"Makeup developed in the 1940s for black-and-white television. [...] The new cameras transmitted images in negative. The company's solution was a makeup that looked highly unnatural in real life but appeared natural when reversed on screen."
That explanation is obvious nonsense. This explanation is slightly less nonsensical, but it still doesn't make much sense to me without an A/B comparison of shots like this and what they looked like on screen, which I've been unable to find.
My mother was taught makeup for B&W TV and film in her theatre makeup classes in the early 60s. My understanding of it pretty much conforms to the second link you provided. I have her old makeup books but they're in storage right now, unfortunately.
Among Mr. Factor's other achievements: "...invented a pie topping that was cheaper than dairy cream and stuck to the face longer."
I hit the same search black hole the last time I wondered about the TV thing. (It probably somewhat predated recording/recording-by-default,* but what, no one with a vintage camera and VCR later on?) But you can find plenty of examples of orthochromatic (vs. modern 'panchromatic') film with similar problems, and I gather motion picture makeup was kind of weird too:
That image is hosted all over the place and I think also suggests how blue eye shadow became a 'look like a movie star' thing, possibly xref'd with ancient egyptian art. (Found tonight on http://1930sglamourandstyle.tumblr.com/ which dates it from the 1920s.)
This probably explains why a lot of modern WWII movies look 'weird' and 'wrong', because we're used to seeing ortho photos but the prop/war nerds are probably nerdy enough to be checking the real artifacts. Veterans must've gotten a weird 19A0 vibe from all the stuff that skewed more to the coconut effect 'for atmosphere' in the '80s-'90s.
Actual before and after shots are impossible -- at very best, you'd have parallax error. And according to the second FA (which actually makes some sense, as opposed to the first FA and its negative camera nonsense), the colors in the makeup used were dependent on the relative sensitivity of the "B&W" camera to such hues.
So, what to do?
Second FA talks about the TVs being 30 line -- that means you can discern a pattern of 30 sets of alternating black and white lines, or about 60 "pixels" worth, on a given raster line.
Second FA also talks about how bad the contrast was, which was the real reason this was necessary to begin with.
Combine these concepts and it becomes clear that it wasn't black and white at all, per se, but a bunch of dull and very blurry shades of grey that aren't really all that related to the actual luminance of the subject, much like a false-color image of Mars from one of the first two rovers that actually made it there (neither of which has an actual color camera).
And the resolution sucks. NTSC is and has always been 480i (plus or minus some VBI and overscan cropping and lack of focus). For appropriately-sucky resolution, let's say that's 200 pixels worth of tall.
Ok. So first we crop it to about 4:3, because that's about right. So, we need to knock the contrast down. A lot. (I forked with the levels until there was almost nothing left.) And scale it down to, say, 60x200. And then back up to its original, for comparison's sake (using some variant of bicubic, because blur is our friend.)
Reduced contrast (five iterations of "contrast -50" in Photoshop), scaled to my interpretation of "30 tv lines" and back. It's less comical, but still very strange:
Same thing, with contrast curves adjusted by hand until it looked adequately terrible/correct:
I think this last one is best representative of the problem that this sort of makeup is a solution to: The girls look like girls wearing appropriate makeup, and not Clown Skeletor. (And remember, early TV barely worked.)
BUT! The photo I'm working with was almost certainly shot on negative silver halide 35mm B&W film, which certainly did not respond to colors the same as early television cameras, and it appears to have been rather over-exposed (which is the opposite of the problem that television cameras had). And then it was half-toned. And then it was printed on paper. And then, and then, and then...before someone scanned it and posted it on the Web, more than half a century later.
But I think it's at least a little bit representative of what they were working to accomplish: A bit blockier to be sure than any analog process would be, and probably too bright (please adjust your set), but I'm trying to illustrate a thought process, not make a career out of this. ;)
Herpderp, d'herp, the reply below was supposed to have been threaded onto this comment.
Are you sure, however, that "60 lines" means 60 vertical lines per raster line, or is it actually 60 raster (so horizontal) lines?
This link here seems to suggest the same: http://www.earlytelevision.org/rca_scanning_disk.html
I don't think NTSC is at all relevant here, because NTSC (as well as PAL and SECAM) were specifically standards for color television, and we're talking about an era far far before that. But yeah, 480i would have been used before NTSC, as NTSC was designed to be backwards compatible with B&W television sets (i.e. it supplements it).
This odd early 60 line TVs, however, may have used a completely different, long superseded standard.
I got the link out of gryazi's post by the way (again very interesting, thanks).
Yep. I'd forgotten completely about the old mechanical televisions. Which makes my contrast-fuckery really not very useful.
NTSC actually started as a B&W standard, though... color was added about 10 or 12 years later depending on how you count. (This is why we have a framerate of 29.97 instead of 30: B&W NTSC is exactly 30, while color NTSC is 29.97 to solve issues with the new chrominance subcarrier at 30. And this curse infects us still today.)
The only thing that NTSC is backward-compatible with is itself. :)
But I'm done goofing with it. I was inspired for a moment, and bored. And the feeling is ... just sorta gone now that it is a new day.
Ah, my bad. You're right, while PAL and SECAM were indeed standards for color television, NTSC was the whole TV standard 8)
If you're this bored, try doing something with the color image over here. Working from the B&W has thrown out the color data and it's not really recoverable unless you can guess. It would have smoothed out the contrast, but probably not quite how you think... and the choice of film might have made it look worse, although my brain can't map to the ortho curve from memory.
Also, pictures from a real 60 line television. And the camera.