditions. The results tell me no. Yes, this is a 'web' reduction, but they're 1600 pixel wide images, which aren't too small. On my monitor, they're equivalent to about a 12x18" print.
Again, this doesn't prove anything about the relative merits of each platform...all it shows is that for non-extreme shooting environments, it's pretty hard to tell any difference between formats at 'normal' enlargements. You can't extrapolate anything beyond that.
I thought 2 was oversharpened, which led me to guess it was m4/3. my experience has been that larger formats can take more sharpening, so I guessed on that rather than the amount of fine detail present.
All three images had default LR sharpening and my standard 'reduction for web - 1600px' action run. Only other post on any of the images was whitebalance adjustment (and I cloned out a snotty - my daughter has a cold).
This exercise really does show how close they are. However, its been my experience that the larger the format, the richer the detail of the skin is. Despite being old tech, the FF IHMO still has a slight edge on color transitions and a tonality of the skin.
I did a few shots at 85mm equivalent with the CV 42.5/0.95 on m4/3, the Hexanon 57/1.2 on APS-C and the Rokinon 85/1.4 at f/2 on FF, but it became painfully obvious since the Rokinon not only has exceptionally smooth bokeh at that range (much smoother than the other two) and at f/2, the Rokinon is far sharper than either of the two manual lenses....it became much more a lens test than how the format themselves render. I could shoot the Oly 45/1.8, Fuji 60/2.4 and Rokinon at f/3.5, and that might be close, since those lenses are fantastic wide open and have smooth bokeh as well (though focusing the Rok on the 1Ds II is SO much harder than on the other platforms...I had to massively focus bracket at the wider apertures).
Would be fun to see if the upcoming PL 42.5/1.2 is really sharp wide open.
As an aside - here are all three formats at f/1.4 - to see the differences in minimum depth of field possible with this framing (personally, in this image, the FF wide open is too shallow...I like both the APS-C and m4/3 DOF...which jives with my stuff...when I shot full frame, I often shot this 50 at f/2):
Jman13 wrote:
Gunzorro - I disagree with all your points after the 'but...'
This was not a test of detail (which is highly lens dependent and has nothing to do with format), this was not a test of noise (mainly because the FF cam is older, and I can tell you right now who wins: for same shutter speed and aperture, it will go Fuji best, then the Canon and OM-D very close...maybe a slight edge to the Canon...for 'handholdability at same depth of field', the OM-D will be best because it can shoot at base ISO in these conditions, while the Canon would need ISO 1600 (though here the OM-D was at ISO 400, two stop slower than the Canon)....
A tripod misses the point, since this was about a test under normal shooting conditions - I'd never shoot a candid portrait like this with a tripod. All three files are sharp...so what's the point?
This test was about rendering, and the 'look' you get from the files. Trust me, all three cameras can resolve tons of detail and all three do a pretty good job in moderate ISO conditions. The Fuji and Canon are a bit smoother at base ISO, the Fuji is cleaner at really high ISOs...that's not the point...the point is simply if the average photographer can see a noticeable difference between images shot with different formats under typical indoor conditions. The results tell me no. Yes, this is a 'web' reduction, but they're 1600 pixel wide images, which aren't too small. On my monitor, they're equivalent to about a 12x18" print.
Again, this doesn't prove anything about the relative merits of each platform...all it shows is that for non-extreme shooting environments, it's pretty hard to tell any difference between formats at 'normal' enlargements. You can't extrapolate anything beyond that. ...Show more →
Random chance said there was a 16% chance you would agree with my points after the "but. . ."
I agree that it doesn't really show anything that a casual observer could distinguish one photo from the other. Certainly there is no intended criticism of your skills and your lovely daughter's patience -- the shots are terrific and each would work for family portraiture. I was simply remarking that for any strong commercial purposes, we'd need to examine the images more closely and rule out variables introduced by attempts to make them seem the same or similar.
Interesting thread. The fact that the results showed a worse than random ability to differentiate based on these shots is a little surprising given how much large sensor bias one tends to find online.
Having shot all three of these formats in similar conditions, I absolutely agree that they all are capable of delivering fine results in this kind of scenario. Most of my favorite photos of my daughter are still the ones I took with my OM-D and PL 25, despite having shot plenty with larger sensor format cameras.
Gunzorro wrote:
I agree that it doesn't really show anything that a casual observer could distinguish one photo from the other.
Yup, that's us ... a bunch of "casual observers".
One other interesting stat tidbit comes to mind for me:
52% of us "casual observers" (37% + 15%) managed to get NONE of them right (of which I was one). Given there were only 2 out of the 6 choices that could yield them ALL wrong, that would only be a 33% @ random. We were at 1.5X of that.
Do we play again with an HDR outdoor infinity scene?
As I have said before, the output from different formats is quite obvious. Do I win a prize? All kidding aside, there are differences but they are very minimal in the examples you have shown. Thanks for the comparison.
