alundeb:
I have circled in some areas where the extra resolution is visible.
You are right - the difference can be seen but it very minimal and I did not even notice it.
The question then remains (you did address it, to some extend in the previous post):
If 5D mkII is a 21 Mpixel camera, is EOS 7D a 18 Mpixel camera?
By the way what is the difference between "old" and "new" standards?
Zenon Char wrote:
Sounds like you had a great trip. Hope you liked our province, eh!.
I just showed my wife the shot of that puppy and she loved it. What a cutie. We're from Manitoba but we have never been up there yet. Seems like the closer you are to a destination the longer it takes you to get there. We do have a cottage in the Canadian Boreal forest and have witnessed spectacular northern light shows, especially in the winter. Driving up we have see Moose, Woodland Caribou, Lynx, Wolf, a lot of Black Bears and an assortment of common critters like fox and coyote.
Yea, the puppy was too cute ... and believe it or not, the photo isn't staged. I.e. just happened to put my backpack down, was taking pictures of another puppy, turned around, and he'd "fiddling" with my backpack!
Canada is spectacular ... as you may have seen on my polar bear trip report, I was also fortunate enough to see the Northern lights.
I just bought my 7D today. I wanted a quick test to verify my hypothesis that 7D's ISO 3200 is sufficiently better than 40D's ISO3200 to justify the purchase. I shot the same scene with 40D in ISO3200, and 7D in ISOs 3200/6400/12800. EF 70-200 f4 L IS handheld, fixed ISO, fixed aperture, auto otherwise, RAW. LR3 beta, enough color noise slider use to get rid of color noise. So all in all a very unscientific test. Short summary: 7D ISO 6400 matches 40D ISO 3200 when 40D is upscaled to 7D size.
Two days ago (Friday) I attended the Israeli 7D launch. I only got to play with it a little but what else can I say but "I'm flabbergasted"? It seems more responsive than my 40D, the viewfinder is visibly larger and when I got home and looked at the pictures I get the feeling that ISO 3200 is like my 40D's ISO 800. Yippee....
Oh, the 17/4 TS and 100/2.8 IS were also there but funnily, the best shot of the day was taken with my favorite current kit: 40D and 120/4 macro TS at full tilt. The old dog can still do well with its old tricks.....
WARNING: If you have no interest in Potential Pixel-Peeping ("PPP"), please skip this post without succumbing to the temptation to post a condescending slap at those who do. I will post some "real pictures" from when I went to "just go out and take some photos!" later.
As advertised, I set up a somewhat-controlled test shot for my 50D and three 7Ds (only keeping one; the first is already on the way back). (The third 7D, by the way, when examining an OOF color chart in DPP like the first and second, showed artifacts somewhere in-between but closer to the second, better body).
I set up the test as described below. The ISO100 shots were illuminated at full manual power (1/1) from a 580exII bounced from an off camera position. The ISO1600 shots were at 1/16. You will notice that the 1/16 shots are a little brighter than the 1/1 shots. I haven't checked to see if the 1/1 was dim, or the 1/16 was bright. There are slight differences in framing, only because my tripod is not spectacular; it was not moved intentionally.
I have not looked at these very closely - only casually in DPP 3.7.2. I did find it interesting how much better the red(-ish) spools of thread looked from the 7Ds compared to the 50D. (Unfortunately, I didn't have blue thread spools in the shot.) Also, unless there's something hinky going on with "zero" noise reduction in DPP, the 7Ds outclass the 50D considerably at ISO1600.
I will be interested to hear any observations y'all expert pixel peepers have to say, and feel free to post any crops, etc, of the files. Here are the raw files:
paulfeng wrote:
I should be receiving my third 7D body tomorrow, as the first and second both have non-IQ related issues. I'm not sure I will have time this weekend, but I may be able to set up a somewhat controlled test shot of 50D vs 7D1, 7D2, and 7D3. My plan is to tripod mount my 100-400 @ 100mm f/8 (we aren't really worried about diffraction, are we?), focus multiple times using live mode AF in live-view (& pick the best) and light the scene with bounced flash, manual power.
I probably won't put a lot of effort into the subject scene; I may put some wine bottles and skeins of yarn on the fireplace mantle. Any suggestions here (easy things I might have around the house)?
mfurman wrote:
Impossible. 5D mkII is barely 1.5 stop better than 40D. You are overly enthusiastic, Yakim.
I don't think so. If you compare captures at a common resolution (instead of at 100%), the 7D handily beats the 40D and all previous APS-C bodies in terms of noise.
Alek Komarnits wrote:
Yea, the puppy was too cute ... and believe it or not, the photo isn't staged. I.e. just happened to put my backpack down, was taking pictures of another puppy, turned around, and he'd "fiddling" with my backpack!
Canada is spectacular ... as you may have seen on my polar bear trip report, I was also fortunate enough to see the Northern lights.
it is a good photo though, really down low perspective at puppy level and then the building way off in the distance behind
garyvot wrote:
I don't think so. If you compare captures at a common resolution (instead of at 100%), the 7D handily beats the 40D and all previous APS-C bodies in terms of noise.
yeah it is better up there but then again it seems pretty clear that the 5D2 also handily beats the 7D and since the 5D2 was measured, at the RAW level, to have only around 1.6 stops better SNR at ISO3200 (normalized, not pixel level) how can the 7D have 1.5 stops better SNR than the 40D?
1.6 stops better than a 40D would mean astonishingly better photon capture which doesn't seem possible, i'm not sure everyone realizes how radical of an improvement 1.6 stops would entail
the lesser banding in the high iso shots might help make it seem better than the better it is
i'm sure it is better, but no way do i believe 1.5 stops better
i also noticed that some places like IR shot 1/3 stop longer exposure to get what seemed to be at best the same light as with a D300s, but that was judging with acr and who knows if that really knows the cutoffs properly sometimes it has not in the past
skibum, I'm hoping you are going to peep those files and tell me that the third body is not appreciably worse than the second. It appears so far that the third body is AF-ing fine, has a clean sensor, and has not presented any odd glitches (fingers-crossed), so it may actually be a keeper.
I don't think so. If you compare captures at a common resolution (instead of at 100%), the 7D handily beats the 40D and all previous APS-C bodies in terms of noise.
It cannot be 1.5 stop. I am talking about "print" measure of noise.
skibum5 wrote:
it is a good photo though, really down low perspective at puppy level and then the building way off in the distance behind
Roger that ... always fun to get down low ... but pretty hard to do with Polar Bears since due to safety reasons, you have to be above their reach! ;-)
alundeb wrote:
I have found the 7D to give the same "pixel sharpness" as my 500D, which is known to resolve essentially equally to the 50D. The 18MP vs 15 MP is not a "numbers game". The improvement is real.
At f8, the advantage of the 7D is muted because of diffraction. When will it ever be accepted that we cannot test the resolution of crop cameras at f8?
These are processed in C1v5, default sharpening. Note that I have not applied any extra sharpening of the 7D image to match the 500D image. Manual focus, unchanged between exposures. The test was repeated 3 times to ensure that the focus was always dead on.
I can't use ACR since I only have Elements 4.
Both 100% crops and equal magnification images are presented, for your convenience. 500D images resized with bicubic sharper. ...Show more →
The effect of diffraction at f/8 appears quite small here, but by f/16 is is quite pronounced. Fortunately, I usually find that f/8 is enough DOF for most landscapes on 1.6X crop.
so 
