I’ve been particularly bad at blogging this semester and I don’t know why.
I had grand ideas for my project but I simply ran out of time trying to record my CME project. Fortunately I learned a fair bit about garageband and recording techniques during this venture.
The project in which I engaged in for this task was the recording of an original song called little bird. I recorded this song in garageband using A Washburn acoustic guitar model D10SLH a single Rhode NT2 microphone through a Scarlet 2i2 USB audio interface. I also experimented with a JAME’S ACOUSTIC PICKUP GUITAR MIKE)
Nearly everyone I had spoken to had assured me that this would be as simple as plugging and playing.
It was not.
The first issue I encountered actually took me some time to resolve. The first few sounds I recorded, when played back through my headphones were for a lack of a better word, crunchy. The only way I could see to alleviate this was to turn the gain down on the input and to reduce the output on the microphone. This would seem like a logical solution. There were 2 issues here, first the signal wasn’t loud enough for me to hear clearly in the playback and second, I wasn’t able to hear the metronome.
I did some research and found out that this is called “clipping”. “Clipping” is when the speaker is unable to handle the audio signal so instead of overloading the hardware (the speaker cones in this case), it caps the level which distorts the signal creating that “crunchy” sound. Generally this is considered not good and for my purposes I have also considered it not good. The solution suggested is to change the level at which the signal is being amplified so the hardware can cope. Generally signals are not amplified so as to leave “headspace” before the signal is “clipped”. Upon learning this I decided that It was entirely possible that the headphones I had been using were either faulty or not calibrated to deal with the signal that my 2i2 was giving. Whether or not my slowly aging Marleys were the cause of this problem or not, I am not sure but I can be sure that replacing the headphones that I was using has solved this issue.
The second major issue I encountered was playing in time.
The obvious solution to this is to use “flexi-time”; the analogue version of quantising. While this is certainly better than nothing, I noticed a few characteristics of this that changed the recording in an undesirable way. First off I noticed a tendency for the attack of notes to be slightly late upon inspecting the signal the received signal. Whether or not this was me simply playing out of time or latency between the sound being picked up by the microphone and it being encoded on to my slow laptop, I don’t know for sure. I did however realise that “flexi-time” tended to favor the loudest part of the sound making that the center of its movement. While in a strictly accurate sense this is correct, this looses the subtlety of guitar sound; you loose part of the sound of the strum and get the sound of the chord instead. Some clever kid working for the development team of garageband has already figured this out so there is a little bit of compensation but the ideal sound would still be the one from doing a single take. Frankly that’s more pressure then I can take right now.
To offset this slightly I thought the most convenient way would be to record a bar or 2 of silence both before and after each take.
A really good reason for this is because the very first sound of that part of the recording coincides with the first beat, therefore if any sound is made before that then it is simply not recorded. This could be as simple as hearing a breath before a singer begins.
Allowing sound after each take, (especially when using the multiple takes function) can also prevent this same problem.
I also found this technique very useful for playing over mistakes. My only concern when doing this is ensuring that the levels remain constant during this time as well. This can be altered by a variety of things ranging from distance from the microphone
I did eventually give up on trying to ensure that each second of sound was precisely in time as this sacrificed the feel I was going for, it was still none the less fun to experiment.
The next issue that I encountered was using auto tune.
I had two issues with this: Firstly, the guitar I am using is not set up for open tunings, nor do I get it serviced by a professional regularly. As such it is not the most perfectly intoned guitar, its not bad but during this process I’ve had to learn how to address this issue. Turns out the neck was slightly warped and as a result I’ve had to adjust the truss rod to compensate for this.
Secondly my time playing trombone in various ensembles especially quartets has left me with a preference for Pythagorean chords over equal temperament. I was unable to find a solution to this problem for the guitar part and have consequentially had to compensate the tuning best I can. The result doesn’t bother me too much, it feels more authentic
The next problem I encountered was to do with adding the percussive guitar part. Up until this point I had been using the contact microphones I had been borrowing from James which produced a great sound but unfortunately I found them (like many other aspects of recording) fiddly and frustrating. I could easily have spent hours trying to find the optimum configuration and unfortunately I just never got there. The
I decided the best alternative would be to play the percussion part with the guitar on my lap. This produced a vastly different sound then what I could usually produce while playing and decided against it. However what this brought to light was the sound of the open strings that were unsettled and had begun to ring during this time. I really liked the sound but it was not the sound that I wanted in my recording. I may use this for a future recording…
The solution to this was actually quite simple however. I wrapped a bandanna around the strings in order to stop them from ringing during the percussion guitar takes. This helped quite a lot because it helped to take the stress off having to play both the percussion and guitar part for part of it.
So there you have it.
I learned stuff I diddn’t know before and hopefully next time when I record it wont be nearly half as stressful.