all are 100% crops it goes 50D/40D/20D and within each camera it goes JPG/DPP/ACR (I did the sharpening for RAWs in CS3 but the de-bayering in DPP and ACR) oh and lens is 135L at f/2.2 (1/3 stop down from wide open):
Hehe, not surprising the 50D looks good, especially since you used Canon's sharpest lens, the 135L, and wide open, to avoid any diffraction.
I still think the 20D looks amazing if you think how old the camera is already! Good times... good times...
IMO 50d looks like the worst, 20d looks like the best. The problem is, your images aren't all the same size and have resizing artifacts.
The 50d shot is slightly overexposed compared to the other two. It also lacks contrast.
Very strange test and since they're not aligned horizontally its impossible to make any sort of eyeball comparison. I don't think this says anything about the relative merits of the cameras and certainly doesn't make a good case for the 50d, which I know is a good camera.
Very strange test and since they're not aligned horizontally its impossible to make any sort of eyeball comparison. I don't think this says anything about the relative merits of the cameras and certainly doesn't make a good case for the 50d, which I know is a good camera.
Well, it gave him something to do. Better than smokin', drinkin' 'n cursing! I give a similar test to new and prospective GFs and, fair or not, it gives me a excuse to toss 'em all out the door. Why not for cameras? Life just ain't fair...
skibum5 wrote:
all are 100% crops it goes 50D/40D/20D and within each camera it goes JPG/DPP/ACR (I did the sharpening for RAWs in CS3 but the de-bayering in DPP and ACR) oh and lens is 135L at f/2.2 (1/3 stop down from wide open):
makron wrote:
When one downrez the 50d image to that of the 20D, the 50D image is sharper. This maybe due to the difference in AA filter strength.
The strength of an AA filter is a little bit of a misnomer. Its not like they come in various "strengths" and the manufacturer "picks" one. The design of the AA filter is based on the size of the sensors photosites, and is engineered to pretty precise standards. This is in fact why they do what they do.
Here's a pretty indepth discussion about them if you are interested.
From your test, image gain from 20D to 50D is a noticable upgrade.
When one downrez the 50d image to that of the 20D, the 50D image is sharper. This maybe due to the difference in AA filter strength.
20D to 40D is not much of an upgrade.
40d to 50d is a noticable upgrade.
40D is noticeable jump from 20D, but naturally it's only a 2MP jump so it's not going to be as much as the 40D -> 50D jump. Not a very good test subject to see detail differences.
we need another test using only DPP-135L in 20D\30D\40D and 50D under bright sunlight, indoor test--availbale light and one with flash and then some shadow detail test. ---- anyone?
orangefirefish wrote:
Hehe, not surprising the 50D looks good, especially since you used Canon's sharpest lens, the 135L, and wide open, to avoid any diffraction.
I still think the 20D looks amazing if you think how old the camera is already! Good times... good times...
I agree that the 20D is not bad at all.
Then I have a question: You say that the 135/2.0 is "Canon's sharpest lens," but on what do you base that statement?
I have both the EF 135/2.0 and the EF 200/2.8 II. The latter is a tiny bit sharper at 2.8 than the 135 is at 2.8. Also, the 200/2.8 II tolerates a teleconverter better and loses less quality than the 135.
I have noticed that many seem to believe that the 135 is the better one, but typically those who believe so have not tried both. I feel that the differences are so small as to be of rather little significance, but if truth is to be spoken my 200 is sharper than my 135.
skibum5 wrote:
not surprisingly 40D has a bit more detail than 20D and 50D CLEARLY has more detail than both, although this goes against DPR detail test.
That would seem to be a quite precise judgement about the relative capabilities of detail rendition between these cameras.
Taylor Barrett wrote:
IMO 50d looks like the worst, 20d looks like the best. The problem is, your images aren't all the same size and have resizing artifacts.
The 50d shot is slightly overexposed compared to the other two. It also lacks contrast.
Very strange test and since they're not aligned horizontally its impossible to make any sort of eyeball comparison. I don't think this says anything about the relative merits of the cameras and certainly doesn't make a good case for the 50d, which I know is a good camera.
umm, but you can clearly read text on the 50D image that you can't on the 20D image.
there can't be any resizing artifacts since these are 100% crops.
i was actualy testing AF, but i took so many shots i found perfectly focused samples, anyway, considering hand-held, near wide-open, flashed, etc. and I can still read stuff on the 50D that I can't on the others I think it's safe to say you can get more detail out of the 50D.
i do agree the posting makes it hard to compare, you really need to save the images to your computer and then put them side by side on your desktop, i didn't have time to cut and paste and build-up all sort of images last night, sorry.
no comparison here, but thought I'd toss up a couple more shots when I was messing around with focusing the other night on the 50D, hand-held, not at absolute peak IQ f-stops, blah-blah-blah, but still looks pretty good per pixel:
