300
Color photo of a man and woman in Dagestan, 1904
(lemmy.world)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 17 Dec 2023
300 points (97.8% liked)
HistoryPorn
4862 readers
235 users here now
If you would like to become a mod in this community, kindly PM the mod.
Relive the Past in Jaw-Dropping Detail!
HistoryPorn is for photographs (or, if it can be found, film) of the past, recent or distant! Give us a little snapshot of history!
Rules
- Be respectful and inclusive.
- No harassment, hate speech, or trolling.
- Engage in constructive discussions.
- Share relevant content.
- Follow guidelines and moderators' instructions.
- Use appropriate language and tone.
- Report violations.
- Foster a continuous learning environment.
- No genocide or atrocity denialism.
Pictures of old artifacts and museum pieces should go to History Artifacts
Illustrations and paintings should go to History Drawings
Related Communities:
founded 1 year ago
MODERATORS
@PugJesus@lemmy.world do you have the source for this image? I'd love to find out more.
This version has been noticeably digitally altered, someone has used a clone or heal tool in the corners:
I assume the original photo or film must have holes or marks on it, they would be interesting to see.
I also have an (unconfirmed) suspicion that this image may have been a black and white photo that has been digitally colourised. It can't have been fully AI colourised as the flowers on the lady's dress are too perfectly coloured (even where they are hidden in folds or shadow). Alas the chroma of the flowers is shaped in perfect circles of pink, even overlapping black areas of the dress (where it's otherwise coloured slightly blue), making me suspect a round brush tool in an image editor:
I can't be 100% certain, there might be some other explanation for this chroma patterning. It's not JPEG (that quantises in square blocks, not circles). Might be some weird optical effects or multiple layers of JPEG on top of each other causing gaussian filtering (if you apply box filters repetitively at different offsets then you eventually approximate a gaussian). Not to mention that the version I downloaded is a .webp (and I have no experience with that format), I suspect Lemmy might have converted it upon upload.
https://www.loc.gov/pictures/item/2018679953/
Thankyou muchly :) Looking now. They have unaltered originals too!
There are 3 separate glass slides for the different colour channels. Ooh. (is it healthy to get excited about this?)
EDIT: This collage of the 3 coloured slides is itself an edited version, but it shows answers to my questions:
They edited the left side with a clone tool to hide the fact some of the coloured slides/layers are a bit faded at the edges. This also explains why the left of the image is yellowish and the bottom reddish. Perhaps those slides were like that their entire life, uneven due to manufacture or developing issues?
Across the entire image are tiny coloured blips and scratches. Most of them were edited out.
The rounded shape of the chroma on the flowers isn't as evident in this version, but it still looks blurry. It's plausible it's an artefact from multiple lossy image encodings POSSIBLY combined with the red channel (glass slide) having worse resolution than the other colours in these areas (?).
This is the kind of quality content we need more of!