Ask the Doctors: The Latest UA Emulations, Part 1
by Dave Berners

Q: Have you worked on any interesting emulations lately?

A: We recently looked at the LA-3A and the Neve 88RS, both of which have several interesting features. Let’s start with the LA-3A.

The LA-3A is an optical compressor that uses the same electroluminescent (EL) panel and photoresistor found in the LA-2A. For both compressors, the EL panel glows according to the strength of the output signal, which causes a change in conductance in the photoresistor. The photoresistor is connected as a shunt in the signal path to reduce gain.

The obvious difference between these two compressors is that the LA-3A is solid state and the LA-2A uses tubes. The makeup gain amp in the LA-2A is a very clean tube amplifier, with a flat response and minimal distortion, so there is not an excessive amount of “tube” sound coming from the makeup amp. Still, the audio path responds a bit differently than the LA-3A, giving each of these two units a distinctive behavior.

The biggest measurable difference between the two units comes in the sidechain amplifier. The EL panel needs relatively high voltage in order to turn on. This voltage can be attained directly in the LA-2A with a tube amplifier. In the LA-3A, the solid-state circuitry does not run at high-enough voltage to drive the EL panel directly, so there is an autoformer (a transformer with shared primary and secondary windings) which drives the panel. As long as everything operates in a linear fashion, the differences between these two technologies will not be huge. For both of these compressors, when a large transient arrives in the input signal, the sidechain amplifier will saturate during the attack period. In the case of the LA-2A, the saturation happens in the tubes. For the LA-3A, saturation could conceivably happen either in the solid-state amplifier driving the autoformer, or in the autoformer itself. Which saturation mechanism dominates would depend on the frequency of the driving signal.

Owners of vintage LA-2As have probably noticed that, with age, the LA-2A requires the Peak Reduction knob to be set to a higher level than when the unit was new

Saturation in the sidechain of a feedback compressor can result in a slew-rate limit on the attack. This means that the gain can only be reduced at a certain rate, no matter how big the signal gets. Saturation also limits the maximum amount of gain reduction that can be achieved with the compressor. For a feed-forward compressor, saturation in the sidechain would limit the amount of gain reduction possible, but would not limit the slew-rate of the attack in the same way.1

Because of differences in the types of saturation in the sidechains, the LA-2A has a lower limit on the slew-rate for its attack than the LA-3A. For small transients, all things being equal, these compressors will have approximately the same attack behavior. But for large transients, the LA-3A can have a much faster attack.

The attack slew rate is also affected by the efficiency of the EL panel. The typical EL panel loses efficiency as a result of age and use, so that an old EL-op can cause a lower limit on the attack slew rate for both the LA-2A and LA-3A. Owners of vintage LA-2As have probably noticed that, with age, the LA-2A requires the Peak Reduction knob to be set to a higher level than when the unit was new. This is because of the reduced efficiency of the EL panel. Moving the knob to a higher level can restore the average gain reduction for a given input signal level, but an old EL-op will result in a lower slew rate for the attack, and can also affect the compression ratio. These differences cannot be compensated for with the peak reduction control.

The LA-3A, in addition to having a different and faster attack, has a different frequency dependence on the amount of compression. The difference is mostly due to the interaction between the EL panel and whatever is driving it (tube versus autoformer). The tube stage has a different frequency-dependent output impedance than the autoformer-coupled, solid-state amplifier.

The release curves for the LA-2A and LA-3A are similar. Both compressors share the same photoresistor, and it is used with the same topology for both units. Slight differences in component values change the compression curves and release shapes a tiny bit between the two units, but, for the most part, the release curve for an LA-2A or LA-3A will depend more on the particular photoresistor being used than it will on the compressor driving the photoresistor. Release times and the program-dependent release properties are therefore similar for the LA-2A and LA-3A.

All in all, the LA-2A and LA-3A are similar enough to be considered brothers, but the differences in attack behavior and audio path are big enough to make these compressors obviously different from each other.

Next month: emulation of the Neve 88RS.

1 In “compress” mode, the LA-2A and LA-3A are purely feedback compressors; while in “limit” mode, there are both feed-forward and feedback components.