Monthly Archives: June 2012

by - 3 Comments

What is mixing ? (part 2)

audio-compression

As mentioned in the previous topic about “What is mixing ? (part 1)”, we defined the mix as a three dimensional world :

  1. Frequency
  2. Dynamic
  3. Stereo

Of course, like in the real world, we can also take into account the time as the fourth dimension, which is absolutely right. But for now, let us focus mainly on the three first ones. Since the frequency dimension has already been covered in the previous section, let us now move on to the dynamic aspect of mixing.

For some of us, in mastering, dynamic is everything… But from a mixing perspective, what does that mean ? Well, it is fairly simple: Some instruments are more dense than others, so they need to be compressed in order to be “competitive” in the mix. Some people (a lot actually) just compress everything to the maximum in the hope that it will sound crazy loud. Well, if it sounds loud, don’t expect me to believe that it sounds right, or either close to be good.

Honestly, if you want it to sound loud, ask your mastering engineer. This blogpost will encourage you to focus on relative dynamic rather that absolute dynamic (or loudness). What I mean by that is that you should make sure that the dynamic between your instruments is making sense, no matter what the dynamic of the whole song is.

That said, as a rule of thumb, start by simply compressing elements that are not dense enough to compete with others. A convenient example of that is vocals compared to brass. The compressor has not been invented to kill the dynamic, but simply to blend the instruments together. Keep that in mind and your mixes will sound better, I promise.

 

by - 141 Comments

Towards a deep understanding of the mastering process (part 2)

Second step: Benchmark measurements methods
Once we identified our key parameters, the question is now : how to measure it ? Otherwise, it’s hard to build a reference and we fall into the subjectivity. At this point, some mastering engineers might disagree and that’s okay. The objective here is simply to propose a different approach, a new perspective. Use it or not but, in both case, consider it. That said, let’s look at the tools we could use for our experiment:
  • Tonal balance can be measured via spectrum analysis
  • Perceived loudness can be evaluated by RMS measurement
  • Stereo can be measured with a vectorscope
What’s missing ? A tool to measure the harmonic content of a mix. We all know what is the problem with exciters. The more you turn the button, better it is… until it sucks ! It is a key parameter that is still based on subjectivity only. For an experienced engineer, it might not be a problem to adjust the “right” amount of harmonic content but, it’s not the case for every newbie home-studio owner that work in an less-than-perfect listening environment.
.
.
Third step: Define a standard range of values by reverse engineering
I really like this part which consists to analyze the key parameters of my favourite best-sounding records. I find it exciting, it’s like finding dinosaur bones. Let’s take the 10 best sounding albums of your musical genre and analyze them, both subjectively and objectively. How does it sound ? What do I like in these records ? Only in terms of sonic perspective, what’s different from other genre ? How the dynamic sounds like ? Does it sound bright, muddy, airy, warm or neutral ? Then:
  • Capture the tonal balance with a spectrum analyzer, then compare each of them with pink, brown and white noise
  • Measure the RMS throughout the song
  • Take a look at the vectorscope; how the mix behave in terms of stereo imaging ?
  • Try to print in your mind the degree of harmonic content.
Especially, if you’re testing similar sounding record, you will recognize patterns very quickly. Maybe only 3 records will be enough give you a very precise idea of how it should sound. More specifically:
  • Tonal balance: Is it closer to pink noise, brown noise or white noise ?
  • RMS: What is the max and min value. Is the range narrow ? Is the average making sense for your needs ?
  • Stereo: Does most records behave the same way ? Do you have a good idea of how it should look and sound like ?
  • Harmonic content: Are you going to make it shinny or raw ?
Personnally, I learned from this experiment that I like tonal balances that are between the brown and the pink noise, RMS values between -10db and -12db, a medium stereo spread with high harmonic content. This is a very very very powerful statement!
See you for the following next week !