One of the main difficulties in setting up a compressor is choosing the time constants Attack, Decay, Sustain, and Release (in short: ADSR). The most important time constants are Attack and Release in the context of gain control, Decay and Sustain are generally fixed and in any event, they tend to play a minor role.The inappropriate choice of time constants is mainly responsible for the familiar pumping or breathing effects, though this effects may be deliberately sought for creative applications. Unless the attack time is set to be very short, fast transient attacks may get through the system, because the compressor response is too slow to get them under control. Threshold and Ratio are easier to set up, but with high compression ratios high frequencies tend to become overcompressed, thus dull sounding.

Usually compressors are set up with reference to the loudest peakl, but then the parameters are optimized for this particular moment only – which means for nearly the complete matierial the time constants are not set optimal. Taking into consideration that each musical instrument has varying attack and release times, depending not only on the character of the instrument, but also on the way it is played, you can see that it is almost impossible to choose a set of fixed parameters that will be correct for an entire piece of music.

In contrast, the DynaMaxx responds directly to changes in dynamics. If, for example, a snare drum peak occurs, DynaMaxx rapidly changes to a very short Attack time so that the snare hit keeps its original transient characteristics rather than sounding "softened". Following signals are processed with longer Attack times to minimize distortion.

If the DynaMaxx is operated in its normal compression mode, breathing and pumping effects jut don't happen. If pumping effects are needed for creative reasons, you have the option either to turn Compress control to maximum and/or switch the unit into Effect Compression mode.

The DynaMaxx re-adjusts the Attack time by continuously comparing the output of the Attack time control stage with the input of Attack time control stage. If the comparison reveals that a fast transient has slipped through the first Attack time control stage, a second, much faster Attack time control stage is immediately activated to catch the peak. DynaMaxx makes use of 12dB/octave filtering instead of the 6dB/octave filtering used in standard compressor side-chains to increase the precision and speed of transient detection and processing.

DYNAMIC INTELLIGENCE: ATTACK TIME AUTOMATION
To understand the DynaMaxx intelligent features, it´s necessary to see what happens when a conventional compressor is used with a fixed Attack time setting. For example, the sound of low bass guitar note can either come in smoothly (especially with fretless basses), or with a very fast transient attack when (slap). If the Attack time is set to minimum (very short), the compressor is able to catch the peak of the transient attack but any closely following notes will suffer increased transient distortion because the control voltage within the compressor rises further as successive notes are processed. This behaviour can only be avoided by setting a longer attack time, but now peaks slip through again. Including a separate Peak-Limiter to catch those fast transients requires a second VCA stage, wich does not necessarily improve signal quality, but even with designs that use the same VCA for both compression and limiting you end up with more controls than necessary.

The DynaMaxx detects very fast transients automatically and re-adjusts its Attack time accordingly. This ensures that no Peaks slip through, but the speed of re-adjustment is so fast that all signals stay within the soft-knee curve for a more natural sound. Attack times can be reduced in the instant of a percussive hit or bass guitar slap. As soon as the peak passed, the attack returns to a longer time constan and both pumping and distortion are avoided.

DYNAMIC INTELLIGENCE: RELEASE TIME AUTOMATION
Again, it is beneficial to look at what happens with conventional compressor designs when the release time is set too short: In this case the compressor will restore normal gain conditions as soon as the peak has passed, and this rapid increase sounds like breathing. Breathing is especially annoying during a part with soft strings or low level layered sounds – whenever a peak comes along, the strings duck down with it, only to pump up again when the peak is over. Sonically the sound image swims and pumps while the subjectively perceived loudness remains at a low level.

The DynaMaxx monitors the average music level. If a loud transient sound occurs, a very short Release time is set and the signal level is reduced to the calculated average music level (not to the threshold; reducing the level to the threshold would again cause pumping effects because low-level signals are lifted in gain). The release time always depends on the difference between the actual peak and calculated average signal levels. If a large difference is detected, the DynaMaxx sets a faster Release time, if only a small level difference is detected, the DynaMaxx will release more slowly.

THRESHOLD AND RATIO
Threshold and Ratio are both set by the Compress control; using a low setting for the Compress control causes only the peak levels to be compressed, because the Threshold is relatively high. For more compression, turning the control clockwise has the effect of lowering the Threshold to include more low-level signals in the processing. As the Threshold is lowered, the Ratio is simultaneously increased, and the fully clockwise position is equivalent to a Ratio of around 3:1. However, the Threshold and Ratio values are not static, but rather vary depending on the attack and level characteristics of the source signal. Peak levels are automatically compressed with a higher Ratio to maintain control over maximum signal levels, but still using an unobstrusive soft-knee compression characteristic.

Double-VCA-Drive
Double-VCA-Drive technology utilizes two VCAs per audio channel, where one VCA handles positive current and the other negative. This way, the control voltage can effectively be halved, but still provides the same gain change as a standard single VCA solution. The benefit of this configuration is that the transistors within the VCAs don´t run into saturation problems, which in turn avoids offset noise, generally audible as clicks and pops. Compared with the acclaimed DBX 2150 VCA, the new THAT 2181 VCA is less sensitive to offset noise, and the Double Drive configuration produces extremely good audio performance, both subjective and measured.

The audio signal is split in two paths, one of which is phase inverted, and after having passed through the VCAs, the two signals are recombined in a differential amplifier. This topography causes a cancellation of negative side effects, mainly distortion and sound coloration, by the mechanism of common mode rejection (CMRR > 50dB) within the differential amp. At the same time the original audio signal is increased in level by 6dB.

One cannot expect the noise floor of the VCAs to be reduced by 50dB, because noise is a non-correlated signal, but taking into account the 6dB signal boost, the noise floor is improved by 3dB. However, the real benefit is the reduction in THD. Although the THD spectrums of positive and negative half cycles are not identical, a significant degree of cancellation nevertheless takes place, and the residual THD is right at the limits of available measuring equipment. The same applies to intermodulation distortion and control voltage crosstalk, both of which are cancelled in the differential amplifier stage.

DE-COMPRESSOR/APPLICATION EXAMPLES

1. To create a stereo-loop with a dynamic difference, try using the DynaMaxx with both channels set separately rather than linked (Stereo Couple off). Channel one, for example, could be set to pump and breath heavily by using Effect Compression and setting the Compression control to 7 or 8. To increase the effect, turn up the Compression control to max or additionally switch in the Soft Limiter.Channel two could be used in De-Compression mode with the Compress control also set to 7 or 8.
   For more dramatic action you can additionally activate the Effect Compression and the Soft Limiter. Compensate for the level changes with the Gain control and the result is amazing! You can either mix this underneath the original loop, use it as an effect or ...
2. If you are using drum sounds from samplers or drum machines, try processing them with De-Compression mode. In addition to adding new life and dynamics to the sounds, individual beats get acoustically shorter, which helps them cut through a busy mix, even at lower levels.