Adult Themes, Disagreement

I'm not totally sure about the amount of edits I made to the rhythm tracks on MB-LP. In some looping projects, repeated mis-timing can be charming, hypnotic even. Like Tune Yards' first album. But for the most part, I chose the cleanest repetition out of 3+ minutes of playing the loop. That's not how it works live, you have to stick with the first one. Or, as I've seen some people do, start playing the loop but only hit record on the second or third pass, once you've settled into the groove. That's more what I did for these songs: pretend I can play them a bit tighter than I can yet live. But what else are records for?

I left vocals for the end of recording. In some cases, I had guide vocals, but for most of them I had no detailed plan for vocals before everything else was tracked and balanced. I did that in the makeshift vocal booth for the Rode NT1a, in the middle of the bare linoleum floor. The only thing out of the ordinary was the presence of a big metal Manhasset music stand right next to the mic, which I had there to hold lyric sheets while rehearsing the songs, but which I kept around for the short plate reverb-sounding reflection it provided.

I think vocal tracking went fairly well, the only thing I wish I wouldn't have done is save a few of the most taxing vocal songs, ENTJ and ENFJ, for the end of the process. My voice was a bit out of shape and after doing 14 other songs multiple times each, my vocal chords were a bit rawer than I would have liked. If I'd spread out the sessions a bit instead of trying to do them all in one block it would have been easier, but I was anxious to dismantle the vocal booth and reconfigure the room into mixing mode.

Apart for the drum EQ I'd applied during editing, I went into mixing with untouched tracks, with no EQ or compression on the way in. So the first step was to try and set the rough balance and see which faders felt unstable. My new laptop has multi-touch, so I used a two-monitor setup with the console view on the laptop to take advantage of finger control.

The most readily accessible of Sonar's compressors is modeled on the 1176, and I wanted to know how to use it properly, so I found an academic article that aggregated opinions of mixing engineers on how to use the 1176 on various sources. They had quotes where engineers suggested settings, and so I tried a few to see what they did. As I went along, sometimes my preferences would change. For much of the initial mix, I hit the compressor hard with the vocals, like 15 dB reduction, but by the end of the mix I was switching songs to ~6 dB reduction with a further 3 dB from an SSL 4K-style compressor on the vocal bus.

For monitors, I rented the 5” Yorkville ones from Long and McQuade. I've now rented the 5”, 6”, and 8” monitors. It's true that the biggest difference between them is the amount of bass, but I found it fairly easy to judge the bass level even on the 5” ones by A/Bing my songs with other material, then referencing my mixes on my parents' home theatre subwoofer. The same is true of the Sony MDR7509-HD headphones I tracked on: the bass is there, just not as prominent as we're maybe used to in modern live settings. So when I actually buy some monitors, I don't think I need to discriminate based on size. One thing I still enjoy about renting monitors is that it puts a time limit on how long I can spend mixing; in this case, a week for 16 songs.

Thankfully I wasn't still fiddling with song structures and comps, so I could just concentrate on making the songs sound ok to me. I got a free spl meter app for my phone so that I could see what level I was mixing at. Turns out about 70 dB at the mixing position, for the most part, with short excursions into louder and softer for forensic work and big-picture perspective, respectively.

When setting the balance and checking for perspective, I increasingly find myself able to just focus on a couple of elements, usually vocals and drums, and making sure these key tracks are audible and consistent as opposed to trying to work on everything at once. I think extending this approach to mastering would be my best next step: putting the album tracks in sequence and focusing on their relative loudness rather than that of the whole tracks, I think this is closer to what I'm listening for when I'm trying to make sure if a record has a consistent volume. Usually I'm just asking, “well, how loud is the vocal”.

I didn't lean as heavily on Boss DD-2 delay-time twiddling on this record as on Little Gwaii. In fact I think it's actually only on ISFP. Instead, I used a combination of delays using instances of Sonar's Sonitus plugin. This usually consisted of a short slapback-type delay and then a longer one for echoes that I would ride the output of, and sometimes the EQ filter as well. I got that idea from reading Mike Senior's dissection of “Paradise” by Coldplay, where apparently the repeats of “para-para-paradise” on the lead vocal go to a progressively wetter signal over the course of the line. I feel like I have more to explore on the front of filter automation.

For only having a week to mix the record, I actually tried a lot of new processing techniques, not the least because the material on MB-LP is so diverse. More on that next.