Wikipedia:Featured picture candidates/Rosie the Riveter (Vultee)
- Reason
- Strong lighting and interesting composition in this photo from the Library of Congress. It's a posed photo, but I believe it has excellent encyclopedic value as a depiction of a real-life "Rosie the Riveter" during World War II. The contributions of African-American women during the war are often marginalized or forgotten. This photo expands our understanding of the "Rosie the Riveter" icon to include African-American women. Very large (8073x6449 pixels). I also like that we have good caption information (including the actual plane being worked on and the location of the plant).
- Articles this image appears in
- Rosie the Riveter, A-31 Vengeance, could also be added to Military history of African Americans
- Creator
- Photo by Alfred T. Palmer, U.S. Office of War Information. Found in the Library of Congress archives and uploaded to commons by Northwesterner1
- Support as nominator (edit 2) Northwesterner1 (talk) 10:06, 30 April 2008 (UTC)
NeutralOppose original - support edit Innovative image of high technical and aesthetic value. Capital photographer (talk) 10:35, 30 April 2008 (UTC)- Oppose not very sharp at this size. Touch it up and bring it down to a manageable size, perhaps?The freddinator (talk) 02:32, 1 May 2008 (UTC)
- Hold on... so you're opposing because it's too big? You can't compare a 52 megapixel image to a 2 megaixel one, just in the same way that you can't look at this on an 87-inch flat screen television and call it unsharp. Remember that downsampling cannot improve image quality; at best, it will do nothing to quality while making the file more manageable. But then again, I don't think manageable is one of the FP criteria. Thegreenj 03:45, 1 May 2008 (UTC)
- The freddinator raises a valid point. There's no reason for this to be 20MB - that does make it unmanageable. Durova puts these sort of things up regularly and they're nothing like that size. And, FWIW, current Rosie the Riveterish FPs: poster, worker, and also similar (two of these are already in the short Rosie the Riveter article). --jjron (talk) 08:18, 1 May 2008 (UTC)
- Thanks for the link to the other Rosie-the-Riveter FPs. All great photos, and I believe the other Rosies are from the same OWI photo series. As I say above, I like the way the nominated photo expands our understanding of the "Rosie" iconography to include African-American women.Northwesterner1 (talk) 08:34, 1 May 2008 (UTC)
- Wikipedia allows the upload of images up to 20MB. Capital photographer (talk) 08:21, 1 May 2008 (UTC)
- No kidding. Your point being? --jjron (talk) 08:23, 1 May 2008 (UTC)
- My point is what is your point? Wikipedia allows images of such a size and FP criteria require high quality. An image of such resolution released under a free licence is an asset, not a burden. Either way, keep it this size or make it smaller, so long as it meets FP size requirements. Capital photographer (talk) 09:32, 1 May 2008 (UTC)
- I have to say, I don't understand the point about manageability. Most wiki users are "managing" images as thumbnails in articles or sometimes as previews on the image page. They don't need to "manage" the full picture at all in most cases. I thought it would be preferable to have the largest image possible in the event that someone out there does want to take it under the public domain provisions and make a poster out of it. Wikipedia has an important role to play in disseminating the highest quality images to the public. However, I'm not opposed if any editor wants to make a smaller image and upload it over the same file name.Northwesterner1 (talk) 08:37, 1 May 2008 (UTC)
- While I would normally favour the highest resolution possible for the reasons mentioned above, in this case the image is unnecessarily unwieldly and large. I've done some tests. Recompressing the image while keeping the quality high results (quality 11 in photoshop) in a perceptibly identical image that is only 10MB. Downsampling the image to 4036x3224 (halving horizontal and vertical resolutions) results in a functionally identical image in terms of detail and the image is only 2.9MB. As proof of the quality after downsampling, I've put together an image containing three segments, cropped from the part of the image that I felt had the most usable detail (the woman's face and the rivet gun). Crop 1 contains the original 8073x6449, crop 2 has been downsampled to 4036x3224 and then upsampled back to original res, and crop 3 has been downsampled to 2018x1612 and then upsampled back to original res. I believe crop 1 and crop 2 are virtually identical in quality and detail. Crop 3 is obviously inferior and I used it only to demonstrate that you can't downsample TOO much. Therefore I think that we could safely downsample it to 4036x3224 as a 2.9MB file instead of 20mb, with no loss of detail whatsoever, yet with the added benefit of increased perceived sharpness and smaller file size. Diliff | (Talk) (Contribs) 12:25, 1 May 2008 (UTC)
- Good explanation, and having experimented with the image myself, I found reducing size was certainly beneficial. I was just a bit perplexed by the emphasis on file size rather than image res. My Edit 1 reflects the reduced size. Capital photographer (talk) 12:41, 1 May 2008 (UTC)
- Your edit has also caused havok with the highlights. The rivet gun has been posterised and has blown highlights (I know the rivet gun had minor blown highlights already but yours has resulted in one third being completely blown), the woman's nose has a minor blown highlight, and the white sheet metal holder is extremely blown now. You need to be a bit more careful of the impact of your edits (or recalibrate your screen!). I like the crop you've used though, but you've introduced too many faults to outweigh the positives of the edit! Diliff | (Talk) (Contribs) 12:56, 1 May 2008 (UTC)
- Tell me about it, I don't know what has happened. I still have it open in PS and it looks fine, I just opened the JPEG I saved and uploaded and it is all wrong. My fault for not checking the saved JPEG before uploading but it shouldn't do that. I will upload a replacement ASAP. Only thing I did different was I used the Save For Web function for some reason instead of Save As, I was doing other work on a website at the time. May have had some presets still active. Even the size is different. I guess its too close to bed time. Capital photographer (talk) 13:38, 1 May 2008 (UTC)
- Ok, uploaded replacement. Because Edit 1 was really a mistake, have replaced with Edit 1 (corrected). I used Save As like I normally do and it came out right. Apologies for that lousy edit. Capital photographer (talk) 13:40, 1 May 2008 (UTC)
- Tell me about it, I don't know what has happened. I still have it open in PS and it looks fine, I just opened the JPEG I saved and uploaded and it is all wrong. My fault for not checking the saved JPEG before uploading but it shouldn't do that. I will upload a replacement ASAP. Only thing I did different was I used the Save For Web function for some reason instead of Save As, I was doing other work on a website at the time. May have had some presets still active. Even the size is different. I guess its too close to bed time. Capital photographer (talk) 13:38, 1 May 2008 (UTC)
- Your edit has also caused havok with the highlights. The rivet gun has been posterised and has blown highlights (I know the rivet gun had minor blown highlights already but yours has resulted in one third being completely blown), the woman's nose has a minor blown highlight, and the white sheet metal holder is extremely blown now. You need to be a bit more careful of the impact of your edits (or recalibrate your screen!). I like the crop you've used though, but you've introduced too many faults to outweigh the positives of the edit! Diliff | (Talk) (Contribs) 12:56, 1 May 2008 (UTC)
- Good explanation, and having experimented with the image myself, I found reducing size was certainly beneficial. I was just a bit perplexed by the emphasis on file size rather than image res. My Edit 1 reflects the reduced size. Capital photographer (talk) 12:41, 1 May 2008 (UTC)
- While I would normally favour the highest resolution possible for the reasons mentioned above, in this case the image is unnecessarily unwieldly and large. I've done some tests. Recompressing the image while keeping the quality high results (quality 11 in photoshop) in a perceptibly identical image that is only 10MB. Downsampling the image to 4036x3224 (halving horizontal and vertical resolutions) results in a functionally identical image in terms of detail and the image is only 2.9MB. As proof of the quality after downsampling, I've put together an image containing three segments, cropped from the part of the image that I felt had the most usable detail (the woman's face and the rivet gun). Crop 1 contains the original 8073x6449, crop 2 has been downsampled to 4036x3224 and then upsampled back to original res, and crop 3 has been downsampled to 2018x1612 and then upsampled back to original res. I believe crop 1 and crop 2 are virtually identical in quality and detail. Crop 3 is obviously inferior and I used it only to demonstrate that you can't downsample TOO much. Therefore I think that we could safely downsample it to 4036x3224 as a 2.9MB file instead of 20mb, with no loss of detail whatsoever, yet with the added benefit of increased perceived sharpness and smaller file size. Diliff | (Talk) (Contribs) 12:25, 1 May 2008 (UTC)
- I have to say, I don't understand the point about manageability. Most wiki users are "managing" images as thumbnails in articles or sometimes as previews on the image page. They don't need to "manage" the full picture at all in most cases. I thought it would be preferable to have the largest image possible in the event that someone out there does want to take it under the public domain provisions and make a poster out of it. Wikipedia has an important role to play in disseminating the highest quality images to the public. However, I'm not opposed if any editor wants to make a smaller image and upload it over the same file name.Northwesterner1 (talk) 08:37, 1 May 2008 (UTC)
- The freddinator raises a valid point. There's no reason for this to be 20MB - that does make it unmanageable. Durova puts these sort of things up regularly and they're nothing like that size. And, FWIW, current Rosie the Riveterish FPs: poster, worker, and also similar (two of these are already in the short Rosie the Riveter article). --jjron (talk) 08:18, 1 May 2008 (UTC)
- (reset indent) I understand entirely that this photo could use some downsampling, but I'm just saying that an oppose based on it being unmanageable is not valid. Thegreenj 20:06, 1 May 2008 (UTC)
- Perhaps I should have just placed it as a comment or made it some conditional statement. I was just saying that the picture could benefit from a size reduction. The freddinator (talk) 02:18, 2 May 2008 (UTC)
- Hold on... so you're opposing because it's too big? You can't compare a 52 megapixel image to a 2 megaixel one, just in the same way that you can't look at this on an 87-inch flat screen television and call it unsharp. Remember that downsampling cannot improve image quality; at best, it will do nothing to quality while making the file more manageable. But then again, I don't think manageable is one of the FP criteria. Thegreenj 03:45, 1 May 2008 (UTC)
- Support all Illustrative and high quality. Thegreenj 03:45, 1 May 2008 (UTC)
- Conditional oppose Short version: Skin tones have a grey cast to them. Tweak the saturation a bit and I'll support it. Long version: Ignoring the racial issues for a moment, the background of this FP on the subject is more interesting. The one being proposed is just up against a boring steel sheet. The skin tones on this one seem slightly grey-cast; some colour adjustment is necessary before it becomes FP (and that's the "conditional" part). From composition and technical quality, this image is not as good as our other FP. However, when we bring back in the race issue, then this is exceptional as a change from the thin, frail white women that were the standard iconography of WWII, and given this was before the civil rights movement proper, serves to highlight a neglected contribution. Hence, comparison with the other FPs is invalid, and we may judge it on its own merits, leaving only the skin tones as a problem. Shoemaker's Holiday (talk) 10:06, 1 May 2008 (UTC)
- Uploaded edited version. Analysed using histogram in PS, corrected using levels, reduced partially washed out areas caused by bare flash bulb, corrected skin tones. Finally cropped it to enhance focus on subject and reflection. Capital photographer (talk) 12:32, 1 May 2008 (UTC)
- You didn't say which part of the image has bad skin tones in your opinion. If it's the face, you may want to consider the possibility that she is wearing make-up that is not entirely flattering to her skin tones, perhaps created for a "white" ideal. If so, it might be encyclopaedic to leave the image as is, rather than try to push the saturation to match our modern-day ideal. Papa Lima Whiskey (talk) 15:12, 3 May 2008 (UTC)
- The bad skins tones affected all visible skin in my opinion. It is a common affliction for images of this vintage, a combination of the films used and the bare bulb flash. The skin gets an odd grey tint with a silvery-satin appearance. I have seen it in many many images of both male and females. The same sort of skin tone issues a lot affect images of people with other skin colors, skin tones are difficult to reproduce for any camera. My adjustments were minor however given it is a feature of the time. Capital photographer (talk) 11:06, 4 May 2008 (UTC)
- Tweak the saturation a bit and I'll support it. Does that mean you now support crop 1? Papa Lima Whiskey (talk) 09:51, 9 May 2008 (UTC)
Support either, with preference for Edit 1.Good, high quality historical photo. Diliff | (Talk) (Contribs) 10:36, 2 May 2008 (UTC)- Support any with preference for Edit 2 after considering discussion about keeping original crop. Diliff | (Talk) (Contribs) 14:12, 6 May 2008 (UTC)
- Support either, but I think the edit is cropped too much. Some cropping is appropriate, but with such a tight crop some of the physical context is lost.--ragesoss (talk) 18:22, 3 May 2008 (UTC)
- Prefer edit 2.--ragesoss (talk) 19:50, 12 May 2008 (UTC)
- Support All Three. This high-quality image is a perfect symbol for the historically significant cultural trends during wartime America. I prefer the second Edit slightly, by the way. NauticaShades 16:39, 4 May 2008 (UTC)
- Support The scan's colour rendition looks spot on to me. It's a very early, large format Kodachrome sheet film capture, the sharpness of which is probably beyond the capability of the scanner to record. Ok, lighting isn't one of Palmer's strong points but he was quite pioneering in actually dragging around those big arc lights to record what are surely the best colour images of the US war effort at the time, and we're very lucky to have them freely-licensed for the encyclopedia. Please, don't crop this or delete the original upload: if files are too big, use them as masters for retouching (which this one needs a bit, I notice) then upload a smaller version for online viewing purposes, linking to the original on the image description page. Best of both worlds. --mikaultalk 11:12, 6 May 2008 (UTC)
- Comment: I agree with mikaul about the crop. I support the retouching in the edit (and I'm very grateful for it, as I wouldn't know how to do it well myself). I also support the downscaling without significant loss in quality (with a link to the original file). But I believe the original framing should be used. The encyclopedic value here is not only in what the image shows but also in the image itself. The Rosie the Riveter article is about women who worked in wartime factories, yes, but it's also about -- you might even say primarily about -- the representation of those women by the U.S. government. From that perspective, the original crop has more encyclopedic value. I would like to upload a retouched, resized photo with the original crop, but I'm don't trust myself to adjust the lighting properly and would appreciate someone else's help.Northwesterner1 (talk) 11:51, 6 May 2008 (UTC)
- I agree with the above sentiments. If this is to be promoted I tend to think it should be in the original version (i.e., uncropped), however for the only option above image page size to be 20MB is silly. I wonder if perhaps Diliff could upload the 2.9MB version he has alluded to above as a decent compromise, and then link to the others on the image page. --jjron (talk) 13:17, 6 May 2008 (UTC)
- Done. It turned out to be 3.4MB with subtle sharpening but not a major difference. Diliff | (Talk) (Contribs) 14:12, 6 May 2008 (UTC)
- Thanks, Diliff.Northwesterner1 (talk) 18:07, 6 May 2008 (UTC)
- Good stuff; and I think that also addresses most concerns other voters have raised. --jjron (talk) 14:18, 6 May 2008 (UTC)
- I was going to close this earlier with some others, but have now commented so much I almost feel like I've voted. In which case I will vote and someone else can close. So Support Edit 2. --jjron (talk) 14:32, 6 May 2008 (UTC)
- Done. It turned out to be 3.4MB with subtle sharpening but not a major difference. Diliff | (Talk) (Contribs) 14:12, 6 May 2008 (UTC)
- I agree with the above sentiments. If this is to be promoted I tend to think it should be in the original version (i.e., uncropped), however for the only option above image page size to be 20MB is silly. I wonder if perhaps Diliff could upload the 2.9MB version he has alluded to above as a decent compromise, and then link to the others on the image page. --jjron (talk) 13:17, 6 May 2008 (UTC)
- Comment: I agree with mikaul about the crop. I support the retouching in the edit (and I'm very grateful for it, as I wouldn't know how to do it well myself). I also support the downscaling without significant loss in quality (with a link to the original file). But I believe the original framing should be used. The encyclopedic value here is not only in what the image shows but also in the image itself. The Rosie the Riveter article is about women who worked in wartime factories, yes, but it's also about -- you might even say primarily about -- the representation of those women by the U.S. government. From that perspective, the original crop has more encyclopedic value. I would like to upload a retouched, resized photo with the original crop, but I'm don't trust myself to adjust the lighting properly and would appreciate someone else's help.Northwesterner1 (talk) 11:51, 6 May 2008 (UTC)
- Support edit 2. I prefer the wider view. Mangostar (talk) 14:37, 6 May 2008 (UTC)
- And as a sidenote: it's great that you can even see her chipped nailpolish. :) Mangostar (talk) 14:40, 6 May 2008 (UTC)
- Support Great pic. I've had my eye on this for months and never quite got around to doing the restoration. The composition is unusually good with a second angle of the subject's face reflected in the airplane she is building, and as another voter noted I've always been intrigued by photos of female factory workers from World War II wearing cosmetics and other feminine ornaments as they do construction work...notice the ring here. Although the colors aren't quite as striking as in some other "Rosie the riveter"-ish photographs, superior compositional elements make it photographically worthy and the subject's ethnic background makes it encyclopedically non-redundant. This gets my vote, and I'm snapping my fingers for not having followed through on the urge to restore this myself. Prever edit 2, but either is okay. DurovaCharge! 22:29, 7 May 2008 (UTC)
- Support. I prefer the wider-field versions, but would support any. Remarkably good color-rendition for early Kodachrome. Very striking image, highly encyclopedic. As an aside, the discussion above is fascinating to this non-specialist. Cheers, Pete Tillman (talk) 05:54, 8 May 2008 (UTC)
- Oppose Clearly not unique. Many pictures illustrate the same topic better, and does not meet FP standards. EgraS (talk) 07:09, 8 May 2008 (UTC)
- Support all Historic, very encyclopedia, and per Durova. SpencerT♦C 16:55, 10 May 2008 (UTC)
Promoted Image:Rosie_the_Riveter_(Vultee) DS.jpg MER-C 12:50, 13 May 2008 (UTC)