This content has been marked as final. Show 8 replies
Sharpening occurs post noise reduction and the Mask has nothing to do with the luminance noise reduction setting which is using it's own algorithm for determinging high fequency nose and trying to mitigate it.
Thanks for the reply - let's see if I've got it. The sharpening mask can show edges of noise artifacts, but since sharpening is happening post noise reduction it doesn't matter as much, although it might result in emphasis of any residual (post-reduction) noise, if any meaningful amount remains.
Uh, yeah...more or less.
>In processing an image with a noisy blue sky, the sharpening mask appears to find the edges of the noise, ignoring the results of the current setting of luminance smoothing, forcing the selection of a higher value for the mask slider than otherwise would be used.
In your situation, it might be best to remove noise reduction and sharpening out of ACR into Photoshop. As Jeff Schewe explained, the ACR sharpening mask is used only for sharpening and does not come into play for NR. Just as an edge mask is used in sharpening to confine the sharpening to the edges, a surface mask can be used in noise reduction to protect the edges. The surface mask is essentially the inverse of the edge mask. These matters are discussed in Bruce Fraser's sharpening book, which was his last work before his untimely death.
Rather than using a surface mask, you could merely use the sky to make a mask. You would invert the mask for NR so that the effect is confined to the sky and sharpen with the mask in place so that the sky would not be affected.
Thanks to both of you for the replies.
My issue with the sharpening mask is that it does not take into account the results of ACR's NR and therefore the sharpening mask can show the edges of noise artifacts. The result is that one has to increase the mask value more than otherwise optimal to remove these edges from the mask.
I have been doing NR using a 3rd party product with a blurred edge mask in Photoshop and then sharpening actions based on those in Bruce's terrific book. With the new capabilities of ACR 4.1 I've been experimenting as to how much I can move to ACR. In general, it has worked out well - for many images the new ACR does a great job and is a noticeable improvement over earlier releases. However, I have been coming to the reluctant conclusion, as you are suggesting, that there are some images that I will simply have to process as before, in part due to the limitations of my camera.
Now, if ACR... a) Allowed the creation of a NR mask, and b) took the results of NR into account while building the sharpening mask.....
But for now I expect the approach will depend on the image - what else is new?
Camera Raw 4.1 absolutely DOES take into account the noise reduction setting while computing the sharpening edge mask.
Thomas: thanks very much for clearing this up. One more question - Does the real time display of the sharpening edge mask (via the Option/Alt key) also take into account the NR setting?
Further to my question. When I adjust the luminance smoothing slider dramatically (from 0 to 100 and vice versa) on a noisy image, while I can see a big difference in the image preview, the preview of the mask (via the Option/Alt key) still appears to show the edges of the noise without change.