Mix Challenge

The biggest challenge for me was managing the microphone bleed while setting up my stereo imaging, but I feel I was able to make it work in the end. The bleed was not an issue for me from any other standpoint, just the stereo imaging part. I would even say that the bleed was helpful in terms of cohesive blending of instruments while I was working on my center channel aka. mono mix.

I shall now document the highlights from my mixing session:

1) Getting a consistent foundation to build on - using Serial Compression on the Bass

Ended up using 3 serially connected instances of FET-style compression, all with mostly similar settings at 4:1 ratio and about 4 dB of gain reduction. The only remarkable difference between instances is that I used different release time settings for each instance as that tended to sound more interesting compared to having all 3 set to the same release time.

I also ended up creating some stereo width on the bass by simply panning the two microphones in opposite directions. This made the stereo mix retain a similar tonal balance compared to when listening in mono, and also reinforced the realism of having a big double bass in the band that fills the room with juicy low end. I even ended up enhancing this width effect with a side signal boost at around 80 Hz on the Bass stereo sub-mix.

Screenshot - Serial Compression on the Bass and Microphone Panning:


A noteworthy challenge was getting the Bass to work with the Stomp-Kick as by default both instruments dominated the 60 Hz range and clashed, so I resolved that by boosting the weaker 120 Hz range by a LOT while cutting the exaggerated 60 Hz range by a LOT on the Bass. I then did the opposite on the Stomp-Kick, albeit to a much lesser degree. Somehow managed to get the instruments agree with each other in the end.

2) Creating stereo width from non-bass instruments

The Guitar seems to be in a supportive role, so I tried simply panning it heavily. However, that did not work well because the bleed also got jarringly sharply localised and gave me a lopsided mix, so instead of panning the entire Guitar sub-mix I created stereo width by panning the two Guitar microphones in opposite directions. Now, because one of the microphones produced a much brighter Guitar signal than the other, I also got a nice localisation effect for the Guitar in addition to stereo width, whereas the bleed component was no longer jarringly sharply localised.

Screenshot - Guitar Microphone Panning:


The Accordion also seems to be in a supportive role, but simply panning it heavily gave me an even more jarringly localised bleed and lopsided mix, so I had to somehow create stereo width for it as well; however, the Accordion was recorded with a single microphone, and therefore I couldn’t use the “pan 2 microphones in opposite directions” method that I had successfully used on the Guitar earlier. I had to come up with a different trick: the good old “delay one channel only by a few milliseconds” trick. Granted, some slight mid-side EQ work was required on the Accordion stereo sub-mix after having applied the delay trick in order to make it blend in again with the rest of the mix.

Screenshot - Accordion Left Channel Delay:


The Backing Vocals were recorded with 2 microphones for each person, so that group was easy to spread out.

3) Compressing and de-essing the Lead Vocals

In order to fight excessive sibilance, I started out by using pre-emphasis of sibilance range before the Compressor, followed by de-emphasis after the Compressor. Still, some excessive sibilance remained, so I did have to strap on a De-Esser as well. The Compressor is set to use a high-pass filter in its side-chain. Ratio is set to fairly aggressive 7:1 and gets me around 5 to 12 dB of gain reduction, depending on the passage. The reason it sounds OK regardless of aggressive reduction is the soft knee and fairly slow release time, and of course the high-pass in the side-chain. As for EQ, there’s a low shelf cut to reduce rumble, slight high shelf air boost, and some micro-adjustments in the “money range”. With acoustic music, I generally prefer to use low shelf cuts instead of high-pass filters to control rumble, as shelves sound more natural for that purpose to my ear, and acoustic music benefits from sounding natural rather than sounding overly processed IMO.

Screenshot - Lead Vocal Effect Chain:


4) Getting a convincing dialogue going between Lead Vocals and the Fiddle - by using Automation

Although there is some occasional Automation on other parts as well, the Fiddle is the most obvious one where frequent Volume Automation was necessary to bring the Fiddle in front occasionally while it was filling in the breaks between Lead Vocal phrases. Great attention had to be paid to make sure the effect wasn’t too weak nor too strong.

Screenshot - Fiddle Volume Automation:


That concludes my list of highlights for this session.