This chapter is from the book
This chapter is from the book
This chapter is from the book
Audio Preprocessing
Audio is a huge part of any production and can make or break the viewing experience. There isn’t typically a lot of preprocessing that needs to be done with well-produced audio, however. Raw captured audio may need some preprocessing (sometimes referred to as sweetening) to clean it up, but this is typically done during the edit process, rather than during preprocessing for compression. When it comes to audio, the job of the compressionist is to make sure that the program’s audio is appropriate for the distribution channel.
Adjusting Volume
Volume is one of the elements easily adjusted in preprocessing, and you can do this in a few different ways. The quickest way is to raise or lower the volume, either by an absolute decibel (dB) amount or by a relative percentage amount. This is another fairly coarse adjustment, akin to turning up or down the volume on your radio. For those needing a little more finesse, there are other possibilities, such as normalization and compression.
Normalization
Normalization is the act of adjusting the audio levels in the content and then raising or lowering the volume of the entire clip so that the loudest sound matches the level you have specified. This is a global adjustment, affecting the volume of the entire track the same way, rather than affecting the relative levels.
Audio Compression
This is a totally different type of compression than what we have discussed so far in this book; with regard to audio, compression refers to a specific type of audio filter known as a compressor. Loud noises in a digital audio track can cause distortion, and likewise, quiet sounds, such as whispering, can be lost. An audio compressor can smooth out these issues in an audio track by removing dynamic range. By pulling down large spikes and lifting up those quiet parts, compression will ensure that the average loudness is fairly constant.
Noise Reduction
Just as with video, there are also noise reduction filters for audio, although these are more often found in professional audio-editing tools, not in compression tools. Unwanted noise in audio tracks is just as bad for compression because noisy video (meaning bits) will be wasted, and the end result will be lower quality than desired. Some compression tools have simple hum-removal filters that will help clean up the audio during encoding, but truly bad audio may need to be preprocessed separately in a professional application such as Avid Pro Tools or Adobe Audition.
Stereo Mixdowns of 5.1 Surround Mixes
Depending on the type of content you normally work with, you may be given source material that has a 5.1 surround sound track and asked to encode it for a destination that supports only stereo audio. Nine times out of ten, if you ask for a stereo mix, you’ll be able to get it because a typical surround mix session produces both the 5.1 and stereo mixdown. It’s almost always preferred to have an audio professional produce mixdowns, but in the rare case where you have only the 5.1 source and you’re the only resource available, you’ll need to know how to properly convert the six channels into two.
The most straightforward stereo mixdown is referred to as Lo/Ro, which stands for “left only and right only.” Sometimes referred to as an ITU downmix, the formula to create this two-channel mix from a 5.1 source is as follows:
Discard the LFE (subwoofer) channel. This channel is of little use to small speakers that will most likely be the destination and will overdrive them.
Leave the left and right channels alone, each continuing to be mapped to Ch1/Left and Ch2/Right, respectively.
Map/pan the center channel to both Ch1 and Ch2 equally while lowering its volume –3 dB.
Map the left surround channel to Ch1 while lowering its volume –3 dB.
Map the right surround channel to Ch2 while lowering its volume –3 dB.
If you are asked to deliver an Lt/Rt (which stands for “left total/right total”), this is a different type of downmix, often referred to as matrix encoding. It is much more complicated to create and is best left to an audio professional.
