top of page

The Vinyl vs. CD Debate, Part 2:  Technical Factors  (September 2018)


Part 1 of this series presented arguments by both vinyl and CD advocates.  This article focuses on the points made under the heading Technical Factors.  One characteristic that all the technical factor arguments share is that they are factually correct.  Nevertheless, their implications are open to challenge by proponents from the other camp, and so I will include counterarguments where they exist (leading with “however” in italics).  Keep in mind that ramifications of some of the points raised may be so minute as to be undetectable.


Points Raised Favoring Vinyl

Quantization Errors in Digital Formats:  As discussed in the article Digital Audio Resolution, there are small differences between a digital value assigned to the amplitude of a wave and the wave’s true amplitude at the instant it is sampled, and consequently sound waves reconstructed from CDs do not precisely represent the original sound waves.  However, with 44,100 samples per second and each sample taking on one of 65,536 values, the differences between the original and reconstructed sound waves are exceedingly small and are unlikely to be detected.


Jitter in Digital Formats:  Jitter is a time disruption occurring in the digital-to-analogue converter (DAC), and the resulting “stutter” in what you’re listening to is very noticeable.  However, jitter is overcome in quality DACs that have an internal clock (i.e., that don’t rely on the computer’s or player’s clock), and so this becomes an issue only occasionally.  (I can personally attest to how jitter can destroy the listening experience.  An audio file I had been working on inexplicably began exhibiting jitter throughout the track, and I thought I had made a terrible mistake somewhere in my processing chain.  I spot-checked several other of the projects’ files, and they too were “infected.”  Fortunately, rebooting my computer had the effect of resetting the DAC’s clock, and all of the audio files again sounded fine.)


Amplitude Exceeding System Limits:  High-amplitude waves that exceed system limits are rounded in an analogue system as they approach the limit but truncated (“clipped”) in a digital one, and the resulting distortion is pleasant in the former but harsh in the latter.  This is certainly true when the amplitude has been increased or excessive compression has been applied without regard to exceeding system limits (see The Loudness War).  However, although clipped waves can occur in digital media, the media themselves are not solely at fault.  Rather, clipping results from inattention or inappropriate recording techniques and should be corrected before the audio track receives further processing.


Frequencies above 20 KHz Must Be Removed for CDs:  The creation of audio tracks for CDs does indeed discard frequencies above 20 KHz, and the reason is that, according to the Nyquist-Shannon sampling theorem (see Digital Audio Resolution), there cannot be any frequencies above half of the sampling rate (which would be 22.05 KHz for a CD’s sampling rate of 44.1 KHz) without aliasing, i.e., fragments of hypersonic waves that appear in the audible range.  A low-pass equalization filter is applied to rid the audio of frequencies above 22.05 KHz, but since such a filter isn’t perfect, filtering begins at 20 KHz.  However, it should be noted that 20 KHz is the upper limit of normal human hearing, and the loss of frequencies above that point would not generally be detectable.  Moreover, sound systems are typically capable of frequencies only up to 20 KHz, so any frequencies above that limit would not make their way to the speakers.


Points Raised Favoring CDs

Frequencies below 20 Hz Are Removed for Vinyl:  The creation of audio tracks for LPs discards frequencies below 20 Hz in order to eliminate artifacts with mechanical origins, notably turntable rumble.  However, 20 Hz is the lower limit of normal human hearing, and the loss of frequencies below that point would not generally be detectable.  As above, sound systems are typically capable of frequencies only down to 20 Hz, so subsonic frequencies below that limit would not make their way to the speakers.


CD Audio Is Not Subjected to RIAA Pre-emphasis and De-emphasis:  As noted in the article Equalization, high frequencies are amplified and low frequencies attenuated before LP stamps are created (“pre-emphasis” in accordance with the Record Industry Association of America convention), and these alterations are reversed (“de-emphasis”) by pre-amps when an LP is played.  The result is that what one hears sounds the same as it did before pre-emphasis was applied, but only if the pre-amp’s equalization circuitry faithfully applies de-emphasis according to the convention.  However, any differences are likely to be minute and will be overshadowed by other differences between sound systems.  Further, the listener typically has control of the player’s tone or equalization controls and can set playback to reflect what he or she personally feels is optimal.


LPs Are Subject to Tracking Error:  When a recording is cut to an LP master for stamping, the cutting stylus is mechanically transported along the record’s radius so that the stylus is always aligned in a perfect tangent with the groove being cut.  Tonearms, on the other hand, pivot from a fixed point, and the cartridge moves in an arc across the record.  Consequently, the stylus is aligned tangential to the grooves at only two points along the radius.  Everywhere else, the stylus tracks at a slight angle, and this difference can affect sound reproduction.  However, this is not a detriment (this is explored in more detail in Part 3, where it will make more sense than it would here).


LPs Are Subject to Inner-Groove Distortion:  As a stylus gets closer to the record’s label, the circumference of the grooves—and consequently their linear speed—decreases.  The vibrations cut into the grooves are necessarily closer together, and this makes it more difficult for the stylus to respond to them accurately, particularly with sound at higher frequencies and higher volumes.  However, mastering engineers mitigate inner-groove distortion by placing quieter songs with lower high-frequency energy at the end of each side.  (And see Part 3.)


CDs Are Physically More Robust than LPs:  Though not quite maintenance-free, CDs do not require cleaning before each playing, they do not acquire a static charge that attracts dust as vinyl does, and they are not permanently damaged as easily as LPs are.  Damaged CDs can be “regenerated” if the tracks have been ripped to lossless audio files.  (I have seen no counterarguments to these points.)


The remaining two of the arguments favoring CDs are based on objective measures characterizing the capabilities of audio media and systems:  frequency response and dynamic range.  These are discussed in detail below, and, to provide greater context, the discussion has been expanded to address media beyond vinyl and CDs.  Please be aware that data portrayed represent manufacturers’ measurements, and that those may have been taken under ideal conditions using professional equipment.  Actual mileage may vary.


Frequency Response

The range between the lowest and highest frequencies that can be reproduced by a system or from a medium is termed “frequency response.”  Those of several media are presented graphically in Figure 1.

PlayItAgainPaul - Media Frequency Respon

Figure 1.  Characteristic frequency responses for different media.


Here are my observations about this graphic.


Comparable Frequency Responses:  With the exception of cassettes and, of course, 78s, all media encompass the range of normal, undamaged hearing (20 Hz – 20 KHz).  In this regard, there is little perceivable difference between vinyl and CDs.


Flatter Frequency Response for Digital Formats:  Ideally, a system’s or medium’s ability to reproduce sound is constant across its entire frequency range, or at least comparatively flat.  In practice, though, there are variances.  Compared to the volume of a 1 KHz signal, the volume of other frequencies produced by analogue media may vary by 3.0 dB, a noticeable difference (at least if you’re paying attention to it).  In comparison, digital media have a much flatter response:  0.5 dB variance.


Cassette Limitations:  Cassettes are limited to 15 KHz at the high end rather than 20 KHz, unless the peak volume during recording is no higher than 10 dB below the maximum.  (One will certainly be tempted to compensate for the lower recording limit by turning up the volume, but doing so will also increase tape noise.)


High Frequencies:  While specially designed LPs and high-resolution audio files will play frequencies higher than 20 KHz (assuming they’re played on an exceptional sound system), there’s no guarantee that one will be able to hear them.  With high-resolution audio, an advantage of capturing hypersonic frequencies is that aliasing is reduced.


Low Frequencies:  Magnetic tape, CDs, and digital audio files are capable of reproducing frequencies below 20 Hz, though you may feel rather than hear them.  The lower limit for LPs was intentionally set at 20 Hz to avoid picking up low frequency “rumble”—mechanical noise generated by turntables.


78s:  To no one’s surprise, the frequency response of 78s is limited.  I included it to illustrate the progress sound reproduction has made since its very early days.  Even when played on a superior sound system, a 78 will still sound like a 78.


Dynamic Range

While the presentation of frequency response data suggests that most media are comparable, that for dynamic range shows marked differences.  Dynamic range is the difference between the softest and loudest sounds that a medium or system can produce, and those for different media are depicted in Figure 2. 

PlayItAgainPaul - Media Dynamic Ranges.j

Figure 2.  Characteristic dynamic ranges for different media.


There are obvious differences between the various media, but two things must be kept in mind before declaring a victor and abandoning the rest.  First, while human hearing has an exceptional dynamic range, we cannot hear sounds throughout that range at the same time.  For example, you won’t be able to hear a whisper while standing next to a rock band’s speaker.  The dynamic range for hearing simultaneous sounds is considerably less.  Second, all media portrayed (except 78s) will handle source audio from a rock band to a symphony orchestra (a rock band has a low dynamic range because the music tends to be uniformly loud).


LPs:  The dynamic range for LPs is 65-70 dB for outer grooves but falls off to around 55 dB for the inner grooves.  With a smaller circumference near the label, the distance for cutting vibrations into the grooves is less, and heavy bass notes and high volumes will interfere with stylus response.


Cassettes:  Cassettes’ dynamic range (50-56 dB) matches that of LPs’ inner grooves, and it matches LPs’ outer grooves with Dolby-C noise reduction (72 dB).


Magnetic Tape:  While magnetic tape nominally has a dynamic range of 60-70 dB, it is improved by about 7 dB with Dolby-A noise reduction and can climb to 110 dB with Dolby-SR noise reduction.


Lossless Digital Media:  The dynamic range of both CDs and lossless digital audio files (96 dB for CDs and 16-bit audio files, 120 dB for 20-bit files, 144 dB for 24-bit files) exceeds that of analogue media by a significant margin.  DACs will likely limit the dynamic range to 120 dB.


78s:  Although I included 78s as an historical curiosity, it stands out because of its degradation from 40 dB to 30 dB with repeated playing.  Shellac resin isn’t the most robust substance, and playing will scrape away portions of the resin, particularly on a gramophone with a steel needle and heavy tonearm head.

What’s Next


While CDs have an advantage over vinyl based on frequency response and dynamic range, the other technical factor arguments aren’t conclusive.  Moreover, perceptual factors in the listening experience have a profound effect on preference, and these are explored in Part 3 of this series.

Back to Paul's Blog & Contents

bottom of page