Mixing Bass | Media
Avoid all the low-frequency pitfalls and learn to achieve the perfect foundation for any mix, with our bass-mixing masterclass...
his page includes additional resources to accompany this month’s ‘Mixing Bass’ feature (www.soundonsound.com/sos/sep12/articles/mixing-bass.htm
). First up are some links to multitrack practice projects, and further down you’ll find a number of audio examples.
Multitrack Practice Projects
Firstly, here are some links to full multitrack sessions which you can download free for practice purposes — I’ve categorised them according to the type of bass instrument featured, and have also included a few notes about the nature of the mixing challenges they present.
Uncle Dad: ‘Who I Am’
A full-bodied single-mic recording this, although lacking in midrange definition and with rather obtrusive mechanical noises such as string snap/rattle. It appears to have been compressed during recording, but a pronounced sub-100Hz resonance has prevented the dynamics processing for solidifying the instrument’s position in the balance very well.
Jesper Buhl: ‘What Is This Thing Called Love’
This single-mic bass recording forms part of a live jazz session by the Jesper Buhl Trio. The part’s wide range, both in terms of pitch and dynamics, requires sensitive dynamics and automation work in general, but you may also require some more specific trouble-shooting processing to tackle a resonance around 100Hz (which occasionally interferes with the evenness of the line) and tame some over-emphatic ‘snap’ from the player’s plucking action. Processing choices may be restricted, however, by the level of spill from the other instruments. This recording was featured in SOS April 2008’s Mix Rescue column, so check that out if you’re looking for some tips.
Will Knox: ‘Cog In The Machine’
This bass mic recording has been captured with very little spill, despite it being part of a live session, although if you want more midrange cut-through then the spill from the kit in particular may become a consideration. The biggest trouble-shooting issue is that the low-end levels are rather inconsistent, and multi-band compression would certainly be on my list if I were mixing this myself, as well as general compression and automation to control the inherently wide dynamics of the instrument.
James May: ‘If You Say’
This multitrack gives you an opportunity to explore the mechanics of dealing with a dual-miked upright bass. Although the polarity/phase match between the two mics is already pretty good, there may be bonus points available if you tweak this further. The main task as regards balancing this instrument in the mix will be to control the rather uneven tone and performance. Some notes have masses of low end, others very little; some notes ring on beautifully, while others fail to sustain properly; and overall there’s rather too much dynamic variation even for an acoustic song like this.
Telefunken Java Jive Session: ‘Song 2’
The second of Telefunken’s Java Jive sessions features another dual-miked upright bass part, although this time with considerable spill from the nearby drumkit. The low-end feels rather overblown to me, with not very much going on in the midrange, and the spill ties your hands a bit if you try to push instrument’s upper spectrum for better small-speaker audibility. This is one situation where a specialised bass enhancer would probably be helpful, by generating more midrange energy based on the instrument’s low frequencies — although you’d be well advised to attend to a few sub-100Hz resonance issues first.
ELECTRIC BASS (SINGLE TRACK):
Leaf: ‘Come Around’
This bass recording dates from SOS February 2007’s Mix Rescue remix, and although it’s musically played, it’s nonetheless rather hissy as well as having a fair amount of fret noise — automation may be required to deal with the latter in particular. The cab’s low-end roll off also robs some of the lowest notes of weight, and it doesn’t in general translate very well to small speakers in its raw form.
The Doppler Shift: ‘Atrophy’
This miked-amp recording is fairly well-controlled dynamically, but is rather too rich in low midrange for some of the arrangement’s flamboyant rock textures, so extra midrange cut-through is likely to be required --however, the fret slides and string rattles will probably need remedial work in that case. In addition, there’s a fairly strong resonance around 100Hz which suits some pitches a lot better than others. In general, expect to do some multing here, so that the bass adapts to fit the big arrangement variations.
Johnny Lokke: ‘Promises & Lies’
This classic metal track has a perfectly respectable miked-amp bass recording, which doesn’t actually need a while lot of processing. The difficulty in this mix is carving enough room out of the other parts for it to fit in comfortably — as I discussed in the SOS January 2011 Mix Rescue column.
Arise: ‘Run Run Run’
Here’s a Reggae production with a single miked-amp bass track that you’ll want to make the most of, given the style. However, you may find yourself hampered a little by the background hiss and lack of midrange information, even though the LF is pretty well-behaved. Low-end information from other tracks also needs to be carefully controlled if you’re going to maximise the bass power in the sub-100Hz zone. For some pointers, check out the Mix Rescue column in SOS November 2011, which deconstructs my own remix of this song.
Moosmusic: ‘Big Dummy Shake’
Although miked-amp recordings can cause problems, DIs can also give rise to mixing headaches, as you can hear in this Mix Rescue multitrack from SOS March 2009. In this case the DI is the only bass track, and has heaps of uncontrolled sub-bass energy on it. Even if you can contain that, though, the dynamics are also rather inconsistent, and there’s a big change in the mid-section sound which will almost certainly require multing.
ELECTRIC BASS (MULTITRACK):
Banned From The Zoo: ‘Turn On Me’
Here you get a nice even DI tone as well as a miked track which provides some amplifier character. The polarity between the two signals appears matched, but you can get a stronger and more detailed midrange if you’re willing to tweak their relative timing as well. The amp recording has pronounced resonances which make mincemeat of the low-end consistency, so this is a case where I’d probably high-pass filter that and rely on the DI’s LF instead.
Jokers, Jacks & Kings:’Sea Of Leaves’
Here you get a bright, overdriven DI signal which provides powerful midrange cut-through, as well as a lovely solid miked-cab signal. The latter, however, rolls off a little at the low end, so you may struggle to keep the power in the chorus if you rely too heavily on that. Getting this indie-rock production’s long-term mix dynamics right is quite taxing, so be prepared to mult the bass if necessary to deal with this. You can read about my remix of this track in SOS April 2012’s Mix Rescue column.
Phre The Eon: ‘Everybody’s Falling Apart’
This is another Mix Rescue project (from SOS August 2010), and features an unusually percussive bass part. DI and miked-amp tracks are provided, and while the former is reasonably workable (as long as you can control the transients with your dynamics processing), it’s very tricky to make any decent use of the latter — it’s very woolly-sounding and doesn’t add anything very attractive to the combined tone. As such, I ditched it and reamped the DI in its place, but that doesn’t mean that you have to do the same...
Traffic Experiment: ‘Once More (With Feeling)’
A very involved and large-scale arrangement, this, so you’re unlikely to reach a successful conclusion mixing it unless you get to grips with the possibilities of multing and automation. The lack of midrange energy here may cause problems if you want to bring out the nice melodic elements of the part, but even if you choose a less upfront role for the instrument you’ll still have to grapple with rather inconsistent levels in the sub-100Hz area, both on the DI and miked-amp tracks.
Young Griffo: ‘Facade’
Here you get two different miked-amp signals to mix or choose between. In addition, there’s the instrument’s well-controlled DI feed, the only niggle with that being some considerable string-slap and fret noise, so careful with piling on high-end EQ boost in an attempt to drill a hole through the massed guitars. Polarity and phase are already quite well matched between the three signals, but a bit of finessing may achieve some small improvements — or alternatively you might choose to abandon the idea of phase-match in search of a dramatic new tone. You can read what I did with these tracks in SOS October 2011’s Mix Rescue article.
This Mix Rescue project from SOS March 2010 features four separate bass tracks, so you’ll need to keep your wits about you if you’re to avoid all the potential pitfalls I talk about in this month’s bass-mixing feature. The main pair of tracks (DI feed and miked amp) are both out of polarity and out od phase, so make sure to deal with that straight away. There’s also lots of fret buzz on the DI to trouble-shoot. The third track is an additional distorted amp layer for the song’s choruses — again, be careful about phase. Finally, a synth during the song’s middle section needs careful handling if it’s not to outshine the bass guitar with its super-solid low end and lacerating HF filter sizzle.
Side Effects: ‘Sing With Me’
You get plenty of high-mid projection from the synth’s strong resonant peak, so the main job here is balancing the synth’s low-end contributions against the kick — both instrument’s play on their own some of the time, so it makes to share out the subjective LF weight between them somehow, but without overwhelming the mix when they play together. I discussed this process a good deal in my SOS February 2010 Mix Rescue article, where I remixed this very song.
AM Contra: ‘Heart Peripheral’
The main stereo synth-bass patch used here has serious level and phase mismatches in the sub-100Hz zone which render it very inconsistent on single-subwoofer systems. Whatever you do to address these concerns (and I’d probably just replace the low end with an additional sub-bass synth layer myself), you’ll need good LF extension on your monitoring to hear what you’re doing — and a decent spectrum analyser may also help.
Girls Under Glass: ‘We Feel Alright’
Here’s another stereo synth with level and phase problems in the bottom octaves, but in addition there’s some kind of sporadic resonance below 100Hz that adds a further unwelcome layer of inconsistency. Even once you’ve addressed those issues, there are still three separate drum loops to fit in with the bass part around, so you’ll need to think quite carefully about which areas of the LF spectrum are best used for which sounds. It was partly for this reason that, when I remixed this track for SOS November 2010’s Mix Rescue column, I decided to mult all three drum loops to gain more control over their kick sounds.
ANiMAL: ‘Easy Tiger’
This UK hip-hop production stands or falls by its bass, which constitutes one of its biggest hooks. There’s plenty of low end on the main stereo synth part here, but there’s also a dedicated mono sub-bass synth track too. Balancing these two together to avoid phse-cancellation inconsistencies is part of the job, but you’ll also likely need to do some multing if you’ve going to get the best out of the long-term arrangement dynamics.
Triviul: Dorothy (MONO SYNTH+MONO SYNTH+SAMPLE PIZZ)
Here’s a track where the main bass part is meant to combine with kick, so they always play together. As such, the bass sound is dependent on the kick, and vice versa. An additional synth layer will complicate matters if you don’t keep its low end in check, and the key-change for the second verse may invalidate some of your EQ and/or balance settings if you’re not careful. Other full-range harp, organ, and synth sounds have the potential to obscure the bass projection too, and the fourth bass track — some sampled pizzicato double-basses — needs careful treatment to prevent it sounding out of place against all the synth parts.
These audio example files demonstrate some of the issues and methods that I talk about in the main article. The captions below give some detailed information about each.
In this file you can hear a bass sound that’s a combination of raw DI and miked-amp bass tracks. The low midrange is suffering badly from phase-cancellation, and unevennesses at the low end are making musical line lumpy.
Switching the polarity of the DI signal makes a big difference to the combined bass timbre, filling out the low midrange and reining in some of the apparent bumps in the musical line. Compare this file side-by-side with the PolarityOutTimingOut example and switch quickly between them to really hear the difference.
Another big tonal change is brought about in this demonstration by delaying the DI signal by roughly a half a millisecond — the midrange becomes more solid such that the timbre carries better through a busy mix. Comparing this file with the initial PolarityOutTimingOut example shows quite how crucial polarity and phase issues can be when it comes to bass sounds.
Here’s a good example of a raw acoustic-guitar recording that has masses of subsonic information that’s of no use whatsoever at mixdown — unless you need your subwoofer to double as a fan! High-pass filtering this at 50Hz all but removes these headroom-munching waveform excursions, and you could afford to advance the filter frequency further within a busy mix too in order to make sure that the low octaves of the spectrum are left to the proper bass instruments.
This mix snippet has a low mid-range conflict between the main bass part and its organ rhythms. As a result, the overall mix tone ends up sounding slightly clouded, and the bass doesn’t come through the mix quite as clearly as it might.
Equalising the organ parts in the DeclutterEQOut example to reduce the degree to which they obscure the bass sound clears up the mix tone appreciably, and also seems to bring the bass part a bit closer to the listener. To hear the differences most starkly, line up the two audio files in your DAW system so that you can switch between them while they’re playing.
Although this funk-rock bass sound displays a fairly muscular attitude from a tonal perspective, it’s not actually that heavy in terms of real sub-100Hz power. Boosting in that region with EQ, however, won’t really improve matters, simply because there’s precious little useful recorded energy down there. One way to add this frequency information artificially is to layer in a MIDI-triggered sub-bass synth patch.
Here’s a simple synth patch designed to add some extra low-octave power to the bass-guitar part in the SubSynthOut audio file above. The patch uses a single square-wave oscillator in combination with low-pass filtering to clear away all but the fundamental frequency and early harmonics. As you can hear, it’s pretty boring-sounding on its own, but that’s as it should be, because it’s meant to add low-end power without subordinating the timbral character of the bass guitar.
This example combines the SubSynthOut guitar recording with the sub-bass synthesizer you heard the SubSynthSolo audio file. Compare this file closely with SubSynthOut and notice that the addition of the sub-bass synthesizer doesn’t really make the guitar sound like a synth at all — in fact, most musicians would be hard-pressed to spot the synth in its owh right at all.
Here’s a section of a mix where the bass didn’t project as much as I’d have liked it to in the small-speaker-friendly midrange zone — particularly during the riffing at 0:06-0:12 and 0:18-0:24. Boosting with EQ in the midrange didn’t achieve anything very palatable, unfortunately, so I decided to layer in a MIDI instrument to provide some additional support.
If you compare this example with the MIDILayerOut audio file, you can hear how adding in a MIDI clavinet part improved the small-speaker translation of this line. It’s worth noting, however, that this addition required its own amp simulation, heavy EQ, flanging, and reverb to melt it well enough into the mix so that it didn’t seriously undermine the character of the underlying bass tone.
In this heavy rock texture, there’s a sense that the bass is a bit detached from the guitars, as well as lacking some warmth. One solution to this is to overdub additional guitar parts to double the bass line, connecting its part more with the other guitars.
These guitar overdubs have been specially tailored to support the bass line without sounding like extra instruments in their own right. This involved restricting them to the bass part’s pitches, editing them strictly to avoid them showing in gaps between bass notes, and EQ’ing them to hide within the frequency shadow of the bass and other guitars. Not a very musical sound on its own, but in context (see the OverdubLayerIn example file) it does its job very well.
This example shows the result of combining the OverdubLayerOut and OverdubLayerSolo audio files — no obvious increase in the apparent number of guitarists in the band, but a definite thickening and ‘gluing’ of the bass tone into the overall guitar texture.
Short attack and release times can have a variety of (potentially) unwanted side-effects on bass, as the following assertively-compressed audio example demonstrates. To start off with I’ve used attack and release time settings of 0.3ms and 70ms respectively, but these are gradually shortened throughout the file, until they reach final settings of 1 microsecond and 1ms respectively. The main things to listen for at the dulling of each note onset’s attack phase and for the progressively increasing distortion as the compressor begins altering the shape of individual waveform peaks.
Here’s a problem that many low-budget productions suffer from: inconsistent low-freuqency levels. Notice, for instance, how the note at 0:07 feels rather thin in the lower octaves, whereas the notes at 0:12-0:13 have much more power, so they jump out at you. Normal compression or EQ don’t deal with this kind of thing very well, so multi-band processing is usually a better bet.
In order to bring a bit more consistency to the low end of the bass part in the LFMultiCompOut audio file, for this example I’ve compressed the sub-100Hz frequency region at a ratio of 4:1, applying around 12dB of gain reduction with attack and release times of 15ms and 74ms respectively. Although the unevenness is by no means completely cured, it’s certainly significantly reduced, and without losing the sampled acoustic bass’s nice attack thump.
Here’s a section of a mix which uses triggered ducking to reduce the bass level by a couple of decibels on each kick-drum hit. The virtues of this are that it allows the kick-drum more clarity and weight, but also reduces the amount of mix headroom required when the bass and kick-drum levels combine. Compare this with the DuckingOut audio file to hear the difference this makes.
For comparison with the DuckingIn file — here I’ve bypassed the ducking action, so that you can hear the brief low-end build-up caused whenever the kick and bass occur together.
In this mix situation you can hear how the subjective bass balance isn’t quite consistent, such that the bass line feels like it’s not holding its position solidly in the balance — even though significant compression is already in action. Furthermore, the inconsistencies aren’t really in tune with the needs of the song structure. The start of the chorus, for example, feels rather undercooked.
Applying some fader automation to the bass part you heard in the AutomationOut audio file locks down the track’s balance in a subtle but appreciable way, and helps the song’s longer-term dynamics feel more inevitable, as you can hear in this audio example. 0