Compressors

Love the squeezebox.
One of the best descriptions I've heard about audio compressors in general is that a good, transparent compressor is like a little guy riding the fader up and down faster than a human ever could. It's such a handy studio tool, but not every type of compressor works the 'mojo' on every type of source.
A good starting place on the subject is to understand the different types of compressor designs.

-Opto gain reduction is accomplished by varying the intensity of a light into a light sensitive resistor (later LA4s) or electro luminiscent panel (older La-2a's T4B module) causing a relatively slow attack and release which is also typically fixed. "Syrupy" and "smooth" are both good descriptions of opto compression. Vocals and bass benefit the most. Teletronix La-2a, Urei La-3a are examples of opto compressors.

-Variable-mu - "Mu" is gain and the compression works by using remote cut-off or re-biasing of a vacuum tube via a sidechain control voltage. As some of the very first compressors ever designed and produced, they continuously increase their ratio the harder they’re pushed. Though faster with attack and release times than opto designs, vari-mu are still slower than VCA or FET compression.
RCA BA-6A, Altec 436b, Fairchild 660, CBS RIIZ, Gates Sta-Level, GE Uni-Level and the Manley "Vari-Mu" (go figure) are examples of vari-mu compressors.

-FET, by definition, is a Field Effect Transistor. Compression is achieved by a FET, which is used as a variable resistor. the FET acts like a resistor whose resistance is controlled by the voltage applied to its gate. The higher the voltage applied to the gate, the smaller the drain-source resistance will be. The greater the voltage applied to the gate of the FET, the less resistance, hence large signals cause the FET to reduce the gain. Urei 1176 is the most obvious of FET compressors, but other FET compressors include the API 525.

-VCA is a Voltage-Controlled Amplifier. The gain is determined by the voltage supplied to a control input. Virtually every compressor technically uses a VCA, whether it is based on FETs, variable-mu tubes, or optical elements. However, in the parlance of audio, the term “VCA compressor” usually refers to a compressor using an integrated-circuit VCA for gain control. Grab and pump are two things done well by VCA compression. DBX 165, 160a

-Diode bridge The compression of a diode bridge circuit is accomplished thus: The diodes in the bridge are forward-biased, and the incremental resistance of the diodes determines the amount of gain reduction. The resistance of the diodes is in turn dependent on the amount of bias current going through the diodes. Bias current is determined by the output of the sidechain control voltage.
At large amplitudes, the audio signal itself can cause appreciable changes in the amount of current flowing through the diodes in the bridge. The audio signal moves the diodes away from their bias points, the incremental gain of the circuit changes. Because the circuit’s gain depends both on the bias currents (determined by the sidechain control voltage) and the audio signal itself.
There is an inherent amount of ever increasing THD in the diode bridge compression circuit which can "soften" audio very nicely.
Neve 2254, Neve 33609 and Vintagedesigns CL1 are all diode bridge compressors.



One defining design element of all compressors is a feed-forward vs. feedback design.
The actual compression circuit, whether a VCA, Vari-mu tube, etc., needs a control voltage in order to respond appropriately to the signal being passed. As is logical to the names, a feed-forward circuit will draw the control voltage from the incoming signal resulting in a greater accuracy whereas the feedback design draws its control circuit from the output signal, thereby letting the circuit breath before applying the control voltage to the compression circuit. Obviously, each of these types of compressions will have their own impact upon attack and release curves depending upon how the control voltage is derived.