Getting the Right Volume Level On Your Audio Files

Getting the right volume level on your finished recordings is important because you want your customer to hear your recordings at about the same level as other commercially produced recordings, and you don't want him or her to be constantly reaching for the volume knob when playing your product.

While not difficult once you get the hang of it, setting the proper volume level is really more of an art than a science. In this article we'll see why.

Volume

Years ago, in the beginnings of audio recording, engineers came up with a way of measuring perceived loudness. They called it the VU standing for Volume Units. This meter ignored short peaks and mechanically averaged the signal so that by looking at the meter, one could get a very good idea of the relative loudness of an audio signal. These meters were built to very specific standards and a VU meter on a tape recorder in Hollywood corresponded to a VU meter at a radio station in New York.

As consumer tape recorders and later recorders were introduced, they all had “VU” meters. These meters were not true VU meters, and the proper setting of levels varied according to the whims of the particular manufacturer.

To further complicate matters, when digital recording was introduced it became necessary to introduce “peak meters and measure the peaks that the VU meters ignored for so many years. Why?  In the days of analog tape recording, tape had the nice characteristic of saturating slowly at high levels and acted almost as a limiter for the quick peaks in an audio signal and compressed them without a lot of noticeable distortion.

But digital recording is an all or nothing situation. It records perfectly up to its peak limit, but over that level the encoder runs out of binary digits and creates severe distortion. Now, some recorders and software handle these peaks better than others, but you definitely don’t want to exceed the peak recording level. That’s why digital recorders and software have peak level meters. Unlike analog recorders, it is never acceptable to go into the red when recording. Fortunately with the high signal-to-noise ratio, it is acceptable to record 9 to 12 db below peak level to leave yourself some headroom.

The downside of peak level meters is that they don’t give a very good idea of how loud one recording is compared to another. For a digital voice recording to sound as loud as a commercially produced program it must be limited and normalized.

When you are done editing, you should normalize the audio. To normalize means to increase the audio level of each track so that the peak level of each track is at the maximum level allowable on the CD.

The goal is to make your CD, video or audio file sound about the same loudness as a commercially produced product.

The difficulty here is that if your recording has a few loud peaks and a low average level, your recording is going to sound low even if normalized. The trick is to keep the average level high and minimize the peaks in the recording.

Open your Normalize function and set the level to approximately - .5db. Apply the function to your file.

One approach to keeping the overall volume level high is to use a volume compressor or limiter. (Please note that Volume Compression is entirely different from data compression such as is used to make an MP3 file) Compressors can make a signal sound louder and more intelligible without raising the peak level of the signal. The compression ratio control adjusts the factor by which the dynamic range will be reduced, and the threshold control sets the level of input signal at which compression will begin.

A limiter is a special instance of a volume compressor, in which the compression ratio is set at 4:1 or greater. The threshold is set so that the signal passes normally up to the threshold point, but after that, further increases in the input signal level produce almost no level increase at the output. Most compressors can be set to act as limiters. Limiters are used to avoid clipping of the signal by preventing overload of the digital encoders or other equipment.

You can use a hardware device that goes between the microphone and the computer or you can use your software package to perform this function after the track is recorded. There are advantages to each system.

If you use a hardware device ahead of the recorder, you are much less likely to overload the recorder and get distortion. This method can also provide the best signal-to-noise ratio for your recording. The down side is that if you apply too much compression, it is recorded that way and can’t be undone.

Applying the processing in the editing system let’s you play with the amount of compression. If you don’t like the sound you can undo the effect. Software compressors and limiters are included with most editing packages, and thus you avoid the cost of buying a physical compressor. The down side is you have to be more careful setting levels so that you don’t get distortion from recording too loud, or noise from recording too low.

Limiting should be applied to all of your voice recordings that are being released on CD or as a digital file. If you don’t, the average level of the recording will be low and the CD or file will sound quieter than commercially produced products, forcing the listener to turn up the volume.

The proper way to apply limiting is to first normalize the track to minus 1 db. as we have done above.

Open the limiter or compressor in your software. Set the limiter so that the threshold is at about -6 db to – 8db. Set the compression ratio for 10:1. Set the attack to about 10 milliseconds (ms.) and the release time to about 200 ms. then apply the effect.

Re-normalize the track. Your volume level should sound louder. Check that it doesn’t sound unnatural. With some experience you will learn how to adjust the settings for the best sound. Read your software manual for the particulars of your device.