What I learned from producing over 50 songs this year

Photo by Mitchell Luo on Unsplash.

The last 525,600 minutes have been full of contradictions for me. The still ongoing Coronavirus pandemic has seen many industries collapse or change beyond recognition, but I’ve somehow managed to find myself working in audio production, a subset of the music industry that seems to have been miraculously spared by the impact of COVID-19. Despite my credentials, I’m more of a doer than a philosopher, so while awaiting the inevitable implosion of the economy, I produced over 50 songs for various clients this year, and I wanted to share some of the lessons I learned from the experience. I hope you find value in them.

  • Don’t be afraid to commit. BAM! Coming at you with good life advice, straight out of the gate — blog over! Seriously now, while as a producer one might like to have the ‘infinite tweaking cheat code' enabled, your computer’s a bit more of a stickler for the rules of physics. Everyone’s loaded a few instances of their favourite soft synth in a session and then quickly realized that the CPU can’t deal with the load in real-time (looking at you Serum, Diva, & Pigments). On key tracks, I use the freeze function (working in Reaper), but if it’s something where I know pitches won’t change and I’m happy with the sound, then I just render-in-place and move on. Option paralysis is a thing in audio production (and it usually stems from insecurity about what one’s doing). Terrifying at first, but I literally never regretted committing a sound in the last 365 days, so I would definitely recommend you give it a try and see how it goes.
  • Build and use a custom mix template. I wish I started doing this years ago, but back then I believed that starting with a blank canvas equalled unhindered creativity. Bottom line: it doesn’t. It just meant I wasted bucket loads of time adding the same reverb, delay, spreader, EQ plugins I use 97% of the time, routing tracks, and grouping everything. Andrew Scheps is famous for having had the same template (more or less) for the past 20 years. Obviously, as in Mr Scheps' case, this template will be ever-evolving, as you might get bored of using a Pultec on your stereo buss and decide that the fancier Gulfoss is the way to go. But that still means you can spare hours that you could be using to enhance another element of the song.
  • Become the tsar of gain staging. We all butcher this in the beginning. One might get everything below 0 dBFS on individual tracks, but the stereo buss cries out for help at +8.6 dBFS. I’ve been paying close attention this year to how audio levels add up and what I’ve noticed is that if you set your levels right, the results will be consistent from song to song. For instance, my mixes invariably hit the Waves J37 tape machine that lives on my stereo bus at -5.6 dBFS. And after using it on 50 songs this year, I can hear if it’s hitting the (virtual) tape too hard. If gain staging is done right, even the master limiter will work less and will be there for tone & feel rather than loudness.
  • Saturation is your best friend. What’s better than a slice of cheesecake? More slices of cheesecake, of course! Because of the history of recorded music and its relationship to tape and analogue circuitry, our ears have gotten used to the sound of (pleasing) harmonic distortion. Distorting and saturating the signal creates a beefier version of the original sound by adding extra harmonics. This is one of those processes that, if used judiciously, can make every source sound better. Moreover, if a tape algorithm is involved, the natural compression that comes along with analogue tape recording will also help with taming transients and smoothing everything out in a pleasing way. Win-win!
  • Parallel compression is often the solution to everything. Compression is awesome, right? However, you do loose a bit of what made the source sound exciting. Parallel compression usually fixes this, levelling everything but still maintaining a bit of the original performance. I started doing this years ago, but only doubled-down on it over the past 7–8 months, and I’m never going back! On vocals, parallel compression can be a wonderful thing, but where I feel this technique shines is on drums and on a parallel mix buss. To fully get this point, try adding a compressor with a fast release time that obliterates the drum group and blend that in with the original signal. Instant excitement! Moreover, as made famous by mixers such as Andrew Scheps and Tony Maserati, the technique can also be applied on full mixes, where you blend in a compressed (sometimes also saturated and widened) version of the entire mix (or the mix-sans-drums) to the full mix. Now that sounds like a record!
  • Get rid of junk in the low end, but don’t forget about the junk in the high end. As sound dealer Jaycen Joshua constantly advises, the more you intelligently chip away at the low end to leave room only for those things that should actually live there, the louder the mix (in Mr Joshua’s case, it seems to be ‘2dB louder than the other guy’). But let’s not forget about the top end! At the end of the day, you want a mix that’s balanced, which implies the top end is also balanced. This one is less about loudness but more about harshness and depth. As the recent SoundCloud mastering service shows, people seem to want bright mixes. But there’s a fine line between bright and ear-piercing. But, more importantly, the presence of top end in sources means that they seem closer to the listener. You might want the vocal to be in-your-face, but if there’s some sort of verby adlib, then that one will probably benefit from a high-cut, just to reflect the laws of physics (again). Spaces (rooms, halls etc.) absorb high-end information, so the bigger the space the less top end it’ll have. One of the most important lessons I’ve learned this year is using high-shelves and high-cuts to create a sense of depth in mixes, placing sounds on the z-axis of the sound-box.
  • A new plugin is actually an old plugin in disguise. This one’s quite fascinating. Given that audio plugins are now extremely affordable and of incredible quality, you can pretty much buy every plugin you’d need in your mixing career for the price of a single hardware Maag EQ. Consequently, it’s extremely tempting to fall into the trap of purchasing the next AI-powered, analogue emulating plugin, especially if you follow some of these companies on social media (kudos to their marketing departments!). I’ve managed to shield myself from this to a great extent, particularly because I’ve managed to buy all my plugins during periodic sales (looking at you, Waves). There are only two plugins that I was excited about but then realized I got less out of them than I thought: FabFilter’s Timeless and Twin 2. As much as I love FabFilter stuff, these two have not been part of my arsenal for more than a few months. Nonetheless, I used them on several client work, so at least it paid off financially. But the truth is one doesn’t really need that many plugins to do their job. Nowadays I rarely use anything else than the FabFilter Pro-Q3 for surgical EQ cuts, for instance, and it’s simplified my life tremendously. I’d rather have a workhorse plugin for each process and only worry about it when it doesn’t work for what I’m doing (which must be less than 5% of the time) than have 25 EQs, 50 compressors, and 100 reverbs that all basically do the same thing.
  • Every moment of the production must be exciting. Comic book illustrator extraordinaire Todd McFarlane often says that artists should make even boring pages (i.e. exposition dumps) interesting. I’ve always liked this idea, as I think it perfectly maps onto music production. The objective in any song is to tell a story with the words — the musical backdrop as well as the mix itself need to reflect that. Getting the intro to pop is key to grabbing people’s attention in the current climate of music-abundance, but keeping listeners engaged through the rest of the track is even more challenging: adding effect-throws, muting parts, automating panning, adding glitch transitions, anything to make each section stand on its own. Obviously, there’s a balance to be achieved here, as you don’t want everything to be OTT all the time. But what I’ve found is that most artists have a great sense of how to achieve constant excitement in their songwriting, so it’s our job to fill in the gaps in the production and mix and enhance their vision.

Ok, so now for a bonus round on dealing with clients.

  • The client is always right. A cliché, yes, but still 100,000% true. Unless you’re collaborating with someone or you’re part of a band, duo or whatever, you’re providing a service. It’s the artist’s song, not yours. It’s their name on the artwork and their legacy. (You’re lucky if you get mentioned in the credits on Spotify. And no one reads those.) If you work in an auto repair shop and someone walks in and requests a new slime green paint job, you’d never paint it candy apple red and argue with the client for not understanding the vision. You could suggest that the red might fit the electric blue seats better, but it’s still the client’s car and they have the final say. (Ok, I’ll drop the car analogy. I don’t even have a license.) In audio land, you can always request not to be credited.
  • Become an expert message decoder. Probably the most important skill I’ve acquired this year. Knowing what a client means and how to then execute on that is essential in any field, but even more relevant in a world where words are ill equipped to describe sound. Examples include: bigger drums = roomier and more compressed (shorter release time); more echo on the vocals = sometimes more delay, most of the times more reverb; louder vocals= more compression or top end on the vocals. But once you get used to the lingo, you realize that most will mean the same thing when using these directions. The key factor here is over-communication. Ask what they mean until you actually get it — it’s your job to understand what they want and to deliver. And if all else fails, ask for references. Sometimes you’ll be surprised what their request actually referred to!

That’s it folks! I’m sure there’re some that I’m forgetting now, but I’ll make sure to make another post should I remember any other gems!

Thanks for reading!