When comparing the technical parameters of vinyl playback with those of digital file playback, digital files are vastly superior to vinyl. In actual listening comparisons, however, things often look different. It is worth examining the contradiction a little more closely. Since every chain is only as strong as its weakest link, it is necessary to look at the main elements influencing the sound in the production and playback chains of both technologies, from the recording of the sound to its transmission to the amplifier of the home stereo system.
Of course, that means comparing apples with oranges. But that’s exactly what we audiophiles do every day when we switch back and forth between vinyl and file playback – just like the fruit connoisseur who, by reaching into the fruit basket, expresses his relative evaluation of apples and oranges.
I. PARAMETERS OF VINYL PLAYBACK
Since 1948, the modern vinyl record has been a round disc-shaped sound carrier made of polyvinyl chloride (PVC), in which the sound information is stored in a spiral-shaped groove whose orthogonal flanks (in the case of the stereo record) map the sound vibration of the signal. It is usually scanned from the outside in by a needle. In this respect, it is not only the elements for recording the sound and producing the record that influence the sound result, but also the elements for scanning and playback.
Graph 1: Schematic tracking of a long-playing record
Graph 2: Tracking the Grooves of a Long Playing Record
a) Sound recording and production of a record:
The information stored on the record (apart from exceptions, such as the direct cut process) has passed through the following stages (assuming an analogue sound source):
- Microphone,
- Analogue microphone amplifier,
- Analogue mixing console,
- Analogue tape recording machine,
- (Audio) mastering machine,
- (Technical) mastering and cutting equipment,
- Galvanic process and
- Pressing machine
Except for the galvanic process and the pressing machine, which only have an influence on the geometrical and surface creation parameters (and thus on the quality of the tracking and playback), all other stages have more or less influence on the acoustic parameters of the recorded sound information itself.
b) Tracking and Playback:
The following are relevant for the tracking of the information on the record:
- the turntable,
- the cartidge and tonearm
- the phono preamplifier
In the end, the vinyl record is a medium which, as the end product of a long chain of production and reproduction steps with greatly differing technical parameters, does not significantly exceed the following parameters on average:
- Frequency range: 30 – 19,000Hz (+/- 3dB)
- Dynamic range: 50 – 60dB (20Hz-20kHz)
- Wow and Flutter: 0.06
- THD: 0.1% (at 1kHz)
- Crosstalk attenuation: 35dB (20Hz-20kHz)
The magnitudes of these data are more important than the exact values for individual components. The data are also not always directly comparable. And of course there is a wide range of quality in each category of equipment – both on the recording and playback side. This is not about the limits of what is technically feasible, but about the data in a representative cross-section of the upper quality segment with which audiophile listeners have or might have experience (the devices used as data references are listed in the appendix).
II. PARAMETERS of Playback of PCM 96kHz/24bit Files
A file is a collection of related data with a defined data format (e.g. WAV, Flac, mp3) in the form of a one-dimensional string of bits that are usually interpreted by an application in byte blocks. Digital music files are – unlike the tracks on a record – files that are independent of their carrier medium. Nowadays, they are mostly stored on magnetic data carriers (e.g. HDDs) or flash memory devices (e.g. USB sticks, memory cards, SSDs). In the future, the carrier medium can change at will without making any difference to the electro-acoustic properties of the file. Since in the digital domain each copy is identical to the original, the degradation of information quality with each copying step known from the analogue domain is overcome. This makes it possible to use the Digital Studio Master of the respective sound recording at the customer’s home.
Graph 3: Masteringscreen of Wavelab
a) Sound recording and production of a digital master:
The information stored in the digital master has passed through the following stages (assuming an analogue sound source):
- Microphone,
- Digital microphone amplifier,
- Mixing console (often software-based) and
- Digital audio workstation (DAW)
b) Playback of a Digital Master:
For the playback of the digital master, only the digital/analogue converter is relevant in terms of sound quality. The sound characteristics of the playback software or the operating system, which many users like to discuss, are – assuming correct system settings – not among the elements influencing the sound quality. The equally passionately discussed influences of power supply and cable quality are undisputed, but not taken into account here (i.e. also not on the analogue side).
In the end, the digital master in 96kHz/24bit format is a medium that, as the end product of a chain of production and playback steps with partly different technical parameters, does not significantly exceed the following parameters on average:
- Frequency range: 3 – 48,000Hz (+0/-3dB)
- Dynamic range: 120 dB (20Hz-20kHz)
- Wow and Flutter: 0.0%
- THD: 0.003% (at 1kHz)
- Crosstalk attenuation: 100dB (20Hz-20kHz)
Digital music reproduction (with at least 96kHz/24bit) is thus vastly superior to analogue playback in all usual technical parameters. In general, one can say that the far greater limitation of signal fidelity is on the playback side. The recording side is and always has been in the past of much higher quality than even the best high-end playback systems of consumers at home were or are capable of reproducing.
Of course, one could argue that the technical parameters cited here for describing sound quality (frequency range, dynamic range, distortion, etc.) are not sufficient to fully describe the sound quality of a recording and playback system. There is some truth in this argument. There are many examples of systems with the same measurement results but very different sound. But the proponents of this argument would then also have to make suggestions as to which parameters are not recorded with the usual measuring methods and what changes need to be taken.
III. Sonic effects
In view of these clear technical advantages of digital playback, where does the subjective preference of many people – above all audiophiles, and we count ourselves among them without reservation – for record playback come from?
a) On the part of vinyl playback:
On the one hand, the differences in sound quality between the digital versions (CD, SACD, DVD-A, download) and the analogue versions (LP) of the same album are mostly due to different masterings and not to inherent differences in sound quality between the various carrier media. As a consequence, the original masters for the different digital versions differ among one another and also from the analogue version. These different masterings are partly due to the attempts to cater for the taste of the respective target audience, but to a large extent they are simply technically necessary, as it is e.g. enormously demanding to modulate a frequency band of the scale of the human hearing in record grooves. These include, among others:
- Cutting distortion (today mostly according to RIAA specification) to deal with the problems of amplitude and space consumption at low frequencies and lack of dynamic range at high frequencies.
- Steep limiting of treble at 18 or 19 kHz to avoid intermodulation problems
- Reducing frequencies below about 500Hz to mono and placing low frequency sound events in the sonic ‘middle’ to avoid problems in the editing process.
- Steep limiting of frequencies below 30 or 40Hz to optimise the time/dynamic balance of the record and avoid tonearm resonance problems later on
The digital version of the same album has usually a lot more leeway than the vinyl version in terms of dynamics, frequency range and stereo information. It should also not be forgotten that even between different masterings of the same vinyl album there can be considerable differences in sound quality. Such different masterings result i.a. from the regional distribution of pressing plants (an album master made in the USA, for example, is pressed from the original tape in the USA and from later generations of the master tape in Asia or Europe) or from – especially in recent years so popular – reissues. However, these differences do not result in a subjective advantage for one of the versions, because there is no reason to assume that the mastering for the analogue version is fundamentally of a higher sound quality than for the digital versions.
On the other hand, the preference for record reproduction is probably based not insignificantly on a kind of ‘misunderstanding’. Interestingly, it is precisely the technical limitations of record production that are perceived by the human ear as ‘better’ or at least ‘more pleasant’. Although audiophiles seek to come as close as possible to the ideal of the ‘original sound’, in reality they move away from the original sound with their preference for record reproduction and towards a subjective sound ideal that is perceived as psycho-acoustically particularly ‘pleasant’ in the environment of their own listening room. The reasons are manifold:
- System-related noise in analogue technology has (in moderation) the effect of increasing spatiality and is perceived as pleasant.
- Harmonic distortions of record playback (even and odd-numbered multiples of the respective frequencies), in which artificial harmonics are created (e.g. more brilliance in the high frequencies), sound more sparkling and livelier. Even-numbered partials (including especially the first octave), for example, are very popular because they can make sounds seem warmer, but also clearer and more contoured – a main reason why certain tube devices are still popular today.
- Smoother clipping behaviour of the record when overdriven leads to softer, more analogue sound,
- The lower crosstalk attenuation of analogue sound recording leads to a less precise stereo imaging, which can sound more pleasing (this is also the reason why, for example, some miking methods are preferred to other, more technically precise ones)
- Dynamic compression makes listening at home much more pleasant. Otherwise, the listener would have to constantly turn up the volume for quiet passages and turn it down again for loud passages.
- The compression effects also give the music more punch (not unlike mastering in the so-called ‘loudness war’).
- In technical mastering (i.e. the preparation for cutting a record), the treble is often steeply limited at 18 or 19 kHz for technical reasons. As a result, the music ‘breathes’ a little less, but seems warmer and more ‘analogue’.
- Non-linear frequency responses in the components of record playback often lead to (and are often deliberately used by the manufacturers as) a tuning of the sound that is pleasant for the listener but alien to the original sound.
- This is confirmed, for example, by Jim Anderson, Grammy-winning recording engineer and professor at the New York University Clive Davis Institute of Rescorded Music: “I think some people interpret the lack of top end [on vinyl] and interpret an analog type of distortion as warmth. It’s a misinterpretation of it. But if they like it, they like it. That’s fine.”
The higher the quality of the components of the record playback system, the smaller any sound distortions. Compared to digital playback, record playback requires an extremely large effort with regard to the playback chain in order to keep the physical deficiencies of the analogue LP as small as possible.
b) On the part of digital file playback:
While the technical peculiarities of vinyl records have an advantageous effect – even if not necessarily in accordance with the original recording, but at least in accordance with the listening psychological perception in the private listening room – this is exactly the opposite with digital file playback. Here, the technical peculiarities of digital playback often lead to a more unpleasant listening experience. In particular, the anti-aliasing filters used in digital-to-analogue conversion exhibit so-called filter ringing: Linear-phase filters have a symmetrical impulse response. This means that they do not only have a natural post-ringing, but they also exhibit an strange pre-ringing that is mirror-symmetrical to the arrival time of the signal, which is just as long and completely unnatural. Both oscillation processes are longer the higher the order, i.e. the steeper the filter is (see diagram 4). Exactly these signal distortions lead to “smearing” in the time domain for short transients. This often makes digital files sound hard, over-transparent and unnatural – as if the music were playing behind a well-cleaned pane of glass.
Graph 4: Symmetrical impulse response at different sampling rates
The ideal solution would be one that uses the superiority of the digital format without its disadvantages. Something like this is possible, of course: Non-Over-Sampling (NOS) DAC technology does not use any anti-aliasing filters during digital-to-analogue conversion and produces a particularly natural (analogue) sound image.
Whether you like vinyl or not, whatever you prefer is fine. We love vinyl. Besides the purely sonic aspects, there are also the physical and ritualistic aspects of handling a vinyl record: Apart from the fact that there are rich record collections built up over decades and with them shorter or longer stories about the respective pieces, there is something very special about searching the spines of record covers on the shelf for just this one record with just this one piece, looking at the large-format cover art, taking the record out of the sleeve , cleaning it and watching the needle touch down on the groove and hearing that special hissing/crackling in the run-in groove before the piece starts …
© Alexej C. Ogorek
