Explore the first season Are You Listening? with Jonathan Wyner:
And get caught up on the latest episodes of season two:
Sign up for our newsletter and get tutorials and tips delivered to your inbox.
Jump to a section below:
When you use a reference track in mastering, your main goal is to accurately compare your work to what’s out there in the world—your favorite recordings or the production decisions that define the genres within which you’re working.
To do this, you need to be able to hear, as closely as possible, the song’s audio the same way it came off the mastering desk—not changed by a distribution service, not changed by cutting it to vinyl, putting it on cassette tape, or converted to MP3 or AAC.
Don’t be fooled by the audio you hear from a streaming service—platforms like Spotify and Tidal have loudness normalization on by default, and some services only allow for normalization on desktop and not mobile.
Yes, music streaming services—the main supplier of the audio we listen to—sometimes modify the audio to give consumers a better listening experience. Streaming services use loudness normalization to make sure the playback level is more or less consistent between songs. You may notice variation depending on low end energy, how much compression there is, and other factors, but in general with loudness normalization on, you won’t get gross differences in terms of level between tracks.
By changing the level at all, if you use that track as a reference for your mastering work, you’re not really comparing it to the level of the master—you’re comparing it to the level of the loudness-normalized streamed track. This is why you want to ensure you have loudness normalization turned off on these services, if you must use a streamed reference track, so you can truly understand what is ultimately going to happen to your master once you’re finished.
The quality of your reference track’s audio is very important. All the different lossy audio codecs used by major distribution services—MP3, AAC, Ogg Vorbis—are used so that it’s easier for your listeners to hear your music. However, these codecs do create artifacts that, while seemingly hard to detect, change the overall tone and add distortion to what you hear.
When you compare your work against a lossy reference track, your reference will be slightly skewed because the lossy codec created some artifacts. If you send your master to a streaming service for distribution, those artifacts may be compounded if loudness normalization is employed by the service.
So how can you tell if lossy artifacts appear in your track? Sometimes seemingly full resolution files—WAV, FLAC—or even pieces of a full resolution file, may once have passed through a lossy codec. Why does this matter? Let’s say, for example, the mix you were sent included an MP3 beat and an AAC vocal file. The signature of the beat’s MP3 lossy codec is likely to appear in your spectrogram display. While this won’t conclusively prove that a mix was encoded with a lossy codec, artifacts from a lossy codec will be evident visually in your spectrogram.
We'll explore what this might look like in practice, and how to identify lossy audio in a spectrogram. Ozone’s Codec Preview tool is perfect for previewing how your master might sound in a variety of different file formats and bit rates.
When we turn on Codec Preview, you'll notice that it seems we've lost all of the energy in the recording above about 15,000 Hz—because, in fact, we have. This is a result of the processing the codec uses to reduce the file size of the track.
To easily see if your track includes artifacts generated via a particular codec, or to see how your track will sound in a particular format, adjust the bitrate of Codec Preview—for example from 128 kbps to 192 kbps—and notice added artifacts in the high end of the spectrum. Switch to 320 kbps, and these artifacts are nearly impossible to notice. You can cycle through bit rates, paying particular attention to any changes in the high end. This is a quick way to check if a file you’re listening to was once a lossy file or had been through a lossy codec.
When we run signals through lossy codecs, we end up with a fair bit of noise that comes from all the filtering that's required in order to reduce the file size—that’s all that added information in the high end of the spectrogram. Let’s hear the difference between a track that has been through a lossy encoder, and a track that hasn’t.
The full track including artifacts
Only the artifacts from the above track
Keep in mind most of what you hear in the artifacts is masked according to the psychoacoustic model the codec is based on. While most of the artifacts are masked, the distortion that is there does influence the sound of your track.
Let’s now hear what the codec is doing to a few specific elements—the sibilance on the S’s and T’s, the high hats, top of the snare drum, anything that’s in the high frequency region—versus the original track.
The full track without Codec Preview
The full track with Codec Preview
We encourage you to seek out something that hasn't been lossy, or at the very least is a very high-resolution lossy file using 256 or 320 kbps.
The Match EQ module in Ozone is especially helpful when working with reference tracks. From here, you’re able to learn the signature of the track you’re working on and apply a sliding amount of your reference track’s target spectrum to your master.
Be sure to play the audio at least 30–40 seconds to give Match EQ the best possible sense of the overall spectrum of the track.
Use the Smoothing and Amount sliders on the right of the module to adjust how much you want your two tracks to sound alike. Do so while you’re listening to the track, until you feel the EQ adjustments suit the master you’re working on. By using the Smoothing slider, you can generalize the change you’re trying to make, rather than trying to make your master sound exactly like the target track.
Here are a few things to consider when preparing to use Match EQ in Ozone:
If you use a track by your favorite artist, take the level of that track down by a few dB, so you don’t try to make up the level all at once before you start in on your mastering adjustments. If you try to make up the level all at once, you may find yourself pushing a limiter too hard, getting too involved with level before you start thinking about tone and tonal balance.
When you use processing like Match EQ, leave yourself some headroom so if you do drive the level up with additive EQ, it won’t result in distortion. Small increments are best—don’t try to make up all the level at once! Give yourself some room to breathe as you try to match your track to your reference track.
Another way to understand the overall balance of your track is through Tonal Balance Control in Ozone. With it, you can measure your track against what a typical spectrum profile looks like for the genre you’re working in.
This is helpful because if there’s something out of balance, you have a moment to reflect—am I making a creative decision I’m happy about? Or is there too much high end? Tonal Balance Control’s supplied curves were created after analyzing thousands and thousands of tracks in each genre to ensure an accurate average, no sourcing of reference tracks required.
A mastering reference track is a great resource for generating lots of creative ideas to use in your practice. We’ve curated a playlist of some of our favorite mastering reference examples. Listen to the tracks and appreciate the craft, the arrangement, the production choices, and every other decision.
We’ve also put together a list of our favorite reference track articles on both mixing and mastering. Be sure to download the free genre-specific Tonal Balance Control curves linked in a few of these articles.
Get top stories of the week and special discount offers right in your inbox. You can unsubscribe at any time.
Copyright © 2001–2021 iZotope, Inc. All rights reserved.