Saturday, January 26th, 2013 | by Adam Kagan
Which sounds better, analog tape or 24-bit digital recording? Los Angeles producer/engineer Adam Kagan set up a session to find out. Listen to solo’d bass, guitar, and piano tracks from session stars Carlos Rodgarman, Grecco Buratto, and Carlitos del Puerto; decide for yourself whether you prefer the pristine sound of digital recording or the “warm” tones of analog tape on electric guitar, bass, and piano.
I started my professional recording career in the early 1990’s, when 2-inch tape was the main recording format for all but the few who could afford a Sony 3348 digital machine. We recorded to either a Studer, Sony/MCI, Ampex or Otari tape machine and mixed through our analog boards to a half-inch tape machine, usually a Studer or Ampex.
Each tape machine had its advantages and disadvantages: the JH-24 (Sony/MCI) could punch like no other and sounded great, but alignment was a pain and the remote was little more than functional. The Ampex machines sounded fantastic, but forget about punching at all. The Studers had the best of both worlds, with good punching and great sound — at a very high price. In contrast, the Otari machines and Sony APRs had modern features, like auto-alignment and built-in synchronization, but were not considered great-sounding machines. You pays your money and takes your choice. [JH-24 photo from Flickr user Ronan_C]
Fortunately for me, I started right on the cusp of affordable digital multitrack recorders, in the forms of the Alesis ADAT and Tascam DA-88s. Pro Tools (then called Sound Tools) was in its infancy, but worked very well for two-track editing. A few esoteric digital workstations, like the NED Synclavier, AMS Audiofile and Fairlight showed up in post-houses and top-level composer’s studios, but were way out of reach for the average person.
For a little perspective, a decent Sony/MCI JH 24 machine could be purchased used in the early 90’s for around $15,000–not including wiring or required maintenance (e.g., several thousand dollars for a new headstack). At that time a reel of 2-inch tape cost almost $200 and gave you 14 minutes of 24 tracks, so a typical 48 track project (one to three songs) would set you back at least $400 in tape costs.
ADATS and DA-88s brought the cost of 24 track digital audio down to $10,000–15,000, and you could buy just one 8-track machine to start with for around $3000. Digital tape was cheap, with ADATs relying on S-VHS tapes and DA-88s relying on Hi-8 tape cassettes. Each one of those formats, and their various spin-offs, had sonic and functional differences, like synchronizers, remotes, sample rates, etc. These modular digital machines revolutionized the home recording scene before Digidesign (now Avid), EMagic (now Apple), Opcode (now gone), Steinberg and MOTU took the lead with computer based solutions.
So here we are today, with me feeling old and telling stories of yore, with everyone all a-flutter with renewed interest in getting back to the “good old days” of recording to analog tape and pressing vinyl. The changeover from all analog recording to mostly digital recording during the late 90’s brought with it lots of bias, superstition and just plain ignorance. It is true that the very first generation of digital recorders, like the Sony F1 and early DAT machines, didn’t sound as good as the state-of-the-art analog machines. However, the low cost and ease-of-use of the new digital machines guaranteed their success. Luckily, pro audio and audiophile users pushed manufacturers to create better sounding converters and better tools to process the sound (now known as plugins).
It is my belief that much of the pain of switching over to digital recording was due to the tools that engineers had mastered for analog recording. For instance, applying EQ and compression (or no compression) to tape to make up for the color that the tape added didn’t sound so great when recording to digital. Bright FET microphones and harsh transistor preamp tones became rounded off in a pleasing way on tape, and by the 100th mix pass, the high-end was rolled off and the transients smeared so much that the final mix sounded phat, warm and fuzzy. It took experienced engineers a minute (or years) to gather their thoughts, re-examine their tools and learn how to take advantage of the clarity, quiet, and unforgiving purity of digital recording. At that point recordists moved towards super-fast, ultra-clean and high-gain preamps and transparent compression. Low cost digital processors stopped using transformers and tubes, which lowered costs and also lowered THD, while widening frequency bandwidth specs from DC to light. We had finally found it: perfect, clean, sterile audio!
Fast forward to now. Today’s great equipment designers came up studying the fundamentals of analog design, and learned digital design and processing during the evolution of digital. This has led to a hybrid approach of using colorful transformers and tubes alongside super-clean analog and digital processing for the best of both worlds. Recording artists like Lenny Kravitz led the way to digital, and then reverted back to analog, showing us the extremes of each medium. I think almost everyone is now satisfied that by using good sounding analog gear and even affordable converters, digital recording is not only acceptable, but can sound downright amazing.
Humans just can’t let things rest, though. Now we want to re-integrate production techniques and sonics that can most easily be found or created using our old friend(?) analog tape, alongside our digital gear. Fortunately, tape machines have not been out of use for very long, and the techs and engineers at top studios still remember how to align and maintain analog machines. Unfortunately, the companies that made 2-inch tape have all closed their doors. Some new companies have done a good job picking up the pieces, and have been manufacturing analog tape for a few years. While the quality and sonics of the new tape is not exactly the same as the best tape of late 90’s, the stock is very good and totally useable. It is, however, expensive, with a reel of half-inch running about $100, and a new reel of 2” tape from ATR or RMGI fetching just over $300.
I, for one, was happy to see 2” tape slowly fade away, and early on I was toting my Pro Tools rig to mix sessions where the studio begrudgingly set up their SSL synchronizer to talk to my Pro Tools rig. Nowadays the Pro Tools rig has taken the place of the tape machines in almost every studio, and a premium has to be paid to use the analog machines, if they are even still available.
So, how do we know when the analog machine will add just the right amount of punch, warmth, fuzziness or just plain magic to our drums, vocals, guitar, bass? Recently, I worked on a project that was recorded at a few different world-class studios, using both analog and digital recorders and used the experience to document some notable differences in the sonics of each recording format. I will summarize each tracking session and give audio examples, where possible for each instrument and format.
Here’s the basic scenario: a Latin pop record with some great players, mostly recorded using acoustic instruments (or amplifiers) in great studios. The drums, played by Mario Calire, were recorded at Henson (formerly A&M) in studio A and were engineered by Elliot Scheiner. If you don’t know who Elliot is, look him up — he has recorded, mixed and produced most of your favorite records, from Steely Dan to the Foo Fighters.
The drums were only recorded to 2” tape, so no direct comparison can be given here, but I’ll include a sample that you can compare against your drum sounds. Keep in mind that Elliot’s drum sound will probably kill yours regardless of analog or digital, but it is a great benchmark to work towards.Dry drum mix (analog tape)
The rest of the album was recorded at Ocean Way in Hollywood, using an Ampex 2” machine and their fantastic Focusrite console. The 2” tape we used was NOS 3M 996 tape stock that Henson Studios sold to us, and we recorded at a level of +5/185 at 30 ips, which is a quiet, but not overly hot level. For reference, most of the R&B stuff I recorded in the 90s was at +9/185, while a lot of rock stuff was recorded at +3/185. Basically, the higher the level, the lower the noise floor and the punchier the sounds. So, drum machines, synths and smooth vocals sound huge with very little tape hiss at a high level and a tape speed of 30ips, while distorted guitars, loud, live drums and screaming vocals sound huge, with a higher, but acceptable, noise floor at a lower recording level and lower tape speed (15 ips). No noise reduction was used on this project.
For the remainder of the instruments, all tracks were recorded buss out from the console and multed (split) at the patchbay to both the tape machine and the Pro Tools interface. The output of the tape machine’s repro head fed a Prism Dream converter and then went to Pro Tools digitally. After each recording pass, the tape tracks were nudged forward in Pro Tools by a bit over 9000 samples (about 10ms at 96khz), which was the delay between the direct signal and the output of the repro head of the tape machine. We recorded into Pro Tools at 96khz, 24bit. Vocals and percussion were recorded at our private studios, so there was no tape used, but the few synth tracks we used were recorded through tape. I didn’t print tape and Pro Tools for the synth tracks because I could spend time playing back midi instruments and adjusting the record level to get the desired tape saturation vs tone for the synths. For the live players, I didn’t have the luxury of time to experiment with tape levels, so I went with my experience for best practice.
WAV Audio Archive
[Download the 16-bit, 44.1 kHz WAV audio files here.]
Bass (acoustic and electric)
WIth drums out of the way, we were onto bass. First we recorded acoustic bass, played by Carlitos del-Puerto. I used a FET 47 on the body and a KM54 on the neck, with both mics about 8 inches from the instrument. Both signals were mixed at the console and bussed to one output.
Electric bass was then recorded using Ocean Way’s custom Jensen transformer DI box into the console and then through their respective tape and Pro Tools paths.
In both pairs of samples, listen for the difference in the low end response, apparent level (thanks to tape compression) and clarity of each path. During mixing, I used only the bass recorded onto tape for all the songs.
Guitars (acoustic and electric)
Guitars were next and the player was Grecco Buratto on both acoustic and electric. Acoustic guitar was miked with a Neumann KM 54 around the 14th fret. After a few minutes, I decided the analog tape was not the way to go with acoustic guitar, so we recorded only direct to Pro Tools. Electric, however, was another story. Grecco came to the session armed to the teeth with amps, cabinets and pedals. We set up four or five different pairs of cabinets in the live room and his pedalboard, amps and switcher were in the control room with us. Each cabinet was miked with three mics (isn’t a big studio wonderful!): an SM57, an MD 421 and a FET47. I usually buss the 57 and 421 to one track and record the more distant 47 as a room mic onto a separate track. For the audio examples, there are a variety of styles and one example also includes the mix processing to show how different the guitar wound up sounding during the mix. For the most part, electric guitars on tape were used for the final mix.
All guitar tracks are the sum of an SM57 and MD421 close mics:
Funk Electric Guitar, 24-bit digital
Funk Electric Guitar, analog tape
Electric Guitar, 24-bit digital
Electric Guitar, analog tape
Electric Guitar, 24-bit digital
Electric Guitar, analog tape
Electric Guitar, analog tape, EQ'd for mix
Piano was interesting. We had two pianos to choose from and we went with the smaller, 7-foot Steinway B, over the 9-foot D. Piano was played by co-producer Carlos Rodgarman. I miked the piano with two Royer R-121s in a sort of XY pair, right above the hammers and focused about halfway down the strings. The setup looks like a “V” with the point toward the strings. The backside of the mics point over the players shoulders. The live room in studio A is huge–over 2000 square feet with 23′ ceilings. With this mic setup, the piano sounded live, but not bathed in reverb. For this type of pop music, I like a bright, punchy piano with good stereo spread and the Royers provide that very well–and still fold towards mono without sound phasey. The ribbons also take EQ well, and I provided a sample of tape, Pro Tools and the processed version used for the mix.Piano, 24-bit digital
Piano, analog tape
Piano, EQ'd for mix
The processed version wound up heavily EQ’d, but in this dense funk mix it sounded great. For some songs, the tape felt a bit thinner than the digital recordings, while for some songs, the mellower tape sound was just the ticket. Each version had its benefits. With the right compression and EQ, I generally find that even a poorly recorded piano can be made to sit well in most mixes–as long as the piano doesn’t have to be featured.
All-in-all it was a good experiment and well worth the extra mile. For the drums and bass, the tape certainly added a punchiness that would otherwise take a bit of processing to achieve. The electric guitars benefited from the rounding off that the tape provided on certain tracks and on others I like the clarity of digital signal.
Acoustic guitar didn’t really work for me on tape, and piano was a mixed bag.
For day-to-day recording I don’t have the luxury of tape, but for the important projects or the special case where a certain color is desired, tape can certainly add its sonic signature. As for printing my stereo mixes to tape vs digital… don’t get me started…
Many thanks to Adam Kagan for conceiving and delivering this piece! Follow Adam online via http://about.me/adamkagan.
If you enjoy challenging long-held assumptions about whether vintage gear always sounds better, you’ll probably enjoy our most famous mic shootout, the U87 vs. the U87Ai.