I shoot 5D and 5D2 alongside each other. I only shoot RAW. I have in camera noise reduction turned off. I am correct in asuming that no matter what NR settings I set on the camera it will not affect me if I'm processing RAW images?
If my asumption is correct then my current observation is that my 5D2 images are slightly noisier than on my 5D.
Am I just stating the obvious, considering the higher megapixels and pixel density of the 5D2?
If you're viewing the 5d2 files at the highest magnification then it's not a fair comparison. I'm sure this has been discussed at length but in short you're seeing the effects of a larger magnification.
No. It is not about magnification. I have a standard workflow for adjustments and sharpening and after 2 - 3 weeks of shooting I think what I'm seeing is more noise in images. I think, perhaps that it might be my sharpening algorithm though. Perhaps I need to use a different set of numbers. For example if I used standard 100,1,1 for sharpening 5D files then I thought that for 5D2 the number would have to be 200,1,1 but if it was below or equal 100 then the overall result should be at least the same as 5D.
I want to add that it is not a showstopper for me at all.
rbraun wrote:
No. It is not about magnification. I have a standard workflow for adjustments and sharpening and after 2 - 3 weeks of shooting I think what I'm seeing is more noise in images. I think, perhaps that it might be my sharpening algorithm though. Perhaps I need to use a different set of numbers. For example if I used standard 100,1,1 for sharpening 5D files then I thought that for 5D2 the number would have to be 200,1,1 but if it was below or equal 100 then the overall result should be at least the same as 5D.
I want to add that it is not a showstopper for me at all....Show more →
What program are you using for RAW conversion? The version of DPP that came with your 5D2 is widely thought to do a much better job of removing noise than the earliest Adobe products that supported that camera. The latest Adobe updates seem to have caught up with DPP.
Brainiac, I'm not talking about 100% crops. My shots are mostly black and white photographs, 1600 iso, low light, high contrast. I have a lot of black and dark areas in my photos. The shadow detail appears more grainy when I work on my session after I used both cameras.
rbraun wrote:
I shoot 5D and 5D2 alongside each other. I only shoot RAW. I have in camera noise reduction turned off. I am correct in asuming that no matter what NR settings I set on the camera it will not affect me if I'm processing RAW images?
If my asumption is correct then my current observation is that my 5D2 images are slightly noisier than on my 5D.
Am I just stating the obvious, considering the higher megapixels and pixel density of the 5D2?
well according to DxO the 5dmkii should be very similar at the pixel but something like a good 2/3 stop better at the image level for 18% noise
Are you saying that you did a print of the same size from each camera, of the same shot, taken at the same iso, and post processed each file appropriately, and the 5D2 came up noisier? If so, I think you did something wrong somewhere. Show us.
No brainiac. I would have to be a pixelpeeper and perhaps I would have to have a lot of free time on my hands to do exactly as you suggested.
I'm absolutely open to admit that I'm mistaken. It is just an observation to my untrained eye. Now, I usually have 85L on my 5D and a 24 on the 5D2 when I shoot. If I need a close up I shoot 85 and if I want to go wide I shoot the 24L or 50L. So I sometimes have shots from the same situation, under similar light but with same ISO and same f stop (2.0).
Brainiac, what would you recommend the sharpening settings to be when going from 5D1. The could be a simple mathematical formula no?
P.S. and I almost never print unless it is to hang on a wall.
The problem is that we are all in the habit of zooming in to 100%. The 5D2 will often look noisier at 100% than the 5D, when in fact a corresponding print is not. So the general impression of using a 5D2 or 1Ds3 is that it is noisier than whatever lower megapixel camera you are used to, but of course when it comes to doing a print, the printed pixels are much smaller, and so the printed effect of the noise is less.
My preferred way to process my 1Ds3 files, and I expect 5D2 files to be exactly the same, is to ouput from DPP with no sharpening and no noise reduction. Then I would noise reduce ONLY the chroma noise in, say, Neat Image, or just using my own Photoshop action, and then sharpen ONLY THE LUMA channel. The easiest way to do that is to duplicate the layer, sharpen to taste (e.g. 300, .3, 1, or 500, .2, 0) and then set the layer mode to luminance. That way colour detail (and chroma noise) isn't amplified, whereas luminance sharpness is.
And don't take him on with a "100% crop comparison between 2 different camera models" argument because he will chew you up and spit you out while writing a book-size post in the process
brainiac wrote:
My preferred way to process my 1Ds3 files, and I expect 5D2 files to be exactly the same, is to ouput from DPP with no sharpening and no noise reduction. Then I would noise reduce ONLY the chroma noise in, say, Neat Image, or just using my own Photoshop action, and then sharpen ONLY THE LUMA channel. The easiest way to do that is to duplicate the layer, sharpen to taste (e.g. 300, .3, 1, or 500, .2, 0) and then set the layer mode to luminance. That way colour detail (and chroma noise) isn't amplified, whereas luminance sharpness is.
Could you explain why you sharpen only the luma channel? I have a ton to learn about post processing and I'm still working out the kinks, but I have never heard of doing that. Alternatively, if you had an article or something about the process that explains the benefits/situations you would use that, that would be awesome.
brainiac wrote:
The problem is that we are all in the habit of zooming in to 100%. The 5D2 will often look noisier at 100% than the 5D, when in fact a corresponding print is not. So the general impression of using a 5D2 or 1Ds3 is that it is noisier than whatever lower megapixel camera you are used to, but of course when it comes to doing a print, the printed pixels are much smaller, and so the printed effect of the noise is less.
...
I agree with your approach, and used a similar approach when comparing my 5D to my 1D3 (which I no longer have) and my 40D. You really have to equalize the final output or you're comparing apples and oranges. For the record, the 1D3 was the winner of my comparison, followed by the 5D and then the 40D.
Whilst I agree that comparing 100% crops is not a great idea for noise/sharpness comparisons there is one counter-arguement - Software deals with pixels so having a sharper, less noisy per pipxel image can help benefit post processing. Obviously more data is going to yield better results over less but it just means that sharpening (particularly) doesnt work optimally with the higher density sensors unless you have really sharp glass!
Luma sharpening using luminosity blend mode is a faster and non-destructive alternative to switching to LAB mode and sharpening the L channel and then switching back to RGB mode. Also don't use USM IMO, unless you are using edge sharpening only, Use something like high pass sharpening and use blend-if sliders or masks to avoid sharpening shadow areas etc.
