What should the order of the modules be in Ozone?
The first module should be an equaliser to correct any tonal balance issues before dynamics processing.
Then some kind dynamics processing, be it a single, multi-band or dynamic EQ.
Another EQ could be next, for enhancements and broad changes.
The brick wall limiter should be at the end for protection against distortion
How do you approach mastering a whole album?
Keep in mind what the focal point of the track is. Think about maintaining consistency from one track to the next, in terms of level and tone. But, ensure there is some discrepancy between songs, a loud track should sound louder than a quieter track and vice versa.
Are there different strategies or rules for processing and mastering songs in different genres?
The first aspect is understanding the genre you’re mastering.
There are no rules outside some essential sound boundaries.
If you understand what your tools are good at and what a genre requires, you can ask yourself:
What’s the sound I’m after?
What changes if any, do I need to make?
What tools do I need to reach for to get there?
A tube processor in the signal path will sweeten the track and add even order harmonics.
What is the crest factor? Is there a crest factor I should aim for?
The crest factor is the difference between peak and RMS level, in a very short-term measurement. It’s the peak to average relationship.
A sine wave has a 3dB crest factor, between the peak level and sustain.
A square wave has 0dB crest factor, as it spends its entire level at the same value peak.
A peak could be a drum strike, and the average a sustain of melodic content.
If you have a broad crest, where the peaks are high compared to the average, then you’ve got something that’s going to sound more transient and more exciting, but you’re also going to lose contact with the sustain, the tonal portion of the track. There is a sweet spot, but not a one size fits all measurement.
Beyoncé’s ‘Drunk in Love’ is an excellent example of a track with a high crest factor that sounds perfect. The crest factor is around 15dB. It’s a high crest, but the track is sparse, that’s a lot of room between the transients. The bass is driving the strong peaks, which then drops to the sustain (RMS) when the vocal is introduced.
If you have a crest factor higher than 8dB, you’ll lose low end and hear some distortion in the high end.
Naturally dynamic music will have higher crest factors. Acoustic Jazz may have around 14dB.
There is a rough relationship between crest factor and genre; you have to think about what’s going on in the music to come up with a number.
How many exports do I need to make for streaming services with different loudness normalisation standards?
It’s unrealistic to make a version for every service.
If you make a single version of your record that will work for loudness normalisation, you should be fine as most streaming platforms use similar numbers.
Many mastering engineers will make a version as hot as it can be, while still sounding great, that will still work if it gets turned down by a streaming service.
Three versions maximum for everything you do:
1. -12 — -10 dB LUFS for streaming.
2. Push the level as hot as you can go for a peak normalisation master for SoundCloud.
3. Then make a version without limiting, or very little for vinyl. Just in case. Vinyl cutters don’t like limiting.