MartinB77: That moment......when you realize that your own pics are nothing in contrast to that one ;-)
Kudos to Rishi - the picture is just stunning.
But hell, Canon should really have spent more research effort in image quality/dynamic range than in pixel count. Looking at the Sony 7R tulip shot from Rishi makes me feel sad for Canon. I wouldn't have expected THAT difference.
Rishi has informed me that the Sony shot was made with a graduated filter.
ThrillaMozilla: The Web site blows the picture up to about 60" wide on my monitor. Why is this a surprise that you can see noise? If you enlarge it instead to 16"x24" size, you have to look fairly carefully to see any noise at all.
Now, I suggest that you all get ahold of that Nikon image and blow it up to 60" wide, and see if you can find any image defects.
Folks, I was mistakenly referring to a Nikon image, which turns out to be a Sony image that Rishi Sanyal posted. He rightly points out that he did not intend to invite comparison, and I regret implying otherwise.
I think the real problem, however, is that two clicks on the tulip image gets you a nearly full-size enlargement.
Rishi Sanyal,If you click through, the image you posted in the article is displayed at 4907x7360, and it appears right on the Web site at about 60" wide, depending on your monitor size. And THAT is a big problem, in my opinion, because the image is terrible and it invites ridicule.
Then you posted a comment with another comparable tulip shot, shot with a Sony (I mistakenly thought it was a Nikon). That one APPEARS to be relatively noise free because it is posted at a maximum resolution of only 1068x1600. The Canon shot also appears to be noise-free if you view it at that resolution.
Horshack,Your Nikon shot also looks pretty bad if blown up to 60". Although the Nikon unquestionably has a lower-noise sensor, both pictures look bad if they are displayed that way.
Yes, that is very good, Horshack, but if you don't mind, I think we should not complicate the issue and stick to a comparison of tulips. And I know Nikon is supposed to have an edge of about a stop or two, but the issue here is that the Canon image is being displayed at about 60" width, which will make any picture look bad.
ThrillaMozilla: So the Canon image is blown up about 60" wide, while the Nikon image is displayed 11" wide, thereby inviting derision on Canon. That's not a nice thing to do. The Web site should be fixed.
No, I don't want a 0.4MP image. Where did you put the full-res. images for both pictures? Links, please.
This conversation is getting split between two threads. Let's maybe keep to one or the other.
Please follow the links and you should see the magnifying glass. Do I have to post screen shots?
Rishi Sanyal, you didn't notice that people are examining the Canon image at huge size? Just click through and then click on the magnifying glass. That's where all the ridicule is coming from.
The Name is Bond: For comparison purposes, just as DxO does it, the DR test needs to be normalised by resizing down to (I think) 8MP.
Yes, that's absolutely right. You have to compare images at the same width and height or you're going to fool yourself. Noise depends on the size of the sample (i.e., the pixel size), and that's why DPR also (optionally) compares normalized noise.
If you think so, then examine the Nikon image under exactly the same conditions and THEN tell me what you think. If you examine the image more carefully and click on the magnifying glass, it blows it up to about 60". You can't do that with the Nikon image. I don't know where you got the Jpeg, but I think most people are looking at the web site.
And this is ISO 100 all right, but it's ISO 100 with dim light. Dim light creates noisy images. By the way, you have to compare images at the same size, and not the same pixel size.
So the Canon image is blown up about 60" wide, while the Nikon image is displayed 11" wide, thereby inviting derision on Canon. That's not a nice thing to do. The Web site should be fixed.
The Web site blows the picture up to about 60" wide on my monitor. Why is this a surprise that you can see noise? If you enlarge it instead to 16"x24" size, you have to look fairly carefully to see any noise at all.