A Technical Review of Vinyl Records

By Chayuti Chetsandtikhun

Last year was a good year for vinyl. In fact, it was the best year for vinyl in a very long time[1]. Enthusiasts, collectors, and some audiophiles praise vinyl for its superior analog sound over CDs or other digital formats. Jack White’s Lazaretto was the top selling vinyl album last year, and Muse’s upcoming album, Drones, is also available as vinyl. Neil Young has said that Steve Jobs, the man who single-handedly launched the digital music industry, preferred to listen to vinyl in the comfort of his own home.  So what should you do? Should you grab a turn table and join the Vinyl Revival? Should you go out and buy vinyl copies of your favorite records to enhance your listening experience? Are your parents’ old collections suddenly valuable again? The short answer is: no. Vinyl as a format is objectively inferior to CD (that is, digital music) in virtually every measure. While CD releases can, and often do, sound worse than their vinyl counterparts, it’s not the fault of the medium but the production process. Despite recent rises in vinyl’s popularity, CDs are still the technically superior medium.

Let’s start with a quick rundown of how vinyl records work and how they differ from everything else today. Sound waves are etched as grooves that squiggle back and forth along the surface of a polyvinyl chloride disc. These grooves are “read” by a needle, and the back-forth movement of the needle reconstructs the sound waves. Vinyl – along with other mediums like tape cassettes – are analog storage formats where changes in sound pressure are directly translated onto physical, tangible objects. Digital formats that dominate the market today – from CDs to downloads to streaming platforms like Spotify – are based on series of discrete data points. Digital formats “connect the dots” between data points, while analog formats translate the curves drawn into the medium to recreate sound waves. The process of digital-to-analog conversion is often brought up by promoters of vinyl, but many arguments used to support vinyl’s superiority are based on misinformation. The argument is that the sound of a violin, for example, in the grooves of a vinyl record “is always near perfect because it is reproducing the frequencies of what a violin really sounded like, not a digital staircase [approximation of the sound waves]”[2]. The “digital staircase” here refers to square waves, a shorthand for poorly reconstructed sound that misrepresents the original recording. One way to think about the allegedly poor quality of digital music is to imagine a violin solo being played with a guitar: the notes are the same but the timbre of the sound is different.

Images like these often show up in articles about analog vs digital sound even though the digital jagged digital signal is not the product of the conversion process. This misconception has spawned a number of other myths, for instance, that digital files like CDs sound harsher, or that digital signals are only approximations of the sound whereas analog is authentic. Image Source.

The argument that digital files poorly reconstruct audio recordings arises from a misconception about how sound waves are reconstructed from digital information. The process of digital recording does not, as some might insist, convert smooth curves of sound waves into a jagged “digital staircase.” While articles explaining the process of digital-to-analog conversion are often accompanied by images of step functions, these images incorrectly show how sound is reconstructed from digital music files. In truth, there is no information between two data points (the horizontal lines in step functions) in digital music files[3]. Various algorithms are used to interpolate the information between points to create smooth curves between data points. Thanks to a clever piece of mathematics – the Nyquist-Shannon sampling theorem – we can guarantee that the interpolation will always result in the correct frequency for band-limited signals (all digital recording is band-limited). On the other hand, vinyl introduces more harmonics and distortion to the recording since the needle’s tracking and the inconsistencies in the grooves always result in imperfections from vinyl records. In audio research, imperfections introduced by vinyl records were a limiting factor in performing “accurate and reliable listening tests on loudspeakers,” and CDs were the solution[4]. In other words, the assertion that digital formats inaccurately capture and reconstruct sound is false, and whatever imperfections it might introduce is virtually nonexistent compared to that of vinyl. As a format, CDs will be able to reproduce sound far more accurately than vinyl.

This video explains the processes of digital-to-analog and analog-to-digital conversion. Specifically, the demonstration from 4:15 to 5:05 illustrates the process of converting analog signal to digital and back to analog again.

Another argument often made by critics of digital music is that vinyl just sounds better. Some say that vinyl sounds “warmer,” that digital music is “harsh.” Some say that vinyl has better reproduction of high frequencies than CDs. Others say that vinyl is more dynamic. Let’s take these one at a time, starting with high frequencies and “warmth.” It is true that vinyl can reproduce higher frequencies than CDs. Mostly. Remember the Nyquist theorem I mentioned? For that to work, CDs can only produce frequencies up to 22 kilohertz (kHz) while vinyl is capable of producing much higher frequencies. The short-lived quadraphonic vinyl can actually reproduce up to at least 45 kHz[5]. Too bad that none of that really matters. First of all, we don’t have instruments that can play a note remotely close to 45 kHz. A high-pitched instrument like the piccolo can only produce notes as high as 4 kHz. Secondly, microphones used in professional recording are typically rated to record up to about 20 kHz, and that’s because humans can only hear up to only 20 kHz. Maybe. If you’re lucky. And also not older than, say, thirty. Otherwise, you can probably only hear up to about 16 kHz as a result of natural degradation of cells inside the ear. So, there’s not really a need for anything beyond the range of sounds that CDs can reproduce. What about warmth, then? Vinyl can only produce those high frequencies if they’re in mint condition because grooves that produce high frequency sounds are the first to wear out. As a result, unless you are comparing fresh vinyl records against CDs, the highs coming out of vinyl records will be dulled, or distorted, or both. In fact, that might just be the reason why vinyl is often described as “warm” and “soft” while CDs are “harsh” and “cold.” It’s because CDs are better at reproducing high frequencies, which can sound alien to those who are used to duller sounds of vinyl, especially if they are comparing the same song in those two formats.

In this video, compare the passage at 2:10-2:25 against 3:55-4:07. (I highly recommend headphones.) They are the same passage from the same song, but the former is extracted from a 1977 vinyl record while the latter is from a 2011 digitally remastered CD. Notice the cracked sounds of the cymbals and snare drum and generally lower fidelity of high frequencies.

It’s not just the high frequencies, either. Vinyl also struggles to produce lower frequencies (especially when they move between the left and right channels) because louder bass will require wider, more dramatic curves for the needle to trace. However, there are two problems with these wider grooves: 1) you can’t record as music on vinyl disk because the grooves take up so much space; and 2) it’s harder for the needle to track the grooves, which leads to more distortion. And have you ever wondered why bass guitars and kick drums always sound like they’re directly in front of you, unlike guitars, which sound like they can be anywhere, left to right? That’s because, a few decades back, bass sounds couldn’t move between left and right channels lest the needle be thrown off the grooves. In order to remedy the first problem, records are made with the bass artificially reduced and highs boosted to make the grooves skinnier. When records are played, the process is applied but with bass boosted and highs attenuated. As for the second problem, there’s no solution, which means that vinyl’s ability to reproduce low frequencies will always be technically inferior to digital formats. So, you might want to hold off on buying EDM records.

T-Square’s “Explorer,” features a slap bass solo at the start of the track. The solo is actually a combination of two separate recordings of different solos with one placed on the left side and one on the right. Before CDs, this intro would have been impossible, or “uncuttable,” because the needle would skip immediately when the song starts. Engineers of the past would have had to combine the two solos and center their sound – rather than panning left or right – a process which would have destroyed the intro’s appeal.

Next up: dynamics. The word “dynamics” refers to the size of the gap between the noise floor and the loudest sound you can produce before it becomes distorted, sometimes referred to as “signal-to-noise” ratio (SNR). For some people, noise floor might be associated with tape hiss or vinyl static; however, many people who listen exclusively to digital music might not have a point of reference because noise floor is virtually a non-issue with the digital format. One of the few ways that you might hear the noise floor with a digital player is to plug in earphones into your phone and crank up the volume as high as possible (without playing music, of course), and you might be able to hear it. This is because digital formats have much better SNR than vinyl or tape cassettes, which makes the noise floor inaudible in most cases. Vinyl records have about 50-75 decibels (dB) of SNR while the vast majority of commercial CD recordings sit at 98 dB or higher. There are two benefits to better SNR, one is that the recorded music can have a much wider range of volume changes (from much quieter to much louder) throughout the songs. To be fair, artists will probably never utilize the whole 98 dB of headroom musically. If they did, you would only comfortably hear parts of the song recorded at 5 dB when parts that were recorded at 98 dB were as loud as a jet plane taking off. This would most likely cause hearing damage. But, while this larger SNR may not be used in its entirety, it all but obliterates the noise floor on CD records, whereas hisses will always be a part of vinyl.

This video shows off the differences between a vinyl record and a CD copy of the same album. At 1:50, you can hear the most obvious difference between CD and the inner groove of vinyl records, which are extremely susceptible to distortion. The condition of the record is also not very good, which further highlights an advantage of digital over analog formats: digital music will never deteriorate over time.

From the technical review I’ve presented, you might assume that CDs versions of albums will always sound better than the vinyl version. In practice, however, things are more complicated. Producing a record is not an exact science. The recording and mastering processes plays a significant part in determining the sound quality of an album. Many CD albums released between the late 80s and early millennium are plagued by poor mastering practices that made them sound inferior to vinyl; however, the production quality of CD releases in the recent years have been much better (most notably, Daft Punk’s Random Access Memories, which is often praised for excellent recording and mastering). Music today – before it is pressed as CD or vinyl – is typically produced using digital equipment, so the grooves in the vinyl are dictated by digital information anyway. This means that some new vinyl records are actually made from digital files, instead of tape (analog) masters. Even reissues of old classics today often sport the phrase “Digitally Remastered,” which means that the songs were digital before it hit the wax. If you go back to the video that compared different versions of Queen’s News of the World, you might be able to hear small differences between them. Whether CD or vinyl release of an album will sound better in the real world often depends on how it was produced, not the medium itself. There is no doubt that the physical limitations and technical capabilities of vinyl and CDs contribute to the way that engineers approach each medium. Maybe those who work on vinyl releases have been around longer, have more experience, and are less influenced by outside forces because of their position. But in the end, there is still no denying that digital music is far more capable than its analog counterpart.

Vinyl is technologically inferior to CDs by every possible measure: its highs aren’t as good, its lows aren’t as good, and its dynamic range isn’t as good. Vinyl records are bulky, susceptible to the elements, and require expensive equipment to be played on. No, Crosley turntables do not count[6]. But, in some cases, vinyl sounds better than CDs, despite being technically inferior. Vinyl owes its superior sound to the people who are involved its production – people who are able to expertly master an album and who ultimately determine its sound quality. But personally, I still think that vinyl is simply not worth the hassle for most people. Vinyl will most likely go the way of fountain pens and mechanical watches, effectively becoming luxury goods (turntables and associated hi-fi audio components have decidedly embraced that transformation). In the end, those who buy vinyl will only be enthusiasts who genuinely like the sound of vinyl or perhaps the rituals that go into listening to a vinyl record: meticulously setting up the tonearm to perfection, taking a record out of its sleeve, setting it on the platter and letting it play while appreciating the movements of the turntable’s mechanism. The resurgence in demand for vinyl is surprising, but this is about as far as it will go. The chains of production that once supported its production are rusted and broken, even those in the industry would agree. The remaining manufacturers of vinyl blanks today struggle to keep up with the demand, and their aging machines are becoming more and more difficult to maintain. Rather than trying to resuscitate a dying medium, I would much rather rally for artists, producers, and record labels to focus their efforts on producing the best sounding music on digital masters.

[1] No one really kept proper records of vinyl sales number until Nielsen started in 1991. Another source is the Recording Industry Association of America (RIAA), but they only track sales numbers from manufacturers while Nielsen got their information from points of sale which is far more accurate. (http://blog.thecurrent.org/2014/02/40-years-of-album-sales-data-in-one-handy-chart/)

[2] http://www.reddit.com/r/vinyl/comments/fiedy/my_attempt_at_creating_a_guide_to_
vinyl_to_answer/c1g5pb3

[3] “Zero-order hold” integral to primitive converters do create the staircase effect, but it either the wrong way to explain how modern converters work or fails to give the complete picture of the process.

[4] http://www.npr.org/2012/02/10/146697658/why-vinyl-sounds-better-than-cd-or-not

[5] Quadraphonic records didn’t actually use higher frequencies to reproduce musical notes, but to store extra audio channels. Stereo has two channels (left and right) while quadraphonic has four channels.

[6] A good turntable must have at least weight adjustment options in the arm to ensure good sound and reduce wear and tear on your records. A decent turntable puts about 1.5 grams of force on your records while Crosley players dish out over 4 grams. Their needles are also made of ceramic (a cheap alternative), which destroys vinyl records even faster.

Homepage Image Source

5 responses to “A Technical Review of Vinyl Records

  1. Pingback: In rotation: 5/4/15 - The Vinyl District·

  2. An interesting read, but do you think that there was a reason behind the increased popularity of vinyls beyond the misinformed claims of vinyls just sound better than digital music? It seems vinyls at least hold a lot of nostalgic appeal to some listeners, almost in the same way that some prefer classic cars even though they may be inferior in both efficiency and durability compared to what’s available now. You did mention that poorer production could lead to less than stellar sounding digital music. Would you say that vinyl promoters are simply choosing, perhaps unknowingly, to listen to the wrong music?

    Like

    • I think the nostalgic appeal of vinyl is certainly a part of its charm and that’s why I think will go the way of fountain pens which are, like you said, “inferior in both efficiency and durability.” I certainly have no problem with vinyl continuing to exist in a niche market. For example, I own a couple of fountain pens which I use exclusively to write my personal journal. It’s far more intimate than writing with a ballpoint or typing into a Word document, but I certainly would not say that the fountain pen is more efficient and durable than ballpoint pens. I certainly wouldn’t try to convince anyone to use them as daily drivers, similar to why I wouldn’t tell anyone to start a vinyl collection in this day and age. Unless they know for a fact that they prefer the vinyl sonic signature (because it would not be my place to argue anymore).

      As for the second question, I don’t think there is a necessarily the wrong music to listen to. Certainly, one wouldn’t ask for a vinyl copy of T-Square song that I referenced, because it’s actually physically impossible. But, the fact that there is poor production on the digital side of things makes it all the more important to show record labels and studios that heavily compressed music is objectively less enjoyable. I’m still puzzled as to why this is still a problem since there are many studies that support my previous claim.

      Like

  3. October 30, 2012 6:29 pmMy Top 15If Upsidedown u turning it.. u getitng love immediately.. )1. Sylvester “Dance – Disco Heat” Fantasy Records 19782. Rampage feat. Billy Lawrence Take It To The Streets Elektra Entertainment 19973. E-Smoove “The Guitar” Henry Street Music (Test Pressing 20014. Ten City “Whatever Makes You Happy” Ibadan Records 19965. Soul Providers “Rise” Azuli Records 20006. Tuff Jam feat. Xavier “Tumblin Down” Mucho Soul 19977. Nathan Haines feat. Verna Francis “Earth Is The Place” (Jon Cutlers Distant Music Mix) Chilli Funk / Wave Music 20018. A.T.F.C. presents Onephatdeeva feat. Lisa Millet “Bad Habit” (Spen & Karizma Mix) Defected Records / D-Vision 20009. Inaya Day “Feel It” (The Down Under Mix) Strictly Rhythm / D-Vision 200010. Full Intention America (I Love America) (Original Sugar Daddy Mix) Simply 12 200211. Aly Us “Follow Me” (Full Intention Mix) Strictly Rhythm 2002 The Underground Solution Luv Dancin Strictly Rhythm 199012. Jovonn feat. Wayne Brown “Mystery” Rena Records13. Lood feat. Donell Rush Shout-N-Out MAW Records 199614. Kenlou The Bounce MAW Records 199515. Kenlou “Moonshine” MAW Records 1995

    Like

  4. Hi, thanks for a great article!

    However; re. dynamic range. Given a 16-bit digital audio-file, even though the SNR is considerably higher, I would argue that the actual available dynamic range (while preserving acceptable sound quality) is far less, not only due to factors of hearing, but also from the fact that for each 6dB of headroom you give yourself, you discard one bit of dynamic depth. I.e., a 16-bit recording peaking at -6dB is, in effect only 15-bit. If that recording also has a considerable dynamic range, you are effectively bit-crushing the lower parts. This is especially true if you’re recording at 24-bit, which seems to be the standard these days, then converting to 16-bit for commercial use. For instance, if the quiet parts are on average -24dB compared to the loudest parts, they are now effectively 12-bit audio. In terms of absolute numbers, this means that while the loudest parts have 32768 available values for recreating the complex waveform (given that they peak at 0dB), the quiet parts have only 4096 available values. That’s a big difference… Any thoughts on that aspect?

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s