How does the dynamics circuit in the OctoPre operate?
The Opto-driven dynamic circuit is not only a limiter and compressor but also is split into two bands to improve sound quality. The circuit works as follows:
There are two independent side-chain control circuits for each channel of dynamics, one for limiting and one for compression. The two circuits are almost identical, except that the limiter side-chain has a fixed threshold (just below 0dBFS, to prevent A/D overload) and the compressor side-chain has a variable threshold set by the front panel dynamics knob. The limiter and compressor side-chains each generate two separate control signals, one fast and one slow, that are then fed to the gain change elements in the audio path.
The gain control elements consist of a pair of Optos, which follow a crossover network that splits the audio into two bands, "high" and "low". The fast side-chain drives the high-frequency opto, whilst the slow side-chain drives the low-frequency opto, the benefit being that less distortion is generated when the limiter reacts to very fast transients. By splitting the response, only the high-frequency part of the waveform is drastically affected, reducing intermodulation distortion, and therefore non-related harmonics, thus providing a more musical effect.
Both the side-chains operate as feedback systems. Thus, the amount of compression and limiting is calculated using the level after the split opto stage. The advantage of this feedback system is that the limiter senses the level after compression has been applied and therefore the limiter only works when it absolutely has to. When lots of compression is being applied, the level is unlikely to cause the limiter to kick in, so therefore the dynamics should be thought of as two discrete dynamics sections as if there were an insert with a compressor first followed by a limiter.
What's the best way to use the dynamics processor?
The thing to appreciate here is that the dynamics can be made to work differently for different instruments, and the response of the limiter and compressor varies for different positions of the pot. So, what is a good setting for one instrument or signal is not necessarily good for another.
The unit works as if one had a compressor followed by a limiter, so one should set up the dynamics knob in a position that is sympathetic to the audio passing through it. The dynamics will exhibit a sharp limiting effect when the knob is positioned just after the off position, (fully anti-clockwise but not off), and that is the intention. It is the last chance to save an over before the A/D. Therefore, the response needs to be fast, representing an audio brick wall. This position works well with transient information, but a different type of control may be desired for vocals or program material because these signals are more complex. For more harmonic and frequency-broad signals, the dynamics knob should be rotated further clockwise, where there is a clearer balance between limiting and compression. The processor works in the same way as one would expect to use external equipment. Would you record a vocal with just a limiter? No, you would use a gentle compression ratio to ride the signal. The limiter is only there for extreme protection, so one should set the knob most of the way around its travel.
The yellow 'comp' LED' is a good guide for getting a starting position of the dynamics knob - for tracking compression try to aim so that the LED is not mostly on or mostly off but somewhere in between. The LED gets brighter the more compression is applied so try to aim for mid-brightness for a starting point and then listen to the signal and vary it from there.
The dynamics control knob also contains an additional make-up control that balances out the level change caused by the compression. This auto make-up gain is handy because, for most applications, the output level to the A/D stays constant whilst further compression is applied. So, any input levels that have already been set should not need further adjustment during dynamics set-up.