Wikipedia:Picture peer review/Trinity explosion, 1945

Trinity explosion, 1945

An icon of the Atomic Age. I recently uploaded a larger version, which is probably why this photo hasn't been nominated before. The New Mexico Museum of Fine Arts in Santa Fe has a print of this photo on permanent display, and I've spent a long time studying it there (and here). The WWI German helmet shape, the One-Eye Monster, Death, the Destroyer of Worlds... An amazing photograph. Berlyn Brixner was the photographer, and some details are there. Pete Tillman (talk) 06:37, 6 May 2008 (UTC)[reply]

Nominated by: Pete Tillman (talk) 06:37, 6 May 2008 (UTC)[reply]

Comments

Great photo - well once I realised it wasn't an old army helmet, and saw what it really was! My only concern would be quality. At this image size, a filesize of 72 KB is pretty tiny. That is probably mainly because the blacks have been largely compressed into almost nothing but pure black. But, geez, I want to say nominate anyway. --jjron (talk) 08:32, 6 May 2008 (UTC)*Note: make sure you replace the original with this in at least some articles if you do nominate. Can I also add that I rather like the scale and time measure that are included on the original that are sadly lacking here. --jjron (talk) 08:34, 6 May 2008 (UTC)[reply]
The more I look at it, the more I think it would probably look a bit more balanced with a bit of the black cropped off the bottom. In the original that works to hold the writing, in this one it's just a big patch of nothing. --jjron (talk) 13:33, 6 May 2008 (UTC)[reply]

Thanks for all the comments. Yeah, I'll stick it in some of those articles -- meant to earlier, but...

Cropping the bottom would be appropriate -- I think that's how the museum's print was presented. Which, incidentally, was unsigned, ims. And we could reference the alternate shot with the time & scale bar on the nomination.

First cut at a crop:

Needs redo on a real image-processor, but hows the composition? Trim the top too? Feel free to take a whack <GG>

Have a look at The Trinity Test: Eyewitnesses. If that stuff don't raise up the hair on the back of your neck, I'll eat my socks.

"Don't fly south." Indeed. In haste, Pete Tillman (talk) 18:52, 6 May 2008 (UTC)[reply]

Have had a try at an edit. Slight straighten, crop, took out some marks, and gave a small sharpen and noise reduction. The quality isn't great, and there's some funny things in the black part, it could be artifacting. Also this version doesn't seem to have the same detail on the explosion itself as the small version looks to have. Where did you actually get this version? I wonder if there's one of better quality around at this sort of size. --jjron (talk) 07:29, 7 May 2008 (UTC)[reply]

I've put up another small version that really highlights the problems in the background. Obviously this is not there as a suggested edit, it's just showing up some issues that are present. Sure there's considerations to do with age, rarity, difficulty of capture, etc, but I think a lot of these problems are to do with this particular digital conversion. For example there's things that are significant enough to be visible on the small version if they were part of the photo, but they're not there. Hmmm... --jjron (talk) 08:03, 7 May 2008 (UTC)[reply]

From http://www.lanl.gov/news/photos/historical.shtml Ah -- these are two different times. The bigger photo is at +0.016 sec. the smaller at 0.025 sec. I'm guessing the background flaws are on the original film -- see the photogs comments, ref'd above.

Thanks for your efforts, which I'll look at in more detail this PM. In haste, Pete Tillman (talk) 18:20, 7 May 2008 (UTC)[reply]

I've uploaded a bit larger version of the 0.025 sec. photo:

and a +/- matching version (captioned, scale bar) of the 0.016 sec. photo:

I looked pretty hard, and wasn't able to find any files bigger than the first version (above), now recaptioned correctly as 0.016 sec. So that's the one we have to work with, to meet the 1000x1000px standard.

I notice that your edit2, above, seems to be bit grayer tone than the slightly-sepia original. Do you see that? And do you think the blotchy background could be cleaned out, perhaps painted over with pure black, which is what the camera would have really seen at 10-20x solar brilliance in the fireball? Brixner did a remarkable job with the tools he had available -- these are 16mm movie stills. Anyway, I agree that the 0.025 sec photo is the more dramatic and better-exposed, but it looks like 540x410px is the largest file online. Thoughts? Thanks, Pete Tillman (talk) 05:13, 8 May 2008 (UTC)[reply]

Ah, OK, now it makes a bit more sense if they're different images. What a shame a bigger version of the 0.025s one isn't available. Yes, I noticed somewhere (I think after I uploaded) that mine had a slightly different tone (though I wouldn't say the original had sepia toning). Not sure why, it could have come in when I did the noise reduction perhaps. I still have my original edit (straighten and crop only) saved as a TIFF, so that should be identical tonings to the original.
I also thought of just cloning the background to remove those flaws, possibly just slightly off pure black; just not sure how much editing should be done :-). When I brought them out in that Edit 2 it looked pretty clear that it wasn't just stuff off the original photo; there looked to be some characteristic jpeg artifacting and possibly posterisation, which is a bit in line with my earlier comment about the filesize being very small for that resolution. There also appeared to some smudges and things like that. However I was also wondering whether some of it was perhaps smoke, and that was one reason I hesitated before removing it. What do you think - should we try removing it all? --jjron (talk) 08:41, 8 May 2008 (UTC)[reply]

Why not give it a shot? That sort of thing is way beyond my skill (& software). Pure black or something close could be counted as restoration (kind sorta...) FWIW, I don't see the background weirdness in Image:Trinity.016sec.jpg, at right. As to tone, I noticed that changing the view angle on my LCD screen, even just a little, significantly changes the tone... Perhaps fiddling with brightness and/or contrast would bring out the fireball details better. I'll fool with it a bit tonight in iPhoto, but so far that program seems to mangle archival files pretty badly when you crop. Which also may reflect my level of skill in using it.... <rueful grin>

As to what happened -- filming a nuclear blast, then sitting around for 60 years -- not too surprising it looks a bit ragged. I wonder if Mr. Brixner is still with us? Still, I doubt that Los Alamos encouraged its photographers to take their work home.... And who knows who digitized the thing, and with what skill? I'll have to ask, next time I'm at the museum in Santa Fe, where they got their file. I'm guessing they borrowed the original, or a 1st-gen copy neg. It would also be interesting to see which of these two they printed. Cheers, Pete Tillman (talk) 18:34, 8 May 2008 (UTC)[reply]

How about this? Image:Trinity explosion2-edit2a.jpg -- all I did was touch up the tint a bit. IPhoto didn't mangle it! And the BG looks OK, though I didn't mess with that. Cheers, Pete Tillman (talk) 22:22, 8 May 2008 (UTC)[reply]

The remaining big problem I see is the overexposed lower left corner, most obviously in the leftmost dust-cloud "skirt". IPhoto doesn't seem to have a "dodge" option to tone this down. Does your software? TIA, Pete Tillman (talk) 23:13, 8 May 2008 (UTC)[reply]

"FWIW, I don't see the background weirdness in Image:Trinity.016sec.jpg"; yes, that's probably right, as with the 0,025sec version. That's why I've been saying that I think it's an issue with the digital conversion on the 'big' version, rather than necessarily being a problem with the photo itself. In fact I rather wonder whether the big one hasn't just been upsampled from a small version, which would explain why it seems to lack the quality of the smaller ones.
Re working on the background, I had a bit of a play with it, but given the level of the problems it almost seems beyond repair, at least to me; it would need a lot of careful work. Re the overexposed lower left corner, I think you want to burn it rather than dodge it. I tried a bit of burning there in Photoshop; it did look a little better, but there's not a lot of information to burn in and it was just a quick try. Your edit2a has a bit of sepia tinge to it now; not really sure what you were trying to achieve with it? Do you think there's a version worth nominating? --jjron (talk) 03:03, 10 May 2008 (UTC)[reply]

"Your edit2a has a bit of sepia tinge" -- I was trying to match the tone in Image:Trinity Test Fireball 25ms.jpg, perhaps not very successfully. Anyway, here's another version, with various fiddles: see what you think...

Uh-oh, think I overdid it on the lower-right corner, trying to cut the lower-left bloom.

"I rather wonder whether the big one hasn't just been upsampled from a small version..." FWIW, I played with Image:Trinity.016sec.jpg a bit -- it definitely has less detail when you blow it up. So who knows.

I guess I'll go back and fiddle some more tomorrow, but to my eye, Image:Trinity explosion2-edit2b.jpg is the best yet. Thanks again for your help with this -- hopefully you can add another "assist" to your stellar FP record! Best, Pete Tillman (talk) 04:44, 10 May 2008 (UTC)[reply]

OK, I don't think it's the tone in 0.025s that makes it look better, I don't really think the sepia helps this one. Re 0.016s having less detail when you 'blow it up', that's the point I'm making about upsampling, it can't add detail that's not there, it can only estimate what should be there.

Anyway, I went back to my original 'cropped only' version and did some different fixes; no sharpening or noise reduction this time, so tone should be more faithful to the original. I selectively adjusted the levels on that troublesome lower left corner, and then another levels adjustment slightly darkened the whole image for a bit more contrast. I then fixed a number of what looked to be flaws on the photo and did a bit of work to improve the background. I think that's about as much time as I want to spend on it. Have a look, see how you think it compares to your latest edit, and I'll leave it up to you. --jjron (talk) 12:37, 13 May 2008 (UTC)[reply]

I think your last edit is probably about as good as we're likely to get -- I like it better than any of mine, including the work files I tossed.

What next? This is all new to me. Thanks again, Pete Tillman (talk) 18:32, 14 May 2008 (UTC)[reply]

What next? If you want to try it out at FPC head over there and create a nomination. If you have any questions about how to do it, just ask, or I'll do it for you if you'd rather. Despite the work done on it I suspect quality may be an issue - but you don't know if you don't try. If it does no good there, it could be a good candidate for Valued Pictures instead (if and when that gets up and running). --jjron (talk) 08:55, 15 May 2008 (UTC)[reply]

Seconder

Nominated at FPC as Wikipedia:Featured picture candidates/The Trinity nuclear test, 1945 by Pete Tillman. --jjron (talk) 14:39, 26 May 2008 (UTC)[reply]