When the term ‘definitive’ is used as a descriptor it implies that the subject or object in question could be viewed as complete, baring the stamp of authority or as a benchmark from which all future iterations would be referenced. In relation to a reissued recording artefact the term has frequently been used in conjunction with terminologies such as ‘anniversary’, ‘deluxe’, ‘digitally enhanced’ and ‘remastered’. In the world of film such reissues are often marketed as ‘The Directors Cut’, re-edited to show alternate and extended scenes. This style of remastering is more than merely digitally repairing and enhancing an original movie reel, whereby fixing damage to frames and enhancing faded colour details due to the processes associated with poor storage and aging could be construed as beneficial to the archiving process.
Although this paper focuses on the remastering of ‘classic’ albums, debate over the definitive edition in music is not just a modern concept reserved for recorded media. The hand-written score has for centuries been the basis upon which scholars analyze, define, and attribute music to its composer and historical period. Studying the handwriting of various editions of the same score show how composers are known to have edited and revised their works over time.
Later, many scores would be embellished with editorial notes on the ‘ideal performance practice’ where academics would declare how ‘classical music’ should be performed. This would include notes on fingerings, pedaling and dynamic markings in a piano score for example (Stockfelt, Brown and Volgsten: 2005, pp. 318-321).
The Fontana edition of Chopin’s Fantaisie Impromptu shown in figure 1 is an example of how various editions could be manipulated containing significant alterations, including optional replacement musical bars and phrases.
Figure 1: Example of alternative phrase by the editor Klindworth.
The desire to archive and preserve musical performances as and when they occurred was a natural corollary to the advancements made in the recording field during the 20th Century.
The sonic limitations of consumer playback systems and the susceptibility to degradation of early-recording media motivated a revolution to find a media storage format of superior quality that was immune to degradation. The digital revolution, which developed during the 1970’s, finally found commercial appeal with the release of the Compact Disc into the marketplace in 1982 (Philips Research online). With this development came a resurgence of sales of artists back catalogues as people wished to experience music, once only available via vinyl or cassette tape, with convenience and a clarity and dynamic range never before achievable in the home environment. Since their initial digitizations a number of popular ‘classic album’ recordings are being re-marketed every few years with the promise of improved sound from a new digital re-master, anniversary re-master or definitive edition remix. As part of the marketing strategy, the music is often repackaged to attract new audiences or to sell the notion to existing fans that the reissue is in someway superior to any previous edition. Andrew J. Bottomley supports this view in his article ‘Rock Music Reissues and the Production of the Past for the Present’ (2015). Bottomley defines such reissues as ‘quality historical’ and explores the notion of authenticity of the complete package in the context of transformations of historical meaning and cultural status.
These subsequent editions often become the only recording source available for purchase consigning any previous editions to the category of ‘rarities’. This marketing practice fuels the debate about sound quality, in particular the issues surrounding the ‘loudness war’. This is the description given to the increasing average level in dB of music production recordings through means of over compression culminating in a detrimental impact on dynamic range (Katz: 2007, pp.167-176).
Within this study, the focus of research is on changes in frequency contour across the audible spectrum between editions, and how historical recordings are susceptible to being filtered through superimposed stylised, fashionable interpretations.
Authenticity and authorship in record production and remastering
In the context of popular music there are many definitions of authenticity. One such definition is that ‘authenticity’ is ascribed to rather than inscribed, that the audience, the consumer or recipient of recordings perceives authenticity. Allan F. Moore categorises authenticity into first, second and third person and largely deals with the question of whom, rather than what, we are authenticating in terms of a stylised performance and its affinity with an origin.
For the purposes of this research, ‘authenticity’ is defined as the quality of accurately reflecting a model or exemplar. In the case of a re-mastering process, the ‘origin’ or ‘exemplar’ could be considered to be the first (or original) master, and in the case where analogue tape masters were later transferred to vinyl, the analogue tape version could be considered the most authoritative version with respect to the recorded artist’s intention, and therefore most ‘authentic’ as ascribed by the recorded artist. What may constitute the most authentic version from the perspective of the end consumer is less clear.
Within the context of this study, the concept of authenticity in record production and remastering will be limited to the expression of the sonic qualities of the source materials and any deviations from them. Authenticity in this case, is more closely aligned with the definitions used in audio forensics, where by there is an original source recording, featuring no changes, additions or deletions, from which any subsequent versions are compared. Details of equipment and components used in the recording process and care to ensure that playback systems are set up accurately also form part of the authentication process. How close those subsequent versions are to the original is a measure of how authentic they are.
The exploration of authenticity and its complex issues such as how, and by what means, listeners engage with music products is beyond the scope of this research. However, it could be argued that there are many initial authors involved in the production of original masters including the songwriters, performers, engineers and producers. The level of engagement of these original authors in any remastering process gives the new edition greater affiliation with the original. In this paper, the notion of authenticity as a measure of proximity to original exemplars is considered through a reflection on the level and nature of engagement by original authors in the mastering process.
The integrity of an audio file can still be intact without being identical to the source, for example, an MP3 audio file may be compressed to contain only a percentage of the original audio data yet is still recognisable as a particular recording. Advancements in digital processing power, higher quality audio components and the ability to digitize at 24bit/192kHz sampling rates have ensured that we can represent an analogue source signal with greater transparency and resolution compared to original 16bit/44.1kHz resolution used for CD. Although never truly equivalent, this advanced digital technology provides better tools to preserve more accurately an analogue recording with its original sonic character intact before any reinterpretations are applied. Any technology used for the recording and reproduction of sound imparts its own character on the content and its application cannot be considered neutral. An understanding of this concept has influenced sound recordists use of traditional and authentic technologies in the production and reproduction of sound for historical accuracy. Examples can be found in the recording of the soundtrack to the film The Kings Speech (2010). Genuine 70-year old microphones originally used for the Royal addresses, were used at Abbey Road studios to record the soundtrack to the movie. In 2014, the music and vocals for the UK TV serialised documentary about the life of the 1960’s UK recording artist Cilla Black were also recorded at Abbey Road using the very same stock microphones used during that period. This sonic fingerprint of a recorded work not only defines its time in history but can also identify the preferences of the creative audio engineers and producers.
One of the most recent examples of an authentic approach to remastering can be found with the reissue of The Beatles In Mono vinyl box set (2014). Abbey Road engineer Sean Magee, supervised by Steve Berkowitz, digitized the original mono master tapes at 24bit/96kHz. They used authentic vintage EQ to cut new lacquers using original mastering notes, cross-referencing with original lacquer pressings. Although some technical elements had changed over the past 50 years, for example the replacing of tubes with solid-state components in the cutting chain, there was clearly a sympathetic approach and an understanding of the historical value in this reissue. Therefore, if this contextual approach can be afforded to a catalogue as significant as The Beatles’, to what degree should sonic alterations of any archive recording be made when a new anniversary edition is re-marketed by any record company/artist?
Audio extracts for analysis
To explore the sonic qualities of various editions of the same product, an analysis and comparisons of a number of audio files from artists including The Beatles, ELO, Yes, Genesis, David Bowie, Kate Bush, Peter Gabriel and Rush were considered. These artists in particular have shown to exert some production control and exercising authorship over their body of work and therefore this provides a strong basis for assuming artistic integrity in their output.
When EMI approached David Bowie with the idea of remixing a number of his albums into 5.1-surround sound, the artist stipulated that only the original producers of those records must undertake this (Scott: 2012, pp.186-188). Therefore, it could be concluded that there was an appreciation of the historical context of these productions, and that the original audio would be entrusted and treated most authentically by those involved in the original process.
The 2010 re-release of David Bowie’s Station To Station (1976) box set, including a 5.1 remix, provided a wealth of detailed information about the original recording process and the inclusion of various other editions ranging from a 180gm vinyl pressing (flat transfer from original source tape), 1985 RCA CD master, 24bit/96kHz flat transfer masters and 2010 remix by the original producer Harry Maslin.
With access to two different vinyl pressings, 30th and 40th Anniversary editions of the album The Rise And Fall Of Ziggy Stardust And The Spiders From Mars (1972), plus a 1998 remix of the song ‘Moonage Daydream’, a presentation and discussion on the variations within these masters with the original producer Ken Scott proved immensely useful in ascertaining which sonic qualities pertain to the most authentic edition of this record.
I had a rare opportunity to meet and interview the world-renowned songwriter and record producer Jeff Lynne in March 2014. This provided a unique insight into the creative decisions behind his production style and shed light on his decision to re-record some of ELO’s greatest hits for a new compilation released in 2012. Using the production sound of the song ‘Mr. Blue Sky’, issued originally on vinyl in 1977, followed by the first digital transfer to CD in 1987, then the 30th anniversary edition issued in 2007, and re-recording in 2012, it has been possible to combine an objective auditory assessment, with primary research from Jeff Lynne, to ascertain which edition is the most authentic production master.
After performing some basic listening tests on examples of music from the various editions of Bowie and Lynne’s back catalogue, a selection of tracks were chosen on the basis that they were rich in sonic content and that there were clear audible differences between them. The initial process included importing the audio into a Digital Audio Workstation (DAW), in this case Pro Tools, so that all of the editions could be synchronised on separate tracks for easy A/B comparison. In order for the highest resolution to be maintained, the analogue extracts were imported at 24bit/96kHz sampling rate. The 16bit/44.1kHz CD files were up-sampled to ensure that 24/96-file resolution became the lowest common denominator quality amongst the recordings. Vinyl transfers were made directly from a Rega Planar 3 turntable, Rega Exact stylus through a Musical Fidelity X-LP phono stage at 24/96 to ensure the analogue files extracted retained as much detail as possible. Headphones were used for monitoring negating any airborne or structural resonances that may affect the stylus pickup. High-resolution audio files were transferred digitally to Pro Tools via Cambridge Audio BD650 universal player.
The subjective loudness between audio files was leveled through a process of A/B comparative listening so that EQ differences would present as the dominant variation. After synchronisation and level matching, the files were subsequently bounced as high-resolution 24/96 .wav files for import into ‘Spek’ spectrum analysis software.
For a detailed look at the spectrograph and the ability to magnify sections for deeper analysis, the software ‘AudioSculpt’ was used. The Voxengo SPAN FFT spectrum analyser was used to represent the RMS and Peak level frequency plots used for comparisons.
In the 1970’s, the period in which the examples under study originated, the vinyl record was still the leading format through which music was marketed to the consumer. General home music centres of that period may not have been of sufficient quality to fully express the detail contained in the grooves. However, continued turntable development has given the opportunity for audiophiles to continue the debate on the subjective ‘superior quality’ of the vinyl format. From a historical context, all studio recordings that were produced until the early 80’s were mastered primarily for vinyl and the original definition of a mastering engineer pertained to the skills required as a cutting engineer. This raises the question as to whether vinyl records could be considered the most authentic deliverable format for how music of that period would most accurately be represented.
‘Golden Years’ from Station To Station (1976) – Analysis
In figure 2, the first and most apparent detail evident in the vinyl spectrograph of David Bowie’s ‘Golden Years’ is that audio information over and above 20kHz is still present and rolls off smoothly above that frequency. Although this is a range beyond the average human hearing capabilities, debate persists on the effect of its presence on the audible range. To test that this information was of a musical nature and not random tracking noise from the stylus or any other electrical artifact, Audiosculpt software was used to isolate all elements above 21kHz and regenerate a new sound file. Importing this file into Pro Tools, normalising and using Waves pitch shifter to pitch down one octave provided audible evidence of rhythmic performance elements mostly created by hi-hats and cymbals.
Spectrograph of ‘Golden Years’ – Vinyl.
Figure 2: Spectrograph of ‘Golden Years’ – CD.
By comparing the spectrograph of the original master flat transferred to CD there is a clear brick wall frequency cut off at the 22500Hz Nyquist frequency.
Although from the same master source tape the vinyl format shows less energy in the High Frequencies (HF), between 5–12 kHz, than its CD equivalent. This becomes more revealing when comparing at the Fast Fourier Transform (FFT) analysis between vinyl and CD masters in figure 3. The FFT graphs show amplitude (y-axis) against the audible frequency range (x-axis). The bold colours represent Root Mean Squared (RMS) levels, and the faint colours represent peak levels. In all of the following FFT analysis charts, the measurements were taken with an infinite hold function across the entire length of the audio tracks.
‘Golden Years’ CD Master.
Figure 3: ‘Golden Years’ Vinyl Master.
It could be surmised that this attenuation of 3dB between 5 – 12kHz would most likely be the contributing factor to the often-cited ‘warm’ analogue quality associated with vinyl. Mastering for vinyl is a highly specialised skill requiring a deep understanding of how to equalize and compress the signals to the RIAA standards for the cutting lathe. Potentially there are a number of stages, each with generational signal losses, in this analogue transference process from tape. In addition, the unique characteristics and quality variations of the vinyl manufacturing process and final product impact on the sound quality. Mastering engineers would compensate with the application of subtle EQ gain for any HF losses due to the polishing process of the master disks (Scott, 2014). However, with the best intentions of the mastering house to reproduce what the artist and producer heard in the studio, providing test pressings for approval, vinyl records would be for many years the closest to a definitive edition of a recording the end listener would experience.
By super-imposing the FFT’s of the vinyl and CD flat transfers, taking 1kHz as a fulcrum point in figure 4 where the frequencies match. There is around a +4dB maximum lift in the HF and -4dB drop in the LF frequency ranges of the CD edition.
Figure 4: Original vinyl master (green) vs Flat Transfer Original Master CD (brown).
When this album was first mastered for CD by RCA in 1985, it was an example of early digitizing processes at 16bit/44.1kHz resolution. Super-imposing the FFT of this 1985 CD master over the original master in figure 5 reveals a subtle ‘smile curve’ between 500Hz and 8kHz. The small boost around 120Hz with slightly steeper LF roll off and lower HF cutoff frequency makes this edition more characteristically analogous to human hearing sensitivities (Newall: 2013, pp. 13-33), and this could explain why, in spite of the many subsequent remastered editions, the 1985 RCA master is often cited as ‘the best sounding version’ by connoisseurs of David Bowie’s discography.
Figure 5: 1985 RCA CD Master (blue) vs. Flat Transfer Original Master CD (brown).
Figure 6 shows the FFT of the1999 reissue, known as the ‘EMI 24-bit masters series CDs’, the key difference appears to be around a -2dB level attenuation of frequencies below 1.5 kHz but with an extended shelving EQ rolling off steeper below 70Hz.
Figure 6: 1999 EMI CD Master (white) vs Flat Transfer Original Master CD (brown).
‘Golden Years’ 2010 Remix
As part of the re-issue box set, Harry Maslin (original producer) remixed the entire album. We could speculate that having produced the original album Maslin would have a sensitive approach in using multi-track tapes and therefore would balance the elements sympathetically to the original. By super-imposing an FFT graph of the new mix over the original master we can trace the line of bass extension, illustrated by the pink area in figure 7, by about +4dB up to 150Hz. The lower-mid region is fairly equal with a small ‘smile curve’ between 1 – 3kHz. There is a +3dB lift on average above 8kHz giving this new mix a sound quality of top end ‘brilliance’ inherent in mixes of the post vinyl era. This is often used as a psycho-acoustic trick to increase perceived loudness.
Figure 7: FFT’s ‘Golden Years’ 2010 (pink) vs Original Master CD (brown).
The graphical representations, although detailed, only show part of the sonic picture, and do not convey mix elements such as stereo imaging, perspective, ambience, clarity and definition. This can only be subjectively analysed upon critical listening.
Production styles during the 1970s
Until the middle part of the 20th Century, record production had long been the domain of big corporations such as EMI working with recording methods established by an elite group of engineers. Popular music culture of the 1960s challenged these recording practices and not only influenced product design, but fostered experimentation in studio design and microphone placement techniques. Simon Zagorski-Thomas (2012) has written about the difference between the US and UK production sound in the 1970s by researching on over 200 recordings. He concluded that a number of factors characterised recordings of this period. More and more independent recording studios were set up during the 1970s, especially in the US. Common studio practices in the US included the use of acoustic treatments to reduce natural ambience, the use of isolation rooms for greater separation, close mic placement techniques, and tight dampened drum sounds to acquire more control over the final balance, thus relying more on outboard processors to replace the lost ambience prevalent in earlier popular recording techniques (Zagorski-Thomas: 2012, p.60).
Recording sessions and listening tests – ‘Golden Years’
The album Station To Station was recorded at Cherokee Recording Studios in Hollywood LA. The studio housed a Trident A Range recording console and the audio was recorded to an MCI JH-24 analogue tape machine. The album was mixed at The Hit Factory in New York on a custom console based around API 550 equalizers and the multi-track was bounced to an Ampex AG-400 2-track machine. The 2010 analogue transfers were made at 24bit/96kHz from a Studer A-80 via an ADA 8 AD converter into Sonic Solutions HD audio workstation.
Through listening to the original vinyl mix, elements of the common practices of 1970’s production style are noticeable. Instruments are close miked and feel very much on a flat sound stage with close proximity to the listener. The instrumental arrangement possesses stereo imaging although there is little front-to-back depth to the mix. In comparison with the various later editions, the digital CD format released in 1985 reveals a timbral difference as represented by the graphic illustrations in figure 8. There is more apparent ‘air’ to the 1985 CD master given by the increased average gain above 4kHz.
Figure 8: FFT Golden Years Original vinyl (green) vs 1985 CD (blue).
Reverb tails and percussive elements are revealed in more detail. By comparing the 1985 CD with the flat transfer master CD produced in 2010, we hear more timbral differences. The flat transfer sounds brighter. Although the EQ curve has more bass extension than the 1985 master, the small EQ scoop between 100 – 200Hz could explain the more defined sounding kick drum in this edition, when combined with a general higher gain above 500Hz, this overall effect gives a sense of more clarity to the mix. By listening to these format editions in chronological order they reveal a rebalancing of the sonic energy from low to high frequencies. Trusting that the original master flat transfer at 24/96 has not been surreptitiously mastered for today’s perception of how production ‘should’ sound, this would make the 2010 hi-resolution edition the closest representation of what the artist and producer would have originally heard in the studio. However, consumers would have originally experienced the music with all the imparted sonic qualities of vinyl. In spite of the marketing hype behind a flat transfer to vinyl from original source tapes, mastering engineers, as previously stated, would compensate for any HF losses attributed to the vinyl format. Through comparisons with alternative vinyl editions of this record these differences in HF content are clearly audible. A question could therefore be posed as to whether a flat transfer to vinyl is truly an authentic representation of the audio, as originally it would have been mastered in a specific way for that format.
The 2010 remix is equally valid as a reinterpretation and a representation of modern production styles; the most noticeable difference is evident in the sound of the drums. The sizzling and bright hi-hats, crisp snare and solid kick drum have been given their own space in the mix. The drums now appear further back in the perceived stereo sound stage and are more dynamically compressed. The lead vocal is of a similar sonic quality to the original master and appears to occupy the same sound stage location as the other editions, but the rhythm section sounds as though it is performing in a much more open environment. This remixed track clearly represents a more modern approach to mixing and production styles and is quite a radical departure from the sound of the original production. Without clear answers as to the motivation for this deviation in the 2010 remix, it is difficult to assume whether this was an exercise in corrective mixing, to reproduce the music in a way that was technically unachievable in 1976, or if it was purely an artistic response to a change of mixing and production style that has evolved over the last 40 years.
Ken Scott – from Abbey Road to Ziggy Stardust
To elicit an authoritative opinion on the definitive edition of a recording the thoughts of the original producer of a work would need to be sought.
Ken Scott, former Abbey Road and Trident engineer, is one of the few people to have sat behind the mixing console with The Beatles and has recently expressed his personal thoughts regarding the release of The Beatles’ remastered ‘vinyl in mono’ collection: “Mono is how the band wanted you to hear their work” . In Scott’s experience of working with The Beatles the stereo mixes were a secondary consideration and were often mixed with very little input from the band members. During the interview conducted with Ken Scott on 14th November 2014 by the author, the question was put forward as to whether he thought that the format with the mono mixes were important. His unequivocal view was that vinyl was best in that case. With CDs still formatted at 16bit/44.1kHz resolution, he thought digital media was not reproducing the music as accurately as a full range analogue source (Scott, 2014). A further challenging question about the authentic edition of The Beatles’ Let It Be was posed. Given that the original 1970 release of the album produced by Phil Spector was not how any of The Beatles had intended it to sound, was the latter ‘Let It Be…Naked’ version released in 2003 more authentic? In this instance Scott thought that it was important that the original remained available for those that deemed it was historically more authentic, however, artistically the ‘Naked’ edition was just as valid an authentic production too (Scott, 2014). The anecdote provided by Scott that George Harrison had discussed the similar removal of Phil Spector’s production elements to his first album All Things Must Pass supports the view that artistic authenticity is subject to change. (Scott: 2012, pp. 373-374). With all of this resurgence of opinion surrounding the most authentic way to listen to The Beatles back catalogue, it seemed equally valid to look at another classic album produced by Scott that has been remastered a number of times since 1972. For example, The Rise And Fall Of Ziggy Stardust And The Spiders From Mars and the appearance on Top of the Pops of the debut single ‘Starman’ both attributed to shooting David Bowie to stardom.
Before looking at the sonic characteristics of the various editions, a question was addressed about a very specific issue surrounding a mixing anomaly within the track ‘Starman’. The Morse code musical bridge between verse and chorus appears to be balanced much quieter in all digital editions mastered post 1980. Scott was aware of this mixing discrepancy but he knew nothing of where the discrepancies had originated. The louder of the two mixes finally reappeared on the latest Bowie compilation Nothing Has Changed (2015), thirty four years after this version was last officially available.
‘Moonage Daydream’ – Ziggy Stardust (1972) Analysis
Using the song ‘Moonage Daydream’ for illustration purposes, a comparison began by using two vinyl editions: the remarketed Direct Metal Transfer vinyl and the standard edition vinyl release. Figure 9 shows two spectrographs and FFT’s of the intro and verse section from an RCA International pressing of the album from 1980, and an EMI direct metal master 180gm pressing from 1997. The spectrograph focuses on the frequency range between 1kHz and 30kHz.
The opening of the song has a fairly sparse arrangement including drums, bass guitar, electric guitar (intro only), 12 string guitars (double tracked) and lead vocal. The zone of particular interest occurs between 2 – 6kHz where the upper spectral components of the cymbal crashes appear denser in the RCA edition. During the remainder of the verse it is the spectral quality of the 12 string guitars, sounding brighter and more crystalline in nature, which exposes the main difference in these vinyl pressings.
Spectrograph RCA vinyl 1980.
Spectrograph RCA vinyl 1980.
Figure 9: Spectrograph EMI vinyl 1997.
FFT EMI Vinyl FFT RCA International Vinyl
Ziggy Stardust, like many works distinguished as a classic album, has been marketed through many different digital editions. There were the first 16bit/44.1kHz CD pressings during the 1980’s, Ryko Disk pressings offering bonus tracks during the early 90’s and later, as digital technologies improved, the EMI 24-bit masters. Yet as a milestone is reached the opportunity to capitalize on an anniversary with a new edition proves irresistible. Whilst true that technological advances have had the benefit of enabling us to digitize and preserve audio at higher resolutions, an analysis was undertaken to test if there were real sonic differences between the 30th Anniversary CD and the (original mastering engineer) Ray Staff remastered 40th Anniversary CD editions.
By error, the 30th anniversary edition flipped the left and right channels of the master, so this needed to be corrected before using spectrograph and FFT analysis tools. Sonically, these two versions are extremely similar. One of the subtle differences appears in the 100 – 200Hz frequency range, affecting the bass guitar’s fundamental frequencies. This can be seen in figure 10 in the super-imposed FFT graphs.
Figure 10: FFT of 30th (purple) vs. 40th anniversary edition (green).
It is worth noting that this track underwent a remix in 1998 by Alan Moulder and Andy Wright. A new mix can bring fresh interpretations to the music as in this case the electric guitar is more prominent with additional musical lines that were never audible in the original. The vocals, although still effected with slap-back delay, are placed in a less reverberant space. However, it is interesting to look at how the overall sonic quality of the mix is treated.
Figure 11: FFT for Moonage Daydream – new mix.
In figure 11, the overall shape looks flatter and wider with a distinct lift of about +4dB around 100Hz when compared to the original master. Audibly this mix gives the acoustic guitar more ‘sparkle’ while the kick drum and bass guitar possess a ‘punch’ more akin to a modern production. When approaching this new mix we could postulate that the engineers would be influenced by modern mixing practices that take advantage of the ability for digital formats to faithfully reproduce an extended frequency spectrum.
When Ken Scott was asked to remaster Ziggy Stardust in 5.1 surround sound, he had a choice whether to keep the sonic quality as close to the original production as possible or attempt to give the music an enhanced modern sound. On the basis that back in 1972 playback systems couldn’t reproduce the bottom end as they do today, Scott decided in favour of keeping the recording faithful to its historical context, by choosing to give the listener an edition as it was originally intended to sound (Scott: 2012, pp. 186-188). Commenting on the fashionable obsession with boosting the ‘top end’ EQ, Scott reiterated that although he had no problem with alternate mixes or sonically varied masters, if ever one edition could be classed as definitive it would be one of the first vinyl pressings, as these would be duplicated from his approved master (Scott, 2014).
To illustrate that vinyl records are very capable of producing a broad, clean spectrum of sound, a brief analysis was undertaken with another of Scott’s productions and remastering of a classic Bowie album, Aladdin Sane (1973). This album was also digitally remastered for a 30th and 40th Anniversary edition CD. A comparison between the 30th Anniversary CD edition and a 180gm vinyl master cut from the original master tapes were analysed by the author. This vinyl edition is not marketed as a flat transfer; therefore, it could be assumed that the cutting engineers used their specialist skills to get the sonic quality as close to the source tapes as possible. In this instance, the vinyl edition (highlighted in blue) in figure 12 shows both an extended LF and HF frequency range compared to the CD (highlighted in red). With the addition of a small attenuation of no more than 2dB for the frequencies between 3 – 6khz, effectively removing some harshness to the sound, this provides evidence of what can be sonically achieved through expert mastering for this format.
Figure 12: Aladdin Sane 30th Anniversary CD vs 180gm Vinyl from Analogue Tapes.
Jeff Lynne – ‘Mr. Blue Sky’
Jeff Lynne, of the Electric Light Orchestra (ELO) and a composer, multi-instrumentalist and producer for his own body of work, is a paragon for considering artistic authenticity in production. Thanks to Lynne’s meticulous attention to detail in achieving his production sound, an ideal case study in researching a definitive edition can be built around his music. Lynne’s personal opinions on the production of his back catalogue are discussed through personal conversations and interviews, and by specifically referencing the recording, production, remastering and re-recording of the song ‘Mr. Blue Sky’, his thoughts helped to build a greater sense of context and evolution of his production style.
‘Mr. Blue Sky’ forms the final part of ‘Concerto For A Rainy Day’ from the 1977 album Out Of The Blue. Lynne wrote the music for this album over the course of two weeks whilst in Switzerland (Lynne, 2014). Instead of making demo recordings Lynne would play the songs to the other members of ELO and explain the parts for the arrangements he had formulated in his mind. The rhythm tracks were committed to tape before adding the many layers of overdubs. The studio that was originally booked for the strings overdubbing sessions was deemed too reverberant making them sound insipid. Lynne suggested they move to Musicland studios in Munich cramming all the musicians into a tight and acoustically dry space with a low ceiling. Again, in keeping with the 1970’s approach to recording techniques, the strings section was recorded by using an overhead stereo pair about 6ft above the ensemble, and double tracking for a richer tonal quality. This would give Lynne the maximum control he wanted at the post-production stage. Lynne often knew exactly how he wanted the finished product to sound, and would drive every element of the process by personally overseeing the mastering sessions, and by adjusting EQ and compression to get the sound exactly as he wished. Lynne expressed a liking for the sonic quality that the inherent nature of vinyl imparted onto the sound of a master, sometimes feeling that a flat sounding mix was given a bit more ‘punch’ by the process. This was particularly the case with the song ‘Mr. Blue Sky’ after hearing it back post-mastering (Lynne, 2014).
For analysis purposes, five different editions of ‘Mr. Blue Sky’ were collated. These editions consisted of the original 1977 vinyl pressing, the Sony (Joe Gastwirt) CD master from 1987, an early 2000’s 180gm vinyl pressing, the 30th Anniversary remastered edition, and the 2012 re-recording.
Knowing that Lynne cared so greatly about how he wanted his music to sound, it could be confidently said that the original vinyl pressing of Out Of The Blue was the closest to a definitive sounding edition and as close to how the artist had intended the listener to hear their music. Figure 13 shows an FFT graph of the frequency contour of the original vinyl pressing.
Figure 13: FFT of Mr. Blue Sky vinyl edition 1977.
By the mid 1980’s, the Compact Disc was expanding its market share, and artists’ back catalogues were gradually being digitized. Joe Gastwirt was responsible for the first digital remaster of ELO’s Out of The Blue album for Sony in 1987. Unlike most other projects, Lynne was not involved in this process so this edition represents an independent interpretation of the original master. We can see clearly from the FFT graph in figure 14 the emphasis that has been placed on boosting the frequencies between 5 – 15kHz by around 6dB. One could speculate that this timbral shift was in some way a depiction of how the new digital media was promoting a more transparent and sonically expanded sound over any previous media format.
Figure 14: FFT Mr. Blue Sky 1987 CD.
In questioning Lynne on his thoughts about the sound of the first generation of digital media and what he thought about the digitization of his own back catalogue, he reported that he initially felt that the CD sounded a bit harsh in the mid range and with regards the first CDs of his own music, and that he didn’t like the sound of the top end frequencies (Lynne, 2014).
It is possible to gain a sense of Lynne’s personal preference for the EQ contour of ‘Mr. Blue Sky’ through an FFT of the 30th Anniversary edition of Out Of The Blue. Jeff Lynne and Jeff Magid produced this reissued edition in 2007 and a very clear distinction can be made over the treatment of the top end from the 1987 Sony CD shown in figure 15.
Figure 15: FFT Mr. Blue Sky 30th Anniversary Edition.
By super-imposing the two graphs in figure 16, it is evident that by comparison the 1987 CD has a high shelving EQ applied above 3kHz with a 4dB scoop centred around 300Hz and a 3dB lift between 70 – 100Hz. Although memories of these remastering sessions were vague, Lynne thought it was certainly possible that this anniversary edition was an attempt to shape the sonic quality more in the direction of the ‘warmth’ associated with the original analogue edition (Lynne, 2014). Listening to and comparing the original vinyl and the 30th Anniversary edition support this hypothesis.
Figure 16: FFT of 1987 CD and 2007 CD.
After many more years of production experience, Lynne decided that he would like to revisit his back catalogue and completely re-record some of the ELO hits. Both in media interviews and in personal conversation, Lynne explained that there were a number of reasons behind taking this radical step. Some of these were performance related, feeling that the tempos were sometimes a little loose. Some were sonic and musical balance related too, feeling that instrumental elements of the arrangement or vocals could be mixed better (Lynne, 2014).
Still favouring the sound of audio through an analogue mixer, Lynne also records in the digital domain with Pro Tools in his home studio in LA. By recording in various rooms within his home, Lynne is able to achieve an ambient sound to suit his particular taste rather than using artificial reverb processors (Lynne, 2014). With digital recording technologies now possessing greater bit depth and higher sample rate resolution, Lynne is able to combine the sonic character of analogue with the timing accuracy, greater functionality and speed of workflow afforded by the DAW (Digital Audio Workstation).
Inevitably when re-recording a track 35 years later it will sound very different. Yet despite the obvious differences of being recorded in a completely different environment with Lynne now playing all the instruments and the sound of his matured vocal quality, there is a surprising similarity to be found between the 30th Anniversary remaster and the 2012 remake.
In both original and new recordings, the production and mixing styles bear the familiar hallmarks of close miking techniques with an upfront sound stage. Lynne did not purposefully attempt to match the sonic quality with the original editions, cross-referencing to check for vocal balance and arrangements only (Lynne, 2014). For all the noticeable sonic differences during listening tests, overlaying the FFT graphs of the previous remaster and new recording shown in figure 17 shows how very close the EQ contours are between the two editions. It could be surmised that the emphasis of Lynne’s production style is on the middle band occupied by the fundamental frequencies of vocals and instrumental melodies.
Figure 17: Mr. Blue Sky FFT’s for 30th (yellow) and 2012 remake (red).
To test this hypothesis, a sample of other recordings produced by Lynne were taken from George Harrison’s 1987 album Cloud Nine and another ELO album Zoom from 2000. These samples in figure 18 also show a tendency for a convex shaped EQ curve emphasizing those mid band frequencies.
Figure 18: FFT’s George Harrison – Cloud Nine (1987) and ELO – Zoom (2000).
‘Mr. Blue Sky’ – the definitive edition?
Music will always be open to new interpretations of arrangement and performance and technology plays its part driving new impositions of sonic quality over many old recordings. Artists have the rights to change their opinions about their work and with the 2012 re-make of ‘Mr. Blue Sky’, Lynne provides an updated artistic interpretation of one of his biggest hits and its existence is equally valid alongside the original.
Discussion: other artists, other interventions
In modern music production, there can be no doubt that a flat transfer of the original master tapes, with very high digital resolution, provides the listener with the closest representation to how the music would have sounded at the time of production and final mastering. It could be argued that this edition is the definitive one. However, the physical properties, technical and human input variables such as the application of EQ and compression involved in the reproduction of vinyl would have altered the sound of the original master. This is the final quality of sound that listeners would have become accustomed to and it could be argued too that this could be classed as the definitive edition. In addition to the already cited example of The Beatles in Mono project, this argument can be supported by evidence of the trend for other remasters being sonically reverted back to their ‘analogue quality’ origins. Mastering engineer Andy VanDette was responsible for the project of remastering the back catalogue of Canadian Progressive Rock trio Rush. He decided to take a sympathetic approach to the historical context in which some of these albums were made. Being recorded in the era of vinyl, Andy VanDette tried to emulate that warmth in the low end, and a direct listening comparison with the 1996 Rush digital remasters by the author confirms the effect of this sonic manipulation.
It is also interesting to note that in 2011, Kate Bush remastered her 1993 album The Red Shoes after finding a ½ inch analogue tape back up copy. Kate Bush takes the opportunity in the new CD liner notes to describe how she was never really happy with the harsh, edgy quality to the original album sound, in spite of the promise of more clarity from digital recording technology. She describes this new analogue master as a ‘truly remastered version’ (Bush, 2011 – The Red Shoes).
Peter Gabriel’s album So has also been through two remastered revisions since the original 1986 release. In both the 2002 remaster and 25th Anniversary edition Peter Gabriel oversaw the mastering process. Gabriel had asked Tony Cousins (2002 mastering engineer) to make significant changes to the low and mid range EQ and apply more compression (Chappell, 2012). This could be considered an understandable strategy in light of the loudness war pervading rock and pop production practices of the time. However, bringing back Ian Cooper (original 1986 mastering engineer) for the 25th Anniversary edition, Gabriel expressed his wish to take the sonic quality back to closely match the original. The analysis of these editions show that the loudness war is still pervasive as can be illustrated below. The graphic display in figure 20 shows the waveforms of the 1986 CD (Red), 2002 CD (Green) and 2012 CD (Yellow).
Figure 20: Waveforms from Peter Gabriel – So.
The significant changes between these editions occur below 300Hz. Figure 21 shows that the original edition has a flatter bass response while the 2002 edition attenuates by 3dB between 100 – 300Hz. The 2012 edition only attenuates by 1.5 dB between 150 – 250Hz but boosts by 3dB all frequencies below 100Hz.
Sledgehammer 1986 (Green) vs 2002 (Brown).
Figure 21: Sledgehammer 1986 (Green) vs 2012 (Blue).
The use of compression combined with EQ also changed elements of the mix most noticeably in the 2002 remaster. Vocal reverb tails became more subdued yet some electronic keyboard elements stood out more prominently. These two elements in particular do sound the same in both the 1986 and 2012 editions confirming that to some extent the desire to revert the remaster closer to the original was achieved.
It could be argued that many of the variations present in the examples researched are products of creative interpretations. However, long after the artists and anyone involved in the initial process have passed, there is a duty to preserve the art of music and its production sound for future generations to access and enjoy. The 2009 remasters of The Beatles catalogue are a good example of the importance of archiving and restoration work whilst being careful to respect the historical context of the recordings. It could be anticipated that if any attempt was made to veer too far from the original masters, Beatles fans would be extremely vocal in their objections. This can be evidenced in reference to online reviews about the Let It Be…..Naked project, in which the original album was remixed and all of Phil Spector’s production elements were removed. The fact that there was artistic contention between The Beatles over the original release makes it more difficult to conclude which edition would be the definitive one (MacDonald: 1994, pp. 262-272).
Restoration, stereo remixing and 5.1 surround mixes have been applied to the complete Genesis back catalogue despite the digital remasters of the mid ‘90’s claims that they were the definitive edition. Currently, the Yes back catalogue is going through the same process under the control of producer Steve Wilson. These CDs are marketed as the definitive edition, and although one may argue that ‘definitive’ in this instance refers to the inclusion of alternate mixes and masters within the CD set, it is incomplete as there are alternate masters not included such as the Joe Gastwirt remastered edition and Rhino remasters. Unlike the Harry Maslin and the Moulder/Wright remixes of earlier Bowie tracks, Wilson’s approach seems to be one of restoration and faithful reproduction. Evidence of tape hiss removal can be heard and graphically shown in figures 22 and 23 using spectrographs of the intro to the song ‘Siberian Khatru’ from the album Close To The Edge (1972).
Figure 22: Siberian Khatru – original master.
Figure 23: Siberian Khatru – 2013 master reduced noise.
Analogue tape dropouts, evident in all previous digital editions, have been repaired.
Steve Wilson’s mix follows very closely the EQ contours of previous editions shown in figure 24 matching the top end of the original 1987 CD and matching more closely the low mids (200 – 600Hz) of the original vinyl.
Siberian Khatru – 2013 Mix (Green) vs. Original CD Mix (Blue).
Figure 24: Siberian Khatru – 2013 Mix (Green) vs. 180gm Vinyl master (Red).
Listeners of recorded music have for a long time been able to control and manipulate how they experience their favourite music. This begins with choices in the quality of playback systems available and being able to control a number of flexible parameters provided by the manufacturers of Hi-Fi’s such as tone controls or graphic EQ’s. Some modern playback systems and in-car stereos also offer dynamic compression and it could be argued that such flexibility should be left in the hands of the listener, rendering the overuse and marketing hype around so many remasters superfluous. The best sounding and most authentic original source recordings – how the artist/producer wanted you to hear the music – have to be the ideal starting point. If the wealth of recordings from the last century is to be accurately representative of the time in which they were made, then the technological frameworks and sonic fingerprints from the recording processes embedded in the music should also be carefully and sympathetically preserved for generations beyond our own. Figure 25 shows one final example, using another of Ken Scott’s productions for David Bowie’s song ‘Life on Mars’ (1971), of how such stark differentials can occur between original sources and a remaster.
Figure 25: Life on Mars 180gm vinyl master from analogue tapes (Green) vs EMI 24bit masters CD (Blue).
The extreme boost of around 8dB between the frequencies of 2 – 5kHz render the almost telephonic sounding CD edition very abrasive and tiring to listen too. Unfortunately, once released into the marketplace this edition is all that is commercially available.
Regardless of any creative validity or marketing strategy in allowing audio to be manipulated in this way, it is imperative that the authentic original source audio must remain accessible.
 Bottomley, Andrew J. (2015) Play It Again: Rock Music Reissues and the Production of the Past for the Present, Popular Music and Society. Routledge. http://dx.doi.org/10.1080/03007766.2015.1036539 Accessed (19/05/15)
Bottomley uses an in depth analysis of the 40th Anniversary edition of The Beach Boys album Pet Sounds to show that major record labels go to great lengths to authenticate the changes and additions made for reissues.
 Moore, A. (2002) Authenticity as Authentication, Popular Music, vol. 21, no. 2 pp. 209 – 223 Cambridge University Press. Moore explores the subtle differences between authenticity as defined by the unhindered freedom of expression and social and subjective authenticity.
 Oxford English Dictionary: 2015, http://www.oed.com/view/Entry/13325.
 Maher, Robert C. (2009) Audio Forensic Examination, Authenticity, enhancement and interpretation. IEEE Signal Processing Magazine Vol. 26, Issue 2. There are 7 tenets used to authenticate an audio recording to ensure that it is admissible as evidence in court. These were adopted in the United States in 1958. Further details can be found at: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4806208&tag=1 (Accessed 19/05/15)
 Koenig, Bruce E. (1990) Authentication of Forensic Audio Recordings. JAES Volume 38 Issue 1/2 pp. 3-33. Further details can be found at: http://www.aes.org/e-lib/inst/browse.cfm?elib=6053 (Accessed 19/05/15)
 In conversation with the microphone custodian during a personal tour of Abbey Road studios on September 5th 2014 revealed the decision to use the 1960’s collection of Neumann U87’s for recording the sound track to ‘Cilla’ TV documentary. An article about the use of the Royal family microphones for the Soundtrack recording of The King’s Speech can be found at: http://www.soundonsound.com/sos/may11/articles/kings-speech.htm
 A review of the process behind the remastering of The Beatles in Mono project can be found at: http://www.analogplanet.com/content/beatles-get-back-mono-and-aaa-vinyl
 Although generally humans cannot perceive sound above 20kHz, research into the effect on brain activity and its emotional impact with the inclusion of supersonic frequencies is ongoing. Further details can be found in the American Physiological Society Journal of Neurophysiology at: http://jn.physiology.org/content/83/6/3548 (Accessed 14/06/14)
 The Nyquist Frequency is the highest frequency that can be coded at a given sampling rate in order to be able to fully reconstruct the signal.
 Online reviews and debates between fans and audiophiles commonly state a preference to the sound of the 1985 RCA edition of Station to Station. Examples can be found at: http://forums.stevehoffman.tv/
 Both in interview with the author and online articles Scott expressed a view about the new release of the Beatles in Mono set with anecdotes from The White Album sessions. One article can be found at: http://beatlesmagazine.blogspot.co.uk/2014/09/ken-scott-beatles-get-back-to-mono.html
 An online interview with mastering engineer Andy Vandette reveals his thoughts and processes surrounding the remastering of Rush Sectors box set.
 Richard Chappell reviews the history and differences between the masters of Peter Gabriel’s So detailed at: http://blog.bowers-wilkins.com/music/remastering-peter-gabriels-so/
 Many reviews are available for Let It Be…Naked most inspiring debate about the artistic validity of such a remix as a review on the sonic quality. Examples can be found at: http://www.rollingstone.com/music/albumreviews/let-it-be-naked-20031120#ixzz2yzTrmbip
Brown, S. & Volgsten, U. (2006) Music And Manipulation: On the social uses and social control of music. Oxford: Berghahn Books
Blauert, J. (2005) Communication Acoustics. Germany: Springer
Fremer, M. (2014) The Beatles get back to mono and AAA vinyl. [Online] June 15th 2014. Available at: http://www.analogplanet.com/content/beatles-get-back-mono-and-aaa-vinyl (Accessed: August 2014)
Frith, S. (1978) The Sociology Of Rock. Great Britain: Constable
Frith, S. & Zagorski-Thomas, S. (2012) The Art Of Record Production. England: Ashgate Publishing Ltd
Hoffman, S. (2012) The Rush CD Mastering Thread. [Online] Available at: http://forums.stevehoffman.tv/search/16793459/?q=rush+sectors&o=date (Accessed: March 2014)
Inglis, S. (2011) Peter Cobin: Recording music for the King’s Speech. In: Sound on Sound [Online] May 2011. Available at: http://www.soundonsound.com/sos/may11/articles/kings-speech.htm (Accessed: September 2014)
Katz, B. (2007) Mastering Audio The Art And The Science. Oxford: Focal Press
Lynne, J. (March 5th 2014) Interview with Author
Lynne, J (September 19th 2014) Telephone interview with Author
MacDonald, I. (1994) Revolution In The Head the Beatles’ records & the sixties, England: Fourth Estate Ltd
Marchese, J. (2010) Review: David Bowie “Station to Station”, [Online] Available at: http://theseconddisc.com/2010/10/11/review-david-bowie-station-to-station-2010/ (Accessed: April 2014)
Newall, P. (2013) Recording Studio Design. Oxford: Focal Press
Philips research, [Online] www.research.philips.com/technologies/projects/cd/technology.html (Accessed: April 2014)
Scott, K. & Owsinski, B. (2012) Abbey Road To Ziggy Stardust. USA: Alfred Music Publishing Co., Inc.
Scott, K (November 14th 2014) Interview with Author
VanDette, A. (2011) Andy VanDette On Remastering 15 Rush Albums, [Online] Available at: http://www.themasterdiskrecord.com/2011/11/andy-vandette-on-remastering-14-rush-albums/ (Accessed: February 2014)
Beatles, The. Let It Be…Naked. [CD] Apple, 2003.
Bush, Kate. Directors Cut. [Vinyl] Fish People, 2011.
Bush, Kate. The Red Shoes. [CD] Fish People, 2011.
Bowie, David. Golden Years. Station To Station. [CD] EMI 24-bit remasters, 1999.
Bowie, David. Golden Years, Station To Station. [Vinyl and CD] EMI, 2010.
Bowie, David. Life on Mars, Hunky Dory. [Vinyl] EMI, 1997
Bowie, David. Life on Mars, Hunky Dory. [CD] EMI 24-bit remasters, 1999,
Bowie, David. Moonage Daydream, Ziggy Stardust. [Vinyl] RCA, 1980.
Bowie, David. Moonage Daydream, Ziggy Stardust. [Vinyl] EMI, 1997.
Bowie, David. Moonage Daydream, Ziggy Stardust. [CD] EMI, 2002.
Bowie, David. Moonage Daydream, Ziggy Stardust. [CD] EMI, 2012.
Bowie, David. Starman, Ziggy Stardust. [Vinyl] RCA, 1980.
Bowie, David. Starman, Ziggy Stardust. [Vinyl] EMI, 1997.
Bowie, David. Starman, Ziggy Stardust. [CD] EMI, 2002.
Bowie, David. Starman, Nothing has changed. [CD] Parlophone, 2014
Gabriel, Peter. Sledgehammer, So. [CD] Geffen. 1986.
Gabriel, Peter. Sledgehammer, So. [CD] Real World. 2002.
Gabriel, Peter. Sledgehammer, So. [CD] Real World. 2012.
ELO, Mr. Blue Sky, Out Of The Blue. [Vinyl] Jet CBS. 1977.
ELO, Mr. Blue Sky, Out Of The Blue. [CD] Epic. 1987.
ELO, Mr. Blue Sky, Out Of The Blue. [Vinyl] Epic. 2002.
ELO, Mr. Blue Sky, Out Of The Blue. [CD] Epic. 2007.
ELO, Mr. Blue Sky. Mr. Blue Sky. [CD] Frontiers. 2012.
Rush, Sectors Box Set. [CD] Mercury/Anthem. 2011.
Yes, Close To The Edge. [CD] Atlantic. 1987
Yes, Close To The Edge. [CD] Atlantic. 1994
Yes, Close To The Edge. [CD] Panegyric. 2013
Video Extract 1. This is a sample of the audio information filtered above 21kHz pitch shifted -12 cents
Video Extract 2. This is a looped sequence from the song ‘Golden Years’. Each loop is displayed with its correlating FFT. Graphically there is a short preview of the up-coming FFT super-imposed over the current loop before an edit cuts to the next audio edition.
Video Extract 3. This is the first verse and chorus to ‘Moonage Daydream’ beginning with the 1980 vinyl edited sequentially through the EMI 180gm vinyl, 30th and 40th Anniversary editions and ending with the 1998 remix each with their corresponding FFT’s.
Video Extract 4. This is an alternating sequence between CD and 180gm vinyl editions of the song ‘Aladdin Sane’ with corresponding FFT’s.
Video Extract 5. This sequence consists of the five editions of ‘Mr. Blue Sky’ with their corresponding FFT’s. A short visual preview of the up-coming FFT is super-imposed before the audio cuts to the next edition.
Video Extract 6. These are the three CD editions of ‘Sledgehammer’ from So each with their corresponding FFT. A short visual preview of the up-coming FFT is super-imposed before the audio cuts to the next edition.
Audio Extract 1. Intro to ‘Siberian Khatru’ original CD
Audio Extract 2. Intro to ‘Siberian Khatru’ 2013 CD de-noised
Audio Extract 3. ‘Siberian Khatru’ original CD including tape dropouts audible between left and right channels.
Audio Extract 4. ‘Siberian Khatru’ 2013 CD tape dropouts repaired.