As someone who has been involved on and off with music recording and audio production over the last few decades, both professionally and privately, Chris Edwards’ article in the October 2019 issue of E&T about the resurgence of recording on analogue tape – Analogue redux: the reel-to-reel rebirth in recording studios – stirred some memories.

I joined Sony’s Pro Audio division as a young engineer in the early 1980s, initially as a service engineer on the company’s new portfolio of digital recorders as well as MCI mixing consoles and analogue multitracks. This, coupled with a subsequent role in which I had product management responsibility for Sony’s digital multitrack recorders, put me at the heart of the digital versus analogue debate.

The marketing messages at the time were all about digital offering a superior sound to analogue by way of standard analogue-specification parameters: wow and flutter should be unmeasurable; total harmonic distortion plus noise (THD+N) 0.001 per cent or better; signal-to-noise ratio 96dB or better. But there was more to it than just objective measurements: there was a very strong subjective side to the argument as well.

I attended many A/B shoot-outs or ‘bake offs’, as they were often called. At one, a respected London mastering engineer told me, “You can tell it’s a digital recording: the bass player and the drummer aren’t playing together”. At that time, I prided myself on having a good set of ears, but I couldn’t for the life of me hear what he was referring to, although that’s not to say it wasn’t there. On another occasion I was babysitting a live classical session in north London. Having promptly installed and aligned the two-channel recorder, the established recording engineer turned to me and said: “You have the transformers [for balanced audio termination] switched in, don’t you? Please switch them out, they dull the violins”.

I firmly believe that a lot of the audio shortcomings and criticism of early digital sound quality lay firmly in the analogue domain: the brick-wall filters used in the anti-aliasing and reconstruction stages, for example. Such was the criticism in this area that many engineers customised their machines by substituting the Sony-manufactured ones with some esoteric imports from the USA that came in ‘G’ or ‘S’ options, standing for ‘gentle’ and ‘sharp’, but even the S had far gentler roll-off characteristics than the Sony-supplied brick-wall filters.

There were also lively discussions about the sound quality of different manufacturers’ digital multitracks. Sony wasn’t the only player and had strong competition from Mitsubishi with its X-800 and X-850 models. Whilst I wouldn’t have admitted it at the time, the Mitsubishi recorders had a better performance at the bottom end, with low frequencies being smoother and more ‘extended’. I could hear this, but it was subjective and no amount of objective measurement would have shown any differences.

There was more than just the sound quality argument when comparing analogue with digital. One big advantage of the Sony digital recorder was its ability to offer new creative opportunities, something I guess today we would call ‘workflow’. The Sony digital machine used its own internal control track, which labelled the recordings on tape down to 1 millisecond segments. This allowed accurate and repeatable drop-ins for overdub purposes to be accurately adjusted and rehearsed with great accuracy.

If you could afford two machines (at its launch the Sony multitrack was around four times the price of an analogue machine) this got even better, as machine-to-machine editing was possible, allowing individual tracks or a number of tracks to be edited with accuracy and repeatability and without any generation loss, as the edited master was a clone. The lock accuracy between machines was so good that a party piece we used to do at exhibitions was to record an 18kHz sine wave on the two outside tracks of both machines and then look at the output of the machines on playback on an oscilloscope.

Both waveforms were rock solid in both time and amplitude domains, something you wouldn’t have got away with on an analogue machine at that time. The waveforms would have been bouncing around all over the place in both time and amplitude, meaning that anything recorded as a stereo pair or, say, the 8-10 tracks used for the drums had to be on the same analogue machine otherwise you would have serious phase errors and, as a result, high-frequency loss or cancellation.

I was involved with the Frankie Goes to Hollywood recordings for ZTT Records mentioned in Chris Edwards’ article. The album was initially recorded on a Sony rental machine and Trevor Horn and the production team were impressed enough with the technology that they invested in two machines of their own.

Unbeknown to anyone at the time, the rental machine had a very slight misalignment in the way it recorded the digital waveform to tape (over-biasing was used even on digital recorders). Problems only arose after they started overdubbing on the new machines, using tapes that had initially been recorded on the rental machine. Clicks and pops abounded and as a Sony service engineer it fell to me to recover the situation. I ended up spending many an hour in some of London’s top studios endeavouring to salvage what I could by tweaking the playback equalisers to minimise the CRC error-correction activity lights as well as listening to the audio playback for artefacts.

This is where the advantage of the internal control track of the Sony machine proved a godsend, as it allowed me to compile a clone of the original, track by track and song by song, with repeatable millisecond accuracy having got rid of the offending pops and clicks. From what I remember, all but two or three of the offending artefacts were successfully removed. As Chris Edwards mentions, the record company decided to acknowledge the existence of these ‘imperfections’ on the sleeve notes, but referenced the SMPTE time code on the master tape and not the absolute running time of the CD.

Fast forward 20 years to the early 2000s and I was still at Sony, but by now I had moved to video. Riding the HD wave, I was tasked with persuading the market to move film production from 35mm to HD Digital cinema-photography. Older and wiser, I chose not to go down the somewhat confrontational route of the way digital was positioned against analogue in my audio days. Instead, any time I was engaged with the production community, I would say, “Look, it’s just a different canvas. Give it a try and see if it works for your productions”.

Hats off to Chris Mara, whose ‘Welcome to 1979’ analogue recording studio is featured in the E&T article. It sounds like he and others are doing the right thing.

Phil Wilton is an IET member. A born-and-bred Cornishman now living and working in North Hampshire, he is coming up to 40 years in the pro audio and broadcast industry, working mainly in product management and product marketing roles across virtually every application area.

As someone who has been involved on and off with music recording and audio production over the last few decades, both professionally and privately, Chris Edwards’ article in the October 2019 issue of E&T about the resurgence of recording on analogue tape – Analogue redux: the reel-to-reel rebirth in recording studios – stirred some memories.

I joined Sony’s Pro Audio division as a young engineer in the early 1980s, initially as a service engineer on the company’s new portfolio of digital recorders as well as MCI mixing consoles and analogue multitracks. This, coupled with a subsequent role in which I had product management responsibility for Sony’s digital multitrack recorders, put me at the heart of the digital versus analogue debate.

The marketing messages at the time were all about digital offering a superior sound to analogue by way of standard analogue-specification parameters: wow and flutter should be unmeasurable; total harmonic distortion plus noise (THD+N) 0.001 per cent or better; signal-to-noise ratio 96dB or better. But there was more to it than just objective measurements: there was a very strong subjective side to the argument as well.

I attended many A/B shoot-outs or ‘bake offs’, as they were often called. At one, a respected London mastering engineer told me, “You can tell it’s a digital recording: the bass player and the drummer aren’t playing together”. At that time, I prided myself on having a good set of ears, but I couldn’t for the life of me hear what he was referring to, although that’s not to say it wasn’t there. On another occasion I was babysitting a live classical session in north London. Having promptly installed and aligned the two-channel recorder, the established recording engineer turned to me and said: “You have the transformers [for balanced audio termination] switched in, don’t you? Please switch them out, they dull the violins”.

I firmly believe that a lot of the audio shortcomings and criticism of early digital sound quality lay firmly in the analogue domain: the brick-wall filters used in the anti-aliasing and reconstruction stages, for example. Such was the criticism in this area that many engineers customised their machines by substituting the Sony-manufactured ones with some esoteric imports from the USA that came in ‘G’ or ‘S’ options, standing for ‘gentle’ and ‘sharp’, but even the S had far gentler roll-off characteristics than the Sony-supplied brick-wall filters.

There were also lively discussions about the sound quality of different manufacturers’ digital multitracks. Sony wasn’t the only player and had strong competition from Mitsubishi with its X-800 and X-850 models. Whilst I wouldn’t have admitted it at the time, the Mitsubishi recorders had a better performance at the bottom end, with low frequencies being smoother and more ‘extended’. I could hear this, but it was subjective and no amount of objective measurement would have shown any differences.

There was more than just the sound quality argument when comparing analogue with digital. One big advantage of the Sony digital recorder was its ability to offer new creative opportunities, something I guess today we would call ‘workflow’. The Sony digital machine used its own internal control track, which labelled the recordings on tape down to 1 millisecond segments. This allowed accurate and repeatable drop-ins for overdub purposes to be accurately adjusted and rehearsed with great accuracy.

If you could afford two machines (at its launch the Sony multitrack was around four times the price of an analogue machine) this got even better, as machine-to-machine editing was possible, allowing individual tracks or a number of tracks to be edited with accuracy and repeatability and without any generation loss, as the edited master was a clone. The lock accuracy between machines was so good that a party piece we used to do at exhibitions was to record an 18kHz sine wave on the two outside tracks of both machines and then look at the output of the machines on playback on an oscilloscope.

Both waveforms were rock solid in both time and amplitude domains, something you wouldn’t have got away with on an analogue machine at that time. The waveforms would have been bouncing around all over the place in both time and amplitude, meaning that anything recorded as a stereo pair or, say, the 8-10 tracks used for the drums had to be on the same analogue machine otherwise you would have serious phase errors and, as a result, high-frequency loss or cancellation.

I was involved with the Frankie Goes to Hollywood recordings for ZTT Records mentioned in Chris Edwards’ article. The album was initially recorded on a Sony rental machine and Trevor Horn and the production team were impressed enough with the technology that they invested in two machines of their own.

Unbeknown to anyone at the time, the rental machine had a very slight misalignment in the way it recorded the digital waveform to tape (over-biasing was used even on digital recorders). Problems only arose after they started overdubbing on the new machines, using tapes that had initially been recorded on the rental machine. Clicks and pops abounded and as a Sony service engineer it fell to me to recover the situation. I ended up spending many an hour in some of London’s top studios endeavouring to salvage what I could by tweaking the playback equalisers to minimise the CRC error-correction activity lights as well as listening to the audio playback for artefacts.

This is where the advantage of the internal control track of the Sony machine proved a godsend, as it allowed me to compile a clone of the original, track by track and song by song, with repeatable millisecond accuracy having got rid of the offending pops and clicks. From what I remember, all but two or three of the offending artefacts were successfully removed. As Chris Edwards mentions, the record company decided to acknowledge the existence of these ‘imperfections’ on the sleeve notes, but referenced the SMPTE time code on the master tape and not the absolute running time of the CD.

Fast forward 20 years to the early 2000s and I was still at Sony, but by now I had moved to video. Riding the HD wave, I was tasked with persuading the market to move film production from 35mm to HD Digital cinema-photography. Older and wiser, I chose not to go down the somewhat confrontational route of the way digital was positioned against analogue in my audio days. Instead, any time I was engaged with the production community, I would say, “Look, it’s just a different canvas. Give it a try and see if it works for your productions”.

Hats off to Chris Mara, whose ‘Welcome to 1979’ analogue recording studio is featured in the E&T article. It sounds like he and others are doing the right thing.

Phil Wilton is an IET member. A born-and-bred Cornishman now living and working in North Hampshire, he is coming up to 40 years in the pro audio and broadcast industry, working mainly in product management and product marketing roles across virtually every application area.