By | February 3, 2007
OK, I know I promised that Iâ€™d have an SEO guest this week, but it is seriously tough getting him pinned down to a definite time and weâ€™re like 3 hours apart in time zones, so hopefully I can get him nailed down in the next couple of weeks. So instead of talking about SEO, I figure weâ€™ll talk a little bit more about audio quality and specifically about compression.
Iâ€™ve told you a little bit about compression in other episodes, but this time Iâ€™m gonna dedicate the whole podcast to it because honestly, itâ€™s a little more complicated than just flipping a switch, and this is something that people tend to get confused easy. Compressors are â€œdynamic range processorsâ€ theyâ€™re also called â€œvariable gain amplifiersâ€ or VGAâ€™s. They affect the dynamic range of sound by varying the â€œgainâ€ or volume of that sound.
If you were recording a concert, there might be a potential for the loudness of that concert to peak around 120dB, and that is very loud. Youâ€™re recording equipment though, letâ€™s say it has a range of 75 or 90dB, and if you go over that, youâ€™re gonna get distortion. What weâ€™ll have to do is be able to turn down the dynamic range so that we get a recording that doesnâ€™t sound distorted. You could do it by manually turning the volume down on the mics but the problem with that is, youâ€™re too slow and by the time you lower the volume, youâ€™ve already over-modulated and recorded, junk. So first, the compressor does the job for us automatically.
In the music world compressors are used on just about anything that has a fluctuating volume, things like guitars, drums, vocals. A simple explanation is that a compressor will make a big change in volume into a little change in volume, so it kinda balances things out.
Most compressors have 4 main controls and they usually have two secondary controls. The main controls are Input Level, Output Level, Threshold, and Ratio. The secondary controls are important for things like voice work or even if youâ€™ve got a noisy guitar amp, and these controls are Attack Time and Release Time.
The Input level is simply how much gain should be applied to the input signal. For example, if you set it to â€˜0â€™, that means youâ€™re getting the exact same volume level that you are sending it. If you set it to -3dB, then youâ€™re reducing the gain of the input signal by 3dB. If you set it to +3dB, then youâ€™re turning the input signal up by 3dB.
The Output level is on the other end of the compressor. The input level is the first adjustment, and the output level is the last adjustment. Once everything has been compressed, youâ€™re setting how much amplification, or reduction, should happen to the signal. So this is another place for you to control the volume into the next piece of equipment which could be a mixer, a PA, or a recording device.
Usually, we donâ€™t want the compressor running all the time. This is where the Threshold comes in. Think of this like the thermostat for your air conditioner. You set it at a certain level and if it gets hot enough, the air conditioner kicks on. Once the temperature gets cool enough, the air conditioner kicks off. Well, the threshold behaves pretty much the same way. You set a volume level where you want the compressor to start working. Any signals that are below that level, wonâ€™t be processed. Once a signal reaches that level, the compressor starts to do its work, based on the next parameter, the Ratio.
The Ratio is probably the trickest part of compression. This is the relationship between the uncompressed signal, and the compressed signal, and itâ€™s described as a ratio. The reason itâ€™s a ratio is because itâ€™s dynamic. Thereâ€™s not just a â€˜setâ€™ level of amplification that occurs. If the ratio was 4:1, then an input signal of 20dB would have an output signal of 5dB. You can figure that out by just taking the input level and divide it by the ratio, so 20 divided by 4 is 5. If your ratio was less, like 2:1, then your 20dB signal would turn into a 10dB signal. A real high ratio, meaning a lot of compression is going on, like 10:1, would take your 20dB signal and make it a 2dB signal. So ratios from 2 to 6 are considered â€˜gentleâ€™ meaning theyâ€™re more subtle, but over 6 would be considered â€˜hardâ€™ and is usually pretty noticeable. For studio and podcasting work, you want to stick with the more gentle side of compression. Start with a 2:1 ratio and see how that sounds.
Another parameter we can set is the Attack Time. This is a number, usually expressed in milliseconds, that tells us how quick the compressor will start to compress a signal that has gone over the threshold. So if itâ€™s set to 200ms, a signal will go above the threshold for 200ms before it gets compressed. The reason for this is to help things sound a bit more natural. You donâ€™t want everything sounding like itâ€™s hitting a brick wall, acoustically. So the attack time is used to flow the compression in.
Letâ€™s say that you start talking and then you stop. The compressorâ€™s gonna stop too because your signal will go below the threshold. If you stop just for a split second and then start talking again, thatâ€™s gonna sound a bit strange because your volume will jump around suddenly a bit. So to compliment the attack time, we have the release time. This is also set in milliseconds and it determines the amount of time for the compressor to turn off after the signal drops below the threshold.
There are a couple of other features that I should mention just in case youâ€™re looking to buy a new compressor, first, you can get processors that include other effects AND a compressor, so thatâ€™s one option. You can also get dedicated compressors. As a general rule, youâ€™ll get better quality with a dedicated compressor. Second, you can get stereo or mono compressors. That just tells you how many independent signals can be run through the compressor. A stereo compressor lets you run two signals, independently, so you could run two mics through it. A mono compressor just lets you run one signal through it.
A feature that is built in to a lot of compressors is a De-esser. This is a selective compressor, it compresses signals in a given frequency, around 5-8k. It does that to soften â€˜Sâ€™ sounds so theyâ€™re not as harsh. Itâ€™s usually a subtle compression, so itâ€™s not something that most people will notice a whole lot.
Aside from that, youâ€™ve two main categories of compressors, tube and solid-state. Without going into engineering terms, Iâ€™ll just say that tube compressors generally have a much warmer sound and a lot of times theyâ€™re used just because of that, they kinda add a presence to the sound. Solid-state isnâ€™t all that bad, it kinda depends on personal tastes.
With compressors, you WILL get what you pay for. The more expensive models can do a lot of compressing and leave the signal still sounding natural, while the cheaper models will sound compressed the more you compress. Prices vary drastically from a couple of hundred dollars to thousands, so see if you can go to a music store and try one out. Youâ€™ll wanna hear how it sounds with strictly voice going through it.
Iâ€™m gonna go ahead and give you three different models to look at in three different price ranges, just so you can get an idea of whatâ€™s out there. For podcasting, as always, donâ€™t spend a ton of money, but get the best you can afford. Compression wonâ€™t make or break your podcast, but it will help your sound quite a bit.
These are all from Musician’s Friend, with which I have no affiliation, but they generally have a good price.