mttran wrote: Sarsfield wrotemttran, not to be a smarta$$ but what are those little samples/comparisons supposed to show? They all look like crap to me.Look like 2 to 3 DR stops differences between 5Dsr and current models
How can you tell that ? The 5Dsr shot looks the same DR as the 5d3 shot, we dont know the lighting or the range of the scene?
Which lens did they use with the 5DSR? Looks like it has quite a bit of CA / Purple fringing and it was not used at its optimum aperture or maybe it was not the sharpest or focused perfectly on the part of the subject that they show on the crops. Something is off optically I think.
Also on the D810 and specially on the 645z crops I see some pretty significant jaggies in some parts indicating much more aggressive sharpening.
Also on the D810 and specially on the 645z crops I see some pretty significant jaggies in some parts indicating much more aggressive sharpening.
They typically use the Sigma 70mm macro for their studio samples on DSLRs (Canon/Nikon), which is an exceptionally sharp lens. I just checked the EXIF and it does say 70mm although the make of the lens is misreported as a Canon 50mm f/2.5 macro.
We are seeing a bunch of "sample" image comparisons without useful information like which lens and what f/stop.
And the lead image sample from Imaging Resource is brain damaged. Who would think it would be a good idea to judge optical characteristics by shooting through glass jars ?!
When the FF images are sampled to the same size (either the 36mp upscaled or 50mp downscaled) I don't see a huge difference either way. I see subtleties, sure, but the image quality is so close and the possible variables introduced by different lenses and different processing make me think there is really not a whole lot to get worked up about either way.
For me, the takeaway is that if you want a high resolution FF Canon body this is still looking like a good option, but if you are invested in a D810 Nikon system you shouldn't lose any sleep over not having a 50mp sensor.
Looking at these results I see zero reasons to move away from the A7R to the new Canon's. This is just one test so awaiting to see others. Maybe will rent the Canon when it's available to do my own tests.
Now that Pentax is smoking hot...just wish it would take my Canon mount lens.
Not much to go on with SOOC jpegs, need RAWs first, then for each body you 'must' at least identify what lens is used; MTF chart for the glass might help, and finally I'd want to see images shot with the absolute 'best' glass from each manufacture. Anything less is moot. And I not a big Sigma fan; if one was actually used.
Tests need to put the 'best Zeiss lens; like the 'Otus 85mm', etc., on the Nikon D810 and Canon 5DSR, absolutely nail focus; which the samples don't appear to accomplish, then pixel peep and make a few large prints, then decide.
Could you tell the difference between the A7R, D810, 5DSR, and 645Z in a blind test, provided files are all normalized to 50MP? (By blind, I mean, you don't know what camera the shots came from). From this test, the only one I could reliably tell apart would be the Pentax, I think.
I agree that the lens probably becomes the bigger factor. It's all sort of moot. If you're printing (which to me, is the ultimate expression) than the difference really begins to evaporate. Any of these cameras with good glass and photographic eye will be a winner.
Compounding the difficulties with such comparisons on the screen...
100% crops from sensors with different photo site density introduce differences from working with different sized areas of the images projected by lenses, and...
... resizing (and interpolating) to attempt to rectify this creates its own issues. (For example, is it better to "uprez" a 36MP image to the same pixel dimensions as the 50MP sample or downrez the 50MP sample to equal the dimensions of the 36MP image? (The best comparison would be prints from both at the same print dimensions, but sharing that output on a screen is problematic.)
As I said, "angels on the head of a pin..." ;-)
Fortunately, all of these options can produce excellent photographs.
I was kind of surprised how well it did at high ISO. I reckon it could take more NR to equal the amount of detail in the 5d3 file and produce a pretty similar file..
sb in ak wrote:
Could you tell the difference between the A7R, D810, 5DSR, and 645Z in a blind test, provided files are all normalized to 50MP?
I don't think that is the question you should ask.
None of this is about looking at 100% crops, it's about actual finished results, either in print or litho, i.e. what these cameras are actually used for and the answer is almost always no, except at very large sizes
dhphoto wrote: sb in ak wroteCould you tell the difference between the A7R, D810, 5DSR, and 645Z in a blind test, provided files are all normalized to 50MP?I don't think that is the question you should ask.
None of this is about looking at 100% crops, it's about actual finished results, either in print or litho, i.e. what these cameras are actually used for and the answer is almost always no, except at very large sizes
I see two reasons to ger these high Rez cameras, print large or crop heavy. If you don't do either, I don't see the need.