In 1926, TV Was Mechanical
(spectrum.ieee.org)122 points by jnord 3 days ago
122 points by jnord 3 days ago
Although it's about wire transmission of photography - which, as pointed out, television is not - it's still well worth watching this 1937 newsreel explaining how it works, mostly because 1) they devote a LOT of time towards explaining the concept of scanning/rastering, which was clearly not widely intuited at the time, and 2) they do it with a brilliant physical analogy, with the incredible pedagogical clarity typical of such 1930s educational videos.
First time I watched that I forgot I had the speed set to 1.5 or something so the already fast-talking 1930s mid-atlantic radio announcer voice got exaggerated to a hilarious degree. Especially funny as the first few lines were about the importance of speed!
So cool, it reminded me a book I once found and never seen again on building analog computers at home with a few bits of magnetic-core memory. That magazine even shows Osram tubes, I didn't know that brand had such long history. One of my desktop toys is a $5 Osram Tungsten bulb that I bought to "watch" the highest melting-point element emitting light.
Also interesting is the electromechanical "tone wheel generator" used in the Hammond Organs:
https://forums.musicplayer.com/topic/155607-a-look-inside-a-...
91 metal wheels with lobes spinning past something like electric guitar pickups to produce sine waves. Their goal was to produce pure sine waves and then combine them via "drawbars" to produce an adjustable sound. 12 different gear ratios and different number of lobes (powers of two) to produce all the frequencies.
The rotating "Leslie" speaker was also cool, as is the electromechanical "vibrato scanner" on later models.
Reminds me of this fully mechanical ancient Greek movie from 2000+ years ago: https://www.youtube.com/watch?v=IW3uaJimMlI
So much was done with mechanical systems back in the day, because they were better understood and comparatively cheaper than corresponding electrics. I wonder if the engineers who designed teletypes or artillery range computers could learn to program, and if they did, would they have any unique insights?
I love this unique insight:
"Fire control computers ... solve ... fire control problems."
https://www.youtube.com/watch?v=gwf5mAlI7Ug
U.S. NAVY BASIC MECHANISMS OF FIRE CONTROL COMPUTERS MECHANICAL COMPUTER INSTRUCTIONAL FILM 27794
Those old educational videos are a lesson in how to give presentations I think. There’s an art to the way they build up a fairly complicated concept step by step. The viewers are starting from zero after all.
The starting step is one that is impossible to misunderstand. From there, go one concrete step to another.
They'd probably take a dependency, slice out the quarter they actually use, and somehow turn it into a cleverly-encoded lookup table or two?
Hopefully, you didnt miss this tangentially moored submission (was your depedendency on time constant re: [VP] sales vs engineers related to that covered therein
No, I was referring to a different time constant: the τ such that the number of times the whiskey priest[0] falls off the wagon[1] occuring at t<τ equals the number of times the decay occurs at t>τ.
So, just like the quantum watched pot never boils, the up-through-sales CEO would, out of habit alone, keep nurturing[2] their company officers — staying in touch (t<<τ) and shepherding them away from personal and towards corporate goals.
(My hypothesis being that an up-through-engineering CEO, not finding this behaviour natural, would have to make an explicit attempt to do so, and hence might experience severe cultural issues dealing with people who can deliver a great deal of value under the right leadership but require steady, if small[3][4], external inputs to keep them in the fold.)
[0] compare https://en.wikipedia.org/wiki/Don_Camillo_and_Peppone#Charac...
[1] pads headcount, orphans mistakes, etc.
[2] for a prospect: staying in touch and shepherding them towards close
[3] the shepherd's crook can be mightier than the prince's cannon. (back when we had Prince-Bishops they used to commission statues of themselves holding both sword and crook, emphasising their ability to use either temporal persuasion or moral suasion)
[4] see also the unstable yet controllable regime for designing fighter aircraft.
If you tour the British cruiser parked in the Thames in London, you can see the artillery computer. It's a triumph of mechanical computing.
There's a long history there. See the Early Television Museum.[1]
It's sad that no Scophony set survives. High resolution and 24 inch screens in 1938.
Full blown color NTSC managed with 6mhz channels! Digital broadcasts have a terrible user experience: Instead of providing the same channel they always did, but to vastly more receivers in much more marginal conditions using advances in digital signal processing and encoding power and methodology, broadcasters were allowed to cut up their frequency into 6 "subchannels", so they crunch the streams to the shittiest quality they can (as low as 3mb/s, DVD is 10!) legally get away with, provide almost zero redundancy in the broadcast, so now TVs that were well in the broadcast area of analog signals are straining to get every last bit out of the digital stream so it can hope to recover a picture.
They did this because 6 channels that most people cannot actually receive over the air gives them more advertisement slots than the 1 previous channel. Good old enshittification.
There is a great model kit made in England for the "Televisor". Got it years ago and it was minutes of fun. Apparently still for sale. The input is audio, got it to run from my iphone at the time (as opposed to using the supplied CD which in itself is Retro technology). I had some plans to build code a video-to-audio converter to run this (or screengrab to audio for that matter).
This reminds me of the equally interesting framing of modern computer failure rates in similar terms to mechanical parts.
This is also a very fun exploration of how many things start mechanical and move to more solid state as they age. I'm curious what major mechanical items are left to move over to solid state? My guess is most things that have changed were instruments in larger things. Gyroscope devices are a fun example.
Very cool. Makes me think of this wonderful scene from Sports Night where William Macy talks about Philo Farnsworth and Cliff Gardner inventing TV.
There were even recordings of it, though they couldn't be played back at the time:
interestingly close publish date to Asianometry's video on Sony's Breakthrough Color TV: https://www.youtube.com/watch?v=TOh3jEJGynA
I have nothing against ublock or other programs for ad block. It just seems like we shouldn’t have to participate in the arms race around internet advertising. I would pay for content (like substack) to avoid an experience where I get popups a few seconds apart.
Several years ago I came across the first issue of "Television" magazine from 1928 and reading it blew my mind in a couple ways. First, the overall tone is remarkably similar to a 1970s homebrew computer club newsletter, including defining what "television" even is (and isn't). For example, We learn on page 10 that "television is not tele-photography."
It's clear from this magazine that early television was the domain of home tinkerers and hackers. On page 26 is a detailed tutorial on how to construct your own selenium condenser cell from scratch, including which London chemist had appropriately high-quality selenium, where to buy copper sheets, mica insulator (.008 thick) and brass bars. Well worth a read: https://comicbookplus.com/?dlid=37097
That analog television not only was prototyped nearly a hundred years ago but then began being deployed at vast consumer scale ~75 years ago is still just so amazing. It's worth understanding a bit about how it works just to appreciate what a wildly ambitious hack it was. From real-time image acquisition to transmission to display, many of the fundamental technologies didn't even exist and had to be invented then perfected for it to work.