Wikipedia:Featured picture candidates/Lassen Volcanic National Park
From Wikipedia, the free encyclopedia
[edit] Lassen Volcanic National Park
A panoramic shot of Kings Creek which appears in the Lassen Volcanic National Park article. The park, home to Mt. Lassen, a volcano which last errupted about 90 years ago, is a lesser known park in northern California, and its article was lacking a picture which truely conveys the magnificent landscape. Oh, and it is taken by me, so fire away.
- Nominate and support. - Dschwen 17:34, 6 September 2006 (UTC)
- Comment - won't vote yet, this is too dark on my calibrated monitor. Needs some curve or level correction, to remove the "sooty" feeling from the greenery. Some downsampling would be nice, too - a teeny bit fuzzy in full size... --Janke | Talk 19:23, 6 September 2006 (UTC)
- I don't get the downsampling. Too fuzzy at full size? Don't look at it in full size! Downsampling always looses information. Always! If anyone needs a downsampled version, its no big deal, create one, but why replace the original with a downsampled picture? --Dschwen 19:27, 6 September 2006 (UTC)
- Oppose Way too dark on my CRT monitor - Adrian Pingstone 20:52, 6 September 2006 (UTC)
- Support It does not look dark on my laptop monitor. I looks like a great picture. HighInBC 22:47, 6 September 2006 (UTC)
- Would support a lightened version. Great picture, and the overexposed snow isn't enough for me to oppose. Plus it's so difficult to not overexpose the snow. --Tewy 23:15, 6 September 2006 (UTC)
- Will support if someone brightens it.Nnfolz 00:38, 7 September 2006 (UTC)
- Oppose for now. Way too dark on my calibrated monitor. I would change to support if a much lighter version was made available. --Nebular110 00:43, 7 September 2006 (UTC)
- Oppose The image is dark on my monitor as well, and somewhat unsharp at full resolution. These might be adjusted, but the uneven polarization across the sky is nearly impossible to correct. The left side is significantly brighter than the right side. -- Moondigger 02:00, 7 September 2006 (UTC)
- Oppose Very blurry. --Midnight Rider 02:54, 7 September 2006 (UTC)
- Which picture did you look at? Sorry, but this comment is just ridiculous. Slightly blurry maybe, I used a G3 to take the pics. But very blurry implies out of focus parts or motion blur, and the pic has neither. Look at the grass, is that blurry, no! --Dschwen 07:44, 7 September 2006 (UTC)
- Blurry - no, but soft - yes. You probably could have safely downsampled without losing much/any detail [Ah, I see you did, sorry]. Also, did you notice you have a horizontally recurring hot pixel/spec of dust in the sky near the top of the frame? Diliff | (Talk) (Contribs) 09:29, 7 September 2006 (UTC)
- Uh oh, dead pixel, I totally missed it! --Dschwen 10:51, 7 September 2006 (UTC)
- Blurry - no, but soft - yes. You probably could have safely downsampled without losing much/any detail [Ah, I see you did, sorry]. Also, did you notice you have a horizontally recurring hot pixel/spec of dust in the sky near the top of the frame? Diliff | (Talk) (Contribs) 09:29, 7 September 2006 (UTC)
- Which picture did you look at? Sorry, but this comment is just ridiculous. Slightly blurry maybe, I used a G3 to take the pics. But very blurry implies out of focus parts or motion blur, and the pic has neither. Look at the grass, is that blurry, no! --Dschwen 07:44, 7 September 2006 (UTC)
- Oppose uneven polarization of the sky. Glaurung 05:44, 7 September 2006 (UTC)
- Impossible to fix. And I've only seen this complaint cone befor, yesterday. Is this a new thing? --Dschwen 08:18, 7 September 2006 (UTC)
- Well, uneven polarization has always been an aesthetic bugbear, which is why it is not recommended to use for wide angle/panoramic photography. Whether it is enough to oppose the photo all depends on the extent of the polarization and the way it affects the sky. Hazy skies are not affected as much and the effect is minimal. Diliff | (Talk) (Contribs) 09:29, 7 September 2006 (UTC)
- Impossible to fix. And I've only seen this complaint cone befor, yesterday. Is this a new thing? --Dschwen 08:18, 7 September 2006 (UTC)
- Support Edit 1. Downsampled image is still high resolution and alleviates most of the faults, but you're right, the polarization is not really fixable. Diliff | (Talk) (Contribs) 09:33, 7 September 2006 (UTC)
- Comment; Sure it is fixable - I'd do it, but I'm busy building steam trains in my workshop now... But, here's how to do it in Photoshop: Use the selection tool with a suitable value, additively select several spots in the blue sky. The, use "select similar", and the whole sky and is selected. After that, you can use an adjustment level or other tools such as gradients to lighten the darker portions of the sky... Greetings, --Janke | Talk 11:56, 7 September 2006 (UTC)
- I think there'd be a better chance of success (more natural looking) if the sky was adjusted with a new layer and a layer mask using the gradient tool on the mask itself. I said it was nearly impossible to fix uneven polarization in my previous comment -- and that's usually true for most images. But given the large, uninterrupted blue expanses in this one, it stands a better chance of success. I could give it a try later on... -- Moondigger 14:58, 7 September 2006 (UTC)
- Well, ok, it is, in theory, fixable in the same way that you can fix a landscape blemished with power lines by removing them. ;) To use a gradient means falsifying the sky by guesswork... I'm not saying I'm completely against the idea in theory, but usually you would end up flattening the sky, removing cloud texture and random variations. As moondigger said, it is as a general rule unfixable but you do have options. Diliff | (Talk) (Contribs) 15:53, 7 September 2006 (UTC)
- "To use a gradient means falsifying the sky by guesswork..." Not if you apply the gradient to a layer mask over a sky which has merely been levels-adjusted, which is why I suggested it. The analogy to removing power lines is applicable to Janke's method, but not to mine. Mine uses the actual sky captured in the image, including minor variations and cloud texture, and does nothing more to it than levels-adjust it. Then we vary the transparency of the layer mask using the gradient tool, such that the levels are unaffected on one side of the sky but progressively adjusted across the sky to the other end. The trick (and the primary problem, really) is making the layer mask adjust at an inverse rate to the change in the sky due to polarization. For many images with uneven polarization, the rate of change is not regular, and the gradient tool cannot simulate the proper rate of change on the layer mask. In those cases the only solution is to just select the entire sky and replace it with the sky from another image or with a paint-bucket dump of sky blue color. But doing something like that is no solution at all IMO, for the reason you mention -- it's akin to changing the actual content of the image rather than simple brightness/contrast adjustments of what's there. This image looks as if it might lend itself to the layer mask method. -- Moondigger 17:27, 7 September 2006 (UTC)
- By mentioning gradients, I didn't mean you should dump a blue greadient over the sky, no, no, use the gradient as a tool (you can adjust the linearity of it, so a good match can be achieved), for masking (or whatever) the original sky. I don't want to spend an hour or more on it, but it can be done, I bet. --Janke | Talk 18:40, 7 September 2006 (UTC)
- Thanks for the clarification. However, even though you can adjust the linearity of the gradient you apply, I believe the change still has to be mathematically "regular" -- either linear or logarithmic. If the uneven polarization is not regular (a possibility given the fact that this is a stitched pano), even small imperfections in the layer mask gradient will be visible. -- Moondigger 18:56, 7 September 2006 (UTC)
- By mentioning gradients, I didn't mean you should dump a blue greadient over the sky, no, no, use the gradient as a tool (you can adjust the linearity of it, so a good match can be achieved), for masking (or whatever) the original sky. I don't want to spend an hour or more on it, but it can be done, I bet. --Janke | Talk 18:40, 7 September 2006 (UTC)
- "To use a gradient means falsifying the sky by guesswork..." Not if you apply the gradient to a layer mask over a sky which has merely been levels-adjusted, which is why I suggested it. The analogy to removing power lines is applicable to Janke's method, but not to mine. Mine uses the actual sky captured in the image, including minor variations and cloud texture, and does nothing more to it than levels-adjust it. Then we vary the transparency of the layer mask using the gradient tool, such that the levels are unaffected on one side of the sky but progressively adjusted across the sky to the other end. The trick (and the primary problem, really) is making the layer mask adjust at an inverse rate to the change in the sky due to polarization. For many images with uneven polarization, the rate of change is not regular, and the gradient tool cannot simulate the proper rate of change on the layer mask. In those cases the only solution is to just select the entire sky and replace it with the sky from another image or with a paint-bucket dump of sky blue color. But doing something like that is no solution at all IMO, for the reason you mention -- it's akin to changing the actual content of the image rather than simple brightness/contrast adjustments of what's there. This image looks as if it might lend itself to the layer mask method. -- Moondigger 17:27, 7 September 2006 (UTC)
- Well, ok, it is, in theory, fixable in the same way that you can fix a landscape blemished with power lines by removing them. ;) To use a gradient means falsifying the sky by guesswork... I'm not saying I'm completely against the idea in theory, but usually you would end up flattening the sky, removing cloud texture and random variations. As moondigger said, it is as a general rule unfixable but you do have options. Diliff | (Talk) (Contribs) 15:53, 7 September 2006 (UTC)
- I'll try the following: Creating a layermask by subtracting red and green channel from the blue channel and adjusting curves (Yields a sky only mask). Then applying a gradien on the mask (from left to right, modulating the mask transparancy) and and adding the unbrightened sky to the brightened image using the aforementioned layermask. This should preserve structure while reducing the polarisation effect. Stay tuned. --Dschwen 20:31, 7 September 2006 (UTC)
- But let me mention that I actually like the gradient in the sky... --Dschwen 20:32, 7 September 2006 (UTC)
- Uneven polarization in the sky can be an interesting effect sometimes, but doesn't represent what skies actually look like when you're viewing the scene in person. It is almost universally frowned upon by professional photo editors, even if it isn't mentioned here often. -- Moondigger 20:58, 7 September 2006 (UTC)
- I think there'd be a better chance of success (more natural looking) if the sky was adjusted with a new layer and a layer mask using the gradient tool on the mask itself. I said it was nearly impossible to fix uneven polarization in my previous comment -- and that's usually true for most images. But given the large, uninterrupted blue expanses in this one, it stands a better chance of success. I could give it a try later on... -- Moondigger 14:58, 7 September 2006 (UTC)
Oppose Per above. If a version eventually address the polarization I might change my vote. But Edit 2 doesn't. --Fir0002 12:20, 9 September 2006 (UTC)
Great, I whish I'd have been told before I wasted my time editing it. Sorry, but this is just ridiculous. Thanks for the constructive input, but if even edit 2 doesn't satisfy the new fad I withdraw the nomination. It's a slap in the face if on the same page pictures with worse issues get nothing but support.
