A CLARIFICATION: My script is missing two important words in describing the stereo groove: "behaves like." When discussing stereo-difference signals, I declared that that's what's on the record - which is arguably true but only when you consider lateral and vertical stylus movements in isolation. The raw L&R channels are still encoded as purely diagonal groove movement, but the phase difference in the the cutting head's actuators ensures that those movements, when combined and equal to one another, only result in lateral motion. The cartridge also has its pickups wired with the same phase arrangement to ensure the L&R channels are in phase with one another when playing a mono record (or the virtual center of a stereo record). Is that a mono signal in the lateral motion with stereo-difference information in vertical motion? Well... you decide!
Note, the same idea was kinda used for video too. Instead of using RGB for transmitting color video, we used YPbPr to transmit the black and white backwards compatible Luma and the difference Chroma channels for blue and red.
This technique is still a thing in digital recordings. In a "joint stereo" MP3 encoding, the sound is encoded as a mono channel and a stereo difference channel. The reason is that it compresses more efficiently and results in smaller files for the same quality.
not the same quality, that's the point: you sacrifice stereo separation to more detailed "mono" components -- only problem is, the stereo difference is not nearly as musical as the mono part, but the encoding does not use significantly different method to encode it, so you will lose a LOT of stereo details (also, some encoders insist using heavy low-passing, at least below certain bitrate mangling the sound even more)
Can you still record two completely different mono track on each R an L channel. Say one song on the left and another in the right and hear them completely separated, like if you isolate and only listen to the L chanel?
huh i didn't know that's what joint stereo is. i thought joint stereo turns mono signal into a fully mono sound while still retaining the remaining stereo channels.
When you started discussing stereo difference channels using Audacity, it was very satisfying. I have a fun story about a radio station that did this by accident, causing me to make a curious detour on a job. Back in 06, I was at university, and also working for the university's media department. They wanted to play the campus radio station over the ceiling speakers in the lobby of the student union, so we put a receiver in a closet and connected the signal. Only -- the ceiling speakers weren't in any kind of stereo configuration; the whole array was driven off the same mono signal, so we had to pass the radio station through a little downmixer. At first I didn't have the part with me, so I just shoved one channel through the speakers to test them, and then I went and got the downmixer to install. When I did that, I found the audio almost totally went away. I checked the wiring and I hadn't messed up; everything was connected the right way. The level indicator wasn't zero, though, so I went from the closet into the lobby with the volume turned up and it sounded like I was underwater. The signal was there, but coming out mud. I tried a different source signal, and it came out fine (after I almost blew out the speakers from nearly forgetting to reset the volume levels). That's when I had the thought, "oh crap, the radio station, they're broadcasting their stereo channels 180 degrees out of phase." One of my roommates had a radio show and worked as a technician at the station so I told him about it. To prove what I was saying, I captured a segment of the station's live Internet stream, which I knew from my roommate was driven off the same mixer as the FM broadcast, and the software I used to capture it was Audacity. All I had to do was combine the left and right channels -- without having to invert one -- and play it back. "I think you've got something wired backwards," I said. We went into the station the next morning and fixed it. They were using analog audio equipment, not digital mixers, and at some point the polarity of one channel had been reversed in the feed. That is the first time in my life I got to fix something by reversing the polarity, and it did satisfy a good chunk of the childhood engineering dreams I had since watching ST:TNG at age 3.
when one of the channel is inverted (which could also happen when speakers are connected in reverse) the center is removed, so i thought i found a karaoke mode (doesn't work for all songs)
Something like that happened to me once. i replaced the cartridge on my turntable and and mistakenly reversed the wires to one of the channels. Took out the center channel until I figured it out.
I plugged in the headphones into the headphone jack, enough that sound will be loud and clear, but not all the way. I could be wrong, but when I did that, I think I may have been listening to the stereo difference channel.
Vinyl record trivia: Monty Python (back in the day) cut an album that had two sides on one side: Each 'side' was cut in one spiral grove right next to another spiral grove. Depending on how you set the needle down, you'd get one track or the other. They had no mention of this on the album notes, just left it for you to figure out. Of course, each track was half the length of a regular album side. Crazy!
I think it was the same album, certainly it was Monty Python, where they recorded into the end loop of the record, at the end of a sketch where a protection racketeer scratches the record. The last few words of the side are "sorry squire, I scratched the record, scratched the record, scratched the record, ... " which repeated ad infinitum until you got bored. The end loop was larger than usual so that the record deck did not realise the record had ended.
…which would have been another approach to stereo. Emory Cooke tried an outer and inner track for stereo with a forked tonearm with 2 mono pickups. It didn’t work that well and neither that nor concentric spirals like the Monty Python record (nor true vertical / lateral stereo) would have been backward and forward compatible.
Techmoan did a video about a "horse race" record; it had four different grooves on the one side, each groove containing a different commentary of a horse race with a different winner.
I remember Tower Records selling a not-as-expensive-as-you'd-think box of CDs of the complete Wagner's Ring of the Nibelung. It came shrinkwrapped with a simple control box with a knob to fade the output from the CD player completely left or right (placed between the CD player and the amp I presume). Instead of the 10 CDs the 15-hour epic opera cycle would normally take up, it only used a paltry 5 discs 😀
Fun fact: I often accidently heard only the stereo difference: When the headphone jack is not plugged in all the way, and both channels have contact but not the ground, you hear only the differences between the audio channels. Happened a lot in the good old days with mp3 players
You accidentally solved an one year long question I've had. I was building a bluetooth speaker. The audio board was stereo but I only had one speaker connected on the L channel. And even if I cut one of the two speaker cables, I STILL had sound! Bad quality, but I had sound! How was this even possible??? Well as it turns out, the low quality sound that passed to the speaker even with one of the two cables cut, was the stereo difference from the two channels. I started questioning everything I knew about physics and engineering but thankfully you answered my question, even after one year.
oh my gosh, i had discovered that myself back in 2018 i think!! i was using headphones with a detachable wire and i bumped onto that same trick by accident. my headphone wire was getting a bit busted tho, when i wiggled it, the audio i was hearing would intermittently switch from full stereo to the stereo difference version and back and forth. so i guess i discovered two ways to hear the stereo difference channel with headphones only! LMAO
Ive done this before by disconnecting wire on my headphones jack. I never quite understood why it works but it was cool to be able to hear hard-panned and buried tracks in a recording. Great video!
Here I am, retired and in my 70s with a lifetime (well almost) in electronics and it never occurred to me to question how stereo sound works on a vinyl record. Thanks for the ‘training session’.
A tip for anyone doing restoration of stereo audio off records. Because dust and crud sits in the bottom of the groove, it proportionally affects the stereo difference signal more than the mono signal. Convert from LR to MS then do your de-clicking etc. on each separately. Be careful not to use anything that might affect the phase relationship between the two signals. Then convert back to LR after processing.
The stylus tip never rides in the bottom of the groove. Styli have the tip rounded off, so only the sides of the stylus are in contact with the groove walls. The very bottom of the groove is typically not in contact with the stylus. One trick you can use with worn records is to use a stylus with different radius than the one that caused the wear, making the stylus ride higher or lower in the groove to find a portion of the groove wall that is unworn. This trick is regularly used by audio engineers transferring rare recordings from old 78s.
I've seen this advice for years, but every time I have tried this in Adobe Audition, it does not produce better results than just declicking normally. Perhaps its declick algorithm is already doing something like that anyway.
I grew up with LPs during the 80s. By then, fine audio equipment was available at a reasonable price. Speakers with fifteen inch kickers were very popular. Listening to the left and right audio was an experience. Even to this day, stereo is one of the greatest inventions of all time. Very interesting video, I enjoyed it. Thanks!
Stereo difference (along with some frequency filtering) is sometimes used to generate Karaoke tracks as the singer is typically placed in the center of the stereo stage while instruments are distributed. There are some Karaoke machine chips with this functionality built-in.
This is indeed true. Anything that is mixed mono will effectively cancel itself out and since most vocals are panned center/ mono they will be "extracted" from the mix when isolating the difference. This is exactly how all "surround" processing works from a stereo source. David Hafler while working for Dynaco came up with a passive analog circuit which could extract the "difference" channel from a stereo signal and then send that signal to dedicated "surround" speakers. Dolby pro-logic is the exact same process just done with digital circuits but functionally identical with the exception that with digital you can alter the timings between channels to compensate for speaker placement and distances.
I remember the sound editor that came with the packaged software included with our sound blaster 16 on our 486 could do that. Though most profession karaoke setups are actually worse than that, often being simple covers done by the production company, often with cut runtime, and sometimes even being performed on cheap synths, like someone's toy Casio keyboard. Lol. I'd imagine this reduces the licensing costs, as opposed to the original studio recording with the vocals filtered out as you suggest. I wish it were the case. Though I have on occasion heard some that seemed to be just that, where you can hear the faint remnants of the vocals if you listened hard enough. And even on a rare occasion, one that sounded like they actually got a hold of the masters and just mixed it without the vocals from the beginning. But sadly, by far, most professional karaoke discs seem to be the crappy covers by their in-house musicians.
A crude version of this can be acheived by connecting the speakers incorrectly. I'm sure there is an audio term for this. I don't remember if it involves connecting both positives to one speaker, or both negatives. Discovered this after accidentally wiring up some speakers the wrong way and suddenly revealed a song I liked had a distinct bit of lyric effect audible this way while nearly inaudible on the normal stereo version. It was quite fun.
That concept would also be used in FM radio broadcast. The combined channel is the main band while the difference channel is a side band. When the signal strength is weak the radio will swap to mono to retain quality.
Yup, and for basically the same reason - the technology existed in mono first, so transmitting the stereo as a difference channel in a sideband kept the transmissions compatible with older receivers. Come to think of it, it's the same thing they did with color television, just with a "color difference" channel instead of a stereo channel.
@@rjhelms Technically on US analog TV it is "color difference channels". The 3 primary colors are encoded as two signals in the sideband channel. The two signals in the sideband are encoded on the carrier in two different phases. Those colors you see are made with more adding and subtracting than if at first obvious.
@@rjhelms Actually, in NTSC TV two 90 degree separated difference signals carried the colour. G-Y and R-Y. Add G-Y to Y to get red, R-Y to Y to get green, then subtract red and green from Y to get blue.
I was taught for FM Stereo, they switch _very fast_ between the left and right channels and TX the stereo pilot signal at a high audio frequency that 90% can;t hear! And also if the pilot signal isn't present, the RX will automatically turn it into mono!
Once you broke out the model everything clicked. Getting to see what the actual effect of the technique was really helped cement how they work at the same time, and how that translates to reading the groove. I remember watching your first video and, while I came away with both tracks being stored at a diagonal to each other, sharing a groove, I couldn't tell you why. I feel coming away from this that I could actually explain why they are at a diagonal and that feels fantastic. The followup about the stereo difference was also really cool and, from a completely different perspective really helped me understand what the record is effectively encoding. "Two different tracks at 45 degree angle to eachother" is hard for me to understand how that translates to a record, but taking that extra step to explain how effectively it leads to a mono track being preserved in the lateral movement with stereo difference in the horizontal it really got it from a technical understanding to an intuitive one. Thanks for another great video!
The implementation of this is elegant AF in its beauty and simplicity. It ticks all the boxes. Maintains compatibility with mono records and turntables (both ways!), yet also performs really well without sacrificing quality/etc to do so.
What a great video, Alec! As a mastering engineer and former lacquer cutter, I think you did a fantastic job explaining the process. The construction paper model was a perfect visual aid! Bravo!
Fun stuff. I'm old enough to remember the coolness of the original stereo records. But honestly, what really WOW'ed me was "quadrophonic" records. That could also be an interesting follow-on to this...
Not to mention "Ortoperspecta", invented by Tapio Köykkä, a three-channel (Front-Center and L + R Rears) reproduction from a regular stereo signal. Center channel was the sum of L and R, while the rear channels reproduced the differential signal.
There were two different quadraphonic systems. There was SQ, a “matrix” system which didn’t really have four separate channels, and there was JVC’s CD-4, which used subcarriers to encode either the rear channels or front/rear difference signals, not sure which. This gave better channel separation. I remember reading a bunch of Sansui brochures at the time. They took SQ and turned it into “QS”, which some kind of dynamic level-based adjustments of the decoding matrix, to give the effect of better apparent channel separation.
Thelma Houston and Pressure Cooker was the first direct-to-disc record that was done without breaks. In other words, LIVE. The only break the band and singers got was the 30 seconds or so before the next song was supposed to start. Also, the break they took between sides of the LP. Well worth a listen!
I had direct to disc records. They were considered the best the format had to offer. CD essentially made them obsolete. Doug Sax of Sheffield Labs had a horrible case of sour grapes about this fact.
@@robertromero8692Doug Sax also at one point worked with James Guthrie on a remaster of Pink Floyd's album Meddle to both vinyl and CD. It's a good remaster.
I was actually listening to a record when I realized this is how it worked! The snares on Todd Rundgren’s A Wizard A True Star are intentionally put out of phase, but panned hard left and right. Listening back to the cd, it sounds like normal out of phase, and in mono it disappears. But on the vinyl, it gets a really weird, unique spatial quality to it as the needle fights with that huge Side channel! I love weird Mid-Side tricks like that, and in other music where listening to it in mono will ‘erase’ certain vocal lines
BTW, fantastic job on describing the audio in the captions! Many people never seem to get that captions are more than just what words are being said...
Loving the inverted credit music haha! Phase cancellation is essentially what we’re listening to. The difference between the sounds that isn’t cancelled. Love the video!
this difference in channels is what hafler used to produce a pseudo quadrophonic sound simply wiring a speaker or two across the positive terminals of the amp output. i have been using this for decades to add depth to the stereo image.
15:46 Complete channel separation on vinyl, utilizing a single needle is what I struggled with until seeing this video. Not only is that clear to me now but so is the misconception that "up & down" is absolutely one channel while "side to side" is absolutely the other. Thank you for this simplified, construction paper explanation. 👊🏻 🇺🇸
Nice touch with the stereo difference version of the closing music! 🤓 Back in the dark ages (‘70s) there was a device called the Thompson Vocal Eliminator which employed this very technique. Since the lead vocal was/is usually panned center, shifting the phase of one channel by 90° would cancel out the vocal, leaving only any reverb triggered by said vocal. I never actually operated one of those, so I don’t know if anything was done to allow the low-frequency sounds of kick drum and bass (also usually panned center) to come through unscathed.
In the 70s when you got bored with a record you lifted the tone arm up to find the little colored wires plugged in the cartridge and switched a pair of connections. Then you got a completely different mix, usually minus the vocals and bass
Yes. They had a variable bandpass filter that you could dial in to only modify only the vocal range...in theory. My aunt had one, and in practice it was pretty much crap. But I suppose it got the job done enough.
I used this process many times in my teen years when transcribing music in my DAW software and reorchestrating it with midi synths. The center cancellation helped to reveal some sounds that were hard to hear in the normal mix. But, yes, it would usually cancel bass too, so EQ’ing was involved before the cancellation process. I must add that it’s not a phase shift of 90°, it’s a polarity inversion of one channel of the musical content, similar in concept to a 180° phase shift of a sine wave added to the original sine wave - cancellation. Common-mode rejection, used in balanced signals like pro microphone signals to cancel common-mode interference that might leak into a long run. Balanced or differential signaling using this noise-cancelling property is also used in HDMI, Ethernet, and other well-known signals.
It's worth mentioning that vinyl stereo is in a way a bit more limited than stereo on another medium like tape or CD. For example, there's a limit to how loud you can have a panned sound. If you have a really loud bass drum hit in just one of the channels, you risk launching the stylus out of the groove. One of the things I was told when it comes to dance music production is that, for vinyl mastering, you absolutely must not pan your kick drum or bassline. CD doesn't have this restriction.
CDs do have a pretty similar restriction (it is called clipping), but it will only result in mangled sound, not jumping forward or backward on the record so its perceived less severe (or by most not noticed at all if it happens).. and this is also one of the reasons Vinyls are considered superior by some people: you cannot mess with the audio too much (increase the volume), or the stylus will jump. on a CD you can turn up the Volume to 11 (as louder is better /s), and all you get is mangled/clipped sound at the peaks. The entire range of a CD is far bigger than a vinyl, but when you restrict yourself to the top quarter of that range during mastering it doesnt help..
That is fascinating I'd never considered that but it makes a lot of sense. We are spoiled these days in many ways and sometimes we don't even realise it!
That's fascinating. So essentially it may be necessary to make completely separate masters of a track depending on the release medium. Which also means that some versions of some tracks can never be released on a certain medium. And in the case of vinyl, due to mechanical limitations no less.
@@unitrader403 The main reason for "clipping" with Audio CDs is simply that when the tracks are mastered, the volume is always increased to make the track louder, given many CDs would be played in noisy-areas, such as in a car, or on a portable-player. If you rip a track from a CD, and then lower the volume in an editing app, you can eliminate most of this
As someone who is deaf in their left ear, this has actually made me interested in getting a mono stylus. I listen to almost everything in mono so that I don't miss anything, and it'd sure be convenient to only need one speaker, rather than two - especially if I have a smaller space to work with in the future. I have no idea whether it'd actually work the way you'd describe with modern records, but it'd be interesting to find out.
Mono and Difference channels, or "Mid / Side", are the key to so many things we use every day -- most of the time, because of backward compatibility with older standards, or where we just want mono compatibility. FM Stereo and analog TV audio both work this way (similar to TV's Y / Cr / Cb color encoding.) Some digital audio encoding formats still use this as an alternative to L/R encoding, for various reasons. MPEG, for example, and probably many of the other perceptual encoding schemes, due to compression algorithms having a particular affinity for correlation. I sometimes track stereo audio (like two guitar or vocal tracks) as a mid and side track, then convert them to L/R using polarity inversion. Again, that helps to ensure the result will work if folded down to mono, but it can also sound interesting (sometimes better, sometimes worse) when there are differences in the takes. This two-channel phase relationship is also one of the core principles to how Dolby Pro-Logic Surround works -- the center channel is from the Mid, the rear channel is from Side, and L/R vs. M/S are steered using some fancy analysis of the signal to determine which direction (front/back vs left/right) is more prominent.
I love this channel because it makes me want to learn about things that I never thought I could have possibly cared about before the video was posted. Thanks for all the great content!
For some reason I am always more in awe of materiel methods of creating magic than electronic ones. The groove in a record has always struck me as remarkable, not only in that it works at all but how exactly it achieves it. Too few people are aware that it is a 90 degree difference in the two channels, and that the groove isn't shaped left and right for the two channels. I admit I thought it was that for the first 64 of my 65 years on the planet.
You thought that in your very first years? I don't remember having any thoughts about how stereo sound was reproduced in a single groove of vinyl until I my third or fourth year at least.
@@reshpeck I have to admit that my father was a Woolworth's record salesman, and all the talk in our house was of stereo separation and cross talk. It actually led to my parents divorcing, although they still spoke over each other.
I think it's in part because mechanical systems required working out everything from end to end and that to make it work you need to think of everything. With electronics and computers in particular, you are dealing with a lot of layers of abstractions in which each layer is relatively easy and changeable. This means the overall system is far more complex, but the bit you need to think about is much more simple. For example, if you are working on signal processing, you don't need to worry about how that signal is produced or even the quantum effect involved in processing that signal. You can simply assume that the layers of abstraction you are using will work as documented.
You can pick literally any kind of subject, and talk about it in so much detail… that is the best explanation about stereo records I’ve ever seen in my life!
Fun fact about the mono and “stereo difference” signals (better known as mid and side): There is a drum micing technique built around this principle! If you take a cardioid mic, and put a bi-directional mic right next to it, but pointed perpendicular from the sound source, you can get stereo drums that collapse perfectly to mono. Simply add the two signals together in one channel, then subtract the bi-directional signal from the cardioid in the other, and you have stereo drums!
You can generalise the technique. Split the cardioid into an omnidirectional and a bidirectional pointed forward and add another bidirectional microphone pointing up. Your resulting set of four signals can represent a full sphere of sound. It's known as first-order Ambisonics. RUclips uses it for the audio component of 360-degree video.
There’s nothing inherent to drums about this. I use mid-side to record piano, and even orchestra. I like it because you can adjust the stereo sound in post.
Very common for video and broadcast from the beginning of stereo t.v. to maintain compatibility with mono tellys. Still used in many budget video mics, but the mix matrix is on the mic so you get L&R out rather than M&S. Some are switchable while others are not.
@@alistercarmichael4990 Still something of a standard for TV and radio broadcasts of orchestral performances. Often with a very very nice Schoeps M/S rig.
You do lose something in mono: the 'side' or difference channel, which is L - R. But you keep the 'mid' or sum channel, which is L + R, giving you a mono downmix: much better than just the L or R channel in isolation.
15:15 that sounds like the “poor mans surround sound” setup. I had a 2nd pair of speakers hooked up from the two positive leads from left and right on my stereo when I was a kid and it gave this interesting surround sound effect from the rear speakers. That reminded me of my setup back in the day.
This was great. It's a subject that many don't ever get to fully understand, and you covered it very well. I retired in late 2020, but one of the last products that I worked on as a product developer was a modern direct drive turntable.
I've been subscribed to your channel for about three years, and I continue to be amazed at how well you explain difficult concepts. This time, you have outdone yourself. I am an audio tech with some 46 years of experience, and now that I'm old, I specialize in repairing vintage audio gear. That said, I believe yours is the best neophyte-accessible explanation of stereophonic record theory I have seen. I was thinking of making a video on my own channel to cover this, but now I think I will just refer people to your video. Fine job, sir.
This concept is actually extremely useful in audio processing too! It’s very common to EQ "mid" and "side" (the equivalent of the left/right and up/down motions) separately, and it’s sometimes useful to do even wackier types of processing to either just the mono or just the "side" signals. If you ever see "stereo difference", this is what that’s doing: separating the stereo L/R signal into a mono signal plus the stereo difference.
The important things about the encoding used to press stereo LPs is that the sideways movement has the highest bandwidth so it makes perfect sense to use that for the average signal between the channels. The same idea is used in digital audio compression, for example MP3 files support "joint stereo" encoding where one of the channels is encoded as delta to the main channel. Because real world stereo music typically has very similar audio on both channels, encoding one channel as difference to the another saves a lot of bandwidth.
14:12 - You can actually generate a "stereo difference" channel with nothing more than a stereo 3.5mm headphone jack. You just have to slowly pull the plug out little by little and once it gets to the right spot, you'll only be hearing the difference between the left and right channels. It's a quick and dirty way to remove the "center" channel.
There's a lot of interesting physical considerations for vinyl mixing/mastering. The bass frequencies have to be center-panned as the low frequencies create such big grooves that they can cause the needle to jump out if they're side-panned. Records undergo RIAA equalization where they're mastered with attenuated low-end freqs so the grooves can be physically smaller and denser on the record, and the attenuation is reversed on playback.
Thanks! I didn't know about that attenuation trick. It's amazing how many clever ideas are used in this sort of technology. I'm also now wondering whether the record sent into space on the Voyager probe was a straightforward mono recording without any clever tricks. I'd assume so, or the aliens might have trouble reproducing the sound properly...
There's something similar done in analog FM broadcasting. In the audio processor or stereo generator, a pre-emphasis curve (with either 50 or 75 microsecond time constant, depending on where you are in the world) is applied to the audio to improve the signal-to-noise ratio on higher frequencies. Then FM receivers apply the inverse of that curve, and it sounds normal.
@@danpetitpas Really? I can't find any mention of that. Both NASA and Wikipedia say they were "phonograph records" but they don't call them "videodiscs", although I know there are images encoded on them. I'm a bit confused now - do you have a link to any further info please?
it’s truly uncanny the way this channel reads my mind. i was literally wondering about this last week. the same thing happened with the percolator video. and the one about the klaxon horns of old school cars. and also when i was wondering about how pulse oximeters work. always within like a week of when i’m pondering the question. truly strange and wonderful
Seeing as how it probably takes at least a week to write, shoot and edit each video, he was probably steeped in thought about the topic when you began thinking about it, so maybe instead of him reading your mind, you were reading his 😀
If you've learned some basic college level linear algebra, this is very easy to understand. Horizontal is the sum of the channels, vertical is the difference. It's a classic diagonal basis transformation
Thanks for giving credit to Alan Dower Blumlein, who did invent the 45-45 method of recording stereo sound to disc. In the late 1950s Western Electric/ Bell Labs tried to claim that they did, but was thwarted when a copy of Blumlein's patent was shown to their representitive. Many many other inventions stil in use in 2024 are attributable to Mr. Blumlein.
This is one of those things I've occasionally found myself wondering about ever since your original series on the subject. Very cool to see it revisited in more detail. Great vid! As a side note, while I'd be the last person to defend the sound quality of vinyl (other than to say "eh, it's not so bad", because I'm just not that picky about audio quality), I do find something charming about the experience of listening to a record. The physical ritual of opening the cover, taking out the record, holding it carefully, placing it on the turntable, starting it playing, flipping it over, etc. puts a level of focus and attention on the music that more modern and convenient methods don't require. I could never be bothered to sit down and listen to an entire album from start to finish without doing anything else before getting a record player. It's kinda meditative. It's not something I do all the time, but it's been really neat to connect with music in a new way... well, I guess it's an old way, but my record player does output to Bluetooth, just to make sure something about my setup annoys absolutely everyone. >=)
By strange coincidence, a few days ago I had a conversation with a chatty old guy while waiting for my takeaway food order. He started off commenting how people can't live without their phones these days (I happened to be checking the news on mine) and went on to how he still takes pictures using black and white film and listens to his LP collection on his old stereogram. Although I've always been into new tech (early internet adopter, dad ran a mainframe computer in the 1970s) I could actually relate with him since my grandad was a rubbish collector and had sheds full of old cameras and audio equipment (my (1970s) amplifier still uses homemade surround speakers I built from salvaged stereogram speakers) and I have a collection on 60s and 70s LPs from myself and my dad. All was going well until he got onto the old cliche about vinyl just sounding so much better and warmer... I nodded and smiled through gritted teeth. Ironically, if he actually embraced the internet instead of being so dismissive, this channel would be SOOOOO up his street!
Unless the record player you got was super expensive I'd be careful with what you play on it. Those cheap ones that come with "All the modern features" like Bluetooth built in usually have an extremely heavy tone arm which causes the stylus to eat through vinyl. What I've seen people recommend as a decent way to not kill vinyls on a budget is to get a used mid to high end direct drive record player from the 80s/90s with a built in phono preamp and plug a Bluetooth adapter into the headphone out. You'll pay a bit more than the cheap record players, but you'll get something that works well, doesn't eat your vinyls, should last for over a decade and you can always upgrade the Bluetooth to the latest standard if some super high fidelity Bluetooth audio standard comes out, not that it should matter that much with vinyl anyway.
I think its time to do a video about (for most of us) our first record player..... The Fisher Price one with the pins! April 1st is coming up after all.
By the way, when digital recording was in the early days, you had to go out of your way to find a vinyl that had been itself recorded digitally. The analog recordings were the standard, while digital had a special label. Nowadays, though, things are flipped with analog being the rare version with the special label, while digital is the standard format without the label. The difference is key because while there are minimal differences between the two types of recordings, analog actually does tend to pick up more of the full audio spectrum, including subharmonic frequencies in the bass region, whereas digital recordings tend to leave that out due to the ear's inability to hear it... but of course you also have to have a really quality-built audio system to *feel* those frequencies in the first place... digital or analog, most vinyl players and speakers connected to them also won't play those lower frequencies either.
Somehow I randomly clicked on CC. I knew you'd added Closed Captioning support but I haven't watched/read it before. I enjoyed the secondary commentary lol!!!
Loved this one. The Aquarius and the 5th Dimension record you had on the table is actually the first record I owned, given to me by my mom. Let the Sun Shine In is my jam 🤩
No matter how many videos I watch on how records work they will never make sense to me and always seem like magic. It just seems like magic that they can easily capture and reproduce unique music and voices.
I have not seen this channel in a while. Glad your back on. I was surprised you did not mention quadraphonic recording. Keep up the great channel. Thanks
I remember wondering how they did this trick when I was young. It blew my mind that they could fit two signals in 1 track. Thanks for making this video!
It really is crazy cool how the record & player came about. The arm itself is also really cool, with the way it rests on the record allows it to play warped records too.
I feel like you'd be a teriffic Intro to Linear Algebra teacher! You've basically described what a change of basis is in a way that is super accessible and tangible.
12:15 This is a very interesting image. I read long ago that you can put more runtime on a vinyl record if you squeeze the grooves closer together in silent parts of the audio. As I have never seen this so far, my imagination of how effective this could be was very limited. Now this image shows me that the grooves are like 3 times as close in those areas, which is quite unexpected for me. Also, I like how you point out the mid-side encoding scheme and its incredible advantages over other solutions.
techmoan has a video about the trimicron record. It boasts 2 hours of music per disc, but suffers from decreased dynamics due to squeezing all of the space out of the disc.
Record lathes are designed to optimize the surface area of a record. A preview head (if cutting analog) or a computer will preview the coming audio signal at least a revolution in advance, adjusting the groove spacing accordingly. Not only does it move quieter grooves close together, it spreads out loud grooves so they don’t intersect each other. If the grooves intersect, then you get a skip or a repeat on playback. A cut lacquer is often inspected with a microscope to ensure no two grooves intersect.
@@xureality You get more grooves on a record by lowering the overall volume level of the recording. The optimal recording length per LP side is 12-18 minutes. If you go beyond that, then you need to start making compromises… either by lowering the overall volume the recording or by reducing the dynamic range of the recording. If you want to hear how great vinyl can sound, get a 12-inch 45rpm disc with one song per side (like a DJ single). The music can be cut loud with lots of dynamics.
@@xureality Also, IIRC, Techmoan found that those records with their skinnier grooves tended to be harder to keep clean as any dirt would stick in the tiny grooves.
The other effect of cutting the grooves too close together is that you get echoing from the grooves on either side due to the pressing process distorting the vinyl. You may be able to get echoing from the master cutting process too? Crank up the volume and you'll be able to hear that on pretty much any record where they've tried to get a lot of music on each side.
Very good description! The left/right difference signal is used for the same backwards compatibility reasons for FM stereo broadcasting. It also has the benefit that we can choose the bandwidth for the difference sub-channel - effectively controlling the amount of separation required
The topic of this video bears relations to Mid/Side recordings, which is when two mics are used to record stereo audio by recording the 'centre' with a cardioid mic and the 'sides' with a figure-of-eight mic facing sideways. You then play both mics on both channels but invert the polarity of the side mic on one side, magically creating real stereo! Also, nice touch on the 'L/R difference' version of the outro tune 😎
Not a real stereo, there is no perfect way to capture it. But it is always mono compatible with a strong mid/side separation. Often much better than a stereo mic. Ideally should be real stereo, but in reality.. it never is.
You can also make it more general. Split the cardioid into an omnidirectional and a figure-8, then add another figure-8 pointing up, and you've got first-order Ambisonics. The four channels of first-order Ambisonics are equivalent to an omnidirectional, a figure-8 pointing left, a figure-8 pointing up, and a figure-8 pointing forwards. With them, you can encode full-sphere surround sound. Decoding it requires a minimum of six speakers, though eight at the corners of a cube would be easier to position around a room. Headphone playback can be done via head-related transfer functions. The spatial resolution isn't very high for first order, but the format can be extended to higher orders with more channels, though the additional ones no longer correspond to any physical microphones. I've heard of audio software plugins going up to third order (16 channels) and games going up to fourth (25 channels), and seen tables of the relevant encoding equations online going up to at least sixth order (which would have 49 channels).
Literally since I was like 15 and getting into audio recording, music production etc this question has been sitting in my brain. Even now 10 years later this question would still nerd snipe everytime I thought about it while playing vinyl. I think I even googled it a few times but just didn't really understand the concepts just from written answers. But THIS video has finally given me a satisfying answer. Thanks mate!
I've always known mastering for vinyl was a more specific process, but after seeing how stereo sound works in a record player, it now makes sense! Awesome video! (Also big fan of the channel in general, this one was especially interesting to me as a hobbyist music producer)
Another interesting aspect of stereo sound is how sound mixing of music has changed over the years in response to changing technologies. Before the introduction of the walkman, music was almost always listened to via a speaker system. Because of this, some music was mixed with a great deal more stereo separation than is common today. Many tracks by The Beatles ("Run For Your Life" springs to mind) had complete (or nearly so) separation of some recording channels. This is fine if you're listening to the music via a stereo speaker setup, because even though some sounds (a guitar or a voice for instance) are only coming from one of the speakers, both of your ears can hear it, just at different intensities, which is how we perceive which direction a sound is coming from in the first place. When listening to these tracks using headphones however, it sounds extremely strange. The sounds that are completely isolated to one of the channels only reaches one of your ears at all, which runs counter to how our brains are used to interpreting sound, so it doesn't quite know what to make of it. The same goes for things like music from the Commodore Amiga. The Amiga had 4 sound channels, two for each stereo channel. This was not generally a problem, as the sound was generally played back through a TV or a speaker system. But listening to original stereo recordings from the Amiga using headphones today gives you two sound channels in one ear and two completely different ones in the other, which makes for an extremely odd listening experience.
On headphones, it basically sounds like something is directly next to your ear making that noise…which is more or less what’s happening. Really annoying when it’s used for two people talking.
Most MOD players outside of the Amiga have a setting to reduce the stereo separation for headphone playback to help deal with that problem. 50% seems to be where I typically end up setting it. They often also have a "surround sound" mode as well, which while the "surround" part is debatable, does make for a more comfortable listening experience.
George Martin did this on the Beatles recordings because he wanted the stereo version to sound the same as the mono version when played back on a mono record player. He was concerned about a 20 db cut in volume because of phasing when a sound was mixed equally on both the left and right channels. Very few people in Britain had stereo equipment when the Beatles started out.
They did use a lot of reverb on the earlier Beatles albums, so it wasn’t too bad. Most tracks had all the vocals on the right, but echoed in the left channel. There wasn’t complete separation of left and right.
the idea of a mono track with a second track describing the difference is called "mid-side" and is used in mastering, mixing, and composing to get better control of the stereo soundstage. mid is just (L+R)/2 and side is L-R. (or sometimes the x2 volume adjustment in on the side instead)
Pretty simple to implement in software, too. M is just the signal from the source and S is the signal weighted by the sine of the azimuth angle to it (or the x component of the direction vector to it). Use that for stereo sound in your game, and you've also got a platform for going to first-order Ambisonics when you make the leap to surround sound.
Yes, like you said in your clarification, the left and right channels are encoded to move the styalus at a 45 degree angle up and down for one channel, and at a 45 degree angle in the other direction up and down for the other channel. At the very same time this encoding can be done a different way but end up with the exact same groves. When skewing the left and right channels on diagonals, this "rotation" behaves like horizontal movement of stylist as being encoded as left + right, and vertical movement of the styalus being left - (minus) right. This makes sense because there is less information in the left - right (difference) signal. This has to be a small signal else the styalis would jump up and out of the groove if that signal was strong. Since the left to right movement of the styalis can tolerate wider movement, it works out well to contain the stronger "sum" of the signals. Either way the groves end up being the same. So some photographs use up/down left/right cartridges and must add and subtract these two signals in the amplifier to get left and right. Other photographs use 45 degreed angled cartridges and thus don't do any addition or subtraction to get the left right channels. If you look closely at a groove that is silent, say, on the left channel and bass on the right, you would see one side of the groove smooth and the other side of the groove waving. This causes the styalis move up and down in only one diagonal for the right in this instance. That diagonal up and down is also a combination of up and down and left and right at the same time, so the result is the same either way: Angled: left 45 angle silent, right 45 angle bass.... or... ((Left and Right: (silent left + Bass right)) PLUS (Up and Down: (silent left - Bass right))) = left silent. And ((silent left + Bass right) MINUS (silent left - Bass right)) = right Bass. Also, on vinyl, the bass is reduced to create smaller bass waves in the groove and the amplifer boosts the bass to get it back (before the tone control adjustment).
Год назад+6
The stereo difference signal version of the outro music puts the cherry on the top.
Thank you for this topic today, it brought a few memories back from my childhood. One of them is from when my father bought an LP cheap (from an antiquary store or similar) and let me learn how sound can be produced by a sowing needle pushed through a piece of paper, resting the needle in the groove of the LP so the sound would resonate from the paper. I think I was at the age of 10-12 at the time. Cool to learn how a simple needle and a paper can produce sound from the groove of a record, even though the needle did deteriorate the record quite a bit during the experience. Oh well, it was a cheap record containing a collection of songs Dean Martin used to sing. Those songs still give me fuzzy vibes of remembering my childhood.
When I was a kid I thought the needle was split down the middle and one side was reading the left wall of the groove and the other was reading the right wall. Audiophiles talking about Vinyl sounding better...it doesn't. But I like Dankpods' take on it: If you're into a hobby, you want to fiddle with it. You can't fiddle with Spotify, you click play and it goes. You fiddle with a phonograph, moving the discs around, changing the speed of a turntable, enjoying all the album art that comes with that big record sleeve. Playing music on vinyl is for participating in the music enjoying experience.
>You fiddle with a phonograph, moving the discs around, changing the speed of a turntable, You will *_never_* get a better sound by doing any of these. The way you "fiddle" with vinyl to get a better sound is "buying more expensive equipment" - the real fiddling is convincing the bank to give you an even larger line of credit (while preventing your wife from ever finding out). >enjoying all the album art that comes with that big record sleeve. Hi-res hi-quality natively-digital art can be just as enjoyable as the big record sleeve, though! Bandcamp has some great covers.
This is such a great explanation of a thing I've been curious about for a long time. I'm a recording engineer, and the way the stylus uses a mono signal combined with a stereo-difference signal to create a stereo image is a lot like the mid/side stereo microphone technique involving one cardioid mic facing the sound source and one figure-eight mic 90° off axis from the sound source. In order to hear the recording played back in stereo you just need to combine the mono track from the mid mic (the cardioid one) with the track from the side mic (the figure-eight one) panned hard to one side and an inverted copy of that same side-mic track panned hard to the other side. Without the mid track they cancel out into silence, but with the mid track there you end up with a very nicely recorded stereo sound
something i love doing in my free time is taking music and just listening to the side information, i sometimes hear things i've never heard before in the mix
I’ll throw you a curve. Look into quadraphonic vinyl LPs. Played with the infamous “Shibata” stylus! Quite interesting and actually produced exceptional sound. However, not 100% analog of course, but still very advanced for the era in which they were introduced. Great video!
The quad recordings used a 30,000 cycle carrier frequency to modulate the rear channels. The problem was the best quad stylus in the world couldn't keep up with the groove and would wipe out the record in about five plays, leaving only the front two tracks.
@@gregrobsn Yeah, I would play my quad LP when I first got it to make sure there were no defects and the second time I played it I would record it to reel to reel. After that I’d file the LP into my collection. Most of my quad LPs only had two plays on them. I run it through a Kenwood KR9940 receiver. I wish I still had that setup.
I did the same and recorded on a 4 channel Akia. I lost most of my equipment during the '94 Northridge earthquake. I was living about 2 miles from the epicenter and the only room that sustained damage was the theater room.
The trouble with flat records which is often overlooked or ignored by analog fetishists is the inner groove distortion caused by the decrease in resolution as the spiral groove gets smaller and smaller and the stylus moves slower compared to when it is playing the outer edge of the disc. A cylindrical format would overcome this limitation with the same audio quality at the beginning and end of the recording
Yup. The inner track has about 66% less bandwidth than the outer track. I think that's why Edison went with a cylinder. But phonograph records were more convenient.
The business with the mono-compatibility of stereo records mastered that way is remarkably similar to how "mid side" recording works, where instead of a Left microphone and a Right microphone, you'll record with a Mid microphone and a Side microphone. This does mean you need to convert the signals into L and R channels for playback, but it does let you do some interesting things. If *nothing* else, a Mid Side recording can be used like a mono or stereo recording, and you can vary the "width of the stereo image" by simply turning the Mid channel volume up or down. But there's *loads* more that you can do with "mid-side processing" like putting different effects on the two channels.
Blumlein was doing other work in Mid/Side stereo, so it makes sense that he'd go for this (incredibly clever and sensible) solution. (also, love the outro easter egg ;) )
playing stereo records on older mono cartridges can be a bad idea as the cartridge mechanics are very 'stiff' vertically, causing wear on the stylus and groove, later mono cartridges were 'stereo compatible', although still mono output ...they'd 'take' the vertical movement, not try to resist it mechanically, just ignore it
That basically caused by the wave length. A lower tone sound have longer wave length. Which makes matching the wave between two channel easier. In the other hand, high frequency sounds have short wave length and hard to cancel out. A little off phase and it's not cancelling but adding up. That is also why ANC headphones hardly do anything to cancel human voice but works perfectly on car noise.
Very well explained. Quick anecdote: Back before multi channel surround encoding and modern AVRs became available, in the 80s some of us hifi enthusiasts would setup "poor man's surround sound" by connecting a second set of speakers (placed in the rear of room) to a 2-channel amp/receiver to only play the stereo difference signal, which would result in surprisingly effective surround sound from a basic stereo source. Give it a try. For the wiring diagram, just google 'poor man's surround sound wiring diagram'.
8:52 - Seems like that could be used intentionally to encode secret data. Say in a spy novel there’s a music producer who is also smuggling information around the world. The McGuffin record sounds like regular music in one way, but a special player running over the groove at a different angle could play another message. The information disks would look just like other records of the same artist, but with a small difference to the label that’s easy to miss unless you know what you’re looking for. Maybe the spy record gets mixed up with others and sent somewhere it wasn’t intended to be, and then story happens.
I remember such a story in a Mickey Mouse comic, but that was some sort of secret groove, only audible with a special stylus and with no deeper explanation how it worked. But yes, the special record was mixed into a batch of normal ones which were then sold. Secret information in the same groove probably would have worked with mono only records, with the secret message encoded in groove depth. I don't really see it working with a stereo record on a stereo player unless you do something more involved like cutting a hidden, very narrow depth-only groove at the base of the normal one, inaccessible to the normal stylus.
@@tz8785 How about this: in the very bottom of the groove, below where the stylus runs, etch digital pits and lands to be read with a laser (encoded like a CD). Tough to detect unless you already know the record has hidden information.
@@matteocristini6221 Something with Claw of Kali sounds quite possible, although I don't recall the exact title and I didn't read it in Italian in the first place.
A CLARIFICATION:
My script is missing two important words in describing the stereo groove: "behaves like." When discussing stereo-difference signals, I declared that that's what's on the record - which is arguably true but only when you consider lateral and vertical stylus movements in isolation. The raw L&R channels are still encoded as purely diagonal groove movement, but the phase difference in the the cutting head's actuators ensures that those movements, when combined and equal to one another, only result in lateral motion. The cartridge also has its pickups wired with the same phase arrangement to ensure the L&R channels are in phase with one another when playing a mono record (or the virtual center of a stereo record).
Is that a mono signal in the lateral motion with stereo-difference information in vertical motion? Well... you decide!
Thanks. I was wondering about the phases since they didn't look right to me in your paper mode.
FYI - Your Closed Captions are one of my favorite parts of your videos!
Note, the same idea was kinda used for video too. Instead of using RGB for transmitting color video, we used YPbPr to transmit the black and white backwards compatible Luma and the difference Chroma channels for blue and red.
Thank you. I thought, either I have a huge hole in my decades-long audiophile knowledge, or . . . Great presentation regardless, as usual.
Also, a cylinder Phonograph has a helical groove, not a spiral one...
This technique is still a thing in digital recordings. In a "joint stereo" MP3 encoding, the sound is encoded as a mono channel and a stereo difference channel. The reason is that it compresses more efficiently and results in smaller files for the same quality.
The most common terms for that are "mid" and "side" channels (and the transform called a "mid-side transform")
not the same quality, that's the point: you sacrifice stereo separation to more detailed "mono" components -- only problem is, the stereo difference is not nearly as musical as the mono part, but the encoding does not use significantly different method to encode it, so you will lose a LOT of stereo details (also, some encoders insist using heavy low-passing, at least below certain bitrate mangling the sound even more)
Can you still record two completely different mono track on each R an L channel. Say one song on the left and another in the right and hear them completely separated, like if you isolate and only listen to the L chanel?
This is how FM radio broadcasts in stereo as well. The baseband signal in order is L+R, pilot tone, L-R, RDS encoding, then supplemental services.
huh i didn't know that's what joint stereo is. i thought joint stereo turns mono signal into a fully mono sound while still retaining the remaining stereo channels.
When you started discussing stereo difference channels using Audacity, it was very satisfying. I have a fun story about a radio station that did this by accident, causing me to make a curious detour on a job.
Back in 06, I was at university, and also working for the university's media department. They wanted to play the campus radio station over the ceiling speakers in the lobby of the student union, so we put a receiver in a closet and connected the signal. Only -- the ceiling speakers weren't in any kind of stereo configuration; the whole array was driven off the same mono signal, so we had to pass the radio station through a little downmixer. At first I didn't have the part with me, so I just shoved one channel through the speakers to test them, and then I went and got the downmixer to install.
When I did that, I found the audio almost totally went away.
I checked the wiring and I hadn't messed up; everything was connected the right way. The level indicator wasn't zero, though, so I went from the closet into the lobby with the volume turned up and it sounded like I was underwater. The signal was there, but coming out mud. I tried a different source signal, and it came out fine (after I almost blew out the speakers from nearly forgetting to reset the volume levels). That's when I had the thought, "oh crap, the radio station, they're broadcasting their stereo channels 180 degrees out of phase."
One of my roommates had a radio show and worked as a technician at the station so I told him about it. To prove what I was saying, I captured a segment of the station's live Internet stream, which I knew from my roommate was driven off the same mixer as the FM broadcast, and the software I used to capture it was Audacity. All I had to do was combine the left and right channels -- without having to invert one -- and play it back.
"I think you've got something wired backwards," I said. We went into the station the next morning and fixed it. They were using analog audio equipment, not digital mixers, and at some point the polarity of one channel had been reversed in the feed.
That is the first time in my life I got to fix something by reversing the polarity, and it did satisfy a good chunk of the childhood engineering dreams I had since watching ST:TNG at age 3.
when one of the channel is inverted (which could also happen when speakers are connected in reverse) the center is removed, so i thought i found a karaoke mode (doesn't work for all songs)
Doctor Who would be proud
And you didn't even have to route it through the main deflector dish. 😜
Something like that happened to me once. i replaced the cartridge on my turntable and and mistakenly reversed the wires to one of the channels. Took out the center channel until I figured it out.
I plugged in the headphones into the headphone jack, enough that sound will be loud and clear, but not all the way. I could be wrong, but when I did that, I think I may have been listening to the stereo difference channel.
Vinyl record trivia: Monty Python (back in the day) cut an album that had two sides on one side: Each 'side' was cut in one spiral grove right next to another spiral grove. Depending on how you set the needle down, you'd get one track or the other. They had no mention of this on the album notes, just left it for you to figure out. Of course, each track was half the length of a regular album side. Crazy!
I think it was the same album, certainly it was Monty Python, where they recorded into the end loop of the record, at the end of a sketch where a protection racketeer scratches the record. The last few words of the side are "sorry squire, I scratched the record, scratched the record, scratched the record, ... " which repeated ad infinitum until you got bored.
The end loop was larger than usual so that the record deck did not realise the record had ended.
…which would have been another approach to stereo. Emory Cooke tried an outer and inner track for stereo with a forked tonearm with 2 mono pickups.
It didn’t work that well and neither that nor concentric spirals like the Monty Python record (nor true vertical / lateral stereo) would have been backward and forward compatible.
Techmoan did a video about a "horse race" record; it had four different grooves on the one side, each groove containing a different commentary of a horse race with a different winner.
Gro🇴ve.
I remember Tower Records selling a not-as-expensive-as-you'd-think box of CDs of the complete Wagner's Ring of the Nibelung. It came shrinkwrapped with a simple control box with a knob to fade the output from the CD player completely left or right (placed between the CD player and the amp I presume). Instead of the 10 CDs the 15-hour epic opera cycle would normally take up, it only used a paltry 5 discs 😀
Fun fact: I often accidently heard only the stereo difference: When the headphone jack is not plugged in all the way, and both channels have contact but not the ground, you hear only the differences between the audio channels. Happened a lot in the good old days with mp3 players
You accidentally solved an one year long question I've had. I was building a bluetooth speaker. The audio board was stereo but I only had one speaker connected on the L channel. And even if I cut one of the two speaker cables, I STILL had sound! Bad quality, but I had sound! How was this even possible???
Well as it turns out, the low quality sound that passed to the speaker even with one of the two cables cut, was the stereo difference from the two channels.
I started questioning everything I knew about physics and engineering but thankfully you answered my question, even after one year.
oh my gosh, i had discovered that myself back in 2018 i think!! i was using headphones with a detachable wire and i bumped onto that same trick by accident. my headphone wire was getting a bit busted tho, when i wiggled it, the audio i was hearing would intermittently switch from full stereo to the stereo difference version and back and forth. so i guess i discovered two ways to hear the stereo difference channel with headphones only! LMAO
I actually learned this by halfway putting my headphones in my switch. SNES games would be missing half the instruments and sounds
Well now I'm going to destroy a perfectly good headphone jack trying to do this...
Ive done this before by disconnecting wire on my headphones jack. I never quite understood why it works but it was cool to be able to hear hard-panned and buried tracks in a recording. Great video!
i've worked at a record store for 8 years now and the technology behind vinyl still blows my mind thank you for doing this
digital wins!
@@thedavesofourlives1 Analog is more fun.
My hearing is degrading and I usually have captions turned on (and hearing aids). I just about lost it with the captions on the Audacity playback! 🤣
Thank you for pointing that out! I never would have checked. And big props to Alec for always putting captions in his videos!
How do you guys get access to the video early? Is it patreon?
@@JoshCarterWeb I believe Patreon patrons get early access.
Thanks for that comment, made me go back and rewatch that section with CCs!
There are a couple channels I always have CCs on for, and TC is one of them. I especially love his jokes during the outro
Here I am, retired and in my 70s with a lifetime (well almost) in electronics and it never occurred to me to question how stereo sound works on a vinyl record. Thanks for the ‘training session’.
its known as the 45/45 out of phase system
I had no idea. Before the internet, these were deep dark secrets
A tip for anyone doing restoration of stereo audio off records. Because dust and crud sits in the bottom of the groove, it proportionally affects the stereo difference signal more than the mono signal. Convert from LR to MS then do your de-clicking etc. on each separately. Be careful not to use anything that might affect the phase relationship between the two signals. Then convert back to LR after processing.
The stylus tip never rides in the bottom of the groove. Styli have the tip rounded off, so only the sides of the stylus are in contact with the groove walls. The very bottom of the groove is typically not in contact with the stylus. One trick you can use with worn records is to use a stylus with different radius than the one that caused the wear, making the stylus ride higher or lower in the groove to find a portion of the groove wall that is unworn. This trick is regularly used by audio engineers transferring rare recordings from old 78s.
"Convert from LR to MS" - could you unpack this a bit? I assume LR = left-right, but I'm having trouble with MS
@@sevenbark MS = Middle (mono) and Side (Stereo difference).
I've seen this advice for years, but every time I have tried this in Adobe Audition, it does not produce better results than just declicking normally. Perhaps its declick algorithm is already doing something like that anyway.
@@IlIlIlIlIxxlIlIlIlIl Quite possibly. I've found it useful in Audacity.
Your closed captions are just A++! Such a small detail that screams attention-to-detail! Love your content!
I grew up with LPs during the 80s. By then, fine audio equipment was available at a reasonable price. Speakers with fifteen inch kickers were very popular. Listening to the left and right audio was an experience. Even to this day, stereo is one of the greatest inventions of all time. Very interesting video, I enjoyed it. Thanks!
Stereo difference (along with some frequency filtering) is sometimes used to generate Karaoke tracks as the singer is typically placed in the center of the stereo stage while instruments are distributed. There are some Karaoke machine chips with this functionality built-in.
I was surprised that this was not mentioned.
This is indeed true. Anything that is mixed mono will effectively cancel itself out and since most vocals are panned center/ mono they will be "extracted" from the mix when isolating the difference. This is exactly how all "surround" processing works from a stereo source. David Hafler while working for Dynaco came up with a passive analog circuit which could extract the "difference" channel from a stereo signal and then send that signal to dedicated "surround" speakers. Dolby pro-logic is the exact same process just done with digital circuits but functionally identical with the exception that with digital you can alter the timings between channels to compensate for speaker placement and distances.
I remember the sound editor that came with the packaged software included with our sound blaster 16 on our 486 could do that. Though most profession karaoke setups are actually worse than that, often being simple covers done by the production company, often with cut runtime, and sometimes even being performed on cheap synths, like someone's toy Casio keyboard. Lol. I'd imagine this reduces the licensing costs, as opposed to the original studio recording with the vocals filtered out as you suggest. I wish it were the case. Though I have on occasion heard some that seemed to be just that, where you can hear the faint remnants of the vocals if you listened hard enough. And even on a rare occasion, one that sounded like they actually got a hold of the masters and just mixed it without the vocals from the beginning. But sadly, by far, most professional karaoke discs seem to be the crappy covers by their in-house musicians.
@@naelblogger7976 Probably wasn’t mentioned in the Wikipedia article.
A crude version of this can be acheived by connecting the speakers incorrectly. I'm sure there is an audio term for this. I don't remember if it involves connecting both positives to one speaker, or both negatives. Discovered this after accidentally wiring up some speakers the wrong way and suddenly revealed a song I liked had a distinct bit of lyric effect audible this way while nearly inaudible on the normal stereo version. It was quite fun.
That concept would also be used in FM radio broadcast. The combined channel is the main band while the difference channel is a side band. When the signal strength is weak the radio will swap to mono to retain quality.
Yup, and for basically the same reason - the technology existed in mono first, so transmitting the stereo as a difference channel in a sideband kept the transmissions compatible with older receivers.
Come to think of it, it's the same thing they did with color television, just with a "color difference" channel instead of a stereo channel.
@@rjhelms Technically on US analog TV it is "color difference channels". The 3 primary colors are encoded as two signals in the sideband channel. The two signals in the sideband are encoded on the carrier in two different phases. Those colors you see are made with more adding and subtracting than if at first obvious.
@@rjhelms Actually, in NTSC TV two 90 degree separated difference signals carried the colour. G-Y and R-Y. Add G-Y to Y to get red, R-Y to Y to get green, then subtract red and green from Y to get blue.
@@dougbrowning82 PAL rotated them so they're equally important and it's that version (YUV) that is used in early digital codecs.
I was taught for FM Stereo, they switch _very fast_ between the left and right channels and TX the stereo pilot signal at a high audio frequency that 90% can;t hear! And also if the pilot signal isn't present, the RX will automatically turn it into mono!
Once you broke out the model everything clicked. Getting to see what the actual effect of the technique was really helped cement how they work at the same time, and how that translates to reading the groove. I remember watching your first video and, while I came away with both tracks being stored at a diagonal to each other, sharing a groove, I couldn't tell you why. I feel coming away from this that I could actually explain why they are at a diagonal and that feels fantastic.
The followup about the stereo difference was also really cool and, from a completely different perspective really helped me understand what the record is effectively encoding. "Two different tracks at 45 degree angle to eachother" is hard for me to understand how that translates to a record, but taking that extra step to explain how effectively it leads to a mono track being preserved in the lateral movement with stereo difference in the horizontal it really got it from a technical understanding to an intuitive one.
Thanks for another great video!
I love that the demonstration was arts &crafts instead of computer animation!
@@MonkeyJedi99 YES!
The implementation of this is elegant AF in its beauty and simplicity. It ticks all the boxes. Maintains compatibility with mono records and turntables (both ways!), yet also performs really well without sacrificing quality/etc to do so.
Ignore the scam reply.
I'd be super interested in hearing you cover quadraphonic records.
I did see a video about quad. Not sure who, You can do a search.
Yes
Upvoted!
seconding
came here to say this^^^
What a great video, Alec! As a mastering engineer and former lacquer cutter, I think you did a fantastic job explaining the process. The construction paper model was a perfect visual aid! Bravo!
Fun stuff. I'm old enough to remember the coolness of the original stereo records. But honestly, what really WOW'ed me was "quadrophonic" records. That could also be an interesting follow-on to this...
The moment he said and showed "stereophonic" I immediately thought of that as well. That would indeed be a very interesting topic.
I would love to know more
Uh, yes please
Not to mention "Ortoperspecta", invented by Tapio Köykkä, a three-channel (Front-Center and L + R Rears) reproduction from a regular stereo signal. Center channel was the sum of L and R, while the rear channels reproduced the differential signal.
There were two different quadraphonic systems. There was SQ, a “matrix” system which didn’t really have four separate channels, and there was JVC’s CD-4, which used subcarriers to encode either the rear channels or front/rear difference signals, not sure which. This gave better channel separation.
I remember reading a bunch of Sansui brochures at the time. They took SQ and turned it into “QS”, which some kind of dynamic level-based adjustments of the decoding matrix, to give the effect of better apparent channel separation.
Thelma Houston and Pressure Cooker was the first direct-to-disc record that was done without breaks. In other words, LIVE. The only break the band and singers got was the 30 seconds or so before the next song was supposed to start. Also, the break they took between sides of the LP. Well worth a listen!
I had direct to disc records. They were considered the best the format had to offer. CD essentially made them obsolete. Doug Sax of Sheffield Labs had a horrible case of sour grapes about this fact.
@@robertromero8692 Technically, all discs before 1948 were direct to disc, until recording on tape came into the picture.
@@danpetitpas True, but the associated equipment wasn't as good back then.
Also, check out James Newton Howard and friends ( some of the Toto guys )
@@robertromero8692Doug Sax also at one point worked with James Guthrie on a remaster of Pink Floyd's album Meddle to both vinyl and CD. It's a good remaster.
I was actually listening to a record when I realized this is how it worked! The snares on Todd Rundgren’s A Wizard A True Star are intentionally put out of phase, but panned hard left and right. Listening back to the cd, it sounds like normal out of phase, and in mono it disappears. But on the vinyl, it gets a really weird, unique spatial quality to it as the needle fights with that huge Side channel! I love weird Mid-Side tricks like that, and in other music where listening to it in mono will ‘erase’ certain vocal lines
BTW, fantastic job on describing the audio in the captions! Many people never seem to get that captions are more than just what words are being said...
Yeah! The words of the script read aloud are usually pretty good image descriptions too
I was literally laughing out loud at them they were so on point and clever.
Thanks for reminding me, I keep forgetting which channels have the good captions or not.
The confusion goes the other way too. I'm always annoyed when caption information finds its way into subtitles.
Oh noes! This is Primitive Technologies all over again. Me watching all the videos one more time not to miss out. 😂
Loving the inverted credit music haha! Phase cancellation is essentially what we’re listening to. The difference between the sounds that isn’t cancelled. Love the video!
this difference in channels is what hafler used to produce a pseudo quadrophonic sound simply wiring a speaker or two across the positive terminals of the amp output. i have been using this for decades to add depth to the stereo image.
The visual aid you made for the angled needle was super helpful for me. I couldn't quite picture it. You make very good videos.
I just want to thank you for putting so much care into your captions 😊
15:46
Complete channel separation on vinyl, utilizing a single needle is what I struggled with until seeing this video.
Not only is that clear to me now but so is the misconception that "up & down" is absolutely one channel while "side to side" is absolutely the other.
Thank you for this simplified, construction paper explanation.
👊🏻
🇺🇸
Wow this is auspiciously timed. I’m literally building my own record lathe right now and studying this technology 🎉 Neat!
Nice touch with the stereo difference version of the closing music! 🤓 Back in the dark ages (‘70s) there was a device called the Thompson Vocal Eliminator which employed this very technique. Since the lead vocal was/is usually panned center, shifting the phase of one channel by 90° would cancel out the vocal, leaving only any reverb triggered by said vocal. I never actually operated one of those, so I don’t know if anything was done to allow the low-frequency sounds of kick drum and bass (also usually panned center) to come through unscathed.
In the 70s when you got bored with a record you lifted the tone arm up to find the little colored wires plugged in the cartridge and switched a pair of connections. Then you got a completely different mix, usually minus the vocals and bass
Yes. They had a variable bandpass filter that you could dial in to only modify only the vocal range...in theory. My aunt had one, and in practice it was pretty much crap. But I suppose it got the job done enough.
Thank you for pointing that out, I didn't even notice the closing music sounded different! 😅 Playing it again it's very obvious it's weirdly off 😁
I used this process many times in my teen years when transcribing music in my DAW software and reorchestrating it with midi synths. The center cancellation helped to reveal some sounds that were hard to hear in the normal mix. But, yes, it would usually cancel bass too, so EQ’ing was involved before the cancellation process.
I must add that it’s not a phase shift of 90°, it’s a polarity inversion of one channel of the musical content, similar in concept to a 180° phase shift of a sine wave added to the original sine wave - cancellation. Common-mode rejection, used in balanced signals like pro microphone signals to cancel common-mode interference that might leak into a long run. Balanced or differential signaling using this noise-cancelling property is also used in HDMI, Ethernet, and other well-known signals.
@@chrismerklin8460 Yes, I misspoke. Polarity inversion. Thanks. 😎
It's worth mentioning that vinyl stereo is in a way a bit more limited than stereo on another medium like tape or CD.
For example, there's a limit to how loud you can have a panned sound. If you have a really loud bass drum hit in just one of the channels, you risk launching the stylus out of the groove. One of the things I was told when it comes to dance music production is that, for vinyl mastering, you absolutely must not pan your kick drum or bassline. CD doesn't have this restriction.
CDs do have a pretty similar restriction (it is called clipping), but it will only result in mangled sound, not jumping forward or backward on the record so its perceived less severe (or by most not noticed at all if it happens).. and this is also one of the reasons Vinyls are considered superior by some people: you cannot mess with the audio too much (increase the volume), or the stylus will jump. on a CD you can turn up the Volume to 11 (as louder is better /s), and all you get is mangled/clipped sound at the peaks.
The entire range of a CD is far bigger than a vinyl, but when you restrict yourself to the top quarter of that range during mastering it doesnt help..
That is fascinating I'd never considered that but it makes a lot of sense. We are spoiled these days in many ways and sometimes we don't even realise it!
That's fascinating. So essentially it may be necessary to make completely separate masters of a track depending on the release medium. Which also means that some versions of some tracks can never be released on a certain medium. And in the case of vinyl, due to mechanical limitations no less.
Just the thought of a loud kick drum casing the stylus to fly out of the groove like it wasn’t expecting it made me laugh
@@unitrader403 The main reason for "clipping" with Audio CDs is simply that when the tracks are mastered, the volume is always increased to make the track louder, given many CDs would be played in noisy-areas, such as in a car, or on a portable-player. If you rip a track from a CD, and then lower the volume in an editing app, you can eliminate most of this
As someone who is deaf in their left ear, this has actually made me interested in getting a mono stylus. I listen to almost everything in mono so that I don't miss anything, and it'd sure be convenient to only need one speaker, rather than two - especially if I have a smaller space to work with in the future. I have no idea whether it'd actually work the way you'd describe with modern records, but it'd be interesting to find out.
A mono stylus wouldn’t do what you’re thinking of, I think. A mono switch would, however.
As a person that I starting a record collection I asked myself this very same question just about a month ago. Thanks for being here to answer it!
Mono and Difference channels, or "Mid / Side", are the key to so many things we use every day -- most of the time, because of backward compatibility with older standards, or where we just want mono compatibility. FM Stereo and analog TV audio both work this way (similar to TV's Y / Cr / Cb color encoding.) Some digital audio encoding formats still use this as an alternative to L/R encoding, for various reasons. MPEG, for example, and probably many of the other perceptual encoding schemes, due to compression algorithms having a particular affinity for correlation.
I sometimes track stereo audio (like two guitar or vocal tracks) as a mid and side track, then convert them to L/R using polarity inversion. Again, that helps to ensure the result will work if folded down to mono, but it can also sound interesting (sometimes better, sometimes worse) when there are differences in the takes.
This two-channel phase relationship is also one of the core principles to how Dolby Pro-Logic Surround works -- the center channel is from the Mid, the rear channel is from Side, and L/R vs. M/S are steered using some fancy analysis of the signal to determine which direction (front/back vs left/right) is more prominent.
I love this channel because it makes me want to learn about things that I never thought I could have possibly cared about before the video was posted. Thanks for all the great content!
For some reason I am always more in awe of materiel methods of creating magic than electronic ones. The groove in a record has always struck me as remarkable, not only in that it works at all but how exactly it achieves it. Too few people are aware that it is a 90 degree difference in the two channels, and that the groove isn't shaped left and right for the two channels. I admit I thought it was that for the first 64 of my 65 years on the planet.
You thought that in your very first years? I don't remember having any thoughts about how stereo sound was reproduced in a single groove of vinyl until I my third or fourth year at least.
@@reshpeck huh?
@@reshpeck I have to admit that my father was a Woolworth's record salesman, and all the talk in our house was of stereo separation and cross talk. It actually led to my parents divorcing, although they still spoke over each other.
I think it's in part because mechanical systems required working out everything from end to end and that to make it work you need to think of everything.
With electronics and computers in particular, you are dealing with a lot of layers of abstractions in which each layer is relatively easy and changeable. This means the overall system is far more complex, but the bit you need to think about is much more simple.
For example, if you are working on signal processing, you don't need to worry about how that signal is produced or even the quantum effect involved in processing that signal. You can simply assume that the layers of abstraction you are using will work as documented.
@@Tampo-tiger Wow, those must have been intense disagreements over some very esoteric subjects
So interesting. But I still think the chain from recording to what comes out my speakers is magic.
CC Closed Captions ftw! Love the added comments XD Great videos like always. This guy is the BEE KNEES.
I never expected Alec to add papercraft to his list of tools to explain with but I am ALL for it!
You can pick literally any kind of subject, and talk about it in so much detail… that is the best explanation about stereo records I’ve ever seen in my life!
Fun fact about the mono and “stereo difference” signals (better known as mid and side): There is a drum micing technique built around this principle! If you take a cardioid mic, and put a bi-directional mic right next to it, but pointed perpendicular from the sound source, you can get stereo drums that collapse perfectly to mono. Simply add the two signals together in one channel, then subtract the bi-directional signal from the cardioid in the other, and you have stereo drums!
You can generalise the technique. Split the cardioid into an omnidirectional and a bidirectional pointed forward and add another bidirectional microphone pointing up. Your resulting set of four signals can represent a full sphere of sound. It's known as first-order Ambisonics. RUclips uses it for the audio component of 360-degree video.
There’s nothing inherent to drums about this. I use mid-side to record piano, and even orchestra. I like it because you can adjust the stereo sound in post.
Very common for video and broadcast from the beginning of stereo t.v. to maintain compatibility with mono tellys.
Still used in many budget video mics, but the mix matrix is on the mic so you get L&R out rather than M&S. Some are switchable while others are not.
@@alistercarmichael4990 Still something of a standard for TV and radio broadcasts of orchestral performances. Often with a very very nice Schoeps M/S rig.
Don't get us started on the ORTF stereo technique…
You do lose something in mono: the 'side' or difference channel, which is L - R. But you keep the 'mid' or sum channel, which is L + R, giving you a mono downmix: much better than just the L or R channel in isolation.
15:15 that sounds like the “poor mans surround sound” setup. I had a 2nd pair of speakers hooked up from the two positive leads from left and right on my stereo when I was a kid and it gave this interesting surround sound effect from the rear speakers. That reminded me of my setup back in the day.
the construction paper diagram was really helpful! that’s a really interesting solution for backwards compatibility
This was great. It's a subject that many don't ever get to fully understand, and you covered it very well. I retired in late 2020, but one of the last products that I worked on as a product developer was a modern direct drive turntable.
I've been subscribed to your channel for about three years, and I continue to be amazed at how well you explain difficult concepts. This time, you have outdone yourself. I am an audio tech with some 46 years of experience, and now that I'm old, I specialize in repairing vintage audio gear. That said, I believe yours is the best neophyte-accessible explanation of stereophonic record theory I have seen. I was thinking of making a video on my own channel to cover this, but now I think I will just refer people to your video. Fine job, sir.
I loved all the attention to audio production detail, and the captions, and the audio technology. This has been one of my favorite videos so far!
This concept is actually extremely useful in audio processing too! It’s very common to EQ "mid" and "side" (the equivalent of the left/right and up/down motions) separately, and it’s sometimes useful to do even wackier types of processing to either just the mono or just the "side" signals. If you ever see "stereo difference", this is what that’s doing: separating the stereo L/R signal into a mono signal plus the stereo difference.
The important things about the encoding used to press stereo LPs is that the sideways movement has the highest bandwidth so it makes perfect sense to use that for the average signal between the channels. The same idea is used in digital audio compression, for example MP3 files support "joint stereo" encoding where one of the channels is encoded as delta to the main channel. Because real world stereo music typically has very similar audio on both channels, encoding one channel as difference to the another saves a lot of bandwidth.
14:12 - You can actually generate a "stereo difference" channel with nothing more than a stereo 3.5mm headphone jack. You just have to slowly pull the plug out little by little and once it gets to the right spot, you'll only be hearing the difference between the left and right channels. It's a quick and dirty way to remove the "center" channel.
There's a lot of interesting physical considerations for vinyl mixing/mastering. The bass frequencies have to be center-panned as the low frequencies create such big grooves that they can cause the needle to jump out if they're side-panned. Records undergo RIAA equalization where they're mastered with attenuated low-end freqs so the grooves can be physically smaller and denser on the record, and the attenuation is reversed on playback.
Thanks! I didn't know about that attenuation trick. It's amazing how many clever ideas are used in this sort of technology. I'm also now wondering whether the record sent into space on the Voyager probe was a straightforward mono recording without any clever tricks. I'd assume so, or the aliens might have trouble reproducing the sound properly...
You say “interesting physical consideration”, I say “analog quality compromise”. Aren’t you glad all that has gone away with digital encoding?
There's something similar done in analog FM broadcasting. In the audio processor or stereo generator, a pre-emphasis curve (with either 50 or 75 microsecond time constant, depending on where you are in the world) is applied to the audio to improve the signal-to-noise ratio on higher frequencies. Then FM receivers apply the inverse of that curve, and it sounds normal.
@@macronencer Those discs on the Voyager probes were RCA videodiscs and it came with a cartridge to play them.
@@danpetitpas Really? I can't find any mention of that. Both NASA and Wikipedia say they were "phonograph records" but they don't call them "videodiscs", although I know there are images encoded on them. I'm a bit confused now - do you have a link to any further info please?
Sound is stored in the walls
This video has made me appreciate vinyl in a way I can't even describe! Thank you for this explanation and this video!
it’s truly uncanny the way this channel reads my mind. i was literally wondering about this last week. the same thing happened with the percolator video. and the one about the klaxon horns of old school cars. and also when i was wondering about how pulse oximeters work. always within like a week of when i’m pondering the question. truly strange and wonderful
I was literally typing that exact same comment until I read yours.
Seeing as how it probably takes at least a week to write, shoot and edit each video, he was probably steeped in thought about the topic when you began thinking about it, so maybe instead of him reading your mind, you were reading his 😀
@@reshpeck 🤯🤯🤯
You're actually living in a simulation, and this is a message from your real self trying to make you wake up.
@@Dremth 😢
My first time watching one of your videos with closed captioning. Now I want to go back and rewatch them all this way. It's like finding Easter eggs!
0:16 was hilarious to listen to from my singular cellphone speaker 😂 Absolutely delightful with the addition of headphones though well done
If you've learned some basic college level linear algebra, this is very easy to understand. Horizontal is the sum of the channels, vertical is the difference. It's a classic diagonal basis transformation
Thanks for giving credit to Alan Dower Blumlein, who did invent the 45-45 method of recording stereo sound to disc. In the late 1950s Western Electric/ Bell Labs tried to claim that they did, but was thwarted when a copy of Blumlein's patent was shown to their representitive. Many many other inventions stil in use in 2024 are attributable to Mr. Blumlein.
This is one of those things I've occasionally found myself wondering about ever since your original series on the subject. Very cool to see it revisited in more detail. Great vid!
As a side note, while I'd be the last person to defend the sound quality of vinyl (other than to say "eh, it's not so bad", because I'm just not that picky about audio quality), I do find something charming about the experience of listening to a record. The physical ritual of opening the cover, taking out the record, holding it carefully, placing it on the turntable, starting it playing, flipping it over, etc. puts a level of focus and attention on the music that more modern and convenient methods don't require. I could never be bothered to sit down and listen to an entire album from start to finish without doing anything else before getting a record player. It's kinda meditative. It's not something I do all the time, but it's been really neat to connect with music in a new way... well, I guess it's an old way, but my record player does output to Bluetooth, just to make sure something about my setup annoys absolutely everyone. >=)
By strange coincidence, a few days ago I had a conversation with a chatty old guy while waiting for my takeaway food order. He started off commenting how people can't live without their phones these days (I happened to be checking the news on mine) and went on to how he still takes pictures using black and white film and listens to his LP collection on his old stereogram. Although I've always been into new tech (early internet adopter, dad ran a mainframe computer in the 1970s) I could actually relate with him since my grandad was a rubbish collector and had sheds full of old cameras and audio equipment (my (1970s) amplifier still uses homemade surround speakers I built from salvaged stereogram speakers) and I have a collection on 60s and 70s LPs from myself and my dad. All was going well until he got onto the old cliche about vinyl just sounding so much better and warmer... I nodded and smiled through gritted teeth.
Ironically, if he actually embraced the internet instead of being so dismissive, this channel would be SOOOOO up his street!
Unless the record player you got was super expensive I'd be careful with what you play on it.
Those cheap ones that come with "All the modern features" like Bluetooth built in usually have an extremely heavy tone arm which causes the stylus to eat through vinyl.
What I've seen people recommend as a decent way to not kill vinyls on a budget is to get a used mid to high end direct drive record player from the 80s/90s with a built in phono preamp and plug a Bluetooth adapter into the headphone out.
You'll pay a bit more than the cheap record players, but you'll get something that works well, doesn't eat your vinyls, should last for over a decade and you can always upgrade the Bluetooth to the latest standard if some super high fidelity Bluetooth audio standard comes out, not that it should matter that much with vinyl anyway.
I think its time to do a video about (for most of us) our first record player.....
The Fisher Price one with the pins!
April 1st is coming up after all.
By the way, when digital recording was in the early days, you had to go out of your way to find a vinyl that had been itself recorded digitally. The analog recordings were the standard, while digital had a special label.
Nowadays, though, things are flipped with analog being the rare version with the special label, while digital is the standard format without the label.
The difference is key because while there are minimal differences between the two types of recordings, analog actually does tend to pick up more of the full audio spectrum, including subharmonic frequencies in the bass region, whereas digital recordings tend to leave that out due to the ear's inability to hear it... but of course you also have to have a really quality-built audio system to *feel* those frequencies in the first place... digital or analog, most vinyl players and speakers connected to them also won't play those lower frequencies either.
Digital recording nowadays is done at such a super high sampling / bit rate. I doubt anything is being missed.
Somehow I randomly clicked on CC. I knew you'd added Closed Captioning support but I haven't watched/read it before. I enjoyed the secondary commentary lol!!!
When he said "that won't stop me from making the connection" I cried
Truly one of the sentences ever said.
It certainly was one of the most sentences.
Loved this one. The Aquarius and the 5th Dimension record you had on the table is actually the first record I owned, given to me by my mom. Let the Sun Shine In is my jam 🤩
No matter how many videos I watch on how records work they will never make sense to me and always seem like magic. It just seems like magic that they can easily capture and reproduce unique music and voices.
This all still blows my mind how simple yet complex records are. Also thanks for having the outtakes.
The construction paper diagram really helped me understand what you were describing. Thanks.
I have not seen this channel in a while. Glad your back on. I was surprised you did not mention quadraphonic recording.
Keep up the great channel.
Thanks
Yet
Yes, how did quadraphonic records work??
That and those early binaural records that played two separate tracks at the same time.
I remember wondering how they did this trick when I was young. It blew my mind that they could fit two signals in 1 track. Thanks for making this video!
It really is crazy cool how the record & player came about. The arm itself is also really cool, with the way it rests on the record allows it to play warped records too.
I'm so glad I was using both my headphones at the very beginning of the video. Thanks for being funny when no expects you to be funny.
I feel like you'd be a teriffic Intro to Linear Algebra teacher! You've basically described what a change of basis is in a way that is super accessible and tangible.
12:15 This is a very interesting image. I read long ago that you can put more runtime on a vinyl record if you squeeze the grooves closer together in silent parts of the audio. As I have never seen this so far, my imagination of how effective this could be was very limited. Now this image shows me that the grooves are like 3 times as close in those areas, which is quite unexpected for me. Also, I like how you point out the mid-side encoding scheme and its incredible advantages over other solutions.
techmoan has a video about the trimicron record. It boasts 2 hours of music per disc, but suffers from decreased dynamics due to squeezing all of the space out of the disc.
Record lathes are designed to optimize the surface area of a record. A preview head (if cutting analog) or a computer will preview the coming audio signal at least a revolution in advance, adjusting the groove spacing accordingly. Not only does it move quieter grooves close together, it spreads out loud grooves so they don’t intersect each other. If the grooves intersect, then you get a skip or a repeat on playback. A cut lacquer is often inspected with a microscope to ensure no two grooves intersect.
@@xureality You get more grooves on a record by lowering the overall volume level of the recording. The optimal recording length per LP side is 12-18 minutes. If you go beyond that, then you need to start making compromises… either by lowering the overall volume the recording or by reducing the dynamic range of the recording. If you want to hear how great vinyl can sound, get a 12-inch 45rpm disc with one song per side (like a DJ single). The music can be cut loud with lots of dynamics.
@@xureality Also, IIRC, Techmoan found that those records with their skinnier grooves tended to be harder to keep clean as any dirt would stick in the tiny grooves.
The other effect of cutting the grooves too close together is that you get echoing from the grooves on either side due to the pressing process distorting the vinyl. You may be able to get echoing from the master cutting process too? Crank up the volume and you'll be able to hear that on pretty much any record where they've tried to get a lot of music on each side.
Very good description! The left/right difference signal is used for the same backwards compatibility reasons for FM stereo broadcasting. It also has the benefit that we can choose the bandwidth for the difference sub-channel - effectively controlling the amount of separation required
The topic of this video bears relations to Mid/Side recordings, which is when two mics are used to record stereo audio by recording the 'centre' with a cardioid mic and the 'sides' with a figure-of-eight mic facing sideways. You then play both mics on both channels but invert the polarity of the side mic on one side, magically creating real stereo!
Also, nice touch on the 'L/R difference' version of the outro tune 😎
Not a real stereo, there is no perfect way to capture it. But it is always mono compatible with a strong mid/side separation. Often much better than a stereo mic. Ideally should be real stereo, but in reality.. it never is.
You can also make it more general. Split the cardioid into an omnidirectional and a figure-8, then add another figure-8 pointing up, and you've got first-order Ambisonics.
The four channels of first-order Ambisonics are equivalent to an omnidirectional, a figure-8 pointing left, a figure-8 pointing up, and a figure-8 pointing forwards. With them, you can encode full-sphere surround sound. Decoding it requires a minimum of six speakers, though eight at the corners of a cube would be easier to position around a room. Headphone playback can be done via head-related transfer functions.
The spatial resolution isn't very high for first order, but the format can be extended to higher orders with more channels, though the additional ones no longer correspond to any physical microphones. I've heard of audio software plugins going up to third order (16 channels) and games going up to fourth (25 channels), and seen tables of the relevant encoding equations online going up to at least sixth order (which would have 49 channels).
Literally since I was like 15 and getting into audio recording, music production etc this question has been sitting in my brain. Even now 10 years later this question would still nerd snipe everytime I thought about it while playing vinyl. I think I even googled it a few times but just didn't really understand the concepts just from written answers. But THIS video has finally given me a satisfying answer. Thanks mate!
I've always known mastering for vinyl was a more specific process, but after seeing how stereo sound works in a record player, it now makes sense! Awesome video! (Also big fan of the channel in general, this one was especially interesting to me as a hobbyist music producer)
Another interesting aspect of stereo sound is how sound mixing of music has changed over the years in response to changing technologies.
Before the introduction of the walkman, music was almost always listened to via a speaker system. Because of this, some music was mixed with a great deal more stereo separation than is common today. Many tracks by The Beatles ("Run For Your Life" springs to mind) had complete (or nearly so) separation of some recording channels. This is fine if you're listening to the music via a stereo speaker setup, because even though some sounds (a guitar or a voice for instance) are only coming from one of the speakers, both of your ears can hear it, just at different intensities, which is how we perceive which direction a sound is coming from in the first place.
When listening to these tracks using headphones however, it sounds extremely strange. The sounds that are completely isolated to one of the channels only reaches one of your ears at all, which runs counter to how our brains are used to interpreting sound, so it doesn't quite know what to make of it.
The same goes for things like music from the Commodore Amiga. The Amiga had 4 sound channels, two for each stereo channel. This was not generally a problem, as the sound was generally played back through a TV or a speaker system. But listening to original stereo recordings from the Amiga using headphones today gives you two sound channels in one ear and two completely different ones in the other, which makes for an extremely odd listening experience.
On headphones, it basically sounds like something is directly next to your ear making that noise…which is more or less what’s happening. Really annoying when it’s used for two people talking.
Most MOD players outside of the Amiga have a setting to reduce the stereo separation for headphone playback to help deal with that problem. 50% seems to be where I typically end up setting it. They often also have a "surround sound" mode as well, which while the "surround" part is debatable, does make for a more comfortable listening experience.
George Martin did this on the Beatles recordings because he wanted the stereo version to sound the same as the mono version when played back on a mono record player. He was concerned about a 20 db cut in volume because of phasing when a sound was mixed equally on both the left and right channels. Very few people in Britain had stereo equipment when the Beatles started out.
They did use a lot of reverb on the earlier Beatles albums, so it wasn’t too bad. Most tracks had all the vocals on the right, but echoed in the left channel. There wasn’t complete separation of left and right.
the idea of a mono track with a second track describing the difference is called "mid-side" and is used in mastering, mixing, and composing to get better control of the stereo soundstage. mid is just (L+R)/2 and side is L-R. (or sometimes the x2 volume adjustment in on the side instead)
Pretty simple to implement in software, too. M is just the signal from the source and S is the signal weighted by the sine of the azimuth angle to it (or the x component of the direction vector to it). Use that for stereo sound in your game, and you've also got a platform for going to first-order Ambisonics when you make the leap to surround sound.
Looking forward to the followup video explaining quadraphonic records ;)
The change to the diagonal basis in the encoding directly corresponds to a literal diagonal basis in physical space! I love that!
Yes, like you said in your clarification, the left and right channels are encoded to move the styalus at a 45 degree angle up and down for one channel, and at a 45 degree angle in the other direction up and down for the other channel. At the very same time this encoding can be done a different way but end up with the exact same groves. When skewing the left and right channels on diagonals, this "rotation" behaves like horizontal movement of stylist as being encoded as left + right, and vertical movement of the styalus being left - (minus) right. This makes sense because there is less information in the left - right (difference) signal. This has to be a small signal else the styalis would jump up and out of the groove if that signal was strong. Since the left to right movement of the styalis can tolerate wider movement, it works out well to contain the stronger "sum" of the signals. Either way the groves end up being the same. So some photographs use up/down left/right cartridges and must add and subtract these two signals in the amplifier to get left and right. Other photographs use 45 degreed angled cartridges and thus don't do any addition or subtraction to get the left right channels.
If you look closely at a groove that is silent, say, on the left channel and bass on the right, you would see one side of the groove smooth and the other side of the groove waving. This causes the styalis move up and down in only one diagonal for the right in this instance. That diagonal up and down is also a combination of up and down and left and right at the same time, so the result is the same either way: Angled: left 45 angle silent, right 45 angle bass.... or... ((Left and Right: (silent left + Bass right)) PLUS (Up and Down: (silent left - Bass right))) = left silent. And ((silent left + Bass right) MINUS (silent left - Bass right)) = right Bass.
Also, on vinyl, the bass is reduced to create smaller bass waves in the groove and the amplifer boosts the bass to get it back (before the tone control adjustment).
The stereo difference signal version of the outro music puts the cherry on the top.
I laughed way too hard at that
Thank you for this topic today, it brought a few memories back from my childhood. One of them is from when my father bought an LP cheap (from an antiquary store or similar) and let me learn how sound can be produced by a sowing needle pushed through a piece of paper, resting the needle in the groove of the LP so the sound would resonate from the paper. I think I was at the age of 10-12 at the time. Cool to learn how a simple needle and a paper can produce sound from the groove of a record, even though the needle did deteriorate the record quite a bit during the experience. Oh well, it was a cheap record containing a collection of songs Dean Martin used to sing. Those songs still give me fuzzy vibes of remembering my childhood.
When I was a kid I thought the needle was split down the middle and one side was reading the left wall of the groove and the other was reading the right wall.
Audiophiles talking about Vinyl sounding better...it doesn't. But I like Dankpods' take on it: If you're into a hobby, you want to fiddle with it. You can't fiddle with Spotify, you click play and it goes. You fiddle with a phonograph, moving the discs around, changing the speed of a turntable, enjoying all the album art that comes with that big record sleeve. Playing music on vinyl is for participating in the music enjoying experience.
I thought the split needle thing too. Only I thought it right up till this video, all these decades.
>You fiddle with a phonograph, moving the discs around, changing the speed of a turntable,
You will *_never_* get a better sound by doing any of these. The way you "fiddle" with vinyl to get a better sound is "buying more expensive equipment" - the real fiddling is convincing the bank to give you an even larger line of credit (while preventing your wife from ever finding out).
>enjoying all the album art that comes with that big record sleeve.
Hi-res hi-quality natively-digital art can be just as enjoyable as the big record sleeve, though! Bandcamp has some great covers.
As a vinyl enthusiast this cleared some misconceptions I had held for years, and indeed questioned said misconceptions as well
This is such a great explanation of a thing I've been curious about for a long time. I'm a recording engineer, and the way the stylus uses a mono signal combined with a stereo-difference signal to create a stereo image is a lot like the mid/side stereo microphone technique involving one cardioid mic facing the sound source and one figure-eight mic 90° off axis from the sound source. In order to hear the recording played back in stereo you just need to combine the mono track from the mid mic (the cardioid one) with the track from the side mic (the figure-eight one) panned hard to one side and an inverted copy of that same side-mic track panned hard to the other side. Without the mid track they cancel out into silence, but with the mid track there you end up with a very nicely recorded stereo sound
Thanks to you I now have an addiction to learning weird and semi obscure things like how how audio recordings have worked through history.
My favorite part was the stereo - difference outro. 😁 I love hearing things that you can't normally hear hidden in the stereo - difference channel.
something i love doing in my free time is taking music and just listening to the side information, i sometimes hear things i've never heard before in the mix
I’ll throw you a curve. Look into quadraphonic vinyl LPs. Played with the infamous “Shibata” stylus! Quite interesting and actually produced exceptional sound. However, not 100% analog of course, but still very advanced for the era in which they were introduced. Great video!
The quad recordings used a 30,000 cycle carrier frequency to modulate the rear channels. The problem was the best quad stylus in the world couldn't keep up with the groove and would wipe out the record in about five plays, leaving only the front two tracks.
@@gregrobsn Yeah, I would play my quad LP when I first got it to make sure there were no defects and the second time I played it I would record it to reel to reel. After that I’d file the LP into my collection. Most of my quad LPs only had two plays on them. I run it through a Kenwood KR9940 receiver. I wish I still had that setup.
I did the same and recorded on a 4 channel Akia. I lost most of my equipment during the '94 Northridge earthquake. I was living about 2 miles from the epicenter and the only room that sustained damage was the theater room.
The trouble with flat records which is often overlooked or ignored by analog fetishists is the inner groove distortion caused by the decrease in resolution as the spiral groove gets smaller and smaller and the stylus moves slower compared to when it is playing the outer edge of the disc. A cylindrical format would overcome this limitation with the same audio quality at the beginning and end of the recording
Yup. The inner track has about 66% less bandwidth than the outer track. I think that's why Edison went with a cylinder. But phonograph records were more convenient.
Glad you addressed the mechanical cross-talk limitations at the end
Two channels in just one groove
Gettin' down just for the funk of it
You and Techmoan are the go to channels for sleep time videos so infomative yet so calming at the same time… thank you both .
Ignore the scam reply.
The business with the mono-compatibility of stereo records mastered that way is remarkably similar to how "mid side" recording works, where instead of a Left microphone and a Right microphone, you'll record with a Mid microphone and a Side microphone. This does mean you need to convert the signals into L and R channels for playback, but it does let you do some interesting things.
If *nothing* else, a Mid Side recording can be used like a mono or stereo recording, and you can vary the "width of the stereo image" by simply turning the Mid channel volume up or down. But there's *loads* more that you can do with "mid-side processing" like putting different effects on the two channels.
Vinyl IS m/s encoded. The lateral channel is mid, and the vertical channel is sides.
I bet most of those MS effects will also work on Ambisonics, what with it being a 3D generalisation of MS recording.
i LOVE when you come out with a new video. That's because all of your videos are perfection and in reality should be used in universities.
Blumlein was doing other work in Mid/Side stereo, so it makes sense that he'd go for this (incredibly clever and sensible) solution.
(also, love the outro easter egg ;) )
I wanna hear how Quadrophonic records were made now
playing stereo records on older mono cartridges can be a bad idea as the cartridge mechanics are very 'stiff' vertically, causing wear on the stylus and groove, later mono cartridges were 'stereo compatible', although still mono output ...they'd 'take' the vertical movement, not try to resist it mechanically, just ignore it
The stereo difference track is trippy! The bass guitar is all ghostly cursed, but the brass is really quite similar to the original…
That basically caused by the wave length. A lower tone sound have longer wave length. Which makes matching the wave between two channel easier. In the other hand, high frequency sounds have short wave length and hard to cancel out. A little off phase and it's not cancelling but adding up. That is also why ANC headphones hardly do anything to cancel human voice but works perfectly on car noise.
@@mmis1000 Thanks for explaining! That's super cool :)
Man I just came across your channel. I love it. Please never stop doing this. We all appreciate it so much!
Very well explained. Quick anecdote: Back before multi channel surround encoding and modern AVRs became available, in the 80s some of us hifi enthusiasts would setup "poor man's surround sound" by connecting a second set of speakers (placed in the rear of room) to a 2-channel amp/receiver to only play the stereo difference signal, which would result in surprisingly effective surround sound from a basic stereo source. Give it a try. For the wiring diagram, just google 'poor man's surround sound wiring diagram'.
8:52 - Seems like that could be used intentionally to encode secret data. Say in a spy novel there’s a music producer who is also smuggling information around the world. The McGuffin record sounds like regular music in one way, but a special player running over the groove at a different angle could play another message. The information disks would look just like other records of the same artist, but with a small difference to the label that’s easy to miss unless you know what you’re looking for. Maybe the spy record gets mixed up with others and sent somewhere it wasn’t intended to be, and then story happens.
I remember such a story in a Mickey Mouse comic, but that was some sort of secret groove, only audible with a special stylus and with no deeper explanation how it worked. But yes, the special record was mixed into a batch of normal ones which were then sold.
Secret information in the same groove probably would have worked with mono only records, with the secret message encoded in groove depth. I don't really see it working with a stereo record on a stereo player unless you do something more involved like cutting a hidden, very narrow depth-only groove at the base of the normal one, inaccessible to the normal stylus.
@@tz8785 How about this: in the very bottom of the groove, below where the stylus runs, etch digital pits and lands to be read with a laser (encoded like a CD). Tough to detect unless you already know the record has hidden information.
@@tz8785 was it "Topolino e l'unghia di Kali?"
@@matteocristini6221 Something with Claw of Kali sounds quite possible, although I don't recall the exact title and I didn't read it in Italian in the first place.