Home » Blog » Making ‘Mongolian Death Worm’: Recording the Music

Making ‘Mongolian Death Worm’: Recording the Music

[13-minute read]
Watch the full musical here:

 

N.B.: For optimal listening pleasure, it is recommended you listen to the audio examples through a good pair of headphones.

 
It may have escaped your notice (because I’ve been keeping very quiet about it) but there’s a new musical in town, and it’s really really great. It’s like Les Misérables but less miserable, like Little Shop of Horrors but bigger and more horrifying, like Priscilla Queen of the Desert but even more desert-based… It’s called ‘Mongolian Death Worm: A Puppet Show Musical’, and it tells the epic story of one man’s quest for truth, justice and meat in a world where even the food is made of cardboard. The steaks have truly never been so high.

As everyone knows, there are two main parts to every puppet show musical: the puppet show, and the musical. Today I would like to shed some light on the latter, and the recording process in particular. How do you take a bunch of songs, all worked out on paper, and bring them to life as fully-formed audio recordings?
 

The Plan

As discussed in my previous post, ‘Making Mongolian Death Worm: The Orchestrations’, my co-conspirator James Ure and I had devised the show’s instrumentation in such a way that we would be able to record most of it ourselves. As well as doing all the vocals, between us we would also record the woodwinds, ukuleles and tuba. As we were living in the same house, it was a simple matter to get together and collaborate on those bits.

This left the keyboard, bass guitar and drums, which we would have to outsource to other specialists. Fortunately, we have friends, and were able to enlist the quite remarkable trio of Dominic Lo, Ben Sarney and Joseph Zbos to help out with each of these parts respectively. They would record their bits in their own homes, then send us the results to combine with everything else.
 

Mock-Ups and Click Tracks

Those unfamiliar with remote recording would be justified in asking “How, when all five band members are recording at different times, in completely different parts of the country, can it all fit together? Surely it would just end up a mess!” Indeed this would be true, if it weren’t for mock-ups and click tracks.

Mock-ups are exactly what you’d expect: rough recordings produced in the computer, to give an idea of how the music will eventually sound. The musicians (and the drummer) can listen to these through headphones and play along while recording their instrument, ensuring that they play in the correct tempo and style. Producing these mock-ups was very easy in this case, because there is a facility for exporting audio files from MuseScore, the software that I used when orchestrating the music. Incidentally, if anyone out there is looking for a set of good quality, realistic instrument sounds to use in MuseScore (or any other program that uses soundfonts) I would highly recommend HQ Orchestral Soundfont. It admittedly has one or two quirks – for instance, the tuba sound is very quiet and the accordion seems to be a semitone out of tune – but overall does an excellent job of mimicking the sounds of real instruments. And what’s more, it’s absolutely free!

Click tracks are also quite self-explanatory, but I’ll explain them anyway. These provide a mysterious ticking noise, that either ‘tick’s or ‘tock’s on every single beat of the music. This ensures that the musicians stick rigidly to the appropriate tempo, and is especially useful during tempo changes or sections when the beat is not audible in the mock-up audio (for example, during long held notes).

Below you can hear a mock-up and a click track in action, in an excerpt from ‘Say Hello To Yellow’.
 

‘Say Hello To Yellow’, mock-up (with click track) vs. finished recording. Note how the click enables the musicians to stay together through the tempo increase.

 

Recording Equipment

It might be of interest to someone to know precisely what equipment was used to record this musical masterpiece, and while I can’t tell you what the other musicians used for their parts, I’m happy to divulge the secrets of my own home recording setup.

For recording individual instruments in the comfort of my bedroom, I used my trusty Blue Yeti USB microphone. This has been my go-to mic for many years – in fact, my only mic until recently – and is just really good at everything. It is also very convenient to use, as you only need to plonk it on your desk and plug it into a USB port and you’re away. This was the approach I used for recording all of the woodwinds, as well as the ukuleles.
 

“Hello, I’m the Blue Yeti microphone, but you can just call me HAL.”
(or alternatively: “I can call you Yeti, and Yeti, when you call me you can call me HAL.”)

 
For the more complex recording tasks (namely the tuba, which needed a long cable and a microphone suspended high up above the instrument, and the vocals, which often required the use of two mics simultaneously) I moved down to the infinitely less comfortable basement, and dug out my Behringer C-2 small-diaphragm condenser mics and Behringer U-PHORIA UMC404HD USB interface. I bought this equipment originally for recording ‘It’s Not Really the Apocalypse – A New Musical’ (available here) last year and have found it incredibly good value for money. Certainly, if you have deep pockets you may be able to get slightly better sound out of more advanced equipment, but the Behringers really do do a great job and are ideal for anyone looking to obtain a high-quality recording setup on a budget.

All this gubbins was of course plugged into my 11-year-old Acer laptop – still going strong after all these years*. The ‘digital audio workstation’ (or ‘DAW’) software that I used for recording and editing all the music was a program called ‘Ardour‘. It’s a little off-the-beaten-track admittedly, but it’s very good and just so happens to be Linux-compatible, which is crucial for me.
 

DIY basement recording studio, pt. 1.

 
Other than that, the only bits of equipment worth mentioning I think are my Sennheiser HD280 headphones that I use for mixing and mastering, and my miraculous homemade pop shields, which I constructed from a selection of wire coathangers and nylon tights, and which are there to stop vocalists forcing too much air into the microphone when pronouncing certain syllables.
 

DIY basement recording studio, pt. 2.

 

The Editing Process

Once the recording sessions are complete, you end up with a whole bunch of audio files, recorded by different people in different places at different times. But how can you take all those individual components and amalgamate them into a coherent whole, so it sounds like a single band performing to the listener in their own home? This is where the process of editing, mixing and mastering comes in.

Before I continue, I have a little disclaimer: Although I am quite well-versed in the basics of audio editing and have a little experience in music production, I still consider myself very much a novice. So if there are any experts reading, I apologise in advance for any inaccuracies, omissions, oversimplifications or downright bad techniques to be found herein!

With that out of the way, let me introduce to you my workstation:
 

The Mongolian Death Worm project file in Ardour 6.

 
This is the interface of Ardour 6 – fairly conventional for a DAW – and as you can see I have loaded up the project containing all of the music for Mongolian Death Worm. It consists of a large number of horizontal ‘tracks’, each containing recordings of a different instrument or character. Each of these tracks can be edited and manipulated individually, and then they can be played back in various combinations to create the sound of a full ensemble.

Altogether, Mongolian Death Worm uses ~62 tracks, as follows:
 

  • Click track (1 track): Used while recording and synchronising instrumental parts.
  • Mock-ups (5 tracks): Used while recording.
  • Chorus (16 tracks): For the 4-part harmonies in ‘The Ballad of the Mongolian Death Worm’ and ‘Finale’ we wanted to mimic the sound of a decent-sized choir. So James and I each recorded all four parts twice, giving a total of 16 voices singing together. I also used these tracks for various other bits of ensemble singing (e.g. the ‘All Beasts Long and Venomous’ section in the ‘Finale’).
  • Barbershop quartet (4 tracks): These I used for the quartet that appears towards the end of the ‘Finale’. There are 4 different vocal parts, but only one voice per part.
  • Characters (11 tracks): Every different character in the musical has their own dedicated track. Some of these – such as Roy’s and Sheriff’s – contain a lot of audio. Others – such as ‘Lips’ – not so much.
  • Reeds (7 tracks): Each woodwind instrument has its own track: piccolo, flute, clarinet, bass clarinet, alto saxophone, tenor saxophone and jaw harp.
  • Keyboard (1 track): Although this instrument switches between several different sounds, I only gave it 1 track. How ungenerous.
  • Bass guitar (1 track): This one is self-explanatory I think.
  • Tuba (1 track): This one, also, is self-explanatory.
  • Ukuleles (2 tracks): This musical uses both ukulele and banjolele, and I gave 1 track to each.
  • Drums (8 tracks): To best capture the sound of a drum kit, it is standard practice to record it using multiple microphones placed all around the instruments. Sadly, however, this does create more work for the poor producer.
  • Pan (1 track): It may only appear briefly for a solo in ‘Every Hero Needs A Villain’, but the pan is an instrument too – well, sort of – and it deserves its own track. Interestingly, the pan solo is the one piece of audio that we recorded live on set, using a microphone placed just out of shot. This was to ensure that the sound would match up exactly with the Witch Doctor’s exuberant on-screen antics.
  • Phone ringtone (1 track): When Sheriff’s mobile phone rings during ‘Steakout’, it gets its own track all to itself.
  • Miscellaneous (3 tracks): During the editing process, I occasionally tidied things up by taking several completed tracks and combining them into one. For example, when I was happy with the sound of the chorus I condensed it down onto 1 track, to save on space and computing power.

Tidying Up
With all of the music in one place, albeit distributed over a number of tracks, the next step is a bit of tidying. This involves listening through each individual sound clip and cutting out extra noises such as creaking chairs, coughs, splutters, grunts, swear words and farts. These are mainly present at the starts and ends of clips, when a musician is either getting ready to record or complaining about how badly they’ve just played.

A related task is that of sorting through multiple takes of a given section of music to select the best one. It is generally best to record at least two takes of everything, so that you at least have a backup if one later proves unusable. Often this process can involve combining sections of several different takes to produce a sort of hybrid. For example, if I record a clarinet part for a song and perform it flawlessly except for one bar – which I got right on the previous take – then it makes sense for me to just edit in that section, rather than waste time re-recording the entire thing for the sake of one small detail.

With all these various instruments and takes floating about, it’s easy to see how quickly the workload multiplies. If I’m working on a 4-minute song which uses 5 instruments and 1 vocalist, that could amount to 24 minutes of audio, all of which has to be checked individually. If I have 2 takes of everything, that time doubles to 48 minutes. So it can take the best part of an hour merely to listen to all the recordings. Now imagine how long it takes to work through a 40-minute album!

Fortunately, with our remote keyboard, bass guitar and percussion musicians, they did much of their own tidying and sent over everything as polished single takes. One minor issue that I did have to attend to though, was a slight background hum in the bass guitar recording for ‘World Without Sand’, presumably generated by an over-enthusiastic amplifier. This was easily fixed using the ‘Noise Reduction’ effect in that ever-popular free audio software ‘Audacity‘, resulting in a crystal-clear clip that would surely make anyone proud.
 

Removing background hum from the bass guitar recording in ‘World Without Sand’, using Audacity’s ‘Noise Reduction’ plugin.

 
Applying Effects
With all the individual bits of audio sounding clean and tidy, surely the next step is to start putting them together, right? WRONG! There is still work to be done on the individual tracks before they can be considered fit to combine!

I am referring here to the various effects that can be applied to each track, to subtly (or not so subtly) alter how they sound. There is no limit to the number of weird and wacky noises you can make in a computer using effects, however for this project my aim was a little more understated. Rather than changing the quality of the music, I merely wanted to help it sound its best while keeping it as natural and acoustic as possible. The equivalent of a tasteful frame round an artwork, or a nice coat of varnish on a wooden thing. Thus, you shouldn’t really notice these effects in the finished product, but all the while they are working away in the background to help bring out the vibrancy of the music.

To give you an idea of what exactly these effects do, I will now detail the workings of my favourites, with some examples to demonstrate how they sound…
 

  • Reverb: This provides the familiar ‘echoing’ effect that occurs when you make a noise in a room with lots of hard surfaces for the sound to bounce off. Anyone who’s ever sung in the shower will know how reverb can add depth and warmth to a sound, and indeed it is important in preventing a recording feeling flat and artificial. The other thing reverb does is add a sense of distance. No reverb is equivalent to someone playing an instrument directly into your ear, whereas lots of reverb makes the instrument sound distant (when most of the sound you hear comes from reflections rather than the instrument itself). Pretty much every track in this project has some degree of reverb added to it.
     

    An excerpt from ‘Every Hero Needs A Villain’, with and without added reverb.

     

  • Compression: This effect serves to smooth out the volume of a track, making the louds quieter and the quiets louder. If over-used it can make things sound a bit robotic by sucking out the dynamic variation, but in moderation it can help improve a track’s clarity and smooth over any rough edges. I have mainly used compression on the vocals, which tend to jump around a lot volume-wise with different verbal sounds, but I have also used it on some of the drum tracks where it can really bring out the characteristic sounds of the kit.
     

    Sheriff’s yodeling solo from ‘Steakout’, with compression to smooth out the volume.

     

  • Equalisation (EQ): This effect will be familiar to many, as it is frequently available on good quality hi-fi systems, digital audio players and car stereos. It enables you to manipulate the volumes of different frequencies in a sound. So if you want to make something sound thin and tinny like an old portable radio, you’ll want to use mainly high frequencies and cut out the low ones. To make something sound dull and muddy, remove the high frequencies and accentuate the low. Incidentally, this is how I achieved the muffled sound of the Witch Doctor shouting ‘I object!’ in the ‘Finale’: By using a ‘low-pass filter’ to remove all the high frequencies from the sound.

    Like compression, EQ can help to bring out the sound of certain drums like the tom-toms and bass. I also employed it on the keyboard and bass guitar tracks, which initially sounded a bit thick and muddy, to make them more crisp, clear and lively.
     

    Using EQ to improve the clarity of the keyboard and bass guitar (in an excerpt from ‘Untitled Worm Song (Part I)’), and to muffle the Witch Doctor’s voice (in the ‘Finale’).

Setting Volumes
With all the individual tracks now sounding clean, tidy, natural, smooth, vibrant and crisp, surely the next step is to start putting them together, right? RIGHT!

This is when you finally get to start playing around with volumes, as you start mixing the tracks to make a clear, natural-sounding blend. The aim is to ensure that nothing is too loud and nothing is too quiet, which seems a simple idea but is actually a bit more complex than it may appear.

For a start, what is meant by ‘too loud’ or ‘too quiet’? Does this mean that every instrument should be equally audible all the time? Perhaps, but not necessarily. Depending on the arrangement, the relative volumes of the instruments may be constantly changing as the musical material is passed between them. If one instrument has a solo then it should really play a bit louder to be heard above the accompaniment, which should be allowed to recede slightly. At another point, an instrument may be playing a background role which is only meant to be barely audible when combined with the rest of the group. Even a single instrument can be inconsistent in terms of volume – the flute, for example, tends to be much louder and more piercing in its high registers compared to its low, where it is naturally quieter. My point, then, is that for every instrument to be equally audible all the time is not necessarily a natural-sounding solution, nor is it particularly musical.

For Mongolian Death Worm, my approach was to find a volume for each track such that they are all at a similar level on average and blend well when playing together. But I haven’t attempted to iron out dynamic variations within each track, so each instrument is allowed to rise and fall in volume, and as such the precise blend varies as different instruments come to the fore. Not only does this lend a natural and musical sound in my opinion, but crucially it reduces the workload as I only have to set the volume for each track once, rather than adjusting it constantly over the course of the whole musical!

Sadly, for the vocals the solution isn’t quite as simple. These actually do need to be heard clearly throughout, but can’t be too loud or they’ll stick out and sound exposed. It is therefore important in this case to go through each song individually, adjusting the vocal levels so that they rise and fall appropriately with the accompaniment.

Before I move on to the final stage of the mixing process, I’d just like to mention one more factor that can make setting volumes a royal pain. That is the remarkable propensity of the ears (in conjunction with their partner-in-crime, the brain) for playing tricks on you. Ears are very good at filtering out noise that is not of interest and focussing attention on that that is, which is hugely beneficial in everyday life but not so much when mixing music. I find that if I’m listening specifically to one instrument to check its volume, it naturally sounds a bit louder because that is where my attention is directed. So it’s tempting to turn it down, only to find that the next day, when I listen with a ‘fresh pair of ears’, so to speak, I’ve made it far too quiet. It seems the human brain really can be a bit too smart for its own good.

Panning
Now that everything is nicely balanced, there is one more stage of editing yet to complete: panning! And no, this doesn’t have anything to do with the frying pan that we used as an instrument at one point in the show. Rather, it’s all about spreading the music out across the listener’s left and right ears. This gives it a real three-dimensional feel, and makes it sound like you’re sitting right in the middle of the ensemble as they’re performing. I’ve often read that if the music sounds good in mono (where left and right speakers play exactly the same thing) then it will sound great in stereo (with panning added).

I have taken quite a simple approach to panning in Mongolian Death Worm, again to give a natural sound while minimising the amount of work I have to do. Rather than detail it all verbally, I think this would be the perfect place for a nice little diagram:
 

Artist’s impression of the stereo panning employed in this magnificent project.

 
Note that the drums are spread out over the whole stereo field, so different parts of the kit sound in different places. The choral and barbershop parts are also spread out widely, to give that expansive epic feel so crucial to a good puppet show.
 

An excerpt from ‘World Without Sand’, demonstrating the use of panning.

 
And that, ladies and gentlemen, is just about it. All that remains is to export each song as its own 2-track stereo file, tweak the volumes, add a little silence on the end, and you have your completed album.
 

Concluding Remarks

I hope you have taken immense pleasure from this glimpse into the murky world of music recording. For anyone thinking of making their own puppet show musical based around a mythical creature, perhaps it has given you some clues as to how to go about it. For those who are just curious, maybe you now have a deeper appreciation of the endless layers of complexity lurking beneath what may appear to just be a daft YouTube video.
 
Whoever you are, and wherever you may be, I only wish that your humps grow tall and your feet grow strong, until we meet again!
 
* Sadly, since writing this article, my trusty Acer laptop has succumbed to old age and had to be replaced. : ( May it rest in peace, one last victim of the Mongolian Death Worm.

Leave a Reply