Aesthetics for Birds

Aesthetics and Philosophy of Art for Everyone

On The Record: An Audio Professional’s Take on Vinyl

| 21 Comments

What follows is a guest post by musician and recording engineer Michael Connolly. It is a response to an earlier piece that argues for the superiority of vinyl over digital audio.

As a long-time recording engineer and musician, I read Tony Chackal’s post “Spin Me Round: Why Vinyl is Better than Digital” with great interest. The analog-digital debate in audio is a longstanding one, and while it is unlikely to be resolved anytime soon, I thought I might be able to offer some background as a longtime audio professional and musician. Recordings are a beautiful mix of technical and aesthetic concerns, and this post will attempt to tease out how to navigate these two framings of music recording, especially with regard to the often-oversimplified distinction between analog and digital recordings.

1         Setting the Scene

1.1         Recording and Playback: A Series of Transformations

Recording does not capture an objective reality any more than a photograph would. Although this might seem controversial, notice that a photographer makes aesthetic choices in lens, camera, lighting, composition, editing, and printing. Similarly, recording engineers and producers make choices at each stage of recording—from selecting and positioning microphones, coaching a certain performance, editing and mixing the recording, and finally mastering the recording onto a consumable format (vinyl, CD, MP3, etc.).

The sound of the original musical performance undergoes many transformations and alterations as it passes through various pieces of equipment and stages of processing. Each step involves tradeoffs and choices, which can be aesthetically or technically motivated. One of the interesting things about recording is that some of these transformations are completely out of the control of the engineers, musicians, or producers. The home playback environment—your speakers, headphones, and other playback equipment—is not in my control as an audio professional. This means that I have to create a recording which will “translate” across a variety of playback environments, and to accept that some tradeoffs will always exist.

As the listener, you’re the final link in this chain. You choose whether to listen on vinyl, CD, MP3 or cassette—and each will sound different. You choose whether to listen with five-dollar earbuds or a $10,000 stereo system—and these will sound different as well. As an engineer and producer, my job is to take a recording as far as I can in the intended direction. After that, your listening environment finishes the job. In a sense, this makes every playback of a recording a unique event; the final transformations occur through your equipment, your ears, and your mindset.

1.2         Analog and Digital Recording

The distinction between analog and digital recording methods can be a confusing one, so here’s the simplest rule of thumb. Digitally recording anything (whether it be a sound wave in the air, the daily temperature, or a photograph) involves measurement, i.e., assigning a value on a numerical scale of some kind (decibels, degrees, or pixel color values). Analog recording methods do not.

Let’s imagine that I want to track the growth of my child. I might measure their height against a tape measure, recording the feet and inches in a notebook, or I might simply draw successive marks on the doorjamb. The notebook records are digital; the doorjamb records are analog (the marks are analogs of, or analogous to, my child’s height).

But analog records cannot be reproduced, processed, or transformed without introducing distortions, sometimes known as “errors.” My doorjamb record is satisfactory until I move to a new house. At that point, I might place a broomstick against the doorjamb, transfer the marks to the broomstick, and then transfer them to one of the new doorjambs. At each step, some inaccuracy is inevitable. Analog records always introduce error when being copied—and often the passage of time alone is enough to change the recording! Broomsticks change length as humidity changes; magnetic charges on tape fade over time; vinyl records warp when stored incorrectly.

In contrast, the wonderful property of digital recording—and the reason that it has become dominant in audio, video, photography, telephony, radio, and other domains—is that, once I have made my initial measurement, the resulting “recording” can be copied and reproduced without generating any further error! These numbers can be transcribed to a second book, sent in an email, spoken over a telephone, all without introducing further distortions. The pages of my book can become yellow and brittle, but if the numbers are legible, they can still be copied perfectly.

Admittedly, digital recordings are not 100% accurate. My tape measure can only measure to the nearest 1/16 inch, so I must round off the measurement when recording it in my notebook. Still, there is still much less chance for cumulative error, and finer-grained scales can decrease even those inaccuracies.

These rules of thumb apply to all recordings, including sound recordings. So, with analog audio recordings, we can expect to introduce additional errors (distortion) each time the recording is copied or played back, while with digital recordings, we can expect an initial inaccuracy when the signal is first measured, but subsequent processing steps will not introduce any unintended errors or distortion. This means that digital recording formats tend to be more accurate—although again I want to emphasize that this is not accuracy to any objective reality, but simply to earlier steps in the series of transformations.

While distortion in analog systems is unavoidable, with digital systems distortion becomes an aesthetic choice which can be added as desired.

1.3         Material and Immaterial Media

My divergence with Chackal’s viewpoint begins almost immediately. He writes,

Digital listening formats are immaterial, and so offer conveniences of portability, efficiency, and expediency. Vinyl records are material, occupy space, need to be properly stored, and require more engagement to operate. The fact that vinyl records are material allows distinctive features to be appreciated and evaluated, which are unavailable in digital formats.

This account conflates analog formats with physical media. A standard view among audio professionals is that analog versus digital representation is orthogonal to the question of materiality. While vinyl records, tape cassettes, Grandma’s 78 collection, and even VHS videotapes are all both material and analog, many material digital formats exist, including CDs, MiniDiscs, DAT cassettes, and DVDs. Further, if a digital file streamed on Spotify can be called an immaterial digital format, then it follows that the all-analog stream of an AM/FM radio broadcast is an immaterial analog format.

Many of Chackal’s claims are clearly identifiable as claims about material recording formats. He writes that the vinyl condition contains four features: auditory, tactile, visual, and epistemic. Of these four, the last three are more rightly conceived as properties of material recordings, rather than analog or vinyl ones.

“Collectors store and handle vinyl with care,” Chackal also writes, “avoiding dust and fingerprints… the inserts invite the collector to engage them—to read the lyric sheet, to apply the sticker, to gaze at the artwork.” It is certainly the case that CD collectors also engage in similar practices to savor their collections—or suffer the consequences, as my collection of scratched CDs can attest.

It is difficult to find any correlation between the method of representation on a recording medium and any of these attributes. Chackal cites the example of the Rolling Stones release Sticky Fingers, whose original vinyl release contained an actual zipper incorporated into the jacket. Clearly, a material digital release of the album could also incorporate this feature. Indeed, in 1993, the Sticky Fingers album was rereleased in CD form with an operable zipper in the CD jacket.

And while listening to a digital stream of the album does not grant the listener access to these material qualities, neither did listening to the tracks when they debuted on radio stations in 1971. Again, these are qualities of material formats, not of analog ones.

Get your own CD with real zipper here.

Interestingly, Sticky Fingers was remastered and rereleased on vinyl in 2015, physical zipper included. Audio engineers almost certainly used digital processing during this recent remastering. (Typically, the (analog) magnetic tapes containing the master recordings are captured (digitally) into a computer, manipulated for sonic effect, and eventually exported into a format which can drive a record-cutting lathe.) So even though a vinyl record is an analog format, this particular reissue, like many modern vinyl records, is a representation of a signal that, at one stage of processing, existed in purely digital form.

But—oh no! Don’t these digital processing steps destroy some warmth or richness in the original recording? Let me assure you that they do not. We’ll talk more about why later.

Collectors may find it distressing that the recording on their vinyl record may once have existed as a digital file. But audio engineers consider this a normal state of affairs. Audio production involves many transformation and processing steps—some routine, but most prompted by aesthetic qualities we hear while working with the music. Here we see two material formats—the 2015 vinyl rerelease which is nominally analog and the 1993 CD release, nominally digital, but both are actually the result of hybrid processes. The vinyl is the result of digital processing along the way, and the CD is a digital capture (measurement) and remastering (more digital processing) of the original analog master tapes from 1971. Both formats are material, contain liner notes, have a fun zipper—and both will be converted to inherently analog sound waves at some point before they reach the listener’s ears.

So while material and nonmaterial recordings offer different experiences to a listener, we shouldn’t conflate these differences with the analog/digital distinction. Virtually all recorded audio from the past 30 years has undergone both digital and analog processing steps, regardless of final medium.

2         Auditory features of “the vinyl condition”

So far, we’ve seen that the tactile, visual, and epistemic qualities Chackal attributes to the “vinyl condition” are in fact qualities of material media formats, analog and digital alike. The remaining quality—auditory—is the one that has long occupied the center of the analog/digital debate. Chackal identifies three words often associated with analog audio: warmth, richness, and depth. Let’s look at these individually.

2.1         Warmth

First, he considers warmth to be a property of records “precisely because they are analog” (because digital formats are compressed and therefore “lossy”), and one that arises primarily or exclusively from physical instruments and physical format.

We’ll return momentarily to his claims about compression. For now, let’s focus on his claims about the two sources of warmth in records: the “play[ing] by a physical instrument upon a physical format”—so physical instruments being played (and perhaps directly captured from soundwaves into an analog format), and the physical interaction of the turntable stylus with the record.

Here, Chackal’s conflation of digital and immaterial becomes important. Many digital systems have material form as well. For example, two of the most popular recording formats of all time are both based on magnetic tape. The first, two-inch analog tape, was (and still is) a hugely popular format for multitrack recording of up to 24 distinct instruments or voices. The second, the ADAT system (Alesis Digital Audio Tape), recorded up to eight tracks of digitally encoded audio onto magnetic tape as 1’s and 0’s. But both systems are material: they capture soundwaves directly into a physical format and employ a recording/playback head physically pressed against the tape. Chackal’s account of warmth would seem to imply that because both systems are physical and physically interact with the recording medium, both should exhibit warmth. But notice that this has nothing to do with whether or not they are digital formats.

Chackal’s claim that warmth arises from physical instruments being played doesn’t seem quite right either. I take his designation “physical” to mean instruments which create sound vibrations in the air directly—another name for acoustic instruments. But not all physical instruments sound warm. Many musicians characterize instruments as existing on a spectrum from cold-sounding (clarinet, silver flute, piano, handbells) to warm-sounding (saxophone, oboe, electric guitar). The primary feature which places instruments along this scale is the presence or absence of certain harmonics, or overtones, in the audio spectrum.

A harmonic is an additional frequency of sound that an instrument emits in addition to the note being played, which is a mathematical multiple of that note’s fundamental frequency. Audio engineers typically understand warmth as the presence, in certain proportions, of these harmonics. Warm-sounding instruments emit these harmonics naturally, and analog recording, processing, and playback can enhance them or even add them to a colder-sounding instrument.

Because Chackal focuses on physical (acoustic) instruments, he also winds up excluding the electric and electronic instruments present on many undeniably warm-sounding recordings. An electric bass, for example, creates a negligible amount of acoustic sound. Its output is only an electrical signal, which is often recorded directly without ever being audible in a room. Or consider Herbie Hancock’s classic album Head Hunters, which contains many synthesizers. These electronic instruments have no moving parts whatsoever—they too create only electrical signals which are captured directly onto the recording medium without ever passing through air. If Chackal is right, neither family of instrument can exhibit warmth. But, as Hancock fans will adamantly tell you, this isn’t so.

I want to emphasize again that distortion is not necessarily a bad thing! I like the sound of analog distortion and the warmth it adds, as do many listeners and other audio professionals, and I take active steps to add it to recordings that I produce. In the 70s, distortion was an unavoidable side effect of using magnetic tape, vacuum tube equipment, and mastering to vinyl—but it added musical qualities to recordings that are still aesthetically valued today. In my digital recording studio, my recording system exhibits extremely low distortion, but I often take additional processing steps to add it for musical effect. Engineers often refer to using special equipment to “warm up” a sound source by adding this distortion.

You, as the listener, may elect to listen to the album I produced on vinyl on a vacuum-tube-based home stereo.  While this technology is considered obsolete in most other contexts, many audiophiles favor this type of equipment because it adds additional distortion to the playback experience.  This distortion creates additional musical harmonics that are not part of the recording I created, but may create a pleasant listening experience for you—and that’s great! An audio engineer’s view is simply that you are adding (desirable) distortion to your listening experience. (As an aside: Like garlic, there is such a thing as too much harmonic distortion. Taken to an extreme, warmth becomes “mush” or “fuzz,” though what counts as “too much” is highly genre specific.)

Not all turntables deliver high-fidelity audio.

Now we’re in position to return to the issue of compression. Compression here refers to a reduction in either the quantity or the numerical precision of the “measurements” that every digital recording comprises. But, contrary to Chackal’s claims, digital audio is not always compressed. In fact, the most popular digital format for more than 30 years, the audio CD, is uncompressed, and it remains the highest fidelity medium ever to achieve widespread consumer acceptance.

Digital compression can be characterized as “lossless” or “lossy.” You can think of lossy compression (such as the MP3 format) as similar to downsizing a digital image, which reduces its resolution for the sake of a smaller file size. Lossless audio compression is the equivalent of making a ZIP file: the data is compressed but is restored to its exact original when uncompressed. Because of this property, there is no difference between lossless compression and uncompressed digital audio—any more than ZIP-ing your term paper will make it shorter.

It is certainly true that you lose audio fidelity when using digital lossy compression. You won’t find me arguing in this post that a low-bitrate MP3 has the same audio quality as a CD. It doesn’t! This fidelity loss is rightly described as distortion—although crucially, it’s not harmonic distortion (not a frequency multiple of the fundamental), and thus doesn’t add warmth to the recording. But it is important to realize that not all digital audio is compressed, and not all compression introduces a loss of fidelity.

In short, distortion and compression are not the same thing. Compression is a property of some digital recordings (but not all, including the CD format). Lossy digital compression introduces undesirable, non-harmonic distortion, while lossless compression introduces none; and distortion can come from things other than compression. In many instances, distortion directly promotes warmth rather than inhibits it.

2.2         Richness

Next, Chackal discusses richness:

Because of record grooves, the sound of vinyl is more open, allowing a greater quantity of features to be heard. The space afforded by the grooves allows one to locate and individuate particular instruments and sounds and observe how they contribute to the music as a whole.

Here, he claims both that the recording contained on the vinyl medium allows for greater differentiation between instruments and sounds, and that the presence of record grooves is the reason this is so.

But if you walked into a typical recording studio (either now or in the heyday of vinyl), you wouldn’t see any machine that could create a vinyl record. The vast majority of recordings available on vinyl began as music on a multitrack magnetic tape, which was then mixed onto a “master” magnetic tape—generally through a process of multiple revisions and a final approval—before finally being fed into a record lathe which cut grooves onto a master disc. Through a series of additional processes, this master disc was eventually converted into a metal stamper capable of pressing vinyl records.

So the creation of a record and its signature grooves is actually one of the very last steps in the production of an album. We know that the original magnetic tapes were “closer to the source” and that they contained a recording which allowed the listener to “locate and individuate particular instruments,” and yet they did so without containing any grooves whatsoever. While these qualities may have been preserved in the final vinyl record, they clearly did not originate from that record’s grooves.

2.3         Depth

Chackal continues, defining depth as “how much of a sound or instrument can be heard” on a recording. He asserts that it is the physical format of a record’s grooves that afford it a greater depth compared to digital formats, which he again conflates with compressed digital audio. But, as with richness, if depth refers to an ability to better hear some aspect of an original audio source, then depth can only be lost—not created—as music is transferred from a magnetic master tape to a vinyl record. That is, the depth must have been present in the original master recording if it is to be heard on the vinyl record as well. And if this is the case, then the record’s grooves as such are irrelevant to the depth in this sense.

Furthermore, if depth cannot be captured in a digital recording, then Chackal will not actually hear much depth on his remastered 2015 vinyl recording of Sticky Fingers. Having been digitally remastered, this recording has “passed through” transformations into and out of digital form and so, on his own view, has lost its depth.

3         Conclusion

So is analog better than digital? Does a record sound better than an MP3 or a CD? The answer is not an easy one. The story is much simpler, however, if we embrace the idea I suggested at the beginning of this piece, namely that the playback of audio recording is the product of a long series transformations. We can accept that a musical performance was transformed through many processing steps to create a master recording. In general, the analog transformation steps add distortion, which many people find pleasant, and the digital transformation steps add very little distortion, unless intentionally designed to do so. Finally, as a listener, your choice of playback system adds additional distortions (even your bass boost knob is a form of distortion!).

In both the 1993 CD and the 2015 vinyl rerelease of Sticky Fingers, engineers made aesthetic choices to alter the sound using a combination of digital and analog techniques. None of the many rereleases sound identical to each other, and collectors argue about which sounds the best. But the differences in warmth, richness, and depth don’t simply stem from the final playback format, but from all of the decisions made by musicians and engineers along the way.

As an engineer, I try my best to produce a recording which makes aesthetic sense for its genre. (Modern metal albums have a lot less warmth than Neil Young records.) And in the final playback step, your choice of equipment puts the finishing touches on the audio you hear. You might wonder which system matches my intent as a producer or musician. My intent was to make a recording which could be enjoyed in many listening environments, so pick the one that works for you. If you want to hear the closest sound to what I heard, listen on CD, using a high-quality solid-state (not tube) amplifier and an accurate pair of headphones or speakers. But if you want to enjoy the experience, use any system that you like.

For what it’s worth, I have a vinyl collection too. I enjoy the format’s materiality, the ample room for artwork, and the ritual of using the turntable. My collection contains many classic records, some modern vinyl releases which were recorded or mastered digitally before being pressed in analog form, and even some albums that I produced for various artists. When I compare the vinyl version of an album to the original master recordings, I can hear the clear effect of the distortions endemic to the vinyl process. The music is fuzzier around the edges. Vocals tend to have more sibilance but less “air”. The added harmonic distortion creates a thickness and warmth to the recording that is not present in the master.

And you know what? It’s a great way to experience music. And on this point, Tony Chackal and I are in perfect agreement.

This was written in response to an earlier piece that argued for the superiority of vinyl over digital audio. See that post here.


Michael Thomas Connolly is a musician, recording engineer, and producer in Bellingham, Washington. He has more than thirty years of recording experience, from cassette decks to computer-based digital recording systems and analog tape machines. He runs the venue and recording studio Empty Sea Studios, and has engineered over 100 full-length albums. He has also performed as a touring musician and appeared on roughly 60 albums.

Splash image credit: Jack Hamilton on Unsplash

Edited by Alex King

21 Comments

  1. In my opinion, this is a fantastic overview of the digital vs. analog discussion. As someone who loved vinyl in the 80s when few agreed with me, I’m now convinced that three main things matter for enjoyable and accurate music reproduction:
    1. The quality of the system. It doesn’t have to cost thousands (if you buy used carefully), but it needs be quality.
    2. The stereo set up including the room itself. This is generally overlooked and it’s as important as the stereo system itself.
    3. The quality of the recording. And this is the biggest challenge if you happen to like some music that isn’t recorded well (and that’s most of it). But I find myself leaning towards better recorded music because it makes a huge difference.

  2. LaserDiscs are not digital. They are analog. They can contain a digital audio track in addition to the analog audio and video tracks, but the format is inherently analog.

    • You’re right – I was referring to the digital audio component but it is a rhetorically simpler case to simply talk about DVDs as a physical and digital media format. I’ll be revising as such.

  3. This was less a screed about digital vs analog than it was a screed attempting to debunk a particular work by another writer. It was a tough read even though it does make some good points here or there within the wall of text.

    It glosses over the digital “distortions”, almost whole-heartedly denying they exist (a digital copy only has one distortion, the first one at time of recording, is false), while magnifying the analog “distortions.” An MP3 is a massive, massive distortion to the music, as are limiters which seem to be employed more in digital recordings than on analogs. (Not that don’t exist on analog, just are less common.)

    Lastly, you cherry-picked The Rolling Stones as an example of something that was re-mastered digitally and then cut on to vinyl with a “so why bother” attitude when there are plenty of releases that stay analog throughout their entire chain. To not even acknowledge that suggests a clear narrative you’re trying to push versus having an honest discussion.

    • Thanks for reading. With regard to digital distortion – the recording-time distortions of timebase jitter and quantization error basically don’t exist in consumer playback environments – in the sense that the artifacts they create are well below ambient noise floors in most listening environments – whereas the analog distortions are often readily audible. THD itself is generally over 1% in a turntable, and less than 0.01% in my CD player. So I don’t think my emphasis on analog distortions is unjustified. Digital systems of CD quality or better are simply record and reproduce a more accurate copy of the input signal – by a lot. Note that this wasn’t necessarily true of first generation digital in the 80s and 90s – it took some time for the DAC and ADC designs to achieve the performance they were theoretically capable of.

      I specifically discuss the case of lossy digital compression, such as MP3, stating:

      “It is certainly true that you lose audio fidelity when using digital lossy compression. You won’t find me arguing in this post that a low-bitrate MP3 has the same audio quality as a CD. It doesn’t! This fidelity loss is rightly described as distortion—although crucially, it’s not harmonic distortion (not a frequency multiple of the fundamental), and thus doesn’t add warmth to the recording.”

      Conflating MP3s with digital recording as a whole is the equivalent of judging analog technology by plastic Flexi-Discs in magazines.

      As for the use of overlimiting – the Loudness Wars happened because digital made it possible, not because it is inherently a function of digital recording. With no needle to jump out of the groove in protest, engineers were free to over-limit recordings to an unpleasant degree. But this is not inherent to the medium, and is more a function of changing trends in the pop world than some statement about digital recording itself. Classical music recordings are a clear example of this, benefiting greatly from the expanded dynamic range and extremely low noise of digital.

      The Rolling Stones case and discussion of digital remastering was not to say “why bother,” but rather to point out that if such a record has warmth, clearly it was not destroyed by its trip through digital processing – refuting the common claim that digital recording cannot preserve the quality of warmth.

  4. When comparing lossless digital recording to a vinyl album, I feel the vinyl just sounds so much more open. For example, if I put a speaker in front of a window that’s open vs closed when listening to music, vinyl sounds like the window is open, digital sounds as if it’s closed. Just a guy who loves music!
    Have a great rest of your day!

  5. Thanks for your excellent piece. I have long thought there is an astonishing lack of technical understanding with vinyl enthusiasts, coupled with an enthusiasm I think stems from hipster nostalgia. But in any case I am in broad agreement with your larger aesthetic claims. I make some related arguments in my paper “Audiophile Aesthetics”; here if you are interested. https://stevenhales.files.wordpress.com/2019/11/audiophile-aesthetics.pdf

  6. Great article, thank you. Beyond CD’s 44k 16bit standard, what is your take on the increasingly prevalent 96/192k 24bit versions that some online sites are selling?

    • High sample rates (96 kHz and above) and bit depths (24 bit / 32-bit float) have definite uses in the recording studio. 24-bit depth translates to an increased dynamic range. While finished, mastered music can easily be fit into a 16-bit dynamic range, during music production, the additional dynamic range of 24-bit audio makes for a more forgiving medium. The analogy here might be a large financial broker on Wall Street that calculates to 100th’s or 1000th’s of a penny. This might be important when making a huge number of transactions per day, or calculating very complex financial formulae in which repeated rounding errors are problematic. For the consumer, 100th’s of a penny don’t matter. If you are listening to classical music, it is possible that a 24-bit recording will allow you to hear slightly ‘deeper’ into the reverb tail of a note before it disappears into the ‘noise floor.’ If you are listening to pop or rock, you are basically listening to perhaps 30 dB of dynamic range anyway, so even 16-bit depth is plenty.

      For sample rate – the math says that 44.1 / 48 Khz sample rates are sufficient to reproduce audio out to 22 Khz, beyond the range of almost all human ears. The reality is slightly more complex – Both A/D and D/A conversion rely on analog filters (implemented as analog circuitry) to limit incoming audio to the allowed frequency range, and to bandwidth-limit the output on playback. The ideal filter is a ‘brick wall’ which rejects all frequencies above the Nyquist frequency of half the sample rate – but making such a filter is very difficult. The task is easier if the Nyquist frequency is above audibility at 48 Khz (for a 96 Khz sample rate.) This simplifies the filter design. So a high sample rate may be desirable for this reason. However, modern D/A converters implement a different architecture which makes this issue less important.

      Practically speaking, it will probably take an extremely quiet listening environment to appreciate the difference in either increased bit depth or sample rate.

  7. Great article. I was familiar with virtually all of the technicals you presented (mechanical engineer with a background in test and signal processing, and also a music recorder and enthusiast). However, I don’t think I have ever seen them presented so succinctly and clearly.

    One additional point that I expected you to make is related to the physical limitations of the media. Vinyl LPs transcribe the sound waves into “bumps” in the grooves of the album. Lower frequencies (bass) create larger bumps. So a recording needs to be “mixed” for a vinyl media; too much bass and the needle won’t track properly. I think the “warmth” and “ openness” attributed to the vinyl media can often be found as an artifact of this “ vinyl mix”.

    Thanks for a great, down-to-earth explanation.

    • That’s a good point. Excessive bass, or stereo information in the bass, is not acceptable for vinyl – the bass is converted to mono below 100 Hz or so. High frequency limiting is also frequently used to address a host of issues including overheating the cutter itself! The relevant test, of course, is to do an A/B comparison of the vinyl record compared to a digital recording made from the _same_ master (with the vinyl considerations in place.) The remaining audible differences (and there will be some) can be attributed to distortions occurring during playback itself.

    • I was hoping on more on this exact point as well. Otherwise a great argument.

      Cheers from Bellingham, WA ^_^

  8. Not buying much of anything here. This is getting WAY over-complicated. The biggest error is the notion that digital recording is somehow inexact because it is based on a number of samples in time. The reality is that a digital signal must be converted back to analog in order for us to hear it. As part of that process, the complete analog curve is reconstructed in the analog domain. We do not listen to just a series of samples. More digital samples will mean that the reconstructed analog curve will be a more accurate reproduction of the original analog curve, but what we are hearing is a “complete” analog curve.

    One of the greatest virtues of digital recording, on a practical level, is increased accuracy of transient response. Music is always changing as it passes thru time, and great transient response captures sound far more beautifully than what happens otherwise. Analog recording methods like tape simply do not do this as well as digital recording can.
    One more comment. Of course there are LPs that can sound better than their CD counterparts. It all comes down to the judgment of the human beings who made the myriad decisions that go into how something will sound, not the media itself. Every recording ever made reflects “the human element” of judgment. That is unavoidable.

    • I believe we are in hearty agreement about these points. It sounds like you are rebutting the often-deployed ‘stairstep’ argument about digital recordings. As you point out, there are no stairsteps in a reconstructed digital recording (because a band-limited signal cannot contain such stairsteps.) But as I stated in the piece, the primary error in digital recording is quantization error, which is independent of sample rate and simply a consequence of having to assign a fixed amplitude value to the signal (the ‘nearest mark’ on the ruler.) This error can be thought of as a noise signal added to the original recording. Luckily, in 16-bit and greater bit depths, this noise signal is very small, and can be made even less perceptible by shifting its power bands to less audible parts of the audio spectrum through special dithering algorithms. Long story short, I never asserted that digital recordings result in a disconnected series of samples on playback – although it is a common myth about digital.

      We also agree about the improved transient response of digital versus analog recording systems. This is, a course, simply another way of expressing that digital systems have less distortion (anything less than the ideal transient response represents a distortion from the original signal – and there are no changes in the time domain without corresponding changes in the frequency domain.) While you characterize a crisp, accurate transient response as beautiful, it has also been called ‘thin’ and ‘sterile’ by those who prefer the ‘slower’ sound of analog recording and playback. And that’s OK! My goal in this piece was only to separate fact from fiction; our preferences are all personal.

    • Transient response? Interesting. So, like sharper attack?

  9. Bravo. Great piece, and extensively indicative of how challenging and inevitably imprecise the task of using language to define audio can be.

  10. Great article and excellent conclusion. I will continue to enjoy music on vinyl, digital and even cassette. All of it is music and however you enjoy it is fine

  11. And yet vinyl has its own special sound, it cannot be transmitted digitally

Leave a Reply

Required fields are marked *.