Commons:Featured picture candidates/File:Molybdenum crystaline fragment and 1cm3 cube.jpg
Jump to navigation
Jump to search
Voting period is over. Please don't add any new votes.Voting period ends on 31 May 2010 at 20:13:39 (UTC)
Visit the nomination page to add or modify image notes.
Info created, uploaded, nominated by -- Alchemist-hp (talk) 20:13, 22 May 2010 (UTC)
Support -- Alchemist-hp (talk) 20:13, 22 May 2010 (UTC)
Oppose -- Whites are blown, greys are posterized. -- Alvesgaspar (talk) 11:28, 23 May 2010 (UTC)
- They are indeed and I still can't understand why... Alvesgaspar (talk) 22:36, 24 May 2010 (UTC)
Question Are we looking at the same image? I see no more "posterization" (by which I assume you mean color banding) than is pretty much inevitable in an 8-bits-per-channel image, and only a few direct reflections of the lights (about 0.1% of all pixels) exceed the dynamic range. —Ilmari Karonen (talk) 22:28, 23 May 2010 (UTC)
- Yes, that is what I mean. In this case I very much doubt it is inevitable as the tone difference between adjacent bands is quite large. I wonder if the original picture shows the same effect. -- Alvesgaspar (talk) 08:39, 24 May 2010 (UTC)
- @Alvesgaspar: "greys are posterized": yes, it is a conversion effect: from 16bit TIFF (my RAW format) to 8bit JPG. Only the background is additional "Gaussian blur" corrected. The background looks allways stained if I use focus stacking. I hope the info is helpfull for you. --Alchemist-hp (talk) 11:47, 24 May 2010 (UTC)
- Sorry, but I still don't understand. The picture has only 5 tones of grey which have the approximate values: 50, 100, 150, 200 and 250. How can that be that a 8 bit image is not capable of a much better tone resolution? -- Alvesgaspar (talk) 15:35, 24 May 2010 (UTC)
- I think I see what you mean now, but you're surely exaggerating the situation. The histogram of the crystal itself is indeed multipeaked, with 5 major peaks at approximately 20, 95, 152, 203 and 248 and smaller ones around 62, 111 and 238 (and at 255 due to clipping). However, many of these peaks are quite broad and have substantial overlap; I assume they simply correspond to different parts of the surrounding scenery reflected by the crystal. (Incidentally, the histogram of the cube is even more clearly peaked, with four peaks around 26, 63, 96 and 126; the lowest mainly corresponds to the top face, the second to the darked part of the front face, the third combines the lighter part of the front face and the darker part of the side face, and the last comes from the lighter part of the side face.) There's nothing artificial about this, it's just a natural consequence of specular reflection combined with a setting with sharp color differences. —Ilmari Karonen (talk) 16:05, 24 May 2010 (UTC)
- @Alvesgaspar: if you are interested then I can linked you the 11 each original images (raw:Canon CR2 = 250MB, or JPG =20MB or 16 bit TIF = 1,5GB.) for the focus stacking job and for your comparison. Please mail me simple. This sample molybdenum looks like in real-life! You see simple a reality sample. --Alchemist-hp (talk) 19:46, 24 May 2010 (UTC)
- But I believe in you, just can't understand the reason. Alvesgaspar (talk) 22:36, 24 May 2010 (UTC)
- @Alvesgaspar: if you are interested then I can linked you the 11 each original images (raw:Canon CR2 = 250MB, or JPG =20MB or 16 bit TIF = 1,5GB.) for the focus stacking job and for your comparison. Please mail me simple. This sample molybdenum looks like in real-life! You see simple a reality sample. --Alchemist-hp (talk) 19:46, 24 May 2010 (UTC)
- I think I see what you mean now, but you're surely exaggerating the situation. The histogram of the crystal itself is indeed multipeaked, with 5 major peaks at approximately 20, 95, 152, 203 and 248 and smaller ones around 62, 111 and 238 (and at 255 due to clipping). However, many of these peaks are quite broad and have substantial overlap; I assume they simply correspond to different parts of the surrounding scenery reflected by the crystal. (Incidentally, the histogram of the cube is even more clearly peaked, with four peaks around 26, 63, 96 and 126; the lowest mainly corresponds to the top face, the second to the darked part of the front face, the third combines the lighter part of the front face and the darker part of the side face, and the last comes from the lighter part of the side face.) There's nothing artificial about this, it's just a natural consequence of specular reflection combined with a setting with sharp color differences. —Ilmari Karonen (talk) 16:05, 24 May 2010 (UTC)
- Sorry, but I still don't understand. The picture has only 5 tones of grey which have the approximate values: 50, 100, 150, 200 and 250. How can that be that a 8 bit image is not capable of a much better tone resolution? -- Alvesgaspar (talk) 15:35, 24 May 2010 (UTC)
- @Alvesgaspar: "greys are posterized": yes, it is a conversion effect: from 16bit TIFF (my RAW format) to 8bit JPG. Only the background is additional "Gaussian blur" corrected. The background looks allways stained if I use focus stacking. I hope the info is helpfull for you. --Alchemist-hp (talk) 11:47, 24 May 2010 (UTC)
- Yes, that is what I mean. In this case I very much doubt it is inevitable as the tone difference between adjacent bands is quite large. I wonder if the original picture shows the same effect. -- Alvesgaspar (talk) 08:39, 24 May 2010 (UTC)
- Yes, it seems to be unsharp, but it isn't. This molybdenum sample is a high reflective, glossy and greasy lustre item with no realy sharp edges. --Alchemist-hp (talk) 19:19, 24 May 2010 (UTC)
Confirmed results:
Result: 12 support, 1 oppose, 0 neutral → featured. /George Chernilevsky talk 06:26, 1 June 2010 (UTC)
This image will be added to the FP gallery: Objects