Pallophotophone-immediate predecessor to RCA Photophone

[Mathew Bailey]

I found two You Tube videos on the Pallophotophone,RCA & GE's predecessor of the RCA Photophone system using twelve optical tracks recorded in an endless loop onto unperforated 35mm film. Some of the recordings are on nitrate stock,while the others are on acetate. I was thinking of modifiying a 35mm Steenbeck flatbed editor by replacing the sprockets & sprocket rollers with pressure rollers & capstans & designing a twelve track optical soundhead for the editor to play back & digitally transfer all twelve optical tracks at once. The You Tube links as follows-https://www.youtube.com/watch?v=rUm_mPizQFk
https://www.youtube.com/watch?v=vZ6NRtD-oAw

[Leo Enticknap]

Could the multiple tracks be played as a continuous sequence (e.g. by the optical pickup assembly moving after each continuous pass of the loop), or did the pickup have to be moved and reset while the transport was stopped, to switch between tracks? Or did it have 12 discrete photocell pickups and you could mix the output on the fly between each and all of them?

[Mathew Bailey]

The way these people have their reproducer built,I believe they have to reset & move the pickup. They ought to have it like you suggested of having twelve photo pickups instead of one pickup head.

[Jim Cassedy]

In the mid 90's , I worked for a time as a network technician for "the phone company', and in the central offices, they had a machine called an Audichron, which was used to read back "out of service" telephone numbers if you dialed one, ("The number you have reached, 123 555-1212 is not in service..." and there was also one connected to a "reverting ringback test" circuit so that installation techs could dial a special number which would read back the number they were calling from, and then ring the phone when they hung up. (This was all before caller ID became common) A slightly different model of the same machine was used on the old telephone numbers you could dial to get the 'time of day' read back to you, and yet another model was used in automated toll offices to tell you how many coins to deposit into a pay phone. (remember them?)

The machine worked something like an old cylinder phonograph. Inside, there was a cylindrical loop of film, about the size of a large Campbells Soup can. On the loop were 11 optical tracks, consisting of one track each of a recording of a woman reading each digit from 0 thru 9, and one track that was blank, except for one part which had a clear rectangle on it, which was used as a 'timing pulse'. A long incandescent light bulb inside the rotating cylinder, was used as the exciter lamp. (It looked like a regular 'fish-tank' lamp to me, but it might have been a special bulb) On the outside of the cylinder, a bracket with 11 optical sensors was used to read the optical tracks. This machine was connected through an interface to either the mechanical or digital central office switch (depending on the office) which was able to 'sequence' the tracks to playback in the proper order to read back a telephone number.

The original machines for doing this used a large magnetic recording drum, but obviously, running 24/7, the drums & mag heads were under constant wear, which degraded performance. The "optical cylinder" had a much longer MTBF, but even those soup-can film loops occasionally broke or got scratched from someone doing maintenance on the machine without shutting it down. The bulb used as an exciter lamp got quite warm and as I recall after a year or so of use, the film loops dried out and became quite brittle and easy to break, usually coming apart at the (ultrasonic) splice that held the cylindrical loop together. The later models of the machine printed the optical tracks on glass disks, about 9" in diameter, which rotated vertically (like the 45rpm records in an old Seeburg Juke Box). This completely eliminated the problem of the film loop breaking, and by putting several disks on a common rotating shaft, one machine could now do the job of several, in roughly the same amount of rack space as one of the earlier a 'single loop' units.

Eventually the optical playback technology was replaced by units using digital audio files. I tried to acquire one the film-loop Audichrons when they were ripping them out. I couldn't get them to part with a whole machine, but they did give me several of the "soup-can film loops" they had as spares. I'd send a picture, but I recently moved and haven't got all my stuff out of storage yet.

[Frank Angel]

Fascinating stuff about old school Telephony. Last night I have a "telephone" conversation on a smartfone. It sounded like what you used to hear on those shortwave broadcasts -- horrible phaseshift swooping in and out level fluctuating up and down by a dozen DB and in the end, audio that was totally unintelligible for at least 1/4 of the time. And my caller said he was on the spanking new 5G system -- you know, the one they are hyping to be the next giant leap forward in technology. Spare me. Just give me a phone connection that is 1) almost never dropped, 2) sound at least as good as an AM radio, 3) works wherever the phone is and 4) is actually designed to be held up againt the ear like the handsets we used for over a century rather than a flat piece of plastic that refuses to stay where you put it. In other words, give me a telelphone that is designed to be A TELEPHONE. Have it connected via copper wires that run back to a central switching office with make the connections using that magical rotating switch technology, and if you MUST have a wireless connection to carry voice audio; get a CB radio or a freakin walkie-talkie; almost ANY older technology will transmit voice audio better than cell phones. Yah, those silly flat things with their identity crisis that don't know if they are cameras, or internet computers or directional maps or a box full of "apps" or text a message delivery system; they can be all those things, just don't make them be mediocre excuses for the telephone as well.

[Mark Gulbrandsen]

Our own Leo Enticknao narrates a documentary about the very beginnings of sound on film.

[Mathew Bailey]

https://www.youtube.com/watch?v=rUm_mPizQFk The Pallophotophone was designed by Charles Hoxie originally as an acoustical-optical recorder but later modified as an electrical-optical recorder.
Both the original recorder & reproducer I believe unless otherwise no longer exist but the current reproducer was made from parts obtained from Ebay. Like I said,I was thinking of taking a Steenbck 35 mm flatbed editor & modifying it with capstans & pressure rollers in place of the sprockets & sprocket rollers,plus a 12 track optical sound head because the films are 35 mm but not perforated. You can hear the full NBC seven note chime sequence in the following You Tube video. https://www.youtube.com/watch?v=rUm_mPizQFk

[Leo Enticknap]

Yikes; that's a blast from the past. De Forest was only really interested in optical audio for movies (ironic, given that he self-styled himself the "father of radio"), whereas Hoxie and several others were trying to develop it for other uses, too, as Matthew describes.

Anyways, the most fascinating aspect of De Forest is the back story that the producers of that documentary cut out entirely. It also largely explains why he failed to interest Hollywood in the technology, where AT&T/Western Electric and RCA eventually succeeded. I'm going to take the liberty of getting into politics a little bit, given that this is politics 98 years ago. When I was researching the development of the De Forest Phonofilm system around 15 years ago, I went through his personal papers, archived at the San Jose Historical Society. He spent around a year and a half in Berlin working on the project in the early 1920s, because the post-WWI depression in Germany made the cost of doing the work a fraction of what it would have been in the USA.

Anyways, from his diaries and papers, it seems that De Forest spent as much time developing admiration for Hitler and Mussolini as he did developing sound technology, and on his return to the USA, wrote and lectured anyone who would listen on how cool this fascism stuff is, and so on and so forth. Needless to say, this didn't go down very well with the Hollywood establishment, especially as De Forest made no attempt to hide his extreme anti-Semitism. That, combined with the fact that De Forest totally burned his assistant, Theodore Case, by trying to take the credit for important refinements to the sound camera that were Case's work, ensured that he eventually remained largely forgotten in mainstream accounts of the history of movie technology, despite having made some of the crucial discoveries (e.g. the use of a weighted idler wheel to eliminate wow and flutter) that made optical sound possible.

Someone really should make a biographical movie about Lee de Forest, though absolutely not the one he tried to have made in the last few years of his life!

[Allan Young]

The later models of the machine printed the optical tracks on glass disks, about 9" in diameter, which rotated vertically (like the 45rpm records in an old Seeburg Juke Box). This completely eliminated the problem of the film loop breaking, and by putting several disks on a common rotating shaft, one machine could now do the job of several, in roughly the same amount of rack space as one of the earlier a 'single loop' units.

Sounds similar to the prototype speaking clock from 1954 in the London Science Museum.

Lots of pictures here: https://collection.sciencemuseumgrou...rototype-mk-ii

[Jim Cassedy]

I recall seeing pictures online in an old broadcasting magazine from the 1940's of several 35mm optical playback units that were were installed at KFI Radio (Los Angeles) to play back radio programs 'because of the superior qualities" of 35mm recording over the transcription disk technology in common use at that time. I don't know how much (or even if) it was ever used. They called it a "film phonograph", which was a term often used at that time to describe 35mm sound only playback machines. It looked very similar to the type of sound followers installed to play back the original FANTASOUND tracks only the soundhead was equipped with a standard RCA sound-head instead of the special FANTASOUND 4-Track heads. ( Think of two, 2000ft magazines and an RCA optical reproducer & motor, usually mounted on an old WesternElectric 3point pedestal. It was basically a 35mm projector, but there was no projector or lamphouse. Only the mags, motor, & sound head)

Here's a photograph of one of the FANTASOUND units (from a 1940's SME Journal Article On Fantasound)
The KFI units looked like this, except they had a standard single channel RCA sound head attached.

Playback.jpg

[Van Dalton]

In the mid 90's , I worked for a time as a network technician for "the phone company', and in the central offices, they had a machine called an Audichron, which was used to read back "out of service" telephone numbers if you dialed one,
*snip*
The machine worked something like an old cylinder phonograph. Inside, there was a cylindrical loop of film, about the size of a large Campbells Soup can. On the loop were 11 optical tracks, consisting of one track each of a recording of a woman reading each digit from 0 thru 9, and one track that was blank, except for one part which had a clear rectangle on it, which was used as a 'timing pulse'. A long incandescent light bulb inside the rotating cylinder, was used as the exciter lamp. (It looked like a regular 'fish-tank' lamp to me, but it might have been a special bulb) On the outside of the cylinder, a bracket with 11 optical sensors was used to read the optical tracks. This machine was connected through an interface to either the mechanical or digital central office switch (depending on the office) which was able to 'sequence' the tracks to playback in the proper order to read back a telephone number.

Actually Cognitronics made the film-loop announcing machines, not Audichron. They came into being when Audichron still held patents on magnetic drum systems, thus they couldn't legally bring one to market so they used film instead. The most advanced models had 32 tracks, for digits and other phrases including a silent track and a reorder signal that trips over itself each revolution (...UMBER-YOUHAVE-DIALED-ISNOT-ONTHEE-NETWORK- -TWO-ONE-M- - -doot doot doot doot doot doodoot doot doot doot.....). Audichron/ETC and Weatherchron always used 96-track drums, and later solid-state machines. Audichron might have had a film machine but if they did it was probably just a one-off or a prototype.

Ma Bell used the Cognitronics film-loop systems as the basis for their rate-quote system used by operators, and ANAC in a lot of places (what your machine was used for) but became especially prominent in the late 70s/early 80s when competitive long distance services started using them. They were often paired with Dan Ray crossbar switching equipment. MCI and SPC (Sprint) were big users of these machines (mostly because they were old technology and they were cheap). Cog eventually introduced a magnetic-drum machine in the 70s once Aud's patents expired, and a digital system using LPC, and all using samples from the same master tapes used to strike the film loops (including the glitchy reorder)!

http://www.evan-doorbell.com/product...nitronics.flac

I'd love to see one of the film loops if you can get a photograph (or scan) of one. It would be especially interesting if one could be opened up flat at the splice, scanned then the samples extracted using FFT or whatever technique is normally used for that, so a set of high-quality samples could be obtained for phreaks building ANAC/intercept emulation into e.g. Asterisk systems.

[Van Dalton]

Our own Leo Enticknao narrates a documentary about the very beginnings of sound on film.

I love your accent, Leo! I've been reading your posts on here for years but this is the first I've heard you speak. I have quite a West Yorkshire "twang" myself, largely affected by my having been raised around my grandparents who came here from Leeds and Harrogate in the '50s.

Does anybody else find it interesting that DeForest apparently foresaw the development of 2-channel bilateral variable-area when he couldn't even get satisfactory results out of his early variable-density system? Look closely just after the cringeworthy scene of his re-enactor shoving the core of film off onto the floor, around 2:45.

[Leo Enticknap]

I think they just used scrap film that they got from any old place. Could have been twin bilateral VA rather than stereo, too.

I never found anything in the De Forest papers at San Jose to suggest that he was ever interested in stereo. AFAIK, Alan Blumlein was the first to develop stereo audio into a usable prototype, both using disc cutters and variable area optical recording. Like so many pioneers of audio-visual technology, he paid the price for being ahead of his time. By the time his research was ready to move into development and marketing, the movie industry had spent staggering sums on converting to (mono) sound, and there was no money for a further refinement that no-one believed would prove a big audience attraction (same thing with 4K projectors today). And this was in the teeth of the Great Depression, so the consumer market wasn't there, either. It would be another two decades before stereo went mainstream, by which time the EMI/Blumlein patents had expired.

Some of Blumlein's experimental stereo films survive, as does a 1939 stereo Abbey Road studio recording of Mozart's 40th Symphony. They provide pretty convincing evidence that Blumlein perfected two-channel recording and reproduction. If he'd started his work a decade and a half later, he'd have likely gotten the credit for it.

[Van Dalton]

I was actually being sarcastic with the comment on their stereo-film flub.

But since we're talking about it: somebody years ago had told me about Western Electric having supposedly developed a stereo version of their variable-density system, but never brought it to market, I guess because they were having issues with its design (probably sound quality problems). Have you ever heard anything about that or were they just full of ?

[Leo Enticknap]

Sorry; should have picked up on the sarcasm.

I've never found anything related to stereo VD. It could work in theory, but I'd guess that you'd need a gutter between the two tracks, or else even tiny misalignment between the two reproduction photocells would cause really gnarly crosstalk.

The difference between VA and VD is like the difference between an MP3 file and a vinyl LP. Area is more idiot resistant at all stages in the mastering, duplication and reproduction chain. With variable density, the sensitometry and densitometry control in the exposure of the negative, and then its duplication through to the release print is really critical. The best VD tracks have (subjectively, to the listener) a far wider range and lower noise than VA of comparable vintage, but the worst sound like a guy with a tea towel stuck in his mouth standing next to several skillets full of frying eggs. VA technology produced a more consistent result in mastering, and which was more tolerant of less than perfect quality control in the striking and processing of release prints, which was why it ultimately won out.