Well, there are simple compressor/limiters that can make effective compression and limiting rather simple. Such as a dbx 166/266/160, Alesis 3630, which are all relatively simple to use. Then there are the compressor/limiters that offer both a choice of RMS & Peak detectors and also variable attack and release times. Along with the prerequisite ratios ranging from 1.1:1 to infinity: 1 and everything in between. Some also offer a short timed hold before release. Some are digital and because of that can allow for lookahead. That means that the detector can operate again reduction element at or before the Event as opposed to after the event which normal analog dynamic range controllers do. Most analog dynamic range controllers cannot change again reduction until after it is sensed.
When you get a dynamic range controller that has variable attack & release times, it allows for changes in density and apparent loudness. Speeding up the release time will create more of a sense of urgency and apparent loudness level. Increasing the attack time will crush any transients. Slowing the release time will allow for a more natural sounding bit of gain reduction control by allowing transient peaks through. Slowing down your release times will make a dynamic range controller less aggressive and obvious. Utilizing low ratios such as 1.5:1 will allow for smoother almost imperceptible gain reduction control. Utilizing higher ratios such as 8:1, 12:1, 20:1 means that for a broader range of decibel input there will be a corresponding smaller range of decibel output. Turning your release times up too fast can create excessive apparent loudness and distortion artifacting that can be used as an effect.
Then there are those different types of dynamic range controllers. Some utilize electronic circuitry to detect the incoming circuitry in which to work upon. There are also the optical types that utilize electroluminescent panels, LEDs and even light bulbs. These optical types and their differences generally dictate the ballistics of the control of the incoming dynamics. Devices such as the $4500 Universal Audio LA 2 don't have any attack or release controls. They are designed to sound smooth and they utilize an electroluminescent panel along with a light-sensitive resistor element to smoothly control all dynamics. The transistorized version that came after that, the LA 3 utilizes the same device but the unit utilizes transistors instead of tubes. That too changes its perceived acoustic signature. Later units such as the LA 4 & 5 utilized LED light sources which have yet a different ballistic nature again in combination with light-sensitive resisters. Whereas DBX units and other similar ones utilize no optical devices. Instead, those devices utilize sophisticated electronic sensors along with VCA (voltage controlled amplifiers) which can vary the amplifiers gain structure utilizing a control voltage derived from the electronic sensors.
So how are you to know which one is right for your purposes? No one can give you that answer. That's why people train as audio engineers to be able to understand the different applications of said dynamic range control devices. Some of these devices even work in reverse and can actually increase dynamic range rather than decrease dynamic range. Some of these devices are known as upward expanders. Some are known as downward expanders which are frequently also commonly referred to as noise gates which is a slightly different type of downward expander. At one point, DBX made a unit called the 163 " one knob squeezer ". This device at a single control. As you increased the control, it increased both the sensitivity, the ratios and associated attack and release times with a single knob. It was generally goof proof for the entry level user. Nothing much to understand except for how much you wanted. And they were great little devices for most general-purpose applications in the studio. They really sucked when you tried to utilize them with a PA system. That's because they didn't have all of the other important controls one finds on more sophisticated devices. To complicate things even further, many dynamic range control devices offer what is known as a " side chain " input. What that allows you to do is to insert an equalizer. The equalizer can then be adjusted so as to affect what the gain reduction control sensor sees and hears. This would allow you to control what the dynamic range control element would work upon. So when you had a signal with excessive low frequency interference coming from bass guitars and bass drums, you could roll off the low frequency sensitivity to the sensor. This would mean that the dynamic range control would only work upon higher frequencies while not changing the actual frequency response of the signal passed through the dynamic range controller. Adding a huge boost on the equalizer in the upper midrange between 3-6 kHz is used to control semblance on vocals. It made the dynamic range control element more sensitive to those sibilant frequencies and so, it would work harder on just those frequencies. In the land of FM broadcasting, TV broadcasting, vinyl record cutting, high frequencies would have a tendency to distort. So in the side chain, you would boost the high frequencies so that the dynamic range control element would work harder to tame excessive high frequency energy while again, not changing the frequency response of the input signal passing through. And that's because FM, TV, record cutting and slower speed analog tape such as cassettes utilized a high frequency preemphasis boost with a corresponding high frequency de-emphasis roll off to create a flat response with less high frequency noise. The high frequency preemphasis in the recording circuitry would have a tendency to overload the high frequencies causing a splattering of distortion. Not only couldn't you have that, the FCC wouldn't allow that. So devices to control that were created for that specific purpose. I modified some of my own limiters to have a high frequency boost in the sensor. This way, I can add all sorts of extended high frequency equalization to the signal to give female singers a nice S E X Y and breathy quality without cutting through your eardrums like a hot knife through butter. It would also allow me to add more air to accentuate certain qualities of other instrumentation without fear of high-frequency overload. OMG! Is this TMI? You bet it is. But you asked. And those are your answers to your questions.
So when you just need some simple dynamic range control, you just need a simple dynamic range controller. When you want all of that other gobbledygook, you're going to be spending $1500-$5000 on a device, of which, you may never quite understand how to correctly utilize. Thankfully, most of what I've described can be accomplished in even some of the simplest and least expensive software. In that way, you wouldn't have to pay one red cent to get what you want. It's all doable in software and we all do it in software because of its convenience and inexpensive nature.
Quite personally, I find that outboard hardware dynamic range controllers during the recording process can actually be quite advantageous to the sound I want to get. Some folks will vehemently tell you to never do that. To record everything without affecting anything and to do it all after the fact in the software. I don't agree with that. I've come from the analog era way back to the earliest of the 1970s and actually earlier. We utilized the dynamic range controllers while recording to analog tape so as to keep the signal much higher from the noise floor. I found no reason to change my technique even when I graduated to digital recording back in 1983. There are still advantages to doing it that way even in the land of 24/32/64-bit recording at 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz, 192 kHz & single bit DSD sample that nearly 6 MHz. And so that comes down to engineering technique which takes time to gain that experience. Personally, I think a lot of folks going to universities and coming out with PhD's in the recording arts and sciences don't know squat. They only know what they been taught by other ill-informed people, that teach it but have never really done it.
So now you know what to do. Right. Check. You are probably completely confused by now? Maybe not? Go for it.
Mx. Remy Ann David