There’s an issue that people who’ve started making music the past 10 years or so think an audio spectrum / waveform = what you hear. The sad fact is that it doesn’t.
One thing I’ve found that seemed to be a key to “professional” audio mixing / mastering people is that they had some kind of experience with real, studio mixers. They didn’t have access to spectrum windows, they mastered by tweaking dynamics knobs and using their ears. They didn’t have techy BS of calling out kHz or Hz numbers like “4,200”, “31.5” or other crap relating to decibel ranges like people using computer software so frequently do. This is because…
Learning in hardware studios has its payoffs
The difference between these two groups (a few years ago) didn’t seem to be that one had better tools thanthe other but that a physical studio taught better, natural audio techniques while people using computer software VST’s seemed to approach mastering like a math problem.
I feel being able to “see” the spectrum of audio does more harm than good for nearly all people who have started out doing software music without experiencing a physical sense of turning dynamic / EQ knobs blindly without seeing a direct spectrum change.
Yes, this is a controversial statement. But it’s true; too many people rely on their eyes instead of their ears. Speaking of ears…
Listen with your ears, not your eyes
The reason spectrum analysis can distort your perception is because spectrum analysis of waveforms are simply not consistent. Why? Because audio waveforms are not consistent to the way the human ear perceives audio.
I happen to use a virtual spectrum analysis device, an RE 60 from Red Rock Sound (mentioned in how to reference audio) that is always on display at the top of the rack (something I’ve seen no one else do for some baffling reason). Here’s what it looks like in action for Department of Corrections.
And, before that, I used to use a BV512 vocoder device in Reason (set to EQ mode) as a spectrum analyzer, doing that since I believe 2003 (the photo of the BV512 is from a 2004 project). That’s 13 years of using a spectrum analyzer on every song, every project, every goofing around session. I do this, first-and-foremost, for the enjoyment of being able to see the audio dance (and of course for utility reasons too that I go into below).
The point of mentioning this is that I have the unique experience of being able to see, constantly and consistently, what the sound looks like. Countless times I see something, visually, that is completely different from what it actually sounds like. Listen to so-called experts on forums and they’ll swear by the numbers of kHz and Hz. But they’re wrong because…
Audio mastering is not a paint-by-kilohertz-numbers book
Mixing and mastering is not a numbers game. Before Reason had a spectrum analysis window, I was able to use my own spectrum analysis devices but I choose not to; I would use the same technique those in major studios of the 70’s and 80’s first pioneered; using my ear and the knobs of the mixer to sweep, listen, compare and adjust sounds.
So many times, spectrum analyzer displays, whether its the RE 60 from Red Rock Sound or a Klerk Teknik DN60, will give the wrong impression of where the sound is or how it should sound. So many times, I’ll have screeching highs that fall in the lo / mid range visually while soothing lo / mid sounding pads will register only on the highest end visually.
Spectrum analyzers; never liars, never honest
The secret to this is that spectrum analyzers are not actually lying to you; ever. It isn’t possible for them to lie to you. If a sound is registering in a lo / mid frequency range, it really does exist in that lo / mid frequency range scientifically. But your speakers will be playing the sound from your tweeters and your ears will perceive it as a high. Why? Because it really is a high-pitched sound, even though the waveform exists in the lo / mid range scientifically.
Audio is a very complicated subject; people can spout impressive sounding numbers all they want, “Pff, I never go above the 4.5k range, anyone operating in the 5k range is…”, no, screw those blowhards. Relying on even the highest end spectrum analyzer is a misleading metric that cannot compare to the human ear (coming from someone using a spectrum analyzer for years).
How to see the spectrum with your ears alone
The best way I feel is to try to dial-in your sound blind. You can peak at the spectrum if you want, but it should be just that; a peak. Go back to listening with your ears.
I use the low pass filter (LPF) and high pass filter (HPF) often and always by ear, never by the spectrum. I dial (physically with a Behringer BCR2000) either a LPF or HPF until I can just start to hear the sound losing characteristics.
Then it’s time to find a frequency, a sweet spot, that I want to accentuate. I do this by enabling the EQ, guessing which area the sound falls into (lo, lo / mid, mid / hi or hi), turn the “Q” knob all the way to the right (narrow) and crank up the dB knob. I then slowly / quickly turn the frequency knob (sweeping through the frequencies).
Sometimes I realize I have to change to another EQ range for the sweet spot. But, once I find a frequency that has the most character of the sound, I then adjust the “Q” knob to what I feel would be right (sometimes narrow, sometimes wide) and then adjust the dB to something appropriate (many times, the dB is raised just a bit).
I then see if I can reverse this process, instead of raising up the dB knob for the other EQ ranges, decreasing it and using the same sweeping method. The goal is to see if there are any problem frequencies that would benefit from being cut.
I taught myself this method through natural intuition; it was only later that I realized masterclass audio engineers were being taught this very method (and had been using it since the 70’s). I point this out because some ideas and methods I preach throughout this site are concepts that haven’t become known or established yet (or to my knowledge) but, this shows my thought process is on the right track.
When spectrum analyzers are ideal
The main reason to use spectrum analyzer is to spot, visually, when a frequency may have one spikey spot that may be responsible for clipping or causing compression / limiting settings to kick in needlessly.
This is what spectrum analyzers were built for. You can look at the frequency and see if there is a spikey frequency that may be causing problems (for me, I simply solo the channel and can instantly see the spectrum on the RE 60).
Your ears are the best analyzers of audio on the market
Spectrum analyzers exist because there can be problem frequencies that are invisible to the ear but will be picked up by audio equipment. Learn to rely on your ears and don’t let others talk you into the foolish kilohertz numbers game.
Later. – MJ