It’s become almost a cliché to say that digital audio recording and reproduction has become clinical, both in its production and in the listening satisfaction derived from the final product. That often high sampling rate 24 bit recordings end up being delivered in lossy formats via downloads is a strange reward for all the engineering efforts made over the years to remove imperfections from the audio production chain.
Yet some of the best recordings of the last 60 years were recorded on, what would be regarded now, as lo-fi equipment and delivered to the listener on vinyl. And what’s happened 60 years on? We’ve become almost obsessed about trying to recreate, through intensive digital processing, the very imperfections which the design engineers spent all that effort trying to get rid of. The emulation of the Studer A800 multi-track machine is just such a case in point.
There is another way!
Way back in 1951 Studer introduced the ReVox brand to market its range of hi-fi products and the ReVox branded T26 tape recorder was launched in 1952. Fast forward (think about it!) to 1967 and the B77 ¼” reel to reel machine was launched as a successor to the incredibly successful A77 which had found its way into many amateur (we call them ‘home studios’ today) recording set-ups, and in fact the machines were so good that they became common in many pro-setups as well.
The B77 model range offered both quarter and half-track versions running at either 9.5/19cm/s or 19/38cm/s. Before anyone gets snooty with cries of ‘it’s not a Studer’. Stop! I visited the Studer factory in Regensdorf many years ago and believe me, Studer worked to only one quality standard whether it was branded ReVox or Studer.
The B77 was designed to be incredibly robust, using diecast parts for the motor chassis, the side bearers, the crossmember and for the head block and the pinch roller arm. This ensured exceptional stability for the precision tape guiding system and the sound heads, as well as for the motors and the brake assembly. The motors were sturdy AC asynchronous types – unbeatable in simplicity and reliability. I’m told that even 40 years on, the motors hardly ever fail.
This direct drive capstan motor with tacho generator and servo electronics ensured outstanding speed stability independent of variations in load or changes in the electrical supply voltage or frequency. As you will see, this speed stability is critical to its use within a multi-track recording set up today.
My own machine is a Mk1 half-track (means 2-track) high speed version switchable to either 19 or 38cm/s and configured with IEC equalisation. There’s no Dolby noise reduction (it was an option) and the inputs and outputs are unbalanced (there was a balanced version, the PR99), but none of this matters at all, when what you are looking for, is the real, authentic, original, sound of tape.
Talking of tape, there are still two suppliers offering the full range of reel to reel tape formats from ¼” all the way through to 2” for the large number of multi-track machines still in use today. ATR Magnetics in the US and RMG/Pyral in Europe. I’ve chosen to use RMG SM911 formulation which my machine was calibrated to but both manufacturers offer quality tape products.
The B77 audio I/O is available on phono jacks on the rear of the machine which can be used standing or ‘lying on its back’ so to speak. Input and output levels are quite low by modern standards but I found no practical problems interfacing to the rest of my recording setup. My machine was setup to read 0VU with 0dBu input (input gain set to ‘4’) and it delivers back the same 0dBu during playback. Maximum input level (before the electronics clip) is around +16dBu.
I had originally planned to track to tape and then bounce straight to digital but I realised that this was pointless as it means more tape use (all those retakes!) for no real benefit. Instead I decided to use the B77 as a track and group insert within my DAW. Any performance loss caused but the A-D and D-A converters is incidental compared to the sonic footprint imparted by the machine and tape combination.
With the machine correctly aligned and ensuring the heads are cleaned before each recording session, frequency response is about 0.5dB down at 50Hz and 20kHz (-3dB at 20Hz). HF response above 20kHz falls away very sharply and is 20dB down at 30kHz.
LF response is fairly independent of recording level but HF response, is heavily dependant on how hard you are ‘pushing’ the tape and is part of the sonic footprint which defines the appeal of analogue tape recording.
One of the other components ‘of the sound’ is the compression which occurs as the tape gets closer to saturation. This is heavily frequency dependant but occurs all across the frequency band at high recording levels. In theory, compression caused by the tape saturating is most measurable (and audible) at higher frequencies, however in practice, with analogue tape machines, it is most prominent at frequencies between 6kHz to 12kHz, peaking on the B77 around 8kHz. Although the tape does indeed saturate more at frequencies above 12kHz, the impact of the fast HF frequency response roll off above 20kHz means that harmonics at 24kHz and above are filtered out.
This tape saturation is a form of compression, and in measurement terms, takes the form of a series of odd and even harmonics (so distortion). To give you an idea of how prominent these are; recording at +9dBu you can measure;
20Hz test audio signal
This measures at a little under 0.5% THD and adds subtle but audible colour.
10kHz test audio signal
This measures at around 2.8% THD but bear in mind much of this is way outside audible territory.
The saturation/compression, especially at mid to high frequencies also results in compression based level reduction. Hitting the tape at the same +9dBu means a compression of around 1dB at 8kHz increasing to around 3dB of gain reduction at 16kHz. Many of you will have heard of this type of audio ‘warming’ based on soft compression caused by valve circuitry when pushed hard, and although the mechanism is different, it shares some of the audio characteristics of soft tape saturation.
I was originally concerned about whether the B77 would have good enough speed stability to be used as an insert device. Those of you familiar with three head tape machines (erase, record, replay) will be aware that there is an audio delay between the point in time at which the recording is laid on tape to the time it is played back through the replay head. This is entirely due to the physical distance between the record and replay heads and dependant on the tape operating speed. With the B77 recording at 38cm/s, this works out at 108.7ms including the DAW A-D and D-A converter delays and any playback buffer latency.
What was critical was whether this delay would be sufficiently constant over the entire recorded track duration. Suffice to say that Studer’s tape speed regulator design was and still is a great piece of engineering and in practice timing delay through the B77 is rock solid.
In practice, all that is required is to insert the same fixed delay into all tracks which are not processed through the B77 and everything is nicely time aligned.
So how does it all work out?
The combination of tailored frequency response, tape saturation, not to mention wow, flutter, and the fairly high tape noise floor, all combine to make a wonderful analogue audio effect for modern digital recording setups. You can control the effect from very subtle to the more obvious, simply by increasing the recording level into the B77. You need to take care not to clip the B77’s input stage but in practice with already recorded tracks its very easy to control the peak level you send out to the tape machine. You can do this in mono or in stereo.
The B77 is mechanically very robust. Electronics and some of the switchgear can be a more problematical. To date I’ve had to change one diode (controls the switching of the tape bias generator) and I’m keeping an eye on one of the record enable switches, but I’ll tell you what – if any of your current ‘state of the art’ digital recording set-up is still working so well in another 40 years – I’d be surprised (software obsolescence will have consigned it to the bin long before then anyway!).
Quality two track machines are available on EBay and from a number of specialists, so really, there’s no excuse for not keeping this wonderful form of recording medium alive for another half century.