Wikipedia:Featured picture candidates/Anna Ternheim

Anna Ternheim
Voting period ends on 22 Jun 2013 at 01:21:59 (UTC)
 * Reason:Fairly high-resolution portrait photograph of the Swedish singer and songwriter with one of her main instruments, the piano (the other being the guitar). A major note on this photograph: I consider the original picture to be damaged—no offense to photographer Benoît Derrier—perhaps due to the unfortunate combination of high ISO and lighting conditions. This picture required extensive and multiple techniques of noise reduction, selectively used in various geometries of the image, including a final scale-down to seal the deal. I found 1625 pixels tall to be about the largest I can keep the image, and I didn't want to bring the photograph down to 1500 if I could avoid it.
 * Articles in which this image appears:Anna Ternheim
 * FP category for this image:Featured pictures/People/Entertainment
 * Creator:Benoît Derrier, photograph; Keraunoscopia, restorative engineering


 * Support as nominator --– Kerαu noςco pia ◁ gala xies 01:22, 13 June 2013 (UTC)
 * Oppose. I appreciate the efforts you went to to try to improve on the original photo but I think the noise reduction is excessive. I'm not sure that any amount of work could really get it up to FP level though, honestly. I'm very surprised that a D300 at ISO 1400 (?) would result in that much noise. &#208;iliff    &#171;&#187;  (Talk)  12:24, 13 June 2013 (UTC)
 * It could've been a poor lens as well? I watched the music videos of her performing (link in image description) and the lighting didn't appear overly dark, so I don't understand all the original noise either. As for the noise reduction, don't FPs prefer almost no noise at all? I certainly had to be excessive in areas to get to this level—the original is practically like looking at a screen print. Ah well. – Kerαu noςco pia ◁ gala xies 15:38, 13 June 2013 (UTC)
 * A good compromise is to noise reduce problematic areas separately. Adam Cuerden (talk) 01:31, 14 June 2013 (UTC)
 * That's what I meant when I wrote "selectively used in various geometries of the image", probably not Featured Article-worthy explanation :) Basically the image can be broken into several shapes between the blacks, whites, jacket, face, hair. Each one got a different NR treatment, either requiring less or more, and sometimes with different techniques (if one way didn't work, I'd try another)... along with spot removal for blatant noise and blur tool for others. The most difficult for some reason was around the neck and especially (not surprisingly) the eyes, which would lose what very little sharpness they had. The irises have different NR than the whites, etc. Maybe I didn't go about it the right way. Without reducing the image's size, though, it was still really noisy. Maybe a 1500 pixel tall version would look a bit better, but probably still wouldn't reach FP. – Kerαu noςco pia ◁ gala xies 01:37, 14 June 2013 (UTC)


 * Oppose as above. I was gearing up to support when I saw the thumbnail, then I looked at it full size... J Milburn (talk) 16:26, 13 June 2013 (UTC)

--Armbrust The Homunculus 01:58, 22 June 2013 (UTC)