Not only do they have more shots with the Nikon D90 than the 50D, but the Nikon shots include pictures of human beings, a popular subject for many photographers
timbop wrote:
you guys are just nitpicking, IMHO. The 6400 fruit shot looks fine at 25%, and it is ISO 6400 of a subject in complete shadow on an APS-C sensor after all.
As for the most usable shot of the bunch, take a look at the ISO 1600 shot of the brass plaque. At 100% with normal USM it looks very good without any additional NR.
Geez
This sums it up. I think sample images do more harm than good when it comes to anticpating the performance from a production copy with a full review. (And the first dozen or so user reviews on your favorite forum).
Way too many factors to judge what the 50D can do from a batch of images done by one guy with unknown in-camera settings using exposure paramters that aren't typically used when showing off what a person and his camera can do.
The sample photos are bad, but that's not the cameras fault. Also who knows what kind of noise reduction settings they had on for these jpegs. RAW and some consistency will be far more telling but these look fine for now I guess. 6400 on a terrible photograph looks fine, so on a great one, I am sure it will be more than acceptable.
DThom wrote:
Not only do they have more shots with the Nikon D90 than the 50D, but the Nikon shots include pictures of human beings, a popular subject for many photographers
I'd much rather see subjects like landscapes and wildlfie that I might actually like to photograph. Besides, people tend to complain about too much sharp detail, like the pores in skin, fine wrinkles, small facial hairs, etc.
RalphJ wrote:
OK. I guess I'm in the minority, as I shoot handheld after dark all the time (40D, 35/1.4, ISO3200) and I also do a lot of theater shooting where a tripod wouldn't help because of the moving subjects on stage.
I understand if others don't want or need 6400 ISO, but it is very appealing to me....
Oh and pls dont forget my leopard in the "shadows" at dusk!
dcmiller wrote:
Look at the banding in the full rez version.
I like how Canon leaves in the luminosity noise. It often looks better in print than a smoothed out image.
Wow!
Yes, I see some banding in 100%, too. What is much less then my D30 would produce at (custom) ISO 1600. So it looks like better results to me even 2 stops higher. At double resolution. Hell! That camera is mine if 5DII is not in my price range.
okay, cats than. Brickwalls seem to have something like a "natural" banding I guess. On the other hand shooting people is shooting a "kind" of fruit, too. Isnīt it?
Here's another ISO 1600 sample from the good folks at DP Review. It came from their review of the EF 50mm f/1.4 lens.
Obviously, there's some softness, as well as some visible noise reduction artifacts. But it's a fully magnified view of 15 megapixels at ISO 1600. I think it looks great.
The ISO6400 sample is not relevant for low light photography. It would have been nice if the DPR samples were in low light conditions. Shoot at f9 @ 1/40s does not qualify as low light in my book to exhibit noise character of the sensor. It would be much more relevant if the light demanded a f2, 1/40s, ISO6400 shot.
