RDKirk wrote:
And we keep concluding that unless there is some extraordinarily wildly variant factor, the larger format wins in image quality....as has been true for more than a century of photography.
Yup. That is my experience and that was the (perhaps not plainly obvious) point of my post.
"I'll expand one point regarding the "larger image circle" issue. This supposed "corner degradation" on full frame is not quite the issue that some imagine it to be."
Well, in principle a good big 'un will always beat a good little 'un, I accept that. However, in practice can you recommend me a Canon full frame wide zoom lens that is sharp in the corners by f5.6 on a 5D2? (for my photography f11 is just too slow hence my emphasis on relatively wide aperture performance). If you can I will cheerfully trade in my Toke 12-24 lens in for it and pension my 40D off.
basically, DOF stuff has been covered by everyone else. I'd like to add one more feature about full frame: the dynamic range would be greater for a full frame sensor such as 5D Mk I, compared to 1.6x crop cam xxD.
This is a great example showing that the depth of field is exactly the same in these two shots. even though the format and the focal length has changed.
Yes the background is more out of focus in the shot with the longer focal length, but the part actually in focus (i.e. the depth of field) is the same.
well for what it's worth I personally didn't find a large difference in high ISO improvement when shooting a 20D side by side with a 5D at a concert. Maybe it's my post processing or something but for what I was shooting the 5D didn't offer me an advantage I felt was worth the investment of going FF.
20D ISO 3200 http://i91.photobucket.com/albums/k320/livelifelivepics/_MG_9162-2.jpg
I know not exactly the same light and subject but I shot with both cameras and when I got home I couldn't tell the difference from one file to the next unless it was DOF related. That was my experience for what it's worth.
Great examples, thanks. Do you do a different amount of post processing for the 20D vs 5D shots? And was there any noticeable difference when you were taking the photos (e.g. viewfinder advantages)?
Mike V wrote:
This is a great example showing that the depth of field is exactly the same in these two shots. even though the format and the focal length has changed.
Yes the background is more out of focus in the shot with the longer focal length, but the part actually in focus (i.e. the depth of field) is the same.
No it's not. That is the case for the same framing on the same FORMAT with different focal length lenses (within reason). Since the camera to subject distance and aperture are the same here, but the longer focal length on the larger format makes for the same framing, there is less DOF in the full frame shot. This has nothing to do with a narrow AOV (since they roughly have the same AOV on their respective formats) creating a more diffuse background, which is the case when you back up and frame the same with a longer lens. In that case, the DOF is roughly the same, but the longer lens creates a more diffuse background due to the narrower AOV. Not so here. DOF really is shallower on the full frame shot.
NumberFive wrote:
Great examples, thanks. Do you do a different amount of post processing for the 20D vs 5D shots? And was there any noticeable difference when you were taking the photos (e.g. viewfinder advantages)?
Same PP on each photo. I used the default noise reduction in LR and adjusted the WB to suit each photo. There is a difference in viewfinders that's for sure. The 5D's is nice and big. Anything but the center AF point was almost useless in low light which kinda makes me scratch my head as to why they wouldn't upgrade the outer AF points on the 5DmkII. The center point works really well though. It has invisible AF points around the center which makes it more sensitive.
Despite there not being a huge difference in noise characteristics between the two in properly exposed areas of an image there is a difference in the shadows and also when you try to recover exposure in PP.
I don't know how noticeable the differences would be in print because I've never printed them out.... also one of the reasons I didn't bother with the 5D, I'm not a big printer.
Jman13 wrote:
No it's not. That is the case for the same framing on the same FORMAT with different focal length lenses (within reason). Since the camera to subject distance and aperture are the same here, but the longer focal length on the larger format makes for the same framing, there is less DOF in the full frame shot. This has nothing to do with a narrow AOV (since they roughly have the same AOV on their respective formats) creating a more diffuse background, which is the case when you back up and frame the same with a longer lens. In that case, the DOF is roughly the same, but the longer lens creates a more diffuse background due to the narrower AOV. Not so here. DOF really is shallower on the full frame shot....Show more →
+1
The part of the railing thats in focus is in focus yes, well you'd hope that wouldn't you but the amount of acceptable focus drops off faster on the FF and is not so far in front either (as it should be) Had the crop shot been shot at 1.6 stops faster the result would have been much closer. (as far as the 9mm equive difference would have allowed)
h_rearden wrote:
^ Exactly. The background is more blurred on the FF shot precisely because the DoF is shallower. The two are inextricably, directly related.
The scenario Mike V described is a logical and optical impossibility.
The DoF is narrower in the FF shot because the focal length is much longer.105mm is quite a bit longer than 60mm.
Nick, coming to a 5D from a 40D and doing a lot of low light work: Your experience that the 5D and a crop sensor show roughly the same image noise at ISO3200 is my experience as well. But at ISO 800 the 5D images are very noticeably better than any crop sensor I've ever used.
Zara wrote:
Nick, coming to a 5D from a 40D and doing a lot of low light work: Your experience that the 5D and a crop sensor show roughly the same image noise at ISO3200 is my experience as well. But at ISO 800 the 5D images are very noticeably better than any crop sensor I've ever used.
Cool, I can't really comment on ISO 800 because to be honest I didn't do much shooting with the 5D at that ISO but if you got some snap shots kicking around comparing the two I'd wouldn't mind seeing them.
Just for fun here are a few 100 % crops,
I won't label them but the exif is intact, see if you can properly label them.
All are ISO 3200
One is a 5d shot pushed one stop, one is a 5D shot, two 50D shots, and a 20D shot.
Having a hard time digging out shots taken in similar lighting with similar exposures of similar subject matters which would allow for a decent comparison. But here is a fairly typical example from the 5D at ISO800. Developed in Lightroom with standard settings (NR to 0)
Yeah, equal comparisons are way better. Posting single pictures doesn't help this discussion IMO - it's just showing that one sensor is capable of something. It doesn't show the difference in IQ between crop/FF.
I like Nick_b's idea of trying to guess which camera took what - if you can't tell whether a shot was taken with a FF or crop, then doesn't that show there is no difference IQ wise?
NumberFive wrote:
I like Nick_b's idea of trying to guess which camera took what - if you can't tell whether a shot was taken with a FF or crop, then doesn't that show there is no difference IQ wise?
For judging the overall IQ and effect, I'd agree. Michael49 did a great job throughout the thread in showing the DoF difference/overall appearance of a FF image vs crop. When it comes to noise, anything other than a controlled test will not tell you that much except, as you said, what kind of potential a camera may have at the pixel level.
Exposure, color temperature, subject contrast etc have a very pronounced impact on noise which makes ad hoc comparisons very difficult. Even in Nick_b's examples above, if you look closely you can tell that lighting and exposure were quite different for the individual images, which I am fairly certain is accounting for a lot of the perceived IQ differences.
NumberFive wrote:
Yeah, equal comparisons are way better. Posting single pictures doesn't help this discussion IMO - it's just showing that one sensor is capable of something. It doesn't show the difference in IQ between crop/FF.
I like Nick_b's idea of trying to guess which camera took what - if you can't tell whether a shot was taken with a FF or crop, then doesn't that show there is no difference IQ wise?
Yes and no, I think. It's silly to argue that the 5D doesn't have better noise control then a 20D or any crop camera because there are a ton of scientific tests on line that anyone can look at and see that there is a difference BUT and that's a big 'but' how will those scientific test results impact YOUR images? We don't shoot under scientific conditions, that's why trying it out for yourself is so important I think. To say you'll get "one stop better" or "ISO ____ on a 5D is like ISO ____ on a 20D" isn't accurate in my experience because so much depends on the quality of light not just the amount. For what I shoot the differences between bodies isn't large given properly exposed files.
