The first assignment in the ‘Recording Techniques’ module for undergrads on my course was to produce a live recording with no post-production, otherwise known as a ‘stereo recording’. This means that we had to find a musician, performer or group to record all at once, in full stereo, with no layering of performances later and with no effects or plugins added after the fact. The only post-production required would be to pan the microphones to their appropriate places before bouncing the audio.
This task was assigned as a test of microphone choice and placement as well as choice and use of recording environment. The implication being that we must choose a place with a nice reverb which is suited to the subject we’re recording. For Example, a choir or string quartet would sound fantastic in a large church, a piano might sound at home in a concert hall and a drum kit might suit a hall with a nice decay.
After being assigned the task, I immediately went to every small local gig I could and approached performers to record them. With many falling through, I was lucky to be introduced to singer-songwriter John Barton just in the nick of time. He agreed to perform a song of his own for me though he didn’t own an acoustic guitar and neither did I, at the time. This lead to the weird combination of singer-electric guitarist, recorded in a small local church near my campus.
Alongside the recording we were asked to include a brief report detailing in part, a description of the piece, space and mic setup and why we chose to record in this way. Here is my report which, alongside the recording, was awarded 72%, a first.
My stereo recording features John Barton, a student at the university and member of the band society, performing an acoustic-style version of his own original song ‘7 Hour Flight’. This was recorded at the Heslington Church with John both singing and playing the electric guitar at the same time. This was recorded through an interface directly to a MacBook running Logic Pro.The final bounce consists of one take, with part of the song cut from the middle to bring it to target length, with the final length being 2:45.
Reasons for choices made
John was chosen to be the subject of my recording mainly for ease and simplicity. Having recorded vocalists and guitars before, I thought that a singer-guitarist would produce a good result with little hassle and John happened to introduce himself via the York band society. The venue was difficult as I have relatively little knowledge of local venues. The church is a building I walk past almost every day and, knowing that churches have irregular geometry, high ceilings and very hard surfaces, seeing it I thought it would have very pleasant acoustics and dense reverb. Overall, I chose practicality and availability over trying to seek out a specific sound. The performer and venue were readily available so I tried to use various techniques to make this pairing work as well as possible.
The main hall in Heslington Church is that of a medium-sized church featuring high, arched ceilings with hard walls and floors. The walls and floor absorb very little of the sound, giving a dense reverb with a long decay time. There are also two slightly raised platforms in the hall which are carpeted. The only access to power in the hall comes from sockets near one of these carpeted sections, meaning that the performer must stand on one of these platforms in order to use a guitar amplifier. Unfortunately this meant placing the performer in a corner next to a large glass door which could produce more early reflections to microphones near it, potentially unbalancing the recording. It did however mean that the performer was stood in a slightly more enclosed spot, giving a slightly smaller reverb than if he were stood in the center of the hall, perhaps making it more appropriate to a singer-songwriter style.
The Recording Setup
I chose to record with three Rode NT2As, used as a vocal spot mic and a spaced pair, and an AKG C414 as a guitar amp spot mic. I also took a Rode NT4 with me to experiment with but found I preferred the spaced pair.
Originally, the spaced pair microphones were placed around 6ft away from one another and 6ft back from the performer. However, between takes, the mics were gradually moved closer to the performer to pick up more of the source and closer to one another to reduce the stereo width causing a ‘hole in the middle’. Ultimately, the microphones ended up about 5ft from one another and 4ft from the source. These mics were set to omnidirectional with no filter or pad, unlike the vocal and guitar spot mics which were set to a cardioid pattern.
The Vocal spot mic was used very much traditionally, being placed with the capsule at the singer’s height and him singing around 1”-2” back, though he did move around a lot during takes. To reduce spill from his guitar (a telecaster recorded acoustically is surprisingly loud) the mic was angled slightly upwards between takes. In the absence of a pop-filter, this also slightly reduced plosives, along with a piece of foam from the mic bag. The C414 used as a guitar amplifier spot mic was placed slightly off-axis around 1 foot back from the speaker cone, roughly halfway between the center and edge, with the mic angled towards the center of the speaker. The mic was later moved closer to the amplifier to give more of the direct sound and moved slightly above the speaker with the mic angled down to further reduce spill.
The Recording Process
After setting up, warming up and allowing time to pack down, we had enough time for 7 whole takes from our 3 hour session, with a few more false starts. The later takes feature less and less room sound and acoustic guitar string sound as microphones are moved and angles away. Since we came without a pop shield, these takes feature many more plosives too. As you might be able to make out above, a small fabric bag and a piece of foam were placed over the vocal spot mic to reduce this.
The singer is also performing as far to his left as he could, as the large glass surface previously mentioned was just out of shot to his right. By getting him to perform further to his left, early reflections into the left microphone from the surface were decreased. Another problem was trying to get John to remain still while recording. Unfortunately, noise from him moving can be heard on the final recording and his voice can be heard moving slightly across the stereo field.
One more technique I tried to get the guitar sounding its best and sitting well in the mix was to use the built- in FX from the amp. Since we couldn’t use plugins in post, I used these FX to add a little bit of compression to help it sit well against the vocal and a little bit of chorus to subtly add a more interesting texture to the sound. The amp’s 3-band EQ was also used to try and shape the tone to remove low end around 160Hz and remove as much of the room mode we found there from the recording as possible.
Song choice was also an issue. Since I had not recorded John before, I wanted the choice of song to be entirely his own so that he was at his most comfortable and would hopefully give his best performance. Unfortunately, I had forgotten that our submissions must be within 2:30 to 3:30 and each take lasted around 4 minutes. This had to be cut down in post-production.
In terms of post production, my spaced pair mics were panned hard left and right with the levels being equal to each other and high enough so that the loudest point of the song is as close to 0db without clipping, though as spot mics had to be brought up these mics were brought down to keep the overall level this way. The vocal mic was panned centrally, to try and anchor the vocal to the center, and slowly brought up such that the effect could be heard but the individual mic wouldn’t really be noticed as independent to the spaced pair by the listener. I found this to be about -5db.
Since the guitar amplifier didn’t move, the spot mic’s main purpose was to broaden the guitar sound. I set the level of that mic to around -14db.
There were only two cuts in the track, all from the same take, just to remove a chunk from the middle and get the length down. The cuts are at 44 seconds and 1:12 and all come from the same take, as shown below. Sections from the whole take which were not used in the final cut can be seen in grey.
Overall, I’m not amazed with the recording but for my first time recoding in such a way, I’m pleased it’s come out as it has. For a piece that was performed entirely freeform with no click track, I definitely like how smooth the edits are and the plosives are much less audible than the earlier takes.
Recording a singer with a guitar kept me close to my comfort zone which made the process easier but the use of stereo in my recording is very limited and fairly boring. Given the chance to do it again, I would rather record a pianist and be able to hear individual notes move from left to right. If I were recording another singer-guitarist in the same way, I would definitely make more of an effort to source an acoustic guitar. Recording the electric guitar with the amplifier did give me some flexibility in using FX on input but the room mics pick up an awful lot of the acoustic sound of the guitar, causing a lot of popping and clicking type sounds which is especially audible in the intro.
Having experimented during the session with microphone placement and eventually finding a balance of direct to room sound I’m happy with, I would also record more takes with the same setup. I was changing mic positions after almost every take which gave me a lot of variety but meant that most takes couldn’t be cut together without being glaringly obvious.
It was an interesting experience and I’d very much like to try it again after learning more about the process.