These audio examples accompany my Mix Rescue feature in SOS July 2019, in which I remixed Daniel Thompson’s track, ‘Every Dream’.
These two files demonstrate how I handled the acoustic guitar part. The original guitar arguably sounds slightly fuller than the processed version on its own, but that’s because it was having to do the heavy lifting in terms of the song’s bottom end and rhythm. At other points in the track, this resulted in unpleasant resonances. The processed version is a touch cleaner in terms of noise, yet perhaps sounds a little thinner in comparison with the original — but mixing is all about context. With the new bass part filling out the bottom end, and the guitar-slap sample lifted from the recording and triggered throughout, to emphasise the rhythm, the guitar part itself could both be made brighter and be panned to one side, the whole idea being to inject some width into the mix, and to create a space (in terms of frequency and stereo width) in which the lead vocal could do its thing.
These three examples show what I did to the vocal. You should be able to hear the reduced noise in 2b compared with the original. This was important, because some of the other processing used to get the vocal to work in the mix thinned out the chesty resonances a touch and boosted the top end, which would otherwise have really emphasised that noise. The third file shows how I used quite long panned and filtered reverbs to give a sense of space and width but without losing the upfront and intimate quality of the vocal. Listen to the full mix for an idea of why I chose the specific panning of the reverbs — it was largely about exploiting spaces between other sounds in the mix. Again, context is king…
These two pairs of clips are taken from the point at which the buzzing noise in the lap steel part became most problematic, towards the end of the song. Notice how the buzz changes in level, which made de-noising really tricky — to remove it satisfactorily with static settings left me with practically no useful signal! I solved this by automating the wet/dry blend of Reaper’s ReaFIR plug-in set to work as a de-noiser, as well as the frequency of a low-pass filter, and the level of the part. These allowed me to tease out the important details to make them audible, while pushing the noise into the background at other times. You can still clearly hear the problem in the processed versions when auditioned in isolation, though it’s much improved — but in the context of the full mix, it works well enough.
The original mix sent to me by Daniel, and my remix, as described in my Mix Rescue feature in SOS July 2019.