Can UA's Studer plug‑in really reproduce the sound of saturated multitrack tape in all its glory? To find out, we tracked down a real Studer A800, a top‑flight studio and a hitmaking band...
California‑based Universal Audio are well known for taking great pains to emulate vintage analogue gear as accurately as possible in their software. Their latest plug‑in takes on perhaps their most demanding challenge yet. A year in development, the Studer A800 plug‑in for the UAD2 DSP platform is officially sanctioned by Studer and modelled after a 24‑channel Studer A800 two‑inch tape machine. UA say that their plug‑in takes tape emulation to new heights, mimicking and reproducing every nuance of analogue tape compression. We decided to put their claims to the test and pit the plug‑in against the real thing, in a classic analogue‑versus‑digital shootout.
The first microprocessor‑controlled tape machine, the Studer A800 marked a new generation of professional multitrack recorders when it was introduced in 1978. Over the years, a lot of classic albums have been recorded on it, by artists including Metallica, Stevie Wonder, Tom Petty and Jeff Buckley, to name but a few. And although two‑inch tape is increasingly scarce and expensive, and tape‑machine maintenance is time‑consuming, studios around the world still rely on the performance and the sound of analogue multitrack recorders. Recording to tape yields a subtle coloration and compression that is often described as 'warm' and 'cohesive', with solid low end and overall presence making instruments glue together in a musical way. No longer limited by track counts in a DAW environment, tape machines are often used purely for these sonic qualities, to record drums, bass and guitars. These are then transferred into a DAW, to combine tape sound with the freedom of editing.
Universal Audio's team of DSP engineers teamed up with AES magnetic recording expert Jay McKnight, and over a 12‑month period, they made sure that the plug‑in faithfully modelled the entire multitrack tape circuit and electronics of an A800 MkIII. To make things even more complicated, they modelled the characteristics of four different tape formulations: 250, 456, 900 and GP9.
At first glance, the plug‑in resembles the real thing, with graphics displaying the spinning reels and VU meter along with a couple of silver‑coloured knobs and the large, characteristic, backlit plastic buttons found on many Studer tape machines. A closer look reveals that the original transport buttons have been replaced with four buttons for choosing the virtual signal path: by selecting Input, Sync or Repro it's possible to listen to the input signal, the signal coming from the sync/record head or that coming from the repro head. The last is the 'full works', modelling the entire signal path going through the input circuits, through the recording head onto tape and then through the reproduction head, plus all corresponding electronics. There's also a Thru setting, which serves as a signal bypass for the plug‑in.
Unlike the real hardware, no tape has to be rewound, and changing tape formulation is conveniently done by turning a silver knob. The choice of tape formulation has a close relationship with the calibration level of the tape machine; as the formulations improved over the years, higher recording levels became possible, thus lowering the noise floor. In normal use, the calibration would correspond to the tape's output level, but some users would under‑calibrate to leave more headroom. With the Cal knob, it's possible to choose four different calibration levels from +3 to +9dB, with correspondingly different noise floors. The speed of the tape will also affect the noise floor, but more importantly it will alter the frequency shift, 'head bump' (the low‑frequency boost associated with tape) and distortion characteristics, with 7.5 inches per second being the slowest setting and 30ips the fastest. For pop and rock music today, the most often-used speed is 15ips, due to its low‑end response and general compression characteristics. The plug‑in's input and output level controls allow you to drive the level to and from tape to your requirements.
Above the Studer logo there's a rather anonymous 'Open' legend. Clicking on it reveals a set of calibration knobs. True to the original hardware, there are controls for adjusting the Bias, HF Driver, Sync and Repro EQ, and Emphasis EQ. A common technique is to increase the bias voltage beyond the ideal setting for the chosen tape formulation, producing a warmer and slightly saturated sound — perfect for drums, bass and electric guitars. The HF Driver is used to tune the HF content of the signal prior to going to tape, to make up for the high‑frequency loss due to bias optimisation. In practice, the Bias and HF Driver parameters can be used to alter the saturation characteristics.
The Sync and Repro EQs adjust the low‑ and high‑frequency content of the signal coming from the tape circuitry. They are normally used to shape the sound to maintain a flat frequency response, but may also be used for creative tonal shaping. Depending on which side of the Atlantic reels were rolling, the pre‑emphasis filtering for 7.5 and 15 ips machines complied with the American NAB or European CCIR standard (also called IEC). Even though the latter is considered technically superior, NAB is still in use for its unique 'sound'. Choosing 30ips tape speed automatically enables the worldwide pre‑emphasis standardisation once determined by the Audio Engineering Society.
When the Auto Cal is enabled, the plug‑in automatically calibrates the controls to the optimum settings depending on tape formulation, speed and emphasis — and I can only imagine how much time this feature would have saved, working with a real tape machine! Another feature many hardware users probably wished for back in the day is the possibility to turn off the noise: simply pressing the Noise button takes away tape hiss and machine hum. It's even possible to adjust the level of the broadband hiss and hum, the latter being, of course, 50Hz when CCIR is chosen and 60Hz when NAB is enabled.
To simulate a two‑inch tape machine within a digitally recorded multitrack project, UA recommend activating an instance of the Studer A800 plug‑in on the first insert point of every audio channel in your DAW's mixer. Thoughtfully, they have made it possible to gang the controls across multiple instances, so that you can experiment with different settings in the project without having to open 20 or 30 plug‑in windows! Of course, you'll need plenty of DSP horsepower to run it on every track in a large multitrack mix. For example, when I used 24 mono Studer A800 plug‑ins on a UAD2 Quad, the UAD Meter showed that 59 percent of the available DSP power was allocated at 44.1kHz, and 79 percent at 96kHz.
In order to get a first impression of what the plug‑in was capable of, I loaded up a basic rock project with drums, bass guitar, two electric guitars and two acoustic guitars. As the manual suggests, I put the Studer A800 plug‑in on every channel and enabled the Gang function. Then I chose the 456 tape formulation and 15ips tape speed, set the calibration to +9dB and raised the bias voltage just a little bit, for some extra tape saturation. Listening to the raw recording with added 'tape' compression just made the tracks sit together a little better; it was a subtle change, but still audible. Using the UAD SSL Channel Strip on most tracks, along with two Lexicon reverbs, I was able to build up a working mix pretty fast, and what struck me was how easily the drums, bass and guitars fitted together. This was particularly noticeable on the toms, which otherwise have a tendency to end up on top of the rest of the drums or tucked in behind the bass guitar. The tape compression made them sit just right and gave the snare a tad more bite, while the kick drum was somewhat rounded off and the low end dampened a little bit, making it fit nicely with the rest of the drum kit.
Both the bass guitar and the two electric guitars sounded slightly warmer going through the plug‑in, and driving the plug‑in pretty hard rounded them off nicely, while the acoustic guitars sounded less 'pokey' with tape compression. The 456 tape type certainly added a sound of its own, while changing the tape speed to 30ips boosted the low end around 60‑100Hz, as well as giving the overall mix some more punch. The difference between the processed and non‑processed mixes was less obvious at 30ips, but the plug‑in still added some kind of glue, making the mix more coherent, and the bass boost is, of course, seductive. Just like in real life, the 900 and the GP9 formulations can take a lot of level without adding any noticeable tape compression, and when I set the plug‑in to 30ips and GP9 formulation and adjusted the low‑frequency Repro EQ, it became pretty hard to distinguish the processed mix from the non‑processed mix, the latter being just a tad thinner‑sounding.
Although it's modelling a multitrack tape recorder, of course there's nothing to stop you using the Studer A800 plug‑in as a master bus processor, to simulate mixing to tape. In this context, across a pretty much finished pop mix, I found the 900 tape at 30ips to be the best‑sounding, adding a little bit of tape compression.
Having dipped a toe in the water, I was itching to put the plug‑in up against the real thing. I got access to a two‑inch, 16‑track Studer A800 at Svenska Grammofon Studion, one of Sweden's finest studios, and studio owner Kalle Gustafsson Jerneholm helped out with the recording, along with in‑house producer Hans Olsson Brookes and tech whizz‑kid Henrik Gustafsson. Our first task was to lift the extremely heavy tape machine down a couple of stair steps and move it into the main control room, a task that even piano movers would have refused to do — according to the manual, it weighed 690 pounds (about 315 kilos). By sliding it on a ladder we were able to get it downstairs and into the control room without breaking any backs. Once it was finally in position, Henrik started recalibrating it and tuning it for Quantegy 456 tape, making the machine slightly overbiased for more obvious tape compression. Hans set up the digital recording session, running 32 channels into Pro Tools at 24‑bit/96kHz via Lynx Aurora converters, clocked by a Rosendahl Nanosync.
Our goal was to simultaneously record 16 channels directly from the mixing console and 16 channels from the repro head of the tape‑maching, thus avoiding the need for synchronisation. The decision to record at 96kHz was made in order to take the converters out of the equation as far as is possible, and to adapt to the workflow of many modern‑day users of tape machines: recording drums, bass and guitars on tape and then transferring them into Pro Tools at 96kHz to keep the sound quality.
Svenska Grammofon is almost like a museum, with a vast collection of vintage equipment and musical instruments, and the main attraction is a 48‑channel Neve 8048 mixing console once owned by Queen. It's packed with Neve 1081, 1066 and 1073 modules, and a lot of classic albums have been recorded and mixed on it. Naturally, we took the opportunity to use it for our recording session, and the track sheet and microphone setup is shown in the box overleaf.
As well as being a studio owner, Kalle is also a songwriter and bass player in one of Sweden's best‑known bands, The Soundtrack Of Our Lives (www.tsool.net), and he managed to round up the entire band to help out with the recording session. Everybody was playing together in the main live area, with the guitar amps tucked away in booths. The drum kit was miked in a pretty conventional way, with two mics on the kick and snare drum, two overhead mics, tom mics and a room mic positioned about four metres away. The two kick and snare mics were summed to one channel each, to save tracks going to tape. The drum tracks received some basic filtering, but no outboard compression was applied; instead, we allowed the signals to hit the tape pretty hard. The bass guitar was, interestingly enough, run through a Roland RE201 Space Echo, adding some tape delay, which gave it a nice sustain, and then went into a Dbx 160 compressor. The Farfisa organ was connected to a guitar amp and the Korg MS10 was recorded through a DI box. The vocals were recorded using an Electro‑Voice RE20, with compression from a Teletronix LA2A.
Standing in the control room one floor above the studio, watching the band play — through a big studio window — was a total blast. The Soundtrack Of Our Lives are a great live band, and rocked just as hard in the studio as they do on stage. We ended up recording just three takes: two songs at 15ips and one at 30ips. Vocal overdubs were added to one song, and it was all done in two hours. As this was just days before Christmas, everybody was happy with the fast and easy‑going recording session!
Back at my place, I embarked on the quest of matching up the directly recorded tracks to those that had been tracked through the tape machine, adding a UAD Studer A800 plug‑in on every channel that hadn't hit the tape. I knew that I probably wasn't going to be able to match the two recordings exactly, due to the fact that the tape machine we used was a 16‑track, while the plug‑in is modelled on a 24‑track tape machine. Furthermore, the plug‑in models the transformerless Studer A800 MkIII, and the tape machine we used was the first model with input transformers.
I went through the instruments track by track, tweaking the settings of the plug‑in and checking the transient response with two level meters. My new Quested main speakers were of great help in revealing the subtle tonal changes that were needed to sonically match the tracks. Tweaking the Bias, HF Driver and Repro EQ parameters, along with some level matching, got the tracks starting to sound convincingly close to the originals.
With a basic match of the raw tracks, I wanted to check how they would compare in the context of a mix with processing applied, so I made a rough mix of the tape tracks, mainly using UAD plug‑ins and some correctional filtering with the built‑in EQ in Pro Tools. Switching to the directly recorded tracks and applying exactly the same processing made me realise that more tweaking was needed, because the low‑end didn't sound exactly the same, but most of all, the mid‑range was a bit harsher‑sounding. So I redid the matching, but this time with the mix processing in place — a process that revealed more tonal differences across the frequency spectrum. This second time around, I ended up using more tape saturation, which made sense, because we were hitting most of the tape channels pretty hard in the original session. There were other differences, too: the low end was somewhat rounder on the tape tracks and the mid‑range a bit softer, probably due to the input transformers on the tape machine. I was able to match these characteristics pretty well by lowering the HF Driver parameter, which backed off the harmonic distortion a little bit, and by boosting the input level and low‑frequency Repro EQ on some tracks. These tweaks effectively made the mix sound even closer to the original.
My main struggle was, unsurprisingly, with the bass guitar, which on tape had a very nice‑sounding roundness, coupled with a solid low end. To get the same tape compression, I had to hit the UAD Studer plug‑in pretty hard, which in turn diminished the low‑end punch around 40‑60Hz. This probably means that a 16‑track tape machine can retain more low‑end punch compared to a 24‑track at 15ips; switching to 30 ips on just the bass track brought the low‑end punch back, but it also changed the overall tape saturation too much, so I decided to go back to 15ips. I had no problem getting a deep low‑end with the plug‑in, but had to accept that it wouldn't sound exactly like the tape.
The two electric guitars — also heavily compressed on tape — had a slightly softer mid‑range than could be achieved with the plug‑in, but the extra mid‑range in the plug‑in versions might equally well be called 'clarity', and in any case the differences were only audible in a direct A/B‑comparison. Likewise, the vocals ended up a little bit more aggressive through the plug‑in, but in this case I didn't find that to be a bad thing.
Having matched the 15ips session to the best of my ability, I opened up the 30ips recording of the same song. Importing the session data and plug‑in settings from the previous session not surprisingly showed the 30ips recording to be more open‑sounding. Everything was more focused and the low‑end was more defined, which on a dense recording like this one was a good thing, though at the same time, the nice sound of tape compression was less audible. Changing the plug‑in to 30ips pretty much made it sound the same as the tape channels, and once I had added some more tape compression to match individual tracks, the two were very close. Somehow, it felt easier to match the plug‑in to the tape at 30ips, presumably due to the fact that the tape compression was less obvious.
I invited the band to make their own artistic adjustments to the mixes, and naturally I took the opportunity to expose them to a blind test. Switching between the two versions I had made of the mixes did reveal some tonal differences, but the consensus was that they were fairly small, and the vote was not unanimous on which was which. In the end, the tape mix was identified as sounding more processed in the low end and a tad smoother in the mid‑range; but on the other hand, the plug‑in mix had a deeper low end and a slightly more open sound. The most interesting thing was that the plug‑in mixes were never identified as 'digital sounding' or harsh; they were just slightly different, which I think really shows the potential of this plug‑in. As the mixing session progressed and we were switching back and forth between the tracks, one of the guys said "Well, it doesn't do anything bad to the sound!” — and given the amount of processing that was going on, that's actually a remarkable accolade. When scrutinising the sound of plug‑ins, there's usually something that gives them away as being plug‑ins, but this is not the case with the Studer A800 plug‑in. When I demonstrated the creative possibilities — adding the often-needed high end in a mix by ganging all the tape plug‑ins and just raising the HF EQ a little bit — it made the band realise what a powerful tool it is.
The Studer A800 plug‑in very successfully mimics the sound of the real hardware, but is also a very powerful tone‑shaping tool in its own right. The Sync and Repro EQs can add a solid low end and a very sweet high end, but the real fun begins when raising the input level to hit the 'tape' hard. Great results can be achieved by experimenting with different tape formulations and tape speeds, along with the Bias and HF Driver parameters. Applying moderate tape saturation will make drums come alive in a nice way, while heavy saturation can inflate them in a cool and creative way. Heavily saturated electric guitars acquire a soft, fuzzy distortion that can be altered by adjusting the Bias and HF Driver. Typical wave‑shaping plug‑ins are more versatile, but can also be a bit flat‑sounding: abusing the Studer A800 plug‑in, by contrast, produces a distortion that feels alive.
When talking about analogue tape, people tend to get a glow in their eyes and go on about how much better music sounded back in the day. Recording on tape certainly makes a difference, but other factors beside tape compression were responsible for that sound. The limited track count meant that adding 40 tracks of vocal overdubs simply wasn't possible, and there were, at best, three tracks for the singer to nail his or her parts. Even if editing could be done by cutting tape, it was a lot easier to just do a couple of punch‑ins until the musicians nailed it, and everything didn't have to sound perfect. Accepting some flaws, it still made the musicians more alert, because what ended up on tape was what ended up on the record. Recording on tape is — compared to working with computers — more of a musical process, because it's easier to focus on what's being recorded when you're not staring at a computer screen. Those fortunate enough to have access to a tape machine and the time and money to keep it in good shape will most likely continue to record on tape.
For the rest of us, not having the space or the financial means to buy an analogue multitrack tape recorder, but still wanting the sound of tape compression, the UAD Studer A800 plug‑in is a great alternative. Will it instantly make your vocals sound great and your guitars sound fat? No, but recording them on tape won't either. A tape recording will make everything sit together in the mix and when the plug‑in is strapped across all audio channels — adding just a little bit of saturation — it will also provide the oft‑needed glue to make the mix more coherent. (It's just a shame that non‑UAD2 users can't get access to it!)
Now and then, plug‑ins are released that really change the way we relate to digital processing, and make the boundaries between analogue hardware and plug‑ins more and more blurred. Using the latter instead of the equivalent hardware no longer necessarily has to do with decreasing production revenues or the music business in decline. Sometimes the plug‑ins just sound as good as the hardware, so why not use them? Universal Audio have gone to great lengths to capture the very essence of analogue tape compression and have done so very successfully. Naturally, it can't reproduce the experience of watching a big tape machine spinning reels and pushing the VU meter into the red zone. What it will do is to provide a glue that makes your mixes more coherent, and if you really commit to the sound, makes mixing a bit easier. It's not like waving a magic wand, but recording on tape isn't either; it's a way of working. The next time you're recording a band, why not tell them that you're using a new tape plug‑in, and that there will only be 24 tracks available for the recording session?
Thanks to the band and everybody else involved in making this article possible.
There are many plug‑ins out there that attempt to model some or all of the characteristics of analogue tape, but I have not seen another one which does so to this level of detail and authenticity.
Svenska Grammofon Studion is one of Scandinavia's leading recording studios, with an amazing collection of vintage equipment, guitar amps and instruments. Located in the centre of Gothenburg — Sweden's second‑largest city — the studio is a creative hub, with resident songwriters such as José González and Robyn producer Andreas Kleerup. The 8000‑square‑foot facility offers accommodation for recording customers and free coffee for around‑the‑clock sessions. The main control room is equipped with a vintage Neve 8048 console once owned by Queen, with discrete 1081/1066/1073 modules, an SSL 4000 console as sidecar and a jaw‑dropping collection of outboard from Fairchild, Pultec, Teletronix, UREI, EMT and Telefunken, to name but a few. Four multitrack tape recorders are available from Studer, Ampex and MCI. The other control room is also equipped with a Neve 8048 console, and both control rooms are located next to the main live room, which is over 30 feet in height and houses an array of classic Neumann microphones and vintage instruments. Two additional control rooms with SSL 4000E consoles are also available for mixing duties, with adjoining vocal booths for overdubs. A recently built mastering room with Weiss outboard and ATC speakers makes it possible to keep all the creative work in‑house.
The studio can be contacted by phone or email, and very impressive gear and client lists are available at their web site.
+46 70 775 97 70.
Grammy Award nominees The Soundtrack Of Our Lives are Sweden's grand ol' ladies of rock & roll. A notorious live band and the Rolling Stones' favourite opening act, they've supported U2 and Robert Plant and played live on David Lettterman and Jay Leno's TV shows. They are touring the US through March and April. The band, from left to right, are Ebbot Lundberg, Fredrik Sandsten, Kalle Gustafsson Jerneholm, Mattias Bärjed, Martin Hederos and Ian Person. For more info, visit their web site at www.tsool.net.
|1||Room mic||Thuresson CM402||Neve 1066|
|2||Kick drum blend||Neumann U47 FET (outside) + Electro‑Voice RE20 (close)||Neve 1066|
|3||Snare drum blend||Sennheiser MD441 (top) + AKG C414 (bottom)||Neve 1066|
|4||Rack tom||Sennheiser MD421||Neve 1066|
|5||Floor tom||Sennheiser MD421||Neve 1066|
|6||Overhead L||Neumann U67||Neve 1066|
|7||Overhead R||Neumann U67||Neve 1066|
|8||Rickenbacker bass||Line + Roland RE201||Neve 1073 with Dbx 160|
|9||Gibson 335 + effects through Fender Twin Reverb||Shure SM7||Neve 1073|
|10||Gibson SG + effects through Fender Twin Reverb||Sennheiser MD421||Neve 1073|
|11||Farfisa organ through Fender Blues DeVille||Neumann U87||Neve 1073|
|12||Korg MS10||Line||Neve 1073|
|13||Lead vocal||Electro‑Voice RE20||Neve 1081 with LA2A|
|14||Lead vocal overdub||Electro‑Voice RE20||Neve 1081 with LA2A|
|15||Choir 1||Electro‑Voice RE20||Neve 1081 with LA2A|
|16||Choir 2||Electro‑Voice RE20||Neve 1081 with LA2A|