Abbey Road Studio’s revered RS124 tube compressor has never been made available commercially... until now.
Chandler Limited are in the enviable position of having negotiated a licensing arrangement with Abbey Road Studios. This enables them to produce outboard gear derived directly from the innovative equipment developed by EMI’s design engineers during the heyday of the British music industry. The studio’s RS124 compressors, which have been credited for much of the sound character associated with early Beatles records, became so popular essentially because of their velvety-smooth and very flattering sound character, and their sublime ability to control dynamic transients without killing the high-end or sounding muffled. They were routinely used for tracking, especially for bass and guitars, as well as for general bus compression, and even in the mastering rooms. The relatively slow attack time worked particularly well with bass instruments, because the instrument’s initial punch and attack was always retained while the dynamic range was brought firmly under control. Apparently it became almost a standard practice during the Beatles recordings to compress the bass in two or three passes to really level the part while retaining the leading attack of individual notes.
Another important facet to the RS124’s sonic character was its ability to warm up the low mids at the same time as slightly attenuating the 3-5 kHz region, reducing any harshness in the source material very nicely — and especially so when pushed towards 20-30 dB of gain reduction. The only real ‘weakness’ in the design was that its response dropped off pretty quickly below about 80Hz — although that might actually have been a useful side effect at the time!
Despite it’s reputation, the RS124 was never made available commercially by EMI and, while various software emulations, and a few hardware ‘homages’ to the design have been produced over the years, Chandler’s new RS124 is the first hardware product produced with Abbey Road’s official approval, and developed with direct reference to the original design notes and the existing vintage units (which remain in regular use today).
Vari-mu
The vari-mu concept was originally developed for automatic gain control (AGC) in AM radios, where the varying strength of the RF carrier signal (due to reflections and interference) resulted in unwanted audio level variations. The vari-mu system relies upon a non-linear transfer characteristic in the valve, such that its overall gain (or ‘mu’) can be varied smoothly by adjusting the valve’s biasing conditions. In the radio AGC configuration, the valve’s gain is controlled by the averaged RF carrier level, removing the unwanted level variations and allowing the decoder to respond only to the wanted audio amplitude modulations. When configured as an audio compressor, the valve’s biasing is adjusted instead by the side-chain control signal (derived from the signal level of the audio itself), but the basic principle is the same.
Many of the RS124’s connections are hand wired.EMI’s RS124 (see ‘The Mods That Rocked’ box) is a remarkably simple design, derived from Altec’s 436B vari-mu compressor, which employs just three valves configured with a differential signal path throughout. The input signal is introduced via a transformer to the two grids of a 6BC8 dual-triode as a differential signal, and this valve acts as the variable-gain (vari-mu) element in the signal path. The anodes of this input valve directly feed into the grids of a push-pull output stage built around a 6CG7 dual-triode valve, and its anodes drive the output transformer to complete the audio signal path.
The feedback side-chain path is built around a 6AL5 dual-diode detector valve, which monitors the output level and generates a control voltage. This is then applied, as a common-mode signal, to the input stage to alter its biasing and thus change the gain. The compression threshold is determined by the biasing of the dual-diode detector valve. As this is fixed in the Altec design, the amount of compression is controlled operationally by adjusting the input level — raising or lowering the input signal above or below the fixed threshold, as required. There is no ratio control, but the ratio has a soft knee and settles at roughly 2.5:1.
Altec designed the 436B with fixed attack and release time constants, but the EMI engineers wanted to be able to adjust the release time, and came up with a simple modification which gave six switchable options. Various other modifications concerned the introduction of constant-impedance input and output attenuators (required for proper matching with other EMI equipment), a facility to check and adjust the output balance to counteract valve ageing effects, and a side-chain stereo link facility (see box).
Chandler’s Interpretation
While Chandler’s modern version of the single-channel RS124 looks suitably vintage and EMI-esque, it’s not quite an exact copy of the original units: it follows in the same vein as most of their Abbey Road product line, in that while it uses accurate circuit recreations these have been subtly and usefully enhanced, to make the product fit better with modern equipment and production requirements. One of the most obvious differences, for example, is the addition of an attack-time switch.
On the rear panel is a jack socket that enables two units to be linked for stereo operation.
The rear-panel connectivity comprises a male and female XLR for the custom–transformer–balanced input and output. A toggle switch changes the output impedance between 200 and 600 Ω, the latter being the most appropriate setting for most situations, with the former sounding a little ‘darker’ in most installations. The quarter-inch TS socket allows the side-chain control voltage to be shared when stereo linking two units — although all the controls on both units must be manually set to the same values to ensure stable stereo imaging. Mains power is connected via an IEC inlet, which is accompanied by a fuse holder. The operating voltage is set at the factory and marked on the rear panel (110-120 or 220-240 Volts AC).
The compressor’s 2U front panel is modelled closely on the original RS124 chassis, with the input and output level controls featuring large chicken-head knobs. The standard model uses conventional potentiometers for these controls, but Elma rotary switches can be specified as a cost-option. (For the review, I was supplied with the more expensive model with switched input and output level controls.) The output balance adjustment facility has been retained, with a momentary switch as on the original, and a recessed trimmer. The typeface of the Dymo-tape labelling for the original EMI modification has also been replicated!
Different Altec compressors were modified by EMI’s engineers and, consequently, each RS124 unit at Abbey Road is known to have slightly different characteristics, particularly in terms of attack times. Chandler’s version allows you to choose between these characteristics, the three red labels around the Attack control being the serial numbers of the original units.
In place of the original’s lone Recovery switch, the Chandler revision has squeezed two rotary switches into the same space. The lower one provides the six Recovery (release time) options, interspersed with Hold positions marked by red dots, while the upper control provides nine attack-time variations, three of which mimic the attack times of specific EMI units (identified by their serial numbers). Apparently, the unit numbered 60070B has attack and release times roughly three times shorter than other models, and is therefore better suited to channel compression duties. While 60050A is slightly faster than 61010B, both models are said to be better-suited to bus-compression applications.
A large gain-reduction meter dominates the centre of the panel, indicating up to 30dB of signal attenuation, while a rotary power on-off switch and associated jewelled indicator lamp sit to the right. The front-panel fuse holder of the original units has been re–purposed here as a rotary switch to engage a ‘SuperFuse’ mode. Not part of the original RS124s, this feature was introduced in Abbey Road’s own plug-in emulations of the processor. When activated, the power jewel light becomes significantly brighter and the character of the compressor changes quite dramatically. Chandler understandably refuse to reveal precisely what this mode involves, but it seems that the release time is radically shortened, probably to match the attack time. Regardless, the result is that the compressor takes on a far more aggressive, punchy, and slightly over-compressed character — something which I suspect is very close to the sound Joe Meek achieved with his own modified Altec 436s.
In Use
Peeping inside Chandler’s RS124 reveals quite a lot of hand-wiring between the power and audio circuit boards, and the front and rear panels’ controls and connectors. The rotary switches are neatly festooned with resistors, and while I’d have liked insulation boots on the mains terminals of the IEC socket, fuse holder, and front-panel switch (and was surprised to find that the mains transformer is bolted to a PCB rather than mounted directly on the metal chassis), the overall construction is to a high standard. The valves are mounted horizontally above a circuit board which hosts all the passive components, and two of the three valves have screening cans. The input and output audio transformers are bolted to the side of the chassis, and a metal screen divides the audio circuit board from the simple power-supply PCB.
Everything felt very solid and reliable in use, but when I first started using the compressor I found I had to adjust the balance trimmer fully clockwise to minimise the audibility of the ‘clicker signal’ and optimise the output balance. This process is very elegant and intuitive, and the review unit didn’t seem to drift much in use, so it’s not a procedure that is likely to need repeating very often. Strangely, though, even with no input signal, the gain reduction meter always showed around 3dB of gain reduction, which is a minor but slightly disappointing alignment issue.
This device is easy to set up and use instinctively — there aren’t many controls — but it doesn’t behave quite as ‘scientifically’ as most compressors. There’s a substantial loss of level through the compressor when the input control is at low settings, but as the input control is raised the output level builds too, along with the amount of gain reduction. There is, in effect, an automatic ‘make-up gain’ function, and at high input/compressor settings the output attenuator has to be adjusted to maintain sensible output levels. I found it slightly frustrating not to have a bypass mode to allow easy comparison of source and processed sounds, but neither the original 436B nor the RS124 had one — I suppose that’s a price of authenticity!
A very unusual aspect of this design is the inclusion of Hold positions between the six Recovery settings. These effectively provide an infinite release time, and thus maintain the current level of gain reduction. The compressor tends to sound most characterful when used with large amounts of gain reduction, so being able to switch to a hold position prevents the gain reduction being ‘dumped’ unceremoniously at the end of a track and producing a rapid and unwanted rise in the noise floor. The operational technique is simply to switch across from the chosen recovery time setting, as the track is playing, to an adjacent hold position just before the last notes.
The same facility can also be used in reverse at the start of a track to avoid the obvious ‘gain snatching’ that could result because of the RS124’s relatively slow attack time. It obviously takes a finite time for a large amount of gain reduction to be built up, potentially resulting in an overly loud initial transient. The operational technique here is simply to run through the opening few bars of the track to allow the appropriate level of gain reduction to be established, and then switch to a hold position: the track can be restarted with a nominal level of gain reduction already in place, and the recovery time switch moved to the appropriate setting after a few beats to apply the required dynamic-range control. (It sounds more complicated than it really is!)
When switching to the SuperFuse mode, the first impression is that the gain reduction is reduced by 5dB or so, compared with the normal mode, but this is really a side effect of changing to a much faster release time. Whatever the underlying technology, the effect certainly makes tracks sound more forward and dynamically exciting.
Verdict
It’s easy to see why the RS124 proved so popular with the Abbey Road engineers, and why it has acquired such a revered reputation over the years. It is an uncannily smooth-sounding but remarkably effective compressor, and those transformers and valves lend a subtle but welcome vintage musical character which is absent from most modern solid-state products. The Chandler RS124 is inevitably an expensive ‘luxury’ item for most of us, and especially so the version with the optional I/O level switches, but boy, what an enjoyable luxury it is!
Mods That Rocked
Recording studios in the late 1940s and early ’50s had to manufacture most of their own equipment, and at the EMI studios their own equipment designs were identified with the prefix ‘RS’, for Recording Studio, regardless of whether they were developed at Abbey Road or at EMI’s research department in Hayes. By the late ’50s, though, several independent audio-equipment manufacturers had been established, so studios started to buy in commercial equipment instead of making everything themselves. The problem for EMI was their peculiar legacy 200Ω interfacing arrangements, which made such third-party equipment unusable without modification. EMI were a very engineering-led organisation, and so all external equipment was thoroughly assessed before the studios were allowed to use it — and any perceived design faults were ‘corrected’!
This was the situation with Altec’s 436B vari-mu compressor, which had been seen and heard in Capitol Studios in America by some EMI recording engineers, who persuaded Abbey Road studios to buy several units in 1959. The Altec 436A had been designed for broadcast and PA applications, and although inexpensive and very effective, it was also very basic with no user controls at all. The B model, released in 1958, added a single user control to adjust the input attenuation (and thus the effective compression threshold), but when assessed by the EMI engineers it was deemed inadequate for studio use! So engineers Bill Livy, Len Page, and Mike Bachelor set about implementing extensive modifications for the company, with the first revised units entering studio service in 1960.
This Audio Precision test plot shows the input-output transfer characteristic with a rising level for various input control settings (0, 2, 4, 6, 8, 10). As you would expect, the compression knee occurs at lower input levels with higher input settings. All the slopes end up parallel, indicating that the ratio tends to settle at roughly 2.5:1.EMI weren’t alone in modifying the Altec 436: several studios around the world made similar changes, and apparently Joe Meek also modified various Altec 436 compressors to give them much faster attack and release times, leading to his trademark heavy pumping sound. Eventually Altec produced a third variant called the 436C, which included both threshold and release-time controls. They also made the closely related 438 model, which combined a valve mic preamp with the vari-mu compressor circuit in the same box.
EMI’s initial modifications to the 436B involved swapping out the American mains transformer for one suited to the UK mains supply voltage, and swapping the input level control for a rotary switch configured to provide a constant 200Ω impedance at all level settings. The switch also allowed accurately repeatable settings, of course, which was considered an important user feature. A second rotary switch was added to adjust the output level, while providing the necessary 200Ω output impedance to interface correctly with Abbey Road’s other equipment.
With the power and audio interfacing arrangements sorted out, the next major modification was to add a release-time control, which the EMI engineers labelled ‘Recovery’. Six options were provided, spanning 0.5 to 1.5 seconds, and this six-way control format persisted in other EMI compressor designs up until the mid-1970s. It has been suggested that the idea of providing six settings was influenced by the facilities on the highly regarded Fairchild 660 compressor.
Although there were just six recovery-time settings, a 12-way switch was used, with the intermediate positions providing a ‘Hold’ function — essentially an infinitely long release time. This was a necessary addition because the compressor was typically used with huge amounts of gain reduction, with as much as 20-30 dB being common. Switching to a Hold position at the end of a track or in silent sections prevented the gain reduction from being ‘dumped’, with an attendant rise in the noise floor.
This AP chart is a variation on the previous one, showing gain through the unit with a rising input level for the same range of input level settings. The transition into compression is more obvious at higher levels, as is the effect of increasing the input level control.Various other passive components were changed as well to improve the noise performance, apparently resulting in a slightly ‘harder’ sound character, and even the wiring and valve bases were changed to improve reliability. Finally, a new front panel was made to accommodate the additional controls, with the RS124 equipment identification marking. Not all traces of its origins were obliterated, though — the original Altec-labelled gain–reduction meter was retained!
The EMI engineers decided not to provide an adjustable attack time, but with different units being rebuilt at different times, they ended up with slightly different attack times, due to component tolerances and other minor build variations, and thus audibly different sound characters. It became quite common for studio engineers to request specific units (identified by their serial numbers) from the Abbey Road stores, where they were kept when not in use.
An inherent problem with the original circuit design was that the balance of the push-pull output stage could drift as the unit warmed up and aged. This resulted in an audible low-level modulation effect, which was particularly noticeable on LF signals. Unfortunately, realigning the output balance required taking the unit out of service to hook it up to the appropriate test equipment, which obviously caused unhelpful interruptions to studio sessions.
Eventually, EMI engineer Len Page came up with a brilliant modification which solved this problem in a very elegant way. He introduced a momentary switch to energise a neon bulb, producing a string of clicks which were fed in antiphase to the input valves. If the studio engineer suspected the output had become unbalanced, realignment was as simple as holding the switch down and adjusting a trimmer (accessed through a hole on the front panel) to minimise the audibility of the clicking. This clever modification was quick and easy to use, required no lab equipment, no change to control-panel settings, and only a very minimal interruption to the session.
In March 1963, a further, final modification was introduced to allow the linking of side-chain control voltages between two units for accurate operation in stereo, and this required the addition of an external link cable, and a rear-panel activation switch.
Pros
- Accurate recreation of the original EMI-modified design.
- Sublimely smooth compression character.
- Custom audio transformers and original valve types.
Cons
- No bypass mode.
- Minor construction details disappoint given the high cost.
Summary
A tastefully but usefully enhanced recreation of the classic EMI/Abbey Road RS124 vari-mu compressor.
information
Nova Distribution +44 (0)20 3589 2530.