The 20th century saw some amazing technological developments. We went from airplanes to the moon. We went from slide rules to digital computers. Crank telephones to cell phones. But two of the most amazing feats of that era were ones that non-technical people probably hardly think about. The transformation of radio and TV from mono and black and white, to stereo and color. What was interesting about both of these is that engineers managed to find a way to push the new better result into the same form as the old version and — this is the amazing part — do it in such a way that the old technology still worked. Maybe it is the rate that new technology moves today, but we aren’t doing that today. Digital TV required all-new everything: transmitters, receivers, frequencies, and recording gear. Good luck trying to play the latest video game on your 25-year-old PC.
It is hard to remember when stores were full of all sorts of audio and video media. We’ve noticed that all forms of media are starting to vanish. Everything audio and video are all streamed or downloaded these days. Records, 8-tracks, cassettes, and even CDs and DVDs are vanishing. However, vinyl records have made a come back in the last few years for their novelty or nostalgic value.
Audio recording on wax, foil, or vinyl was more or less the same process perfected by [Thomas Edison] (or, perhaps, people who worked for him) back in 1877 although the flat records we think of didn’t appear until around 1890.
The principle is simple. Air pressure from sound cuts a groove into the recording medium. A piezoelectric stylus (or later, a stylus with a dynamic element) traced the groove and reproduced the same sound. Amplify it, and the phonograph is in business. You might enjoy the gramophone [SynthDan1] restored in the video below.
Stereo
By the 1950s, the hackers of their day were building or buying “hi-fi” equipment, gear that sounded better than the poor-quality audio spewing out of record players and AM radios of the day. Eventually, companies would roll out stereo recordings. But the records didn’t look any different, and they would still play on a standard (mono) record player. How is that possible? No, it isn’t two separate records like the vintage player at the top of this post found in the Museum of Technology in Paris.
We could explain it, but it is more fun to let [Bob Banks] from RCA explain it in this vintage advertisement.
https://www.youtube.com/watch?v=f8y9lNGh2bM
The Real Story
Pretty impressive special effects for the time. [Bob] did oversimplify a few things, though. First, the groove can have a vertical component and a horizontal component. But the resolution on the vertical isn’t nearly as good, so that means one channel would be disadvantaged. Instead, the two tracks in the single groove are spun 45 degrees so that each channel has some horizontal and some vertical component.
[Bob] wants you to think RCA invented this, although he never actually says that flat out. In fact, [Alan Blumlein] of EMI patented the scheme back in 1931. The first commercial stereo records, which were not from RCA either, would not appear until 1957.
Because of how the groove was rotated, the movement of the stylus horizontally was the combination of the left and right tracks — the same as the mono signal. The vertical motion carried the difference: the left channel minus the right channel, or L-R.
That’s how a mono record could play back normally on a stereo player. The horizontal motion on the track will reproduce the same sound on both channels. Conversely, a mono stylus reading a stereo track would only pick up the horizontal part of the track and play both channels together. Unfortunately, many mono players didn’t move up and down very well and could wear a stereo record, so users learned not to play stereo records on mono players, even though it would work. Of course, that assumes you have the same-sized grooves. Older records had wider grooves and wider needles.
If you want to see how a stereo cutter works, check out [EpicenterBryan’s] video below.
Compatibility
In a market where Elvis Presley was still selling 78 RPM records because his fans couldn’t afford new record players, this compatibility was very important. We imagine [Alan Blumlein] would be horrified to see how we routinely tell everyone to throw away their tapes for CDs and their CDs for digital music. TV was the same. Making a signal with color that black and white sets could still receive was quite the marvel (and a topic for a future Retrotechtacular). The idea of making everyone throw out their sets for new ones or buy government-subsidized converters would have been poorly received, indeed.
We can’t help but wonder if we are doing all we can on compatibility. Do we really have to trash operating systems and CPUs every few years? Do we really need to double the memory in our phones every time our contracts run out? Or is it a clever planned obsolescence ploy? As people who create things, how are we doing on compatibility? We’ll see how history judges.
Featured image by [ParameterBond], Public Domain
You can (sort of) 3D print records using very high resolution 3D printers. You could, in theory, upgrade the software to use it to produce 3D printed, stereo vinyl records. That would be neat to hear.
Just print the filament and talk to it as it’s coming out of the nozzle. You would then need a pretty sensitive playback machine but the sound must be there. :)
Yeah, might just be simpler to tell it to print a spiral and modulate the extruder motor with audio.
I sense a whitehat presentation about listening to a recorded conversion using someone’s 3d printed object by detecting tiny changes in position and density.
That would be awesome and relatively easy to attempt to get some kind of data. Could record several printe while “quiet”, and then several prints while some loud music is playing in the background. Post the resulting files (preferably *.wav), someone would figure out how to extract useful data.
P.S.: Mythbusters tried one myth involving clay pottery and the straw pressing against to attempt recording sound, but IMO they half-assed that one.
Relevant links to previous hackaday articles:
http://hackaday.com/2012/12/21/3d-printing-records/
http://hackaday.com/2014/08/12/cutting-records-out-of-cds/
http://hackaday.com/2017/03/11/burn-music-on-to-anything/
VVS
I was about to post the exact same thing.
The “French Model HMV 101 Picnic Gramophone Playing Frank Sinatra” video looks like CRAP.
ahhhh… thanks for posting. Funny and to the point, though this video should be shared more and more because the message isn’t really spreading enough.
Unbelievable, the main picture is not photoshopped!
:o)
http://www.douglas-self.com/MUSEUM/COMMS/auxetophone/auxetoph.htm#gau
Yes, but I cheated. The system is to play two platters one after the other, not R+L stereo. Still.
Ah, that explains my first thought: this could never work for stereo. You would have to synchronize the start and the speed to within a fraction of a millisecond to perceive it stereo and not two interfering signals sources.
Then again, engineers might have come up with a mechanical solution for that…
Gear coupled tables with a second, offset pin to align the records. Use a similar setup for cutting the masters and all would be synchronized throughout production.
*sits back and waits for someone to do this just to do it*
And it seems that the twin turntable system is nor intended for stereo but rather to allow a quick changeover between disks to maintain continuity of a soundtrack of a film. A sort of DJ system then, with a high power PA as well!
Oops, I’m a slow typist!
I always loved the album cover artwork from the LP’s in the old days….
They killed the super CD. They would rather wait for stereo during the birth of rock and roll. They wouldn’t use the new electrical recording process they invented and missed electrically recording the great tenor Caruso. Most of The Decade ’65-’75 has been mastered in 4 channel surround but is not available. They want to encumber our devices with all kinds of prevention stuff. Do I still do business with them? No. As an artist there are many alternatives to their business model.
What many consider the “music industry” is really just the “units of recorded media” industry. This industry didn’t exist til a little over 100 years or so, it greatly disrupted the sheet music and live music scenes, and together with radio put most musicians out of work.
Now that the physical media has diminishing importance, they are less important to the industry. payback’s a b1tch… Though as noted they are struggling mightily to put locks and fences around online music.
Hey, now there’s Earthlike planets detected in a 15 light year range, we should point out to these “Atlases” that they can pay SpaceX to build them a nice big spaceship (Spaceshap!!!) to go and form a new capitalist utopia…
… name it Golgafrinchan Ark B.
this is a good plan.
Then there were the Cook Binaural records with two sets of grooves. Two properly spaced cartridges on one tonearm rode in the grooves giving half the playback time per record side but in “stereo”.
Joint stereo is really cool. You can extrapolate the method to any number of dimensions provided you have the bandwidth.
Isn’t joint stereo just changing your choice of eigenvector = rotating what axes you record along?
Of course, very little music is now actually in stereo. The engineers record each instrument and vocalist on one track, then mash the outputs together, blending a little more left than right for one instrument, a lot more right then left for another, etc. Not stereo at all, just mush. If you ever get a chance to play one of the old “direct to disk” LPs, recorded with two mikes, you will be amazed by the difference in sound.
Foo Fighters is still recording the correct way.
There’s alot of mush yes, but a lot of effect processes will mess with time differences as well as level differences, so it’s still “stereo”, albeit synthetic.
Some engineers and artists like to track acoustic instruments using classic stereo miking techniques, and yes direct stereo recordings can be very realistic and musical. I particularly like making binaural recordings.
What? Most drum sets in modern music are reproduced in stereo… overheads, room mics, panning of close mics. Guitars are panned to represent how they would appear in a live setting. If you mean most music isn’t typically tracked 100% in one room using a stereo microphone setup, you would be correct. It is more practical to use a combination of techniques to convey the impact of a particular piece of music. You can’t say the end product isn’t “stereo” just because it wasn’t tracked with two mics in a room, that’s ridiculous.
Operating systems might be trashed every few years, but the standards within them are not. Windows is known for keeping suboptimal technical choices purely for the sake of backwards compatibility. One well known example is the length of file names, which is ridiculously short for modern days standard. The same applies to a rights management/Active Directory. Even shifting from 32 bit to 64 bit computation did not break much and is used happily side by side even today.
CPUs are trashed simply because running them is more wasteful than binning them. I love running old hardware for the sake of it, but if it is pure calculations per watt you are after, they age incredibly quickly.
Microsoft ditched 8×3 filenames with Windows 95, but maintained backward compatibility with older FAT file systems by having a holder 8×3 name linked to a long name in the new long name table. Problem was, they PATENTED the long filename algorithm so that nobody could use long file names in FAT unless they paid a royalty, which pretty much nobody did. This is why your camera still uses 8×3 filenames even though your computer using the same file system doesn’t have to.
There were several 3rd party addons that provided long file names to DOS and Windows 3.1x, but all but one used systems not compatible with how Windows 95 and later do them.
What was “The One”? A bit of software with the ‘does exactly what it says on the tin’ name of “Long File Names for Windows 3.1 and 95”. Boring, precisely descriptive names like that make it extremely difficult to find.
What it does is add true Windows 9x long name support to Windows 3.1x and when installed in Windows 95 (any version) it adds long name support for 16 bit software that would normally show the alternate 8.3 names. I assume it also enables 16 bit software to write to files with long names without wrecking the long name. IIRC it also worked on networks so a Win 3.1x computer could properly work with long name files shared from a Windows 95 system. Don’t know if it would work for 16 bit software on Windows 98 or any later 32 bit version. It’s available here. http://www.mdgx.com/files/LFN402.ZIP
One need only look at the die area, convolution, and power draw involved in the modern x86_64 instruction set carried to maintain binary backwards compatibility all the way back to the 8086 to see A) that we still do this at least sometimes and B) that it doesn’t come for free…
Two channels per recording? At one point, they were packing four!
I think we had a Q4 tape setup, big grundig machine like double stacked briefcase. It had one speaker in detachable lid, one in base unit and DIN outputs available to plug in 2 more speakers. Think there were only 2 Q4 recordings for it we had, portions of Holst’s Planet suite and maybe excerpts from Vivaldis 4 seasons. They ran faster than stereo and only had about 30 mins.
I think the machine had a remote on a wire, that had been stepped on or something and was all taped up and temperamental…… as was the mic that came with, might have been dual or quad ribbon, shaped like a cob of corn on a mini xmas tree stand.
In fact there were several competing matrix systems, SQ and QS being the ones that got the most press. After waiting over 6 months for Thijs Van Leer’s “Introspection” album to not arrive I gave up on all of it as a bad joke.
RCA did, however, develop the ‘winning’ colour/color TV system. Amazingly CEO David Sarnoff directed his engineers to develop a system that was backwardly compatible with existing black & white TVs – a much harder engineering feat than developing a colour-only system from scratch.
Our engineers and standards authorities could learn from this.
In Australia in the last few years we were forced to buy digital TVs or tuners, only to find that most of these were SD and not compatible with HD/Widescreen digital. And in the last year or 2, broadcasters have started using an MP4 standard so rendering many of the first-generation HD boxes obsolete.
Don’t start me on Apple…
In the USA, there were some analog HDTV systems developed that were backwards compatible with the old NTSC system. One I recall carved out the bandwidth needed by sort of randomly making single frame ‘holes’ in the red, green or blue colors. The person who wrote a magazine article about it claimed to not be able to notice them while watching a program at normal speed. Another concept filled in the sides on 4:3 shows with patterns generated by what the inventor called a “woodgrain chip”.
But in the end, the Advanced Television Standards Committee decided to chuck analog and go for an all digital system. They’d already stuffed color (circa 1964) and MTS audio (development starting circa 1975 but full deployment taking until the early 90’s) into the old monochrome and monaural NTSC system. ATSC still uses the same bandwidth per channel. Keeping the same channels (except for low VHF) allowed for keeping the same broadcasting and receiving antennas and some of the existing NTSC broadcasting equipment. Conversion costs could have been much higher.
What biting the bullet and abandoning compatibility with analog receivers brought was subchannels, lots and lots of them. In the USA, a station can stuff in as many subchannels as they care to degrade the picture quality of each, without paying any additional licensing fees to the FCC. (Some run 5 or 6 480p subchannels and no HD ones.) That’s why 1080p broadcasts here are extremely rare, no room for many, if any, subchannels. Most HD broadcasting is 720p.
IIRC in Canada the stations have to pay a separate broadcast license fee per sub channel, so some stations have none and will use the bandwidth to broadcast in 1080i or 1080p.
Despite being mostly finalized way back in 1996, ATSC has had some stuff added while maintaining backwards compatibility. http://atsc.org/newsletter/atsc-2-0-bridge-to-3-0/
ATSC 3.0 is going to break some stuff. Notably, it will support HEVC which should enable higher video quality with the numbers of subchannels currently being used. Very likely we’ll see some stations going up to 10 or 12 sub channels while keeping the same not so great picture quality. More channels = more airtime for advertising. Which Smart TVs will get a software upgrade for ATSC 3.0? Any of them that can currently play HEVC videos off a USB stick should be *capable* of being upgraded – the issue is will Samsung and the other companies bother to produce the upgrades for their older, discontinued models? Here’s one of the few cases where I’d be in favor of government intervention to require companies to support products that are possible to be updated. Perhaps via a program similar to the CECB coupons. Get a coupon to exchange for a software download for your Smart TV.
Records aren’t dying, they’re on a mad rise – vinyl LP’s are actually for sale in regular supermarkets here in the UK, something I’ve not seen for probably 20 years. Not specialist record shops, not music stores, regular supermarkets!
Hah, just got round to watching the living stereo one, love the subtext of…
“Stay in with your best girl for Victrola and chill.”
:-D
Now it is “hang out and watch Netflix”
Probably goes all the way back to “Come back to my cave, I’ve got fire! “
Thanks for sharing, very informative post.