Mastering Tutorial Part 2: The Secret Notebook of a Mastering Engineer (Advanced)

Audio Suitability for Mastering

You are now one step away from seeing the four tools discussed so far be put into action in a live mastering session. But just before we get to that, now is a good time to go through a checklist to ensure a premaster is ready for the final mastering stage. This checklist is based on the audio having no glitches and errors at this point.

1. Amplitude and Headroom - Is it clipping?

Ideally, the highest peaks should be approximately 3dB below the maximum level.

The digital realm, unlike analogue has very strict rules governing waveforms exceeding the allowed maximum level, or 0dB as it would appear on your master fader inside your DAW. Digital technology works in precise values. If more level is pushed into a digital system than it can handle, it will be unable to capture that part of the signal, resulting in a flat line rather than the natural curves of a waveform. This is known as clipping, which can sound quite nasty, and can even damage the high frequency drivers (tweeters) in speakers. A mix that clips before mastering should be re-bounced by the mix engineer to fall within the allowed headroom before mastering can commence.

2. Signal-to-Noise Ratio and the Noise Floor - Is it too quiet?

Ideally, the highest peaks should be high enough to be approximately 6dB below the maximum level.

Recording systems, both digital and analogue, are not perfect systems. They often generate their own low-level distortions or ‘artefacts’ which can be caused by many things. For example the hum and whine of electronic components adding to the signal, or slight inconsistencies when an analogue input like a microphone is converted to digital information. This layer of undesirable information is known as the 'noise floor'. Tape hiss is a good example of a sound that makes up the noise floor.

The relationship between the noise floor and the audible signal that we actually intend to hear is known as the 'signal to noise ratio'. If we are dramatically raising the level of a song during mastering, then we are also raising the level of the noise floor, and the quieter the mix before mastering then the louder we have to make it. This means that the noise floor becomes ever louder. Ensuring that the song is at a good level before mastering gives us a good signal to noise ratio, and helps to mask undesirable noise.

3. Sample Rate and Word Length - Is there enough data to work with?

Check that the files are at least 44.1kHz sample rate and 24-bit word length (word length is sometimes referred to as bit depth).

In terms of a PCM Wave File or in fact any other professional digital audio file, the minimum quality required to produce a good master is a Sample Rate of 44.1 kHz (CD quality) and a word length of 24-bit (beyond CD). Hopefully the sample rate will be higher than this, since usually the more frequent the sample rate (say, 48kHz, 96kHz or even 192kHz as opposed to 44.1kHz), the more harmonic detail is present in the audio files and the better they will hold up to processing in the mastering chain.

16-bit files are okay, but since they have fewer memory addresses dedicated to dynamic range (and so contain a less detailed ‘map’ of the various loudness levels in the mix) they tend to respond less favourably to dynamic processors such as compressors and limiters.

4. Capture - Is the entire mix present in the audio file?

Check that there is a small amount of silence before and after the audio information begins and ends. This is to ensure the DAW that ‘bounced’ the audio file was able to properly capture the beginning and end of the audio, which is a surprisingly common issue.

5. Master-Bus Processing - Was the mix hit with an overall compressor or limiter before reaching the mastering engineer?

Check that the song to be mastered has a good dynamic range and has not been over processed by master bus compression.

Some mix engineers prefer to add a slight amount of compression to the overall mix in order to add punch and/or cohesion before the mastering process. When the effects are very subtle, this should not  be a problem. But if the audio is in an excessively compressed state before reaching additional mastering compression and limiting, it can lead to lacklustre audio that has little punch or definition.

6.Tonal Balance.

Is the mix of a high enough standard for the mastering process?

You should now assess the mix. You must ensure that the individual parts that make up the mix are balanced and clearly defined. Is the bass synth too loud and masking the kick drum? Is the snare drum too quiet? Are the high frequencies of the strings section creating resonances with the high frequencies of the hi-hats? Is the overall mix too boomy in the low end or too brittle and harsh with treble? Although mastering is able to address some of these issues it is preferable that they aren’t present in the first place. The fewer processes required at the mastering stage to achieve the desired sound, the better the end result.

If you put a good mix through the mastering process, you will produce a good master. Conversely, if you put a bad mix through the mastering process then its sonic shortcomings become even more apparent.

In the advanced section there is chapter dedicated to how a mix should ideally sound to produce a professional and commercial sounding master.

This concludes the first half of the mastering tutorial course, 'The Audio Mastering Blueprint'. The next section is made up almost entirely of videos. I will master a track in full before your very eyes using the techniques and methods discussed up to this point.

After the live mastering session is when it really gets interesting - this is where I reveal some of the most closely guarded mastering secrets.


David Eley - TGM Audio

Next page...  Introduction to the live mastering session...