Last year I wrote a blog entry about using Magic AB to reference mixes or masters against commercial recordings. You can see the original post in the side pane but having reviewed it, I realised I had left out one vital piece of info; making sure the reference material was appropriately level matched for accurate comparison.
Magic AB seems to have become the standard for referencing, which makes a lot of sense given its ease of use, as well as the ability to load multiple tracks independently of the session and save presets etc. It makes referencing fun (an achievement in and of itself) but unfortunately, the RMS and peak metering on Magic AB can leave you in the dark when trying to frequency match commercial music against our own.
If you are a member of the production advice forums, you will have heard Mastering engineer Ian Shepherd talk about how crucial it is to have level matched reference material, and how even half a dB difference in volume can affect your judgement when trying to match frequencies.
Over the years his videos have helped me enormously and you can check out his production advice site here.
LUFS (loudness units relative to full scale) is the measurement system used to gauge the overall, or integrated, loudness of audio. It takes an average reading over time and provides an integrated, short term and momentary level reading, something that can’t be deciphered manually using traditional peak or RMS meters, which Magic AB has.
This means that level matching in the plugin can only be rough, and not as accurate as if you used an LUFS meter to level match reference tracks to a set target loudness, before importing them into the plugin for AB’ing.
So, here is a short video showing how I level match my reference material. I master to a target level of -16 LU, but you can set it however you like. The most important thing is that the integrated loudness levels are matched to your material when referencing.
I use the LUFS plugin from Klangfreund, which has great metering and auto level adjustment (as well as an extremely handy grouping feature), all for $49 (there is also a $24 version with a few less features). I use this particular meter because the auto adjust speeds up my workflow, but a meter alone is good enough and something like Toneboosters EBU Loudness meter is a pretty affordable option. As long as it works according to the specifications laid out in ITU-R BS. 1770 it’s fine.
I use auto adjust, but you can do it manually by taking a reading (using the offline render like in the video) and adjusting according to your target.