Analogue v digital

Analogue v digital recording


BASF tape- green leader tape at the start, red leaer tape at the end

My friend Graham had a tape recorder when we were fourteen. It had a small square grey microphone which was permanently attached on a wire. When not in use, you rolled up the wire and tucked the mic into a small compartment on the side. He could place that little square mic in front of his grandfather’s large and ancient radiogram and record the Top Twenty show on the Light Programme on Sunday evenings (in mono) which we would then listen to all week while playing snooker on his scaled down table. In 1961 and 1962, home taping on open reel was growing.

New Musical Express 2 November 1962

In working with tape I go back to the days of splicing blocks and razor blades and magnetic tape. Splicing things together by cutting pieces of tape and sticking them together (with splicing tape) was literal and painstaking.

My experience was recording spoken voice for English Language teaching and we had a small studio with two Revox A-77 open reel tape recorders plus a Uher Report for ‘field recordings’ outside the studio.

A Revox A-77 was a popular ‘semi-pro’ machine, and you can see one on the 1975 LP sleeve for The Basement Tapes.

Screenshot 2020-08-01 at 14.30.08

The Basement Tapes: Bob Dylan and The Band, CBS 1975 double LP with Revox A-77

Not that The Basement Tapes were recorded on an A-77, I hasten to add. Nor were all The Band tracks recorded there.  The Basement Tapes were recorded on a quarter track Ampex tape recorder at 7.5 IPS, and the Complete Basement Tapes box set has photos … they were standard domestic 7″ tape reels.

Garth Hudson: It was a quarter track machine, meaning it had two tracks one way and then you turned it over and there were two tracks the other way, so leakage from one track to the next is possible. I think on most of the tapes, I only recorded in one direction.
Interview 5 November 2014, “Something Else” online.

They started out with premium brand Scotch, but continued onto tapes from local stores … Shamrock, Pure-Tone, and Village Silver. When the restoration project started for the box set, the Scotch reels had survived best.

After The Basement Tapes I guess the most famous home recording was Bruce Springsteen’s Nebraska which he recorded as a demo at home on a TEAC 4-track cassette recorder. He had intended to re-do them in the studio with a band, but decided to release them just as they were. Recorded at 1 7/8 IPS cassette.


There were three domestic tape speeds … 1 7/8, 3 3/4 and 7 1/2 inches per second (IPS). Studios ran tape at 7 1/2 IPS then at 15 IPS. Eventually Frank Zappa recorded Uncle Meat at 30 IPS. The faster the tape speed meant the more tape the signal covered. The wider the tape was also increased the area covered, which enhanced fidelity. So domestic tape was 1/4″ wide, then you had 1/2″ and 1″ tape. Again, domestic machines were four track – two in each direction for stereo. Semi-pro machines which we used for spoken voice in the early 70s were two track – just one direction. Again the more tape you covered the better.

Not only that, but it was easier to splice a 15 IPS tape with accuracy than it was a 7 1/2 IPS tape. The wave covered a greater length of tape.

Splicing tape and splicing block

The other main use of splicing for me was sound effects for stage plays. We used open reel tape, and spliced coloured leader tape between sound effects, so that you could cue them up accurately by hand, and also the machine would stop when it reached the next piece of leader tape. Because we did sketch shows, it was a task redoing and re-ordering the sound effect tape every week, so I switched to Memorex advertising cassettes – just two minutes a side with no leader tape, one sound effect on each. They were designed for adverts on radio stations.


In the studio, with mere 4-track recording machines, material had to be recorded from one track to another with ever increasing hiss and generation loss. George Martin’s Beatles recordings are masterpieces of layering. Strawberry Fields Forever took 45 hours spread over five weeks with multiple takes and overdubs.

The vocal tracks on Queen’s Bohemian Rhapsody took three weeks with 180 separate overdubs. Even with 24 track mixers, they ended up with an eighth generation tape. Brian May said you could see light through the tape because it had been passed over the recording heads so many times,

If you didn’t want to splice and overdub, you did it again. There’s a bootleg cassette of The Band doing multiple takes of We Can Talk from Music From Big Pink. If someone slipped, they started again. Then they listened back, and if they thought they could do it better they did the whole thing again.

Bob Dylan on the other hand was fond of doing one take and living with the result.

Digital recording was a boon for people like Stevie Wonder, who could play every instrument himself. He could overdub without increasing hiss or degrading the signal. The Secret Life of Plants was all digital.

I’ve watched music recording in the digital age, where a singer replaces bum notes a sound at a time. You’d never hear the difference.

But it IS different, and not because of the technology.

I’ll use spoken voice recording examples. The studios we used switched from open reel tape to Sony digital DAT tape, then to ProTools on a Macintosh computer. With DAT, the fear of a tiny thin tape breaking meant they used two identical recorders locked together. The tape carries the digital ones and zeros rather than an analogue wave form. I had a DAT player because Listening copies came on DAT in the early 1990s.

DAT tapes

DAT tapes in case you’ve never seen them. They’re tiny. Listening copies dated 1992 to 1996

DAT was doomed as a recording format as hard disk drives took over, and also it became easy to run off CDRs as listening / test copies. There was no point in commercial releases because Sony was developing MiniDisc at the same time, releasing the format in 1992. Recordable CDR defeated MiniDisc in turn.

Audio recording has been affected by improved technology, and mainly this has been a negative on performance. Years ago, the producer would listen to a reading before recording, and because it was so fiddly to edit the tape, it was easier to re-record any mistakes than play around with scissors, or rather a razor blade and cutting block. The result was that even on a short six-line ELT dialogue, the actors would read it five or six times before producer, engineer and author (notice which one comes last) were agreed that it was a good take. By the time they’d read it six times, the actors had more or less memorized the script. They’d also played around and tried different interpretation and emphasis.

Nowadays, the recording begins immediately on the first reading, and if there’s a problem on a word, the actor can just say the word again and in seconds the engineer will have clipped it in. This will happen with all fluffs and throat noises and sibillants, so that mainly what you get will be a first reading.

In contrast, thirty years ago I spent most of an afternoon watching two actors record a pastiche ‘Yes / No contest’ many many times. It was perfect to me by around the fifth reading, but we went on to do a dozen more. I asked the producer (Terry O’Neill) why. The first line in the dialogue said, ‘You’ve got thirty seconds …’ and we kept going until the actors timed the game at exactly thirty seconds. I protested that no one would be counting, but was told that wasn’t the point. Nowadays that couldn’t happen. If it was thirty-four seconds or twenty-seven seconds, the engineer would digitally tweak the running speed and clip pauses until it was thirty seconds.

A CD  has a technical maximum length of eighty minutes and if the reading were to come to eighty-two minutes, the engineer would simply speed the recording up enough to eliminate two minutes. As a listener, you wouldn’t notice. No one is going to produce a six CD audio book with ten minutes on the sixth CD. They would speed up and manipulate pauses until it fitted on five.

When twenty-four frames per second feature films were shown on TV at twenty-five frames a second (PAL) no one ever noticed that they were slightly faster. (On American NTSC it’s thirty frames a second, and some frames are doubled randomly to process twenty-four frame a second movies … it’s less speeded up, but not as smooth). So no one will notice three or four minutes taken from an eighty-four minute recording by increasing the speed. But they will notice ten minutes taken from a ninety minute recording, and this ‘gabble’ factor can be present on audiobooks. An author attending the recording session will make no difference to this. The dirty deed will be done later.

You lose the actor or musician’s creativity. A spoken voice example:

We’d written an ELT dialogue about a boy going off to university, and his younger sister was trying to scrounge items from his room; posters, CDs, plants etc. At one point he says ‘What about these magazines?’ and she replies, ‘No, thanks. Not those magazines.’

We were still working on open reel tape in those days. So there was no “punching in” and we did the dialogue five or six times. On the next take, when the actor read the part, she wrinkled her nose in disgust and said, ‘Ergh! No, thanks! Not those magazines!’ and everyone fell about laughing in the studio. The producer laughed too, but said, ‘You’d better read it straight.’ Because I was there, I could intervene and say, ‘No, she’s improved it. It was dull before. Now it’s funny. Keep it.’ Allowing actors to improve your work may appall most writers. Scriptwriters are used to it. Just smile and take the credit.

On one of the last sessions we did, I got to the studio and our producer was on the phone. He was ill. The actors were booked. I said, ‘So what do we do?’ and he said, ‘You’ve seen it enough. You produce.’ These were Pro-tools Mac days, and I said to the engineer, ‘I’m going to do it my way. I won’t be punching in single words or singe lines, I’ll retake. And even if it’s perfect first time, I’m going to do two or three more takes.’ He was appalled. The actors were very happy to re-take, rather than sitting puzzling over The Guardian crossword (which is what spoken voice actors do) while the engineer fiddled about with the computer. We finished the day half an hour early, rather than the 90 minutes early of the two previous days. I knew some of the cast, and they told me they had felt far more involved than usual especially as I told them the language teaching aim of each dialogue before and discussed it with them.

I’m going to extrapolate that to music. Some musicians love the seat of the pants energy and commitment of a single live take. But mainly there is a great deal of punching in to correct stuff.

Some vinyl fans find digital recording ‘cold’ and ‘inhuman’ and find analogue on vinyl or tape more ‘natural.’ Vinyl has had compression applied to the signal, otherwise deep bass or sudden piercing lead guitar might cause the stylus to jump in the groove. This may create a more pleasant listening experience, and one we are also used to hearing.

Digital should have a wider frequency range and some listeners complain about excessive bass and treble. A producer told me that he had spent a decade recording live bands in BBC studios, and stressed that to him, live music sounded exactly like a digital recording. He agreed that was often harsh, but he maintained that vinyl may have a nicer sound because of expert application of compression) but disagreed that it was ‘realer.’