Mastering for today’s playback formats

How loud should I go?

What should your target volume be when mastering, and should it be different for each platform? It’s a common question.

Well, we wouldn’t recommend using a target volume as such. Instead, balance the songs against each other, then push the loudest bits up into your limiting, but stop before it starts to sound worse. This will give you the optimum loudness potential (ie, as loud as it can go without sacrificing any quality) for your music. If possible, this should be the version you release, in all formats.

However, sometimes you might run into issues with a specific format. A dynamic master might end up a bit quieter than other songs on a streaming platform, for example. In this case, you need to make a value judgement: is it better to limit a bit harder and keep one master for all formats? Should you create a louder version? Or is it better to just allow it to play quieter on that platform? Ultimately, there’s no correct answer, and you need to use your judgement. Target loudness levels used by streaming platforms are susceptible to change at any time. As people increasingly use the same devices for music as for film and TV, target levels will probably continue to trend downwards towards the lower levels used by the broadcast industry. Allowing your song to play quieter on one platform may be the best long-term solution, as the issue will go away if that platform lowers its target.

You may also need different technical formats for different platforms. If you need to convert sample rate, you should do so before dithering to 16-bit for CD, so it’s often best to render 24-bit files from your mastering project, then convert them in separate steps.

However, when different sample rates are required, we often like to upsample to 96kHz before mastering (a free quality upgrade for all your dynamics and saturation plugins!) and render out a 96KHz master, which can then be downsampled as required.