Dynamic range is essentially the distance between the loudest part of the song, or audio file, and the quietest part of the song or audio file, so say, for today, let’s talk about the dynamic range of the human voice rather than the dynamic range of a song, just for a change, to try and understand dynamic range in a slightly different example.
In the dynamic range of the human voice, the loudest thing that the human voice is capable of is a shout, and the quietest thing the human voice is capable of is a whisper. The dynamic range represents the difference between the loudest thing achievable by the source, and the quietest thing achievable by the source.
If I wanted to compress the human voice, the first thing that I would do would be decide where to place my threshold. The threshold determines the point at which anything that passes over the threshold becomes affected by compression. To put that another way, the threshold will … we’ll set the threshold at average speaking voice, in this example. Anything that passes over average speaking voice, for example, a shout, that would pass over the threshold, by a whisper wouldn’t pass over the threshold.
The next thing I would do would be, when using a compressor to compress the dynamic range of the human voice.
The next thing I would do would be apply a ratio. What this means is that anything that passes over the threshold point, which is average speaking voice, will be affected by this ratio. For example, a shout would pass over the threshold point, and then be affected by these ratios here, that I’ve set. Would it be affected by a 2 to 1 ratio, or a 4 to 1 ratio here?
A 2 to 1 ratio will half the signal over the threshold point, and a 4 to 1 ratio will quarter the signal over the threshold point. Your shout will be reduced by a half … your shout will be halved or quartered, depending on which ratio has been affected by … as a result, you can see here this red line which shows our compressed signal. This reduces the dynamic range of the overall conversation, the overall human voice, and here, the slightly blue line, the 2 to 1 ratio that has been reduced as well … the dynamic range has been reduced as well by a slightly less amount.
As a result, the entire conversation is slightly quieter than the uncompressed signal, so what you would do to compensate this would be apply a makeup gain, so you turn it up, the compressed signals to match the original uncompressed signal, so you match the lost volume.
What this really means, for example, let’s focus on this 4 to 1 ratio for a moment, this red line. The distance between the shout and the whispers has been reduced. What this will, mean in real world terms, is some of the shouting has been slightly toned down to welcome some of the whispers, all a bit more forward in your focal point. Your conversation will feel a little bit more exciting. Your whispers will come forward a bit in the conversation, so the conversation will be a bit more upfront and in your face, overall, at the expense of dynamic range, and this is essentially what we wish to achieve with the compressor. You sacrifice a bit of dynamic range to achieve a more powerful or more upfront sound.
I’d like you to consider this, however, for a moment. What if we were to apply extreme compression settings to the human voice? In this example, I’ve set the threshold point much lower, so the threshold is down set near the whispers, near the quiet parts of the song, and the ratio is much more extreme. It’s 8 to 1. That means that anything that is larger than a whisper will be affected by compression, of this ration 8 to 1, which is quite a high compression ratio. Anything louder than a whisper will be compressed by quite a lot as a result. We can see here this slightly darker blue line.
The dynamic range of this particular conversation, this theoretical conversation that we’re talking about, has been greatly reduced, so, when we turn up, withy makeup gain, to make up for lost volume, this on the left-hand side here, is our original uncompressed signal, and this on the right, this much shorter line is our signal that has been compressed quite drastically with the 8 to 1 ratio, and the low threshold point. What this means is, the distance between the shouting and the whispering is much less, so, we’ve sacrificed some of the severity … the sincerity of the shouting to encourage the whispers to come forward.
Consider for a moment what this would mean emotionally, if you were having a conversation that has a very, very reduced dynamic range, like this, and the whispers were almost as loud as the shouts, consider what that would mean in the real world. The guy who’s screaming and shouting would have almost as much impact as the person who’s whispering, so you’re going to end up with a much less emotive and expressive and articulate conversation, if this was the case, because your expression has been sacrificed in order to achieve loudness overall.
In the real world, if your whispers are almost as your shouts, there would be no subtlety, no secret, in your conversation, and your screams will have much less impact than they would before, if it was left with a bit more dynamic range intact, with a bit more sonic integrity than squashing it to this degree.
It’s exactly the same for music. If you over compress something, then you begin to lose some of the natural expression of the piece of music. It’s not going to sound powerful. Your song … you’re sacrificing some of your louder, more important sounds for the benefit of some of the quieter sounds. On a piece of music, you’re not going to end up with as an expressive or emotionally enjoyable piece of music. You’re going to end up with a much more fatiguing sound, if everything’s fighting for volume and fighting for attention, because it’s been pushed so hard. That’s something to consider when applying compression in mastering. The overall desired effect is to bring things forward a little in the mix, and help the quieter sounds become a little bit more involved, at the expense of some of the unnecessary loudness of the peaks, but you don’t want to overdo it.
You don’t want to compress to a point where you’re losing some of your expression and your emotional joy of the song. You want the song to … you want the quiet parts to complement the loud parts, and not to compete, essentially. That’s something to keep in mind, when applying compression at the mastering stage. If in doubt, err on the side of caution, I would recommend.