Project Reflection
Wedding Song
For my next project, my services were elicited by a friend Brendan who has taken the bold step and tied the knot with his now lovely wife, Kate.
He wanted to record an original song he had written for her and give it to her on the day of the wedding.
I was very excited to help out and I booked the studio and locked in a date.
He would be recording an acoustic guitar and his vocals. I wanted to capture the essence of what the song was; intimate and whispery, so I did some research as to best match the microphones with the instruments.
Eventually I settled for a Royer 121 ribbon microphone; I had used it before on vocals and the quality was very warm and thought that might help with the intimacy side of it. However, I wanted to use it on the acoustic guitar to try capture that same warmth and get the definition of the strumming/picking.
My mic choice for his vocals was the AKG C414. Again a microphone I have used many times before; its a great mic and always does what you need it to.
The set-up was easy and following in the footsteps of Bobby Owsinksi, I used the extra time I had from not having to set up millions of instruments to step into the room with Brendan and experiment with the placement of the Royer. I put on a set of headphones, had him play and moved the position of the microphone until I had the sweet spot, aimed just right of the bridge. Back in the control room, I high-passed some of the extreme low frequencies — it sounded too boomy and bass heavy, certainly as a result of the proximity effect. I had to really get close to his guitar with the Royer to capture a good signal. This makes sense I guess, as the Royer can tolerate high SPL levels.
As I’ve found with singer/songwriters, there was no known tempo to the track but after some experimentation we settled for 134 bpm. We eventually recorded to a click track as this would help me later in editing.
We spent a considerable amount of time on the guitar takes, before a quick turnaround to begin setting up the vocal recording.
I turned to the C414 here and the results were good. Unfortunately, time was running short so we cruised through the takes quite liberally, spending less time than was desirable.
However, we did collect some solid takes before heading out to lunch to reward ourselves for jobs well done.

The Mix — Guitar
The first thing I did was compile the best takes in ProTools from the playlists I had recorded. Once I was happy, I ported everything over into Ableton where I have a more extensive set of editing tools.
The first thing I noticed was how noisy the Royer 121 is, or how poor the signal to noise ratio was. It’s actually something I can recall happening the previous time I used it as a vocal mic, then having to fine tune a gate to get it sounding just right.
Here though, I turned to iZotope and lowered the level of the noise. The trade-off was between a clean sound with its character removed on the one extreme, and a noisy sound with something extra on the other. I went for the clean sound as I felt the noise didn’t really add anything desirable. Also, on reflection, the recording was heavy on the fret slides. This is obviously due to the Royer being placed so close to the guitar meaning all the intricacies were picked up. I wanted the intricacies, but those fret slides had too much time in the spotlight for my liking. If I were to do the recording again, I wouldn’t have used the Royer on the guitar.
Also, there were some quite severe resonant frequencies which poked through during the bridge sections of the song. Working predominantly with electronic music, this was the first time where it was truly necessary to pick up an EQ and notch out the frequencies. I had always watched videos of this technique being offered but never found it necessary to use until now.
I selected a band, adjusted the Q value to its highest (so the band was as narrow as possible) and boosted the gain to its maximum value. I then begun sweeping through the frequencies starting around 1–4 kHz.
I was really quite impressed with this technique in action as on some frequencies, the resonant frequency completely takes over so as to create a sustained note, drowning out the rest. Obviously I didn’t want this and so inverted the gain, notching out the frequencies until it felt right. I did this another 5 times until the harshest offenders were removed, careful not to remove too much of the guitar’s character while doing so.

I also used a gate and some compression on the guitar. The gate was to keep me from having to manually remove noise pre and post takes. Fab-Filter’s gate is excellent as it shows you visually, the level of noise, so I could set it quite accurately with minimal fiddling. I sued the compressor simply to tame the extreme peaks and keep the level consistent. The compression was mild though, a ratio of 4:1 and a threshold of -15 db.
In total I had four guitar tracks, the “bones” or skeleton which formed the majority of the guitar, then also a chorus section and 2x overdub tracks. I sent all four tracks to a common reverb which helped glue the sounds together and create a bit more space as the room in which they were recorded is a ‘dead’ room with almost no noticeable, room reverb. The reverb I used was subtle though as the guitar needed to feel close and not washed away in some cathedral preset. The preset I used was “acoustic guitar studio” so this felt like a match made. I put a 24 db pole, high pass filter after the reverb and cut everything below 220 Hz. I’m not a fan of having bassy reverb and again, I wanted the reverb to just sit in the background, in the air as it were.
For the guitar overdub layers, I set up a delay on a return track. Theses layers consist of short call and response plucks. I set up a delay only on these as they were short enough to afford them a delay which again would be subtle but help add depth to the mix. I wanted to enhance the call and response effect though so I placed an auto-pan after the delay so it wasn’t static. I set the phase to 111 degrees so the delay begun on the left, as if in mono and ended on the right. I set the rate to 4 bars so that the delay tail would move gradually through the stereo field matching the cadence of the call and response.
The Mix — Vocals
For the vocals I had two tracks; the bones and harmony tracks. I sent both to a reverb return track. On it I placed a stereo reverb, set to an incredibly short decay time; 0.13 seconds. This was done to help place both layers in the same space and glue them together, but it also gave the layers a lovely sense of space which the original vocal recording didn't have. It did this though, without it being obvious that there was even reverb applied; felt rather than heard.
On the reverb bones track, I used the same preset as on the return track but here, the reverb was in mono. It was also set to have a decay of 6.53 seconds.
I put the reverb in mono here, as I only wanted the stereo space to be created at the return track. I wanted to use this mono reverb as something you would hear and not merely sense. As the reverb tail was long, I set the mix value to around 9%, so that the original vocal is favored heavily over the reverb. However, at the end of phrases, I automated the mix value to sharply ramp up to 50% so that the end syllable would “hang” and fill up the spaces. Once the singing resumed, the mix value would return to its original 9% and become just perceptible. This is a different way of achieving something I have already spoken about, namely side-chain reverb which I cover in this blog post.
Below is a snippet of the hanging reverb in this song. Note the hang on the words, “again” and “stay.”
Next, I set up a delay to achieve something similar; be subtle when Brendan is singing and become more obvious and fill in the spaces when he isn’t. For this, I set up a delay on a return track and automated the amount of the return (i.e. the volume of the delay). I placed an auto-pan after the delay to achieve the same results as the overdubbed guitars. The rate was the same (4 bars) but the phase was set so that these delays begun on the right and swung left. This should hopefully keep the delays out of the way of each other whilst enhancing the stereo field ever so slightly.
Finally, I went to work on the vocal harmonies. The trick here was to make them sit underneath the main vocals while adding extra weight to the chorus. I used another reverb set to mono to add presence and I used a filter delay which is a cool way of creating stereo width. I reigned in the width though as the lower registers needed not be sent to the extremities, just far enough out to create contrast.
Finally I sent this all to a bus of sorts and printed (recorded) the result down into one stereo file. I then applied some saturation to warm everything up before adding some gentle limiting to catch any peaks on the way out.
In Summary — Reflections
I would not use the Royer 121 on the guitar again as it created more problems than it needed. I don’t think it was the right choice for the job but in saying that, I still got a good result after some processing in the edit. Also, session time management was poor again, as we had far too much time for the guitar and far too little for the vocals. I think this is partly down to Brendan taking extra time initially settling into the recording as we had to scrap a number of recordings before making progress.
However, despite seemingly rushing through the vocal takes, I absolutely loved how they turned out. His voice has lovely natural power and the C414 did us both massive favors. The recording along with the processing really make the vocals a highlight for me and a reason to keep returning to have another listen.
Please enjoy.
