If you honestly think that you should just view them at the 100% mag. level, then there\'s no changing your mind.
You can change my mind provided your evidence is clear and scientific (can be reproduced by others). I am open minded individual and I can admit that I am not infallible. I would hope that you are the same way.
Several Professional Camera review sites such as dpreview.com and popphoto.com compare camera noise by looking at the pixel level. So far it sounds like you are ignoring the evidence provided by professional reviewers who have been in the industry for a while. They have been measuring noise for a long time and they have consistently been measuring it on the pixel level. Check out the reviews, you can clearly see the noise and how each camera is tested on the noise level.
Camera noise isn\'t some kind of magical phenomenon. Noise is an inherent property of digital imaging sensors. Noise happens when the photo sensor is unable to resolve an accurate reading in its color channel. This can be due to many reasons. Here are two reasons: 1. ISO gain - The higher the gain in a color channel the more likely noise will show up in that channel. 2. Exposure - If the photo sensor is unable to accurately sample a color, it will likely give an erroneous reading. This is why correct exposure or even overexposing your photo slightly is likely to produce less noise. More photo sensors are able to get a more accurate sample thus less noise is produced. Again, all of this is happening on the pixel level. Noise happens at the pixel level.
What do you do when you are adjusting the noise reduction slider in ACR, DPP, Lightroom or Noise Ninja? Its no coincidence that Lightroom 3 and Noise Ninja have a 100% preview window which allows you to see the noise and acurate get a view of how much noise is being removed as you move the slider.
RobertLynn wrote:
The FACT of the matter is, the 7D downrezzes to do any native resolution the 40D does.
Another FACT is that to do an even comparison, you MUST uprez the 40D\'s output to the 7D. If not, you\'re NOT comparing apples to apples. More like an apple to an apricot.
I have to disagree with you on this. Yes, you can resize a 7d file to 40d size. In doing so you are changing things. You are making the noise smaller and less noticeable. This is not an accurate method of measuring camera noise. Noise happens at the pixel level. Pixels must remain the same size (meaning no resampling) in order to compare apples to apples, pixels to pixels, noise to noise.
RobertLynn wrote:
When you can show me an image taken with identical settings from BOTH cameras, neither manipulated aside from resizing, in both PRINT and in online display, I will simply refuse to believe that in output the 7D has more noise. The 7D maybe at the PIXEL level does, but if you uprez the 40D file to the 7D file, the 40D file loses a lot of detail, and in fact becomes less usable than the 7D file, even a full stop of ISO above it.
An online comparison can be done and has been done. (I have already provided linked evidence) A print comparison is a whole different kind of test and has nothing to do with this debate at all. Nothing was said about prints. The original question was directed towards comparing noise levels between the two cameras. This has to be done on the pixel level or otherwise you are skewing scientific results.
Dec 28, 2009 at 04:19 PM
Previous versions of OO7MIKE's message #7937086 « 7d low iso noise vs 40d »
