Intro
At the end of last academic year, I recorded a modern jazz live concert in Kettle Yard here in Cambridge. The work was a part of the assignment for Live Recording Practice for which I required only one piece from the concert, which I processed and handed in. The rest of the material stayed intact though, patiently waiting for me to find some time to work on it - so here we go, I decided to finally have this concert mixed and done.
Short note on the recording
I have already written a lot on this recording (contact me through comments if you'd like to see a logbook), so just shortly:
Band: String quartet, double bass, drums, guitar.
Microphones: NT4 (stereo) hanging from above (later called 'overheads'), 2 AKG C414s behind and above string quartet (you can see one of them on the pic, later called 'strings'), 2 Rhode NTKs above and behind drumkit (later called 'drums'), and AKG D112 (kick drum mic) directed on the guitar's speaker (later called 'guitar').
All recorded using Fostex HDD recorder in 24 bits, 48k.
Session 1:All processors and effects I used and set during this session were tweaked quite roughly and the levels go up above 0 often. I plan to keep them just as guidance for later precise mixing.
Mixer set-up
I sent all paired microphones to buses to be able to apply effects on stereo pairs instead of just applying them to single tracks. This also enabled me to control pan of each track separately. I also created 3 groups for each stereo pair so that I would automatically select 2 tracks in the arrange window.
Sample delays
I started off with close looking at the differences in time between different microphones. I decided to use overheads as a reference for the stereo image, and using sample delays I corrected the position of the remaining microphones so that the sound waves would be exactly in line with the overheads, thus minimalising cancellation and clearing the sound from delays caused by sounds getting to microphones in different moments. Here is how the measured delays looked like - as you can see, the difference in various microphones is quite big. I kept the differences between paired microphones intact (I just delayed two mics using one delay on the bus), as the time differences between left and right channel are important part of the stereo perception.
Sample delays on all tracks
Guitar
The sound from D112 was great. As it was picking up from the guitar amp, the levels were rather flat. To increase difference between background noise when the guitar was quiet and the rest, I used expander. Then, I compressed it with threshold just around peak levels and gained it so that the signal would be much stronger. Attack on both processors is set to 0ms, as I want it to affect as much signal as possible.
Processors applied to gutar track
Overs
These were just compressed. The levels on the recording were not great, so the threshold is quite low, but still it doesn't apply to sounds that are not very loud.
Compressor used on Overheads
Drums
To bring up the drum sounds only, I expanded everything that was louder than -11dB with ratio 0.5:1. Then, instead of compressing again, I just used Adaptive limiter to make sure that the drums will not exceed 0. I adjusted the gain so that the limiter would just affect really loud parts.
Processors used on Drumkit busSession 2:In this session, I started working closely on the first track of the concert (actually, the second one - because of lack of time before the recording, we couldn't set the levels up, so we had to do it during the first track).
Cutting the regions
To avoid unnecessary and unclear mixing of sound from microphones not actually in use, I cut the regions so that when the song starts with strings introduction, only string microphones are used:
Arrange window for the beginning of the first piece
Pan automationBecause only one stereo pair was used at the beginning of this track, I panned in centrally (27 to the left and right). But for later mix, I wanted strings to be on the left, so I automated pan on both lef and right mics to go where I wanted them to go:
Pan automation on stringsFinishing touch
...at least for this session. It took quite a while to work out a way to EQ drums properly. The setting pictured below works quite well - metallic sound has been removed by 7.5dB dip at 2.2kHz. Additional cut at 235Hz elimiated unpleasant vibrations, and thanks to bass cut, I lost most of the noises.
And for now, I think that's it! Here's a bonus pic of the mixer setup:
