Now there is something Bartosz Ciechanowski should write a blog post about!
I also think something important is missing from this explanation: the vertical/horizontal solution would not have been backwards compatible in two ways: (1) A mono needle would not have been able detect the vertical movement, resulting in a missing channel. (2) A stereo needle able to play such a record would not have been able to play a classic mono record, because the mono signal would have only been translated into a single channel (the left or right channel), and the other channel would have remained silent.
(1) would've made stereo records unattractive for customers who already owned an existing mono player, and (2) would've made stereo players unattractive to customers who already had a substantial mono record collection.
With the 45 degree solution, existing mono players were able to play stereo records (the horizontal movement is exactly the sum of the two channels), and stereo players were able to play existing mono records. For this to work, the left and right signals were recorded in opposite phase. A really elegant solution, which somehow reminds me of the cover of GEB [0].
PS: if you are interested in cutting-edge record groove technology, the "Füllschriftverfahren" invented by Eduard Rhein might interest you. It was an early compression technology for audio. The method was based on earlier work by the London-based Columbia Graphophone Company, but their work was never used in practice. Basically, before this invention, the spacing between the grooves on a record was fixed, with enough margins so that large amplitudes would not cut into neighboring grooves. Rhein build a machine that dynamically spaced the grooves based on the maximum local amplitude, allowing much smaller groove margins for quiet parts of an audio file, and therefore increased information density. This nearly doubled the running time of typical records.
Sadly, I only found an extensive description of this technology in German, including original patents [1]. But the figures are self-explaining.
"When Stereo records first appeared circa 1958, it was important that they were NOT played on existing mono equipment, as the two playback modes were incompatible with regards to wear and tear. A mono player would damage the stereo record primarily because a stereo record required the stylus to wiggle up and down as well as left and right. Mono only required left and right, and mono players of the time had little up and down capacity.
"Mono record players (or more specifically cartridges and stylii) then incorporated the ability to play stereo records and modern, of the time (late 1960's) players could therefore play a stereo record without damage. At the same time mono records were phased out, so people needed to be assured they could still purchase records (which were only available in stereo) and play them without causing damage.
"It has nothing to do with the record, everything to do with the playback equipment.
"As a further point stereo records were limited by phase - they needed to ensure that at various critical frequencies there was no difference in phase between left and right (bass is critical), otherwise the stylus might jump out of a groove. Some digital formats have recordings that contain much out of phase information - creating a holographic type impression. If these contain low frequency sounds out of phase, then these need to be modified prior to be prepared for release on a record. "
This is why there are still mastering engineers who specialize in vinyl. See also https://badracket.com/vinyl-mastering/ for more vinyl-specific mastering concerns.
Yeah the phase thing is quite a pain, I always thought it was just for really low stuff, but a band I was helping out had a master rejected by the pressing plant for excessive out-of-phase audio even when the passages in question weren't particularly 'heavy'.
Since then I've been quite paranoid about it before sending masters off, and it's surprising how many mixes have a load of out-of-phase stereo. It can usually be tamed with some M/S EQ but then you miss the spaciousness in headphones...
Sorry for delay in replying. Two main causes in my experience:
1. as a deliberate effect, either on its own or as part of a reverb; similarly, some synth sounds have this 'built in'
2. as a consequence of multi-microphone recording, where two microphones are used on the same source but are positioned such that the channels end up significantly out of phase
The trouble is that inter-channel phase is one way in which stereo 'works' so it's all a bit of a compromise... this is why there are various tried and trusted spaced stereo microphone techniques that give a sense of spaciousness without introducing too much out of phase content (e.g. ORTF) and often these give a more pleasing result that the coincident-pair (XY) techniques that minimise phase differences. Often in situations where there's lots of sources being recorded at once you can end up with problems. The usual rule of thumb is to ensure that if two microphones are picking up a single source, the more distant one is at least three times the distance of the closer one to the source. A quick way to check this stuff with headphones when positioning microphones is to sum them to mono, reverse the polarity of one of the mics and move them around to minimise the bass cancellation.
> (1) A mono needle would not have been able detect the vertical movement, resulting in a missing channel. (2) A stereo needle able to play such a record would not have been able to play a classic mono record, because the mono signal would have only been translated into a single channel
I think the way they do it with FM radio is: one channel is A+B and the other is A-B. The A+B channel is the "mono" channel for backward compatibility. (Then they add/subtract to get left/right: (A+B)-(A-B)=B; (A-B)+(A+B)=A.)
"For this reason, the left (L) and right (R) channels are algebraically encoded into sum (L+R) and difference (L−R) signals. A mono receiver will use just the L+R signal so the listener will hear both channels through the single loudspeaker. A stereo receiver will add the difference signal to the sum signal to recover the left channel, and subtract the difference signal from the sum to recover the right channel."
Other systems were considered by the FCC, but this one won (see article for the details)
If you're interested in this sort of thing, the book Defining Vision is a great discussion of the history of HDTV.
I don't recall why or how, but the A-B channel is of lower bandwidth, or otherwise more likely to drop out. Hence when you're starting to lose FM, you may find it dropping to mono.
The phase shift is really useful if you’re a DJ playing out on a loud system with a good chance of mechanical feedback.
i.e. there’s a chance your cartridge will pick up sound from the speakers.
The feedback will be picked up exactly the same on both the L and R channels so if you combine them to make mono the inverted channel will cancel out the feedback on the non-inverted almost perfectly.
The stereo image isn’t usually useful in a venue setting anyway so there’s not much downside.
> Rhein build a machine that dynamically spaced the grooves based on the maximum local amplitude, allowing much smaller groove margins for quiet parts of an audio file, and therefore increased information density. This nearly doubled the running time of typical records.
This is one of those things I never gave much thought but is super facinating. The technology connections youtube channel just did a video about this, too: https://www.youtube.com/watch?v=3DdUvoc7tJ4
I was just going to post this. Others have given some of the math and technical explanation below but if you would like to see this in action with a bit more of the history behind it all, I highly recommend this video.
I frequently find myself wanting to think more critically and guess about things like this. If I had thought a bit further before reading (one needle, but what is the groove like?) I would gotten to assuming that it was one dimensional, but then stepped from there to it actually affording two dimensions.
Does anyone have recommendations for reading or exercises to improve this kind of thinking?
When I was a kid I used to wonder how everything worked and would disassemble things to find out. Started to take apart those disposable cameras people used to use for vacation photos. Learned the painful way what a capacitor is and how it discharges enough electricity to make that camera flash in a hurry.
So my recommendation is electro-shock therapy. Self administered of course.
I have pretty good mechanical sense & experience and can reason through how things work. I guess I just want to get better?
But this applies to thinking that gets you into a rut working on a bug. Sometimes es I find that if I would have stopped, backed up, and thought through the whole stack more critically, I would have gotten the fix faster.
What age range is this channel's target audience? The slow pace and sophomoric description of 2 channel audio at the beginning makes me thing it is aimed at a young audience.
I am also 30 or 40 years old. I'm Schrödingers millenial.
I quite enjoy the delivery of technology connections. A nice change of pace imo, compared to other youtubers that edit out every millisecond of quiet time between words and sentences.
But I'm not watching the videos to learn new stuff either
I absolutely love this channel even though I find the guy's delivery excruciatingly slow and his style very annoying. I watch it on 2x to solve the first problem and just cringe through the dorkiness.
Yeah, the dorkiness can be a bit too much sometimes. One time he made a dorky joke, and then made a meta-joke addressing how dorky that joke was, and it was extra cringe...
I think the simple explanations are just that guy's style. He often has oddball videos about stuff you've probably seen but never thought about in depth - from lava lamps to retroreflectors.
He gets quite a few views and educates a lot of people. No need to look down on it - if it isn't for you thats ok.
As someone who is younger than HN demographic, it is consistently funny to me how many on Hacker News seem to think that 20-30 year olds are teenagers, whenever a YouTube video targeted towards that demographic is posted. Here's a thought that might frighten you: The oldest Zoomers (the first generation where being exposed to Windows XP is optional) are now 26.
I suspect you're not really inferring this from the sophomoric descriptions, as there are plenty of science/engineering documentary films dating back to the 19th century that are aimed at non-researcher adults and use simplified easy-to-understand descriptions, but rather gen Z stylistic choices like using the word "wiggly" to describe the groove of a vinyl record.
It's not playback speed that's slow. It's the sophomoric definition of stereo. It's the absolute dragging of the feet to deliver that definition that isn't even a good description of stereo.
I was recently in a large shop that sells books, music, etc. and I was amazed at the number of "LPs" (as I used to call them) of music old and new. The one thing that I really like about that format is... cover art. Cover art kind of survived the CD but the large size of a 12" record sleeve means you get real art.
I asked someone way younger than me (some students of about 19/20) why they buy records and they expressed how much they like the physical format and owning the music.
1. It's a different listening experience than the "song shuffle on Spotify" that dominates now. It's more a more active and focused listening experience. You put an album on to listen to that album from start to finish.
2. I want to support musicians by buying their albums. However I dont like buying digital music because it doesnt have the permanence a physical object has. I dont like spending money on what's basically ctrl-c ctrl-v files onto my hard drive. CD's are too flimsy. Vinyl has heft.
3. I'm not into the album art that much but a lot of my guests are and like to thumb through my records and look at the art.
4. I'm a hipster contrarian. When vinyl started to get too popular I started buying cassettes too
2. I want to support musicians by buying their albums. However, I don't like buying vinyl because it's horribly inconvenient, and the sound quality isn't great (clicks and pops), and requires special equipment that needs too much space for my apartment. On top of that, every time you play it, it degrades the sound quality, and even handling the sleeve wears it out since it has no plastic protective layer: well-used vinyl albums are dog-eared. CDs aren't flimsy like vinyl, and will last indefinitely if you take care of them, and their album art, while about 1/4 the size of a vinyl album, has the same aspect ratio and looks like new since it's kept in a "jewel case" (except for some newer albums). On top of all this, ripping a CD into a FLAC file is easy if you still have an optical drive (you can buy a USB-connected portable one for $20).
1. It's a different listening experience than the "song shuffle on Spotify" that dominates now. I just select the album I want to hear in my music player (which has my whole library on my laptop's HD), click "play", and let it play the whole album. Skipping songs is easy if I want to do that, but nothing is forcing me to, so I usually don't, and just listen to the album start-to-finish. For some odd reason, a lot of people seem to not have the discipline to do that.
Records for sure beat longevity of most media. I got a 1979 record from a friend. One side was full of dust. I guess it was parked that way on record player for few decades. After simple gentle wash and dry it was as good as new! Of course, if it's scrached it's not good. And if you listen it hundred times it will degrade, but it will still work.
as someone that was a vinyl DJ in a former life that saw the transition from vinyl to CDs to where it is now, vinyl is absolutely my favorite precisely because of the physicality of it. it was the closest to feeling of playing an instrument. learning the touch and feel of picking up the stylus cartridge without the infamous screeching/scratch sound, how light/heavy to touch the platter/spindle/label to push/pull/drag to speed up or slow down the record to keep line up the beats were all things that reminded of playing an instrument (I played woodwinds) with how much air to use, how hard to hit the reeds, and the various other physical things to play. there was also the fact that it had to be played and not programmed that i appreciate from playing vinyl.
as a medium for just sitting back and enjoying some tunes at home, it's gotta be the worst! LPs might give you 3,4,5 songs per side, but you still would have to get up and flip it or find another LP. singles like 45s or 12"s were even worse as it was one song at a time. a 12" at 45rpm only has somewhere around 10mins of recording. an LP at 33rpm might have 20-ish minutes. so that's a lot of interruptions in the music flow. if you really enjoy that manual switching out, then i'd suggest just taking the next step and being a DJ constantly mixing track to track.
> LPs might give you 3,4,5 songs per side, but you still would have to get up and flip it or find another LP.
There are actually some record players that can play a stack of records at a time. That's why some double-disc records are pressed as ADBC instead of ABCD - the player will play the front side of each, then flip, then play the back side.
They're pretty expensive, though. It made more sense when vinyl was the predominant media format.
> If you really enjoy that manual switching out, then i'd suggest just taking the next step and being a DJ constantly mixing track to track.
Funny enough, I actually do spin vinyl, and this is my least favorite part of listening to records. I guess they occupy different modes in my brain. If I'm actually trying to mix, swapping out is part of the fun. But if I'm trying to sit back and listen, getting up from the couch is an annoyance.
For me it's the other way around. Striming or local digital collection trigger my "let's skip to next song" impulse. While with LP I'll listen to entire album. So vinyl is for me medium of choice for sitting and listening at home.
This is where the cassette or CD are great mediums for listening to the full album (using the cassette's autoflip mode).
recently, i started going back through music collection to listen to full albums instead of the streaming track to track as you've mentioned. it was a very nice change of pace from just hearing random tracks continuously. it seems to be popular enough that even youtube has Full Album titles. the playlist assembling full albums are rather annoying for those albums with no gaps between tracks, but are better than nothing if that's all that's available. so now it's best of both worlds where it is streaming but still the full album experience.
"Music lovers clearly can't get enough of the high-quality sound and tangible connection to artists vinyl delivers," Glazier said, "and labels have squarely met that demand with a steady stream of exclusives, special reissues, and beautifully crafted packages and discs."
I bought (and long ago gave away, alas) an LP copy of the Jethro Tull's "Thick As A Brick" because it opened up to a 12-page newspaper, while the CD had a smaller abridged copy.
A few years back I had the privilege of working for the New York City edition of the Red Bull Music Academy, and, perhaps unsurprisingly, a large portion of the lectures and general discussion focused on things related to vinyl.
One thing I remember standing out to me was a mention by a vinyl mastering engineer that he would manually roll up the bass level on the EQ over the length of the side as the record was cut. He explained that this was because on most record players the tonearm is more or less parallel to the groove at the start of the side and by the end of the side is less so, which decreases bass response, requiring more bass energy in the signal to compensate.
I don't think that a constant roll-in of EQ would be helpful. On playback, the stylus is actually parallel to the groove at two positions, both in the middle of the disc, not at the edge, as depicted on a cartridge alignment protractor.
So you start playing a record and the stylus has a small lateral load bias to one side, has a zero crossing, an small opposite side bias, another zero crossing, and then finishes up with a high lateral load bias on the inner groove and all the female vocals sound terrible due to the extreme sibilance distortion.
The classic mid-side mic technique uses two capsules. The first (middle) is a cardioid or omnidirectionsl facing the subject, the latter (side) is a figure-8 turned 90 degrees. The capsules should be as close to each other as possible to avoid low-frequency phase errors. The signals are converted to L+R via a simple Mid/Side network:
I thought that was where the article was heading when it mentioned that horizontal was better (and the unstated fact that mono was horizontal), but it still took me a minute to realize that the 45 degree pickups are equivalent to midside. In (digital) mid-side you encode M=L+R and S=L-R. With this, if you breakdown the horizontal and vertical components:
HL = +L/sqrt(2) VL = +L/sqrt(2) L = sqrt(HL^2 + VL^2)
HR = +R/sqrt(2) VR = -R/sqrt(2) R = sqrt(HR^2 + VR^2)
H = HL + HR V = VL + VR
= (L+R)/sqrt(2) V = (L-R)/sqrt(2)
Note that VR is negative and HR is positive because the sign of the pickup is inverted in the diagrams.
It's also kinda related to how NTSC works - the color signal is at a frequency and phase such that it falls inbetween consecutive horizontal lines (e.g. pixels), so that when played on a black and white TV that isn't aware of color, the chrominance (as opposed to luminance) isn't visible since it occurs when the beam is advancing.
I’m not sure I agree with that explanation of the color subcarrier.
The color subcarrier is modulated to a higher frequency, but on an old B&W TV, there are no pixels, only lines. The color subcarrier can appear as a pattern superimposed over the picture if it is not filtered out with a low-pass filter—and indeed, B&W TVs made after the advent of color television contain such a filter, which can be removed if you want to use the TV as a higher-resolution monitor for your home computer (ask how I know...)
The two chroma channels are, however, “rotated”, similar to the way mid-side encoding is done. You take a YUV signal, you extract the chrominance UV channels, and then you rotate them to IQ channels. The I and Q channels are then quadrature modulated with different amounts of bandwidth assigned to the I and Q channels. The I channel gets, like, 3x as much bandwidth.
Broadcast NTSC did actually modulate the color signal into a "comb-shaped" frequency spectrum that was overlaid onto the complimentary "comb-shaped" spectrum of the B+W signal, because they didn't want to have to re-allocate higher broadcast channel bandwidths to be able to support color (and therefore have fewer channels available). The comb-shaped nature of the spectrum is an artifact of the discrete lines the picture signal is made up of, and the pause between each line when the electron beam reset itself and prepared to draw the next line. The fact they were able to figure out how to do color in the same bandwidth in a backward-compatible way is actually kind of insane, and I've seen it mentioned somewhere as one of the, if not THE most impressive engineering achievements of the 20th century.
Yes, that’s a good point. And then in NTSC, the UV is further rotated to produce IQ, which are given different amounts of bandwidth in the color subcarrier.
Aside from the contents, let's all take a moment to admire the simplicity of this page. A few animated pictures and everybody got right away what's going on. Absolutely amazing!
Hm, I thought that stereo vinyl records are recorded in a way that vertical groove profile is V=L+R and the horizontal represents H=L-R. It's backward compatible with older mono gramophones that read only L+R and stereo players do H+V=2L and H-V=2R to decode.
Also, while writing this, I realised that it's exactly the same as in the article. Just another way of thinking about it.
My great 'aha' moment of 'technology is actually sometimes really simple when broken down' was realizing that magnetic sound recording is just running past a speaker coil at a constant speed while holding something long and ferrous next to it. To play it back do the same thing but plug the coil into the input.
DPLII was based on the fact that if you invert one channel of a stereo signal, it sounds effectively the same. So, they had some way of encoding front and back signals as some sort of symmetry/anti-symmetry between the left and right signal.
Later Fosgate worked with Roger Dressler at Dolby and Dolby's engineers converted the analog circuit to digital. See these two posts from Dressler for additional detail:
This is an amazing example of moving forward without breaking changes ... We could probably do better with our software too. Another interesting example of creating a compatible signal is the addition of chroma information to the black and white NTSC television signal. The original signal was modulated using VSBAM and the chroma signal was added to (predominantly) the high frequencies of the channel.
Another reason why mono is better for 12" dance records (another being avoiding stereo image/balance effects in clubs; and if the PA setup is mono anyway then you'll get a real mono mix rather than a summed stereo mix.)
I imagine stereo could help a bit with phase effects between speakers, but they would likely still be an issue in multi-speaker setups.
I assume you mean actual binary encoding into the vinyl?
Now that's a hacker idea if I ever heard one. I assume the way to do it would be to use a modem the same way the internet operated over dial-up. Listening to the vinyl would sound the same as when you accidentally picked up a phone on the same line, a static-y kind of squeal.
It does make me wonder what the maximum practical bitrate would be -- more than phone lines with their severely limited frequency range, that's for sure. Could it handle CD-quality FLAC audio?
It also makes me wonder how you'd manage records wearing out. What type of distortion occurs, and how might that affect how you'd spread bits over the spectrum, or would it affect which error correction code you'd use?
I doubt it, since CD-quality audio is already objectively better (for human hearing) than analog audio on LPs. The only thing LPs are better at, sound quality wise, is extension into the ultrasonic range of 20-30+ kHz (which we can't hear anyway) but the SNR and distortion metrics are much worse in the ultrasonic band than the audible range.
I'm not sure if this is a named principle, but it seems intuitively obvious to me that on any particular encoding medium (magnetic tape, vinyl groove, etc.), you can encode "more" data in an analog way than digital, since with digital encoding you need to be able to make distinct symbols onto the medium, and represent any signal fluctuation with a series of such symbols; whereas with analog encoding a tiny fluctuation in what's recorded to the medium can correspond to a tiny fluctuation in the signal. Of course the tradeoff is that digital data is much more immune to distortion from imperfections in the medium.
If the same track width and pit sizes on CDs were used to encode audio in an analog way like LaserDisc does video (the continuous distance between pits being modulated by the signal), no doubt it could encode well into the ultrasonic range and surround audio channels via modulating them into different frequency bands. But it would have its own characteristic "surface noise" and "pops and clicks" just like vinyl.
> it seems intuitively obvious to me... you can encode "more" data in an analog way than digital
It doesn't seem obvious to me, because I'm not talking about encoding distinct symbols into the medium. I'm talking about using the entire available frequency range to encode digital information the way a modem does. And I'm talking about using lossless compression as well. And as you say, if an LP can encode ultrasonic information, that opens up all sorts of extra space for information that is useless in analog.
It may very well be the case that an LP couldn't record CD-quality audio, but I do very much wonder what kind of bitrate would actually be achievable. A dial-up modem is 56 kbps, using the frequency range 300 hZ to 3.3 kHz. CD audio is 1,411 kbps, but vinyl frequency range is from 7 hZ to 50 kHz, which is 16x wider than a dial-up modem. And you can pack much more data into higher frequencies.
So on the face of it, digital CD-quality audio seems like it might be very much achievable in theory on vinyl. But again, when you account for error correction and degradation and particularly distortion, I wonder if you could actually get it reliably in practice.
Hmm that's a good point, that you can use the ultrasonic bandwidth of the LP, which is useless for analog audio, to increase the bitrate. I'd say it's a bit generous to give an LP a bandwidth of 50 kHz, especially in the inner tracks though. I think what I was getting at would apply where the bandwidth of the analog signal you want to encode is as wide as the bandwidth of your media though. Take compact cassettes, let's say they have a bandwidth of 18 kHz and a dynamic range of 60 dB. If you want to encode an arbitrary audio signal with a bandwidth of 18 kHz and with 60 dB dynamic range, you could only do that in the usual analog way. But if you wanted to encode a digital telephone signal (say at 8 kHz bandwidth, 30 dB dynamic range) maybe you could use the 18 kHz cassette bandwidth to encode that signal digitally.
And if you could truly do analog manipulations of infinite precision real-valued quantities, you could beat a Turing machine. Analog precision eventually hits the so-called Heisenberg limit, below which further information does not exist.
You can indeed encode digital data onto a vinyl and then play it back meaningfully to a computer - http://boginjr.com/it/sw/dev/vinyl-boot/ is somebody making a vinyl boot disk and using it to boot DOS.
I'm vaguely aware of a band that had a track titled something like 8600 baud that was basically instructions to decode a secret message. Not quite what you wanted, but it proves the principle of digital-on-vinyl is possible. Another similar demo is a long forgotten digital VHS format, which wrote binary video data to an old fashion VHS tape. IIRC it had some outrageous capacity for number of hours of video it could store. So very doable though storage density and bandrate would need to be calculated.
Trivia: well into the 1970s, promotional copies of hit songs were released to U.S. radio and record stores on vinyl 7-inch 45 RPM format, one side in mono, and one side in stereo.
They work exactly the same way. All of the popular quadraphonic formats work modulating or encoding the four channels together into two signals which are cut just like a stereo record.
There were/are a great many schemes for achieving the multichannel encoding/decoding all with inherent advantages, disadvantages, and compatibility with legacy equipment. Dolby Pro Logic (and related successor tech) were probably the most successful multichannel tech that people will be familiar with, though that particular scheme came after the heyday of vinyl.
I don't think is widespread enough to be "a thing", but it sure does exist. Techmoan made a video on it, but, AFAIR, he had to use modern decoder, so I'm not sure how it's made.
I also think something important is missing from this explanation: the vertical/horizontal solution would not have been backwards compatible in two ways: (1) A mono needle would not have been able detect the vertical movement, resulting in a missing channel. (2) A stereo needle able to play such a record would not have been able to play a classic mono record, because the mono signal would have only been translated into a single channel (the left or right channel), and the other channel would have remained silent.
(1) would've made stereo records unattractive for customers who already owned an existing mono player, and (2) would've made stereo players unattractive to customers who already had a substantial mono record collection.
With the 45 degree solution, existing mono players were able to play stereo records (the horizontal movement is exactly the sum of the two channels), and stereo players were able to play existing mono records. For this to work, the left and right signals were recorded in opposite phase. A really elegant solution, which somehow reminds me of the cover of GEB [0].
PS: if you are interested in cutting-edge record groove technology, the "Füllschriftverfahren" invented by Eduard Rhein might interest you. It was an early compression technology for audio. The method was based on earlier work by the London-based Columbia Graphophone Company, but their work was never used in practice. Basically, before this invention, the spacing between the grooves on a record was fixed, with enough margins so that large amplitudes would not cut into neighboring grooves. Rhein build a machine that dynamically spaced the grooves based on the maximum local amplitude, allowing much smaller groove margins for quiet parts of an audio file, and therefore increased information density. This nearly doubled the running time of typical records.
Sadly, I only found an extensive description of this technology in German, including original patents [1]. But the figures are self-explaining.
[0] https://i.stack.imgur.com/OKBvZ.jpg
[1] https://grammophon-platten.de/page.php?530