Basic Audio Mixing
This paper offers an introduction to the basic principles of audio mixing within digital audio software.
The principles of mono and stereo audio are summarised in this short screencast.
Multi-track mixing is a very subjective process that requires skill that is acquired through experience. This introductory guide is written for those who wish to conduct very basic mixing of small multi-track projects and who have little or no knowledge or experience of the subject.
A mix is the ‘putting together' of audio files to create a combined and finalised mix file. With the creation of a new audio resource, this could be in the form of separate audio files of spoken word, adding background music to voice recordings, or blending in special effects sound files with other source files.
It discusses the fundamental aspects of mixing in a DAE, such as volume, pan and automation, and what mixing down to mono, stereo and surround actually means. This guide also advises on how to finalise your mix within a DAE, the starting blocks of the mastering processes.
What is mixing?
Mixing is simply the layering together of separate sounds so that they ‘fit' in the sound field upon playback. Commonly the aim is to create the illusion that all the sounds were recorded together, or if not that the timbre of the sounds compliment each other and can be heard clearly. Effects and processing techniques are usually applied when mixing to enhance the quality of individual sounds and the overall ‘mix'.
There are three types of resultant file that can be mixed to; mono, stereo and surround formats, which increase in file size respectively. Files within a project may be either mono or stereo. For a definition of mono, stereo and surround file formats, see the advice document File Types and Compression.
A monaural (mono) file consists of one-channel of sound. Upon playback it contains all the audio in the central balance of the sound field. That is, if one were to listen to a mono file on headphones all the sounds would appear together in the middle, with no left or right pan positioning. Monaural sound production was common in early recordings but has since been surpassed by stereo and surround production.
The perception of monaural sound.
Mono mixing is useful when only one source has been recorded, i.e. a recording with a single microphone, or when use of stereo perception is irrelevant or not required for delivery purposes. This could be to keep the file size to a minimum for efficient web delivery or when delivering a one-channel recorded spoken word file.
Stereophonic (stereo) production uses two channels which represent the left and right output of a system - such as speakers or headphones. Sounds can be panned (positioned) at locations within the sound-field to create a natural perception of how we hear sound. The two channels are commonly combined to give a stereo file.
There are many methods of stereo recording which are discussed in the advice document Microphone Techniques, but mixing to a stereo file allows us to position mono and stereo files within a project to any location ranging from far left to far right upon playback through speakers or headphones.
The correct speaker placement for monitoring stereo recordings.
Stereo mixing allows you to create a stereo ‘image' - a spread of sounds - which are akin to how we hear sounds in our day-today environments. For example if you were listening to a conversion between two people, it is likely that they will be facing each other standing apart. Two single channels can subsequently be positioned to represent this natural occurrence which helps us distinguish the location of each person in this example.
Surround sound mixing
Surround sound is a multi-channel format to create a 3D immersive listening experience where the listener perceives sound in a 360 degree sound-field.
There are many formats of surround sound, the most common being 5.1, where the 5 relates to the number of speakers positioned (see diagram 3), and the 1 to a sub-woofer which handles low frequencies from all the channels. Many electro-acoustic compositions are mixed to surround formats and can sometimes require up to and beyond 16 separate speaker channels for playback.
A correctly aligned 5.1 speaker array.
Note that the aforementioned sub-woofer (LFE) position is not indicated. This is due to the human lack of perceiving sub-bass positioning and can therefore be positioned anywhere in the room.
To mix in surround sound requires a speaker array as shown in diagram 3 above and only the more sophisticated DAEs offer surround mixing features Further encoding software is then required to create the finalised surround sound files.
Mixing volume levels
Mixing the volume levels within a project can be relatively straight forward when working with a small number of audio channels within a project. It may be the case that some audio files have a wide dynamic range - this means they could be quiet in places and very loud in others - which will need to be taken into consideration. To adjust channel levels some DAEs provide a db meter for each channel with a fader for adjustment (this may be found in the main edit window or the mixer window). This is shown in diagram 4 below.
A digital volume fader controller
When working with digital audio 0db is the threshold where any signal above becomes clipped (audibly distorted). Care should be taken to make sure that this does not occur.
A master level meter, which displays the combined output of all the channels should be monitored and adjusted correctly so that this too does not exceed 0db.
In cases where a DAE does not have a mixer window, a control for the level will certainly be located elsewhere. Commonly it is found to the side of the channel in the edit window.
It is possible to adjust the volume within regions of specific areas. For the beginning and end of regions it is advisable to use a Fade tool - advice on this can be found in the advice document Basic Editing - making volume changes within a region can be done through automation, discussed later.
For stereo and surround mixing, it is possible to adjust each channel to a position in the sound-field. Each channel in a DAE commonly has a panning controller which allows you to position the channels output relative to the main output format.
A stereo pan control lets you control how much of a channels signal is fed to the left or right output channel. The amount sent to one channel is relative to the amount reduced in the other. When fully to the left side, the signal is sent only to the left output channel and vice versa.
A pan control allows you to control the amount of audio signal sent to the right or left channels.]
What is automation?
Automation is the control over channel specific parameters at any given location of your arrangement. Automation comes at different levels of sophistication with DAEs, but primarily most offer simple volume and pan control over single tracks.
For example, a mono channel of a car driving past a microphone will sound more ‘realistic' if the effect started off fully panned left and moved across from left to right during the sound. This can be done simply using automation. For further information about platform specific automation refer to the instruction manual provided.
In more high-end DAEs automation offers control over routing levels and plug-in parameters and as such requires a high processing speed and amount of RAM.
The orange nodes indicate changes in volume over time. These are remembered during playback and the volume control is subsequently automated.
What are plug-ins?
Plug-ins are software versions of hardware effects units and dynamic processors, from filters and distortion to reverberation and equalisation. Different formats of plug-ins are compatible with different DAEs with no common format available. Some often come bundled with a DAE whereas more sophisticated plug-ins are available to purchase separately.
Plug-ins can generally be applied to individual channels within a DAE, across a bus or even over the entire mix. Care and thought should be applied when using plug-ins and further advice can be found in the advice documents regarding use of digital effects and processing.
Should I Normalise my audio files?
When mixing your audio project you may be presented with the option to normalise your files or your final mix. Normalisation increases the amplitudes within an audio file so that it peaks at 0db, utilising all of the headroom possible.
The process can be useful for making entire files louder but also increases the level of any undesirable noise within a file. To do so the value of each sample is multiplied by a set amount. However, small errors begin to occur when mathematical functions are applied over samples and as a result dither is applied to mask these errors (for information on dither see the advice document An Introduction to Digital Audio).
Diagram 7 shows a stereo audio wave before normalisation, in 7a, and after normalisation, in 7b.
It is advisable to avoid normalising audio files, unless you are confident in your approach, as in some circumstances it can radically cause detriment to your files. Also you will find that as your skills and knowledge around mixing increase, you will find that there should be less of a need for it. Audio files should be initially recorded with enough headroom that there is no distortion but not too much headroom that the overall level suffers. However, in some cases of digitising analogue media normalising can be useful if you are ingesting at a low gain level in an attempt to avoid any input distortion
I have finished mixing in the DAE, now I need to ‘mix down' all the individual files
When you have perfected your mix within the DAE you will need to export or ‘mix down' all of the channels to your chosen output format. This feature may be termed ‘bounce', ‘export', or ‘share'. Note that when you ‘save' or ‘save as...' it is the project file you are saving and not the mix you are creating.
Generally you will be offered a common dialog box which allows you to customise the format you are saving to. Take care to select the correct file format, sample rate, bit depth or bit rate for your delivery method when mixing down your files.
Some DAEs will ask you to select the timeline region you wish to export. Make sure you enter the correct range of time that you wish to be exporting. One your mix has been rendered, it is strongly advisable to always playback and listen to the finished file to check for any mistakes or errors which may have arisen.
Published in: Creating