Top Production Mistakes: Part 1

Stop Repeating the Same Mistakes.

In my experience running Berklee's Electronica Club (soon to be called the Electronic Production Club), I've noticed many recurring mistakes producers make. In the past, I would often redirect them to the internet for help. However, I have recently been surprised by the lack of accurate online information. Many essential tips and tricks are buried deep in the pages of blogs and forums, beyond the reach of people looking for general directions that will significantly impact their tracks. With this in mind, I've compiled a list of the top 3 mistakes I hear when listening to the demos people send me.  

1) Lacking Basic EQ "Reflexes"

Here, I am not talking about specific sound processing techniques but general mix hygiene techniques that should become as automatic as putting your hands out when you trip. This is why I call them EQ "reflexes."  

- Low Cutting (almost) Everything

Without a sub, an EDM track has no real presence in a club or festival environment. These sub-frequencies create the physical vibrations the audience feels and reacts to. If you were to insert a spectrum analyzer on many different tracks, you'd realize that - in most cases - the sound you're processing contains frequency content in the 20 Hz - 100 Hz range. If you're looking at the sub or kick, that's fine, but for any other sound (hats, synths, effect sweeps), these frequencies are not perceivable or fundamental and do nothing but muddy up the low-end. Make sure you high-pass/low-cut these sounds with a four-pole (or steeper) filter. Or more than one… How can you be sure when to low-cut and when not to? If unsure, place a low pass on the channel and isolate these low frequencies. Now that you hear them, are they essential to the sound? Solo it with the subchannel. Are they interfering? Here's a frequency analysis of a high hat with a high pass around 500 Hz. As you can see, there is no longer energy present below 100 Hz.

This is the same sample without the EQ. As you can see, although it is quiet, there is some frequency content below 100 Hz. You wouldn't expect to see this because it's tough to hear. Nevertheless, this content is present in most sounds. Accumulation of low-frequency content will interfere with the sub and take energy away from it.

Here is the original and high-passed version. You'll notice that cutting away these low frequencies has little to no perceivable effect on the isolated sound.  

~Original

~High-passed

People are generally aware of how important it is to low-cut most channels. However, I rarely hear about the importance of doing the same thing at the other end of the spectrum. In an electronic production, you'll probably have white noise or a cymbal layer in the high end (10 - 20 kHz). There is no denying that these frequencies are essential to many sounds. However, leave them untouched; the overall mix result will be sharp and potentially jarring. Take a four-pole or steeper high-cut/low-pass filter and lower the cutoff frequency until you feel you are removing characteristic, important frequency content. Then back off a little. Solo your white noise layer/channel. Then, see if the initial sound still sounds too thin or if the cut enables the white noise layer to appear better.  

- Problem Frequencies

We've talked about treating the sound at each extremity of the frequency spectrum, but there are also things you can do to clean it up from the inside. Take an EQ band with a bell curve and add a significant gain at a high "Q." "Then move the bell around until you hear a strange resonance or area that seems out of place. Compare completely removing the problem band with a notch setting to simply lowering the band's gain and decide whether the removal or attenuation is more suitable. Then, switch the band on and off (A-B) with the sound playing to see if the change is favorable.  

2) Too Much Mud!

- More Dry

In production lingo, 'm'd' 'usually refers to unintended, unwanted interference in the low-mid and low-end of your mix, making it seem crowded and damp. A common source of mud is the wet effect tails from reverb or delay that are difficult to distinguish and, thus, mold with different sounds, taking the focus away from the frequencies that need to be there. Before applying reverb or delay, imagine what you expect it to do. Do you want a long ambient tail? Do you want to add somebody with a short decay? Once you've subconsciously figured that out, see if you have space in your mix to make that wet signal audible. Then, listen to all the elements while adjusting the dry/damp and find the driest setting at which the effect is apparent and meliorative to the track.  

- Return Tracks

Another way to apply these effects more precisely is to use return tracks. Return tracks enable you to treat the wet signal exclusively. This means you can EQ your reverb tails to eliminate any unnecessary frequency content. It also means you can physically place your wet signal in a different part of the stereo field than the dry signal using panning or other stereo modeling tools. This can help it stand out without muddying up the original sound. This image compares an untouched reverb tail and an EQed one. You can see from the bottom line that the EQ has removed a lot of frequency content in the highs and lows. Return tracks enable you to do this and use reverb more effectively.  

Listening to the lead synth with the reverb tail EQ'ed and un-EQ'ed, you'll notice how much low end and high-end frequency content it is adding. This should be a red flag. In context, those frequencies would interfere with the sub and muddying up the cymbals.

~Synth with EQ

Note that I'm using an excessive dry/wet setting so that you can obviously hear the wet signal.  

- Gate

A final mud-reducing precaution would be gate sounds so they only play when needed. Often, recorded tracks will have noise between intervals of the sound you're trying to capture. For vocals, this noise could be feet shuffling, breaths, or traffic going by outside, and although it may be very quiet, it adds unnecessary frequency content to your track, mud. (See blog: Noise Reduction: Gates)  

3) Under-Mixed Drums

Creating the drums for a track takes a lot of critical listening and patience. The individual hits have to sound compact by themselves and organic as a whole, and they also have to say good with other elements. Under-mixed drums that don't take a central place in the mix are typical among learning producers and take a little work to solve.  

- Sidechain

Sidechain compression is the main reason drums sound so powerful in relation to the other elements in the best EDM mixes. Basically, this type of compression causes sounds to duck (drop in level) when a sidechain signal (key) is triggered. This sidechain signal can be the output of any track in SoundBridge. In my example, I use a ghost key - an external sidechain source whose output you don't hear, like a control signal. It is usually very short, at unity gain, but muted. The compressors it is routed to often have strong compression settings to ensure enough room for the drums. The sidechain signal should be present on every kick and snare hit.

As you can see from this image, my sidechain signal is a white noise clip about as short as the drum transients. I've aligned it with the kick and snare to compress the synths at the right moment. The advantage of using a ghost-key instead of the drums themselves is that, with a ghost-key, the ducking will still happen even if the drums aren't playing, creating that cliché pumping effect.

On the channel that I would want to duck, I would place a compressor. It will receive the control signal from my sidechain channel and instantaneously compress the entire output signal.

~Here is the track with the lead getting side-chained

~And without the sidechain

Your kick and snare will likely have crucial harmonic frequency content around 100 Hz and 200 Hz, respectively. Try busing (grouping) similar channels like synths and basses and EQing them with notches at these respective frequencies. You'll create a gap in the spectrum that will only be satisfied by the characteristic frequencies of your main drum hits, creating substantial space for them.  

- Keep Them Dry

Your drums have to remain pretty dry to stay punchy. Ideally, it would be best to have no reverb on the low end. Nevertheless, it can be placed on snare, hat, or cymbal layers but is applied in moderation. Think about it. If you want your snare to have a splashy white noise tail, you're better off layering a white noise splash rather than lengthening it with a reverb tail.