Original piano composition: “Brand New Day“, 23 Apr 2006. Length of recording: 2min 24sec. MP3 file (565KB).
NOTE: Audio output is rather soft. The recording sounds better with headphones or via an amp. The volume for this piece increases progressively, so watch your ear-drums or speakers! 🙂
How it was done:
(click on the image to see additional notes)
This is what my “home music studio” looks like. Basically you need (1) a musical instrument, (2) a microphone, and (3) a computer/ software to record and mix. No expensive equipment required (the piano and computer hardware notwithstanding). The microphone was part of a headset. The software used for recording and mixing was freeware from audacity.sourceforge.net.
The microphone was placed at the top of the piano. I started playing a few notes to test the recording volume. In this case, I set it at around 0.4 DB (according to the Audacity mic input control).
Afternote: Upon hindsight, it might’ve been better to record at the highest possible input level setting and then adjusting the DB setting afterwards. After exporting to MP3, I realised the output volume was too soft. I went back to the source recordings and boosted the DB levels but it came out distorted. Overall, the output volume is still my “archilles heel”.
The finished piece comprises of about 5 to 6 separate recordings (main tracks and also fills) which were later pieced together. It’s very much like the Layering technique in Photoshop — you just put one track over (or after, rather) the previous one, taking into account the start and end points.
This was a boon to an amateur player like me, who tended to make mistakes very often (I probably drove the neighbours crazy playing the same parts over and over…) Without the ability to selectively copy and save the tracks, one mistake in the recording would meant that I had to start over again — imagine the frustration when you hit one wrong key in an otherwise perfect piece.
However, recording in bite-sized tracks also presented some challenges. This being a one-person operation, I had to repeat the process of record, play, check recording playback, and do it again… so much so that my tempo got a haywire (i.e. some tracks were played in faster tempos, which meant I had to play them again).
To make the layers blend better, I used Fade In and Face Out effects for selected tracks (otherwise the tracks might sound like they ended very abruptly).
Click on the image to see notes on what I mean by layers and channels.
In brief, this was what I did:
- Recorded in Mono
- Copied the Mono track to another two tracks, and set them as “Left” and “Right” channels respectively to get the stereo effect.
Some interesting recording techniques discovered during the experiment:
- I learnt that overlaying the Stereo tracks with the Mono track created a fuller sound, as compared to only the Stereo tracks.
- This technique was discovered by chance — if you time either the left or right channel to start a fraction of a second slower than the other, it creates a surround-sound effect.
- If you’re like me, not a technically-skilled player, where all your repeated tracks sound more or less the same, you can break the monotony by playing the track as a separate layer with variation in stereo effects or adjusting the DB setting which gives it a slightly different sound. Kinda like what Mike Oldfield did for Tubular Bells.
Total time taken = approx. 7 hours
I’d say it was a Sunday well spent. Of course I realise the final output isn’t something to shout about. It won’t win any Grammies, that’s for sure. But it was my first time trying out the mixing and layering techniques. Also, considering that this was, by my standards, a major composition piece that’s recorded and mixed together in less than a day, I’m pretty pleased with myself.