One of the most controversial topics in the audiophile universe is the digital versus analog debate. After the introduction of the compact disc in the early 1980s, the sales of analog music formats (LPs and cassette tapes mainly) declined steadily until 2007, when there was a revival of interest in vinyl. Since then, the market for vinyl LPs has seen a double-digit percentage rise each year, whereas CDs are gradually being replaced by music streaming, such that the value of LPs sold in 2020 exceeded that of CDs for the first time since the 1980s. Even the compact cassette tape is making a comeback, with the recent resumption of blank tape manufacturing.
There are several reasons for the revival of vinyl LPs. While some audiophiles claim that LPs sound better than digital formats, sales growth is being driven not by audiophiles (who only represent a small fraction of the record buying public and are mostly more mature adults), but by young music lovers getting into the format for the first time. Music has become a commodity, something that is so easily available from streaming sites, and this allows music lovers to acquaint themselves with music from years past. My son listens pretty much to the same music as I did at his age; he is no longer confined to what is on the charts. He can explore and decide for himself what he likes. The revival of interest in the music of times past also stirs up the desire to own the music in the formats that people used to have in that era. I can still remember the thrill of opening a copy of Pink Floyd’s The Dark Side of the Moon that I bought with the money I earned at my summer job. There is no such thrill when I click a song to play on Tidal. The whole package, with the illustrated jacket, the black disc, the printed lyrics and posters instill a pride of ownership. As audiophiles, we should thank these consumers for giving a reason to the record companies to continue manufacturing vinyl LPs.
Going back to the question of which is better, digital or analog, this is not an easy question to answer and depends on the perspective of the user. If you ask a professional (recording and mastering engineers), you will probably hear a completely different answer than if you ask an audiophile. It is not so much due to the differences in how these two groups of users evaluate sound quality, but due to their very different experiences in using these two technologies. The experience of a typical audiophile is often limited to compact discs, SACDs, high-resolution PCM and DSD file playback, streaming, and vinyl LPs. For the professionals, it is the high-resolution formats (nobody records in Red Book format anymore) with the associated hardware and software, versus analog tape and the associated analog hardware. While audiophiles only care about the sound quality of the end product, professionals have to take into account the production process.
I am not a professional, but I have been making recordings for many years as a hobbyist, so I have some idea about the production process. Music production nowadays invariably involves multiple tracks, and digital technology has made this infinitely easier. All the mixing and editing can be done in post-production with a digital audio workstation and computer. A large variety of plug-ins are available to apply different effects to the sound. Expensive hardware is no longer necessary; there are plug-ins that emulate the sound of famous vintage microphones, plate reverbs, compressors and so on. And the changes made are fully reversible, whereas a bad splice of the analog master tape can become a disaster. It is like the difference between using a typewriter and Microsoft Office.
The danger is in relying too much on post-production and not paying enough attention during the recording process. During the early stereo era, sessions were recorded onto two or three-track tape recorders. Some companies such as Mercury only used three microphones, and the three tracks from the microphones were mixed down to stereo during mastering. Companies such as Decca that used multiple microphones would mix the tracks in real time into stereo during the recording session. This meant the balance engineers had to get everything right during the expensive recording sessions, as there was no way to remix the tracks afterwards. Microphone placement was of paramount importance. After the introduction of multitrack analog recorders, mixing could be done during post-production (and Dolby noise reduction, introduced in 1965, aided in the process). However, tape was (and still is) expensive and editing must be done manually (by cutting and splicing the actual tape!), giving the engineers the incentive to get everything right during the recording session. As my partners and I always record in analog (with digital as backup), we are well aware of these pitfalls.
However, I have attended professional multi-miked recording sessions where one microphone was used for each player of an orchestra, placed casually and without regard for phase cancellation. The idea is that everything can be corrected during post-production, which is actually not true. The natural acoustics of the recording venue and the perspective of the orchestra can never be re-created by simply mixing the individual instruments together. This might be one of the reasons why recordings made 60 years ago still sound better than many modern recordings, despite the technological advances that have happened since then.
Old analog recordings often sound better because of how the music business is run today. In the past, large labels had their own recording teams with highly experienced recording and mastering engineers, along with an apprentice system to train the next generation. The engineers were intimately familiar with the recording venues and produced consistently excellent recordings. During the heyday of the music business, labels were able to make good profits from record sales. Nowadays, the revenue stream from sales of physical media has dried up, and the income from streaming is miniscule. Recording projects are often outsourced to the lowest bidder, and artists sometimes have to pay for the recordings themselves. Nobody can afford to take on projects such as Decca’s Wagner Ring cycle.
Another reason why early stereo recordings are often better is that the record-buying public in that era cared about sound quality. Buying a stereo system involved a significant financial outlay, and there was no distinction between “audiophile” and consumer equipment, at least not until the Japanese companies entered the market with mid-fi and mass market products in the 1960s and dominated it in the 1970s. In other words, anyone buying LPs or open reel tapes in those days was what we would now call an audiophile. All major classical labels were in effect audiophile labels, and sound quality was a major selling point, in addition to the quality and reputation of the artists.
Music nowadays is mostly played on smartphones, car sound systems and computer speakers. The number of people who still sit and listen in front of a stereo system is very small. Music is therefore mastered in such a way so as to optimize the quality when played through these modern means of listening. That means compression is used so that soft passages can be heard even in noisy environments outdoors or in a car, and equalization is used to compensate for the limited bandwidth of these devices. This obliterates the dynamic shading and tonality of the music when played through a high-quality stereo system.
This is not to say that there are no high-quality recordings being made nowadays. Many small independent labels still produce recordings with sound quality in mind, using the latest high-resolution digital technology. Ironically, some engineers feel that passing a digital recording through analog tape makes it sound more natural. This might have to do with the higher noise floor of analog tape. This noise mimics the background noise of natural acoustic environments, whereas the almost noise-free background of digital recordings actually sounds unnatural. There are now plug-ins that add tape noise, tape saturation and other analog artifacts to digital recordings!
Many people have offered opinions as to why digital recordings do not sound as good as analog in their estimation. As I have very limited technical knowledge of digital audio technology, I will not comment on the merits of these arguments. Through the monitoring system we use during recording sessions, switching from the live feed to high-resolution digital, especially DSD, sounds indistinguishable to me. However, during playback at home, the tape often sounds more dynamic and natural, but this could be due to the quality of the playback equipment, as I have not invested anywhere near the same amount on my digital front end as on my analog front end.
For the audiophile, comparing analog and digital often comes down to a comparison between LPs and CDs or high-resolution digital formats. Again, the quality of the respective playback equipment matters, and for LPs, proper set up of the record player is a must. The question is, do LPs represent the best analog has to offer? LPs have a lot of inherent limitations. The linear velocity of the groove decreases towards the center of an LP, and the lower velocity at the end of a side leads to an increase in distortion and makes tracking more difficult. For symphonic music, it is often the end of a piece that has the greatest dynamics, right where the groove velocity is the lowest. Compression (dynamic range limiting) is therefore often necessary to prevent mistracking. Longer pieces require narrower grooves to fit onto one side of an LP, which again can require compression.
The whole process of LP production involves multiple steps, with potential for sonic degradation at each stage. LPs that are made with new stampers sound better than those made with worn out stampers. Background noise is a function of the quality of the pressing process and of the vinyl material. An off-center spindle hole will cause pitch instability that is more evident with certain instruments such as a piano. Only when all the stars are aligned will one get a perfect record. Digital recordings, on the other hand, are always consistent. They sound the same whether you have played them once or a thousand times. Whereas music with limited bandwidth and dynamic range, such as a folk singer with a guitar, might sound better on an LP, a Mahler symphony will almost certainly sound more dynamic on high-resolution digital, given the superior signal to noise ratio and dynamic range of digital recordings compared to LPs.
There are around 60 recordings for which I have both the LP and a copy of the master tape (mostly copies of the production or safety masters). Most of these are Decca, EMI and RCA recordings from the late 1950s to mid-1970s. In no case is the LP superior. In over half, the quality gap is wide, and all the LPs that sound close to the tapes are modern reissues. Certain prominent magazine reviewers past and present have touted the superiority of first pressings, and some of these now cost an arm and a leg as a result. Examples include some RCA Living Stereo “Shaded Dog,” (Nipper, the RCA dog, is pictured against a shaded red background; later “Plain Dog” pressings have a plain red background), Mercury Living Presence and “wide-band” Decca LPs (so called because the silver band on the label that says “Full Frequency Stereophonic Sound” is wider).
In my experience, these old LPs rarely live up to their reputation, which I don’t find surprising. Vinyl record production technology has advanced by leaps and bounds since the late 1950s, so it would not make any sense that these ancient LPs should be better than those reissued today, unless the master tapes have significantly deteriorated. The original issues were also made in larger numbers, whereas modern audiophile reissues are made in far smaller quantities, with smaller production runs from each stamper to ensure more consistent quality. Rather than spending the money on these vintage collector’s items, why not spend the money on reissues to support today’s manufacturers and ensure they will continue to be available in the future?
So, here are my conclusions. Assuming the quality level of the playback equipment for digital and analog is comparable, I would go for a digital format if the original recording was in digital. It makes no sense to me to produce an LP from a digital source (except for DJs who use turntables for scratching). For music that was originally recorded in analog, the choice comes down to the type of music. For music that is large scale and dynamic, I would go for a high-resolution digital remastering as long as it was done correctly, in order to avoid problems associated with LPs such as end-of-side distortion, compression and noise. For other forms of music, it comes down to the quality of the LP pressings versus the quality of digital remastering. Given a choice, I prefer the DSD format. Other than any simple “splicing” or editing that might need to be done, DSD must be converted to PCM (usually in 24-bit, 352.8kHz, also called the Digital eXtreme Definition or DXD format) for editing before re-converting back to DSD. Whether this causes any appreciable loss in quality is debatable. For conversion of analog materials to DSD, it is best to do the remastering in analog domain before conversion.
Dealing with music originally recorded in Red Book CD standard (16-bit, 44.1kHz) is another matter. Early digital recordings suffer from a loss of low-level detail. In an article by a recording engineer about his early experiences with digital recording, he talked about the way he heard the steps of the recording artist as she entered the studio; on the analog tape, he could also hear the reverberation following each step, but on the digital recording played back at the same level, he could only hear the feet striking the floor but not the reverberations. This loss of low-level information is what makes early digital recordings sound unnatural and less dynamic when compared to analog tape. Early analog to digital converters had an effective bit depth of only 14 bits even though 16 bits were specified. This gave a dynamic range of 84 dB, and if overloaded would result in highly unpleasant non-harmonic distortions. The Nyquist limit (the highest frequency that could be encoded without aliasing, which is half of the sampling frequency) of 22 kHz is just at the limit of the audio band, thus requiring steep anti-alias filtering before digitization. These steep analog filters can introduce amplitude and phase non-linearities as well as ringing. The eventual adoption of oversampling allowed the use of more gentle filter slopes. Unfortunately, the loss of low level detail and the artifacts introduced by anti-alias filters cannot be undone during remastering. We therefore have a decade’s worth of recordings that will always remain problematic.
I have bought very few CDs over the years; most of my digital music collection comes from converting LPs and tapes to DSD, and from high-resolution downloads. For standard Red Book listening, either from CD rips or streaming services, I prefer real-time conversion to DSD128 during playback using Audirvana software.
There are many recordings made during the golden age of music performances, decades before the digital era, featuring artists such as Furtwangler, Walter, Kleiber, Callas, Oistrakh, Kogan, Du Pre, Cortot and Richter to name just a few. On the rock music front, most of the important releases from the Beatles, Pink Floyd, Led Zeppelin, Hendrix and other greats were made during the analog era, not to mention many classic jazz and blues recordings. Some of these have been remastered into new LP and digital releases, but we are at the mercy of the mastering engineers, since many of the original artists are no longer with us and cannot ensure that their original intent is properly preserved. This is why many people still seek out the original LPs. Some of these recordings are now being released in open reel tape format, mostly 1:1 copies from master tapes (copies made at the original playback speed rather than high-speed duplication, which is used for expediency but can create sonic degradation) and without additional manipulation. In my view, this is the ideal format for preserving recordings from the pre-digital era. I will further discuss this new trend in future articles.
Header image courtesy of Pexels/cottonbro.