From the Microphone to Finished Media

In the lead up to my end of year showcase I am in the process of finalising compositions, recording them and producing them ready for release. In this post I am going to go through the process of creating my work step by step.

Once the composition is finished, the first step in turning that into music that can be listened to and enjoyed, is to complete the score. Often I will have composed music just by scoring it or scored it as I have gone along bu this doesn't mean that the score is finished. Small details will often be left out my scores as I work on them and some there were notated might need to be removed. Additionally, scoring the music in an appropriate way is important. In the case of the piece I will be taking about today, I needed a score somewhere between a classical recital score and a condensed folk score, as strumming every rhythm like a robot on a solo mandolin piece would be boring to listen to and close to impossible to either notate or play, yet the more intricate sections need to be notated exactly.

Once I have a completed score, this needs to be printed out for practice and learning. In the case of this composition, I needed to practice the more technically difficult sections to a point where I could perform them clearly, however being 8 minutes long, it would be un-necessary for me to be able to recite the entire piece without the score for recording when I can read the score as I play.

When I am able to play the piece confidently and articulately, and as I think it sounds good, I move straight to recording. Recording a Mandolin is something that took me quite a while to get right. Recording any instrument is something that often comes with practice as they all require unique attention in terms on mic placement. Often the mandolin is recorded with a single mic pointed at one of the f holes of the instrument, however it is more often than not an accompaniment, and not a solo instrument. Hence I decided to use a slightly more complicated (albeit still very simple) setup with 2 mics, to capture the instrument in stereo sound. Being quite a small instrument, I didn't want this to be absurdly exaggerated, so I set up my Rode NT-1 (large diaphragm condenser) close to the f hole of the instrument to capture the rich warm tones of the instrument, since larger diaphragm mics are often suited to warmer and softer tones. I also set up one of my Rode NT-5 mics closer to the fingerboard to catch more of the intricacies of the performance such as the sounds of my fingers on the fingerboard as well as more of the overtones of the instrument. This setup can be seen at the head of the article.

Once I have recorded the music I begin working in my DAW: Studio One 3 Pro. There are a number of jobs to do here. First and foremost I have to stitch together the best version of the recording. For a piece of that length, it is far quicker to record in sections and spend 5 minutes fiddling round with audio events and cross fades to put a few bits of audio together than it is to keep trying until you nail 8 minutes of music flawlessly.

The next job is to adjust the levels in each section of the music and create a bounce of the chosen recordings of each section in one audio event. This is what I then work on to create the final mix. One problem with home recording is that you are never going to have the sound isolation of a professional studio, so I first clean up the audio with a noise reduction plugin. This works through something called a Fourier analysis of the signal which essentially isolates all the notes and their overtones in the signal and reduces the volume of all the bits in between.

I then decide if the track needs to an EQ, which gives control of the volume of different parts of the signal. For this mix, I did several things with the EQ. Firstly, I applied a low cut, and lowered the frequencies below the instrument's note range as each note or chord can sound boomy otherwise, I didn't want to loose this completely as it is the only instrument and hearing that articulation is in my opinion, appropriate for a solo recording, however this would be much more aggressive in a multi-instrumental mix. I created a small boost to the frequencies near the top of the instrument's range as these are quieter. Lastly I applied a high frequency shelf, which boosts the high overtone frequencies of the instrument to make it sound brighter. This visualisation of the EQ shows how the track is well balanced within our audible range.

I also used a reverb to fill out the sound a bit more, this helps the music feel, for want of a better term, more alive, and like it is in a normal space, as opposed to a recording setup where the aim is to capture no sound other than that straight off of the instrument. A big problem with mixing the mandolin is it's aggressive decay within it's typical ADSR envolope, i.e. once the note is played, the sound dies off very quickly. A way around this is to apply compression, but this can be easily overdone and is more suited to mixing pop music than classical music where the aim is to hear the detail of the instrument. However in want of not overdoing the compressor, I used a limiter plugin which prevents peaks from clipping but increases the volume of the entire signal equally.

Once the music is mixed I work on some artwork for the piece. Keeping in style with my current work, I create a multiple exposure style image with the following parts:

1: A picture of the score

2: The instrument(s) it is for

3: A photograph that for some reason, I think suits the piece.

In the case of this project, the images I selected were the same as the image of the score above, this image of the tailpiece on my mandolin:

And this image taken by a reservoir in Jersey:

These are edited together in the software "GIMP" or GNU Image Manipulation Program, which is a fantastically good quality open source program much like Photoshop, except it is completely free!

A text layer is then added with the title of the piece and my name. Lastly I create a version of the artwork in the centre of a 1920 by 1080 black rectangle to use as the image in the video form of the piece.

I then setup the video version of the piece in the world's most intuitive software, Blender, an open source 3D modelling, animation and video editing program that can do just about anything you are smart enough to make it do, however exporting in the right format on your first time will require a tutorial or two unless you are a member of Mensa.

Whilst the video is rendering, I also upload the audio file to my SoundCloud and fill in the required information, and in the case of unreleased music like this, set it to private. Once the video is exported, it gets uploaded to my YouTube where I either publish it straight away or schedule it for release at the desired date and time.

I will be releasing this piece as part of my AMATA showcase on the 10th of may. Tickets are available at at https://amata.ticketsolve.com/shows