I've been mastering music for over 20 years now, and in that time I've worked on hundreds of albums in just about every musical genre you can think of - and probably some that you can't. I've heard thousands of music mixes under studio conditions, and surprisingly few have been perfect. But that's OK - because mastering is about ironing out imperfections, as well as adding the final polish.
But sometimes I get tracks where it's difficult or even impossible to get a great result due to mistakes or shortcomings in the mix. In these cases I'll recommend a remix to the client in order to sort out the problem. Sometimes that's not possible, and I have to go ahead and try to use all the tricks at my disposal to get the best result I can, but it's not an ideal solution.
So my aim here is to point out the Top 10 most common Mix Mistakes - and tell you how you can avoid making them in your mixes, because better mixes mean better masters!
Before we get started, let’s get some terminology out of the way.
When I’m talking about ‘organic’ recordings, I mean the more traditional ‘band’ recordings where more than one person has been playing at the same time, in a studio or home studio, and recording real instruments and vocals through microphones and DI boxes. If I refer to 'EDM (electronic dance music)' or ‘programmed’ recordings, then I’m talking about mixes that are mostly created ‘in-the-box’ with loops, samples, virtual instruments, etc., using a microphone only for vocals. Of course in reality it's not one or the other - between the two extemes there's a whole spectrum of overlapping going on with different genres borrowing different elements and techniques from others.
I’ll be mentioning ‘Mid/Side processing’ later on, so I’ll just explain how it works. Normal Stereo audio is split into two channels, Left and Right. But it’s also possible to convert it to M/S, where Mid equals the sum of the L & R channels, and Side is the difference between the L & R.
What that means in practice is the Mid channel contains, to a greater degree, things that are panned into the centre of the stereo image (often the lead vocal, kick drum, snare and bass). Whereas Side contains, to a greater degree, things panned to one side or the other (eg. backing vocals, drum overheads & toms, guitar parts, stereo keyboards, etc.). The Mid and Side signals can be processed independently, then recombined back to Left & Right.
If I talk about ‘rules’ I guess I’m really talking about conventions - the way things are usually done. The conventions can vary according to where you are on the 'Organic <---> Programmed' spectrum that I mentioned above. Generally speaking, the more 'Organic' the style is, the more you might go for a natural sounding mix. Whereas at the far end of the 'Programmed' scale, you can do crazier things with the sound because it's not natural and was never meant to be.
The rules and conventions are there to be broken, and that's how music evolves. It's a creative process where you are free to experiment all you like in order to achieve the result you want. Even so, it’s good to know what the rules are before you start breaking them. As Banksy is fond of telling us, he learnt to draw before he started doing his street art. Which is probably why his stuff is so much better than most of the others.
So here we go...
OK, I’m cheating a bit here because this is about what happens before you even get to the mix.
I’ve had people phone me up about mastering their tracks, worried about technical stuff, like should they send me the mix at 96 or 192K; asking me will I be EQing through vintage Pultec EQs, and which dithering algorithm I’ll be using to create the final 16-bit master, because someone said on a forum somewhere that these things are of vital importance.
Then when I get their tracks I find they’re playing out of tune, they’re not tight rhythmically, and the singer can’t sing! OK, I'm exaggerating a bit, but the message is clear...
- Get your priorities right - sort out the basics instead of worrying about stuff that’s going to make very little difference to the end result. Sort out any issues of tuning or timing or bad performance before you start mixing - or preferably before you start recording!
It’s a good feeling when you think you’ve got your vocal level sitting just right in the mix. But remember, you may have only got it right for that particular point in the song! I’ve heard a lot of tracks where the lead vocal is fine in the verse, then it gets buried in the chorus because there’s more other stuff going on in the track at that point.
Of course, it’s not just about vocals - it might be a guitar solo where the level has been kept at the same level as when it was playing the rhythm chords - instead of being pushed louder in order to highlight the solo.
That’s what I mean by ‘set and forget’, where you’ve set a level and you leave it there for the whole track. With today’s DAW-based recording, there’s no excuse for it. Mixes can be instantly recalled and tweaked, and level automation is fairly simple. So it's a trial and error approach, really. Tweak then listen - until every element is at the right level - for every section of the song!
One more point about mixing a lead vocal. Remember, you know the words already, but someone hearing your track for the first time doesn’t! You may think the level’s OK, but it may be too low for someone who doesn’t already know the words to pick them out. Get a second opinion from someone who’s unfamiliar with the track.
You could try these listening tips to help get it right quicker...
- Get a pair of crappy radio-type speakers in your studio specifically for checking mix balance. If you can’t do that, try going out of the studio, leave the door open and listen from the next room. Get your ears tuned in to listening for anything that leaps out of the mix too much or should be pushed louder.
- Listen to your mix on a few different systems in different environments (like in the car).
- Do it with fresh ears - make sure every element is at the right level - for every section of the song.
This carries on from No.2. With beginner mixers it’s often the lead vocal that’s too loud, but it can apply to anything in the mix. Whatever, if it’s only slightly too loud, then the compression that is often used in the Mastering process can go a long way to solving the problem.
Sometimes I can use Mid/Side processing to get a result. Like once I was working on an album where the first track featured guitar solos which were mixed too loud - in fact they very nearly obliterated the rest of the instruments! On this occasion I was lucky - one of the solos was panned extreme left and the other extreme right. Using processing that split the audio into Mid and Side, I was able to tame the solos by lowering the level of ‘Side’ in relation to the ‘Mid’. The result wasn’t ideal, but it was better than it was before. But on other occasions we’re not so lucky and a remix might be the only answer, so get it right - follow the advice in No. 2 above.
Also, there’s another type of “too loud” problem where one of the elements of the mix eats up all the headroom of the mix, even though it may not actually sound outrageously loud in the mix. In programmed hip-hop or electronic dance music it’s often the kick drum, or in band recordings it can often be the snare. Often I can spot this in the waveform - the body of the wave form will look thin and spindly relative to the peaks, which will be sticking out close to the digital zero,
I recenty worked on a track where every snare hit was creating a peak in the waveform . One of the objectives of mastering that track was to get it louder as a whole to match the other tracks I’d already mastered. But I was running into the problem that my normal dynamic processing could only go so far because that snare was eating up my available headroom, and the track was sounding ‘thinner’ than the others as a result. In the end I was able to sort it out, but it took extra time trying out various strategies, and involved using more processing than I would normally want to use.
- If you see something that's making your main stereo output meters go wild when you're setting up the mix, try searching the individual channels of the multitrack to locate offending peaks - that would be the snare or drum overheads in my example. You can then apply some compression or limiting to the channel(s) so that the peaks are under control before they get to the mix.
This crops up quite a lot during mastering and it’s a classic beginner error. Imagine a situation where the guitar is too bright, but the lead vocal is too dull and needs brightening up. That’s an example of what I call ‘EQ Mismatch’.
Because I’m dealing with the stereo mix I can’t treat the two things separately (unless I get lucky and I can use Mid/Side processing). Also, guitar and vocals occupy a similar range in the frequency spectrum. So anything I do in this case to improve one is likely to have a detrimental effect on the other.
Under those circumstances, I’d suggest a remix, but if that’s not possible I just have to use my judgement to get the best result.
- So check those mixes like I said before in No. 2. Try and get things sounding similar in regard to the Dull/Bright spectrum.
When I first started in recording, a studio would probably have one source of reverb, a spring or (in a posh studio) a plate. The big London studios might also have an echo chamber. Later on there would be digital delay and reverb in the form of rack-mount outboard effects. Even then, the likelihood was that a mix would have only one variety of overall reverb, which could be applied to each channel on the mixer to a greater or lesser degree via an aux send. Any special reverb effect on a specific instrument would have to be printed to tape during the recording.
These days with DAW-based plug-in processing it’s possible to have a different reverb setting on every track in the mix. Possible, but not a good idea - too many conflicting reverbs can make it difficult to achieve a coherent sense of space.
- If it’s an ‘organic’ recording and the goal is to make it sound like a performance taking place within a specific type of space, then it’s probably best to have just one overall reverb sound as an aux send/return that establishes the ambience of the performance, plus maybe only one other reverb at any one time for spot effects - like on a guitar solo or a certain vocal section. But like I said, rules are meant to be broken!
OK, imagine recording a synth keyboard with a R/L stereo output, and one of the output leads has a jack that's been wired back to front - meaning one side will be 180 degrees out of phase with the other. The average listener might not notice it when it’s one element in a busy stereo mix. I’ll notice it, because to me it’ll sound like the keyboard has a stereo image that is unnaturally wide with a huge ‘hole’ in the centre. Switch the monitoring into mono, however, and the keyboard will almost disappear as the two out-of-phase signals try to cancel each other out. And even the average listener will notice that!
That’s an extreme example of an out-of-phase element in a mix, where the two sides of a stereo track are 180 degrees out of phase - and to be honest I’ve only heard that a few times in my mastering career.
Less extreme cases of this can be caused by overuse of a ‘stereo width enhancement’ effect. The normal stereo image field is more or less between your speakers. If you’re using this effect on a stereo track and you're hearing the sound coming from points wide outside the two speakers, then, in addition to sounding unnatural, it’ll probably be subject to cancellation if played back in mono.
Less extreme still, but more common, are cases of ‘comb-filter’ cancellation caused by time-delay differences in multi-mic setups, or ‘mic plus DI’ setups. Let’s say you have a bass guitar being recorded direct with a DI box plus a mic picking up the speaker sound. The mic sound is going to be delayed from the DI sound by a certain number of milleseconds - according to how far away the mic is from the speaker. A similar thing can happen with a multi-mic setup where the mics are at different distances from the sound source.
- To stop it happening you can experiment with mic positions to get the best sound before you record. Or if you’ve already recorded, you can use a phase-adjustment plug-in on your DAW which allows you to adjust the phase and/or timing of one signal relative to another to get the best sound (e.g. Voxengo PHA-979).
- If you have a ‘mono’ switch in the master output section of your hardware or software (in-the-box) mixer, use it to check for out-of-phase elements causing phase cancellation.
A classic mistake of the inexperienced mixer is not realising that you can place sounds within the stereo image using the ‘pan’ controls on the channel of your hardware or software mixer. It doen’t happen often, but in my time I have received a good few mixes where everything’s squashed into the centre with zero width!
On the other hand, it can sound strange if things are panned extreme right or left. I had a track recently where the lead vocal had been double-tracked all the way through and mixed with one panned hard left and the other hard right. To me it just created a disturbing psychological effect that was at odds with the happy disco vibe of the song!
Basically, the idea is to create an illusion of space and separation between the different instruments by means of ‘panning’ them spatially to different positions within the stereo image. Additional careful use of Reverb can reinforce this idea of hearing a performance taking place in a particular space, even though it might really be made up of different things recorded at different times, possibly even in different places.
- The usual rule is to keep Kick, Snare, Bass & Lead Vocal in the centre, and pan anything else to a greater or lesser degree. But rules were made to be broken, right?
- However if the recording is going to be cut for vinyl, for technical reasons you should absolutely definitely keep Kick & Bass in the centre.
A great mix shouldn't start at one level and just stay there till the end. I'm not just talking about volume level - I'm talking about the density of instrumentation and the levels of excitement, interest, intensity and emotion. Where are you going to go if you pack your mix with a kitchen-sinkload of stuff right from the off?
It's all about building up impact for the listener and retaining their attention. You don't have to use everything you've recorded, and you don't have to use it all at once. Let it build!
It's always been part of the classic pop song formula, but it applies to most genres. For example, contemporary EDM producers are great at it, because they know from their DJing how to work the audience. Whatever your musical bag, your music has to be interesting - even when excitement isn't the main aim!
- Start sparse right from the beginning then build, or maybe bang in on the intro and then bring it right down for the first verse so you can build up from there. Build in waves so each chorus has more impact than the verse before it. then break it down so you can build it up again to end on a climax!
- And let's suppose your music is not about sheer excitement - keep things interesting and surprise the listener with new sounds, new parts, or different effects - like the cliche 'filtering the vocal on the breakdown section'. Go mad, use your imagination, then have a listen the next day to make sure it all still works and makes musical sense!
I have actually said that to clients. You may be thinking, “OK, so how can a mix be too good?”
Once I was doing a Trial Track so the client could make sure I was the person for the job, and the track in question was well recorded - the individual instruments sounded great, their levels were well controlled, and the balance between them was perfect. Trouble was, it just sounded a bit bland!
I’ve always approached my Mastering from the musical viewpoint, rather than the technical. So I notice things about dynamics, the ebb and flow of a track, and how it’s affected by the relationships of the different sections of a track to each other.
In this case, I encouraged my client to go back and look at how he could inject dynamics into the track, let it build up as it went along, and allow a few ‘jagged edges’ to protrude out of the mix instead of keeping everything under tight control.
Taking that on board, he revisited the mix for that track, I mastered it up for him, and he was more than pleased with the exciting result - and went on to apply the same philosophy to mixing the rest of the tracks. A few tweaks further down the line and we had a great album!
- Recording your tracks can be a solitary experience, especially if you’re doing it in your home studio. It’s always good to get a fresh perspective from an experienced Mastering Engineer who can see the big picture. If you need it, I'm always happy to listen to a mix and give my feedback.
We're talking about digital clipping here. It sounds nasty and is very difficult to remove once it's there. It can happen to an individual sound if you overload the channel A/D converters when you record it, and it can also happen to a mix if you overload the master stereo output of your DAW mixer when you're doing the mix.
Recently I was sent a dance music track where every kick-drum beat was accompanied by nasty distorted crackling. He'd obviously not been checking the meters on the master output whe he did the mix, and every kick beat was enough to overload it.
The thing is, he may not even have heard it at the time, even if the meters were going in the red! That's because if your DAW is doing its internal processing at 32 or 64-bits, it may be able to store and play back the overloaded data without distorting. But when you bounce it down to 16-bit 44.1K and play it back it can sound horrible, So this client should have noticed the problem if he'd checked the mix file before sending it to me.
But digital clipping can be quite easy to avoid...
- Don't overload the A/D converters when you're recording into a multitrack channel of your DAW. Allow some margin and keep an eye on the meter to make sure they're not going into overload.
- Don't overload the master stereo output of your DAW mixer when you're doing the mix. Start the mix with the individual track faders somewhere in the middle, and the master output faders at the zero position - that's the default position about two-thirds up (not all the way down!). Then if you find your perfectly crafted mix is overloading on the master output meters, you can fine-tune the master faders down so that the mix is peaking just below digital zero.
- When you've bounced the mix down to a WAV or AIFF file, always check the file before sending it off for mastering. It's amazing how many people fail to do this!
True to my policy of always over-delivering, my Top 10 goes up to 11. So I'm going to finish by giving you the dos and don'ts of sending your material for mastering.
- First of all never mix straight to MP3 or AAC format - they use data compression which will damage the audio quality of the mix you've worked so hard on.
- Always mix to a full-resolution uncompressed file format, I recommend WAV or AIFF files in any of these formats - 16 or 24-bit resolution, with 44.1K, 48K, 88.2K, or 96K sample rates.
- Don't mix with compression or limiting over your master stereo output buss. Leave that to the mastering engineer. Mastering engineers hate getting mixes that someone has already tried to master! If you really want to get an idea of what it might sound like with some dynamic processing, then by all means do a treated version of the mix for yourself, but send the untreated version for mastering.
- Audio files are huge - way too big to email, so use a service like WeTransfer, YouSendIt or DropBox to send the files. Or if you prefer snail mail, you can always burn the files onto a CD-R or DVD-R, or copy them onto a USB stick.
- Make sure to keep lots of backup copies of your mixes. If you want MP3 or AAC files you can always make them by converting from one of these backups.
- I can also handle mixes on older formats like Audio CD, DAT and MiniDisc. I've even got analog Reel to Reel and Cassette.
There are some good 'Mix Mistakes' lists worth looking at out there, and I've chosen three of them below. There are others, but be careful - I watched one YouTube vid by a dance music producer which was fine, but he was suggesting ways of working that might be OK when you're mixing EDM, but are absolutely NOT OK for other types of music - so I left that one out because I'm trying to give you info that can apply to most types of music.
Here are my favourites out of what I've found online...
- Sound On Sound magazine did an excellent feature by Mike Senior based on his experiences of judging a mix contest back in 2011...
- SOS 'Mix Mistakes' feature
Thanks for reading. I hope the information here has been useful and will help you in mixing your own music!