You are here

Mix Rescue: Flags | Media

Mixing & Recording Advice By Mike Senior
Published March 2011

These audio files accompany the Mix Rescue article in SOS March 2011 (/sos/mar11/articles/mix-rescue-0311.htm). You'll get the best results by downloading them and comparing them in your DAW.

The following audio files are available in both WAV and MP3 format. Both the original and the remix projects are also available for download here.

Download| 201 MB
Download 732 MB
Download| 936 MB


This is a section of the raw stereo overhead recording from Flags' multitrack files. Because the mics had been placed in the 'textbook' position over the cymbals the balance and tone of the snare drum are compromised. Notice how much louder and brighter the cymbal crashes (and indeed the ride-cymbal and hi-hat stick noises) are than the snare drum. Because most of the naturalness of the snare sound usually comes from the overhead mics, it proved very difficult to create anything like an appealing snare sound for the remix.


Here's the same section of the raw mono room-mic recording from Flags' multitrack files. The kick and cymbal sounds are much louder than the snare, which indicates that the snare drum itself wasn't appropriately selected to balance with the rest of the kit instruments. Using a louder snare drum would probably have made the mixing job a lot easier.


This is the over-snare close mic from the unprocessed multitrack files. Because the cardioid snare microphone was placed too close to the drum's head (the band were afraid of catching too much spill from the rest of the kit), it spotlighted only a small patch of the drum's top head, giving a very unrepresentative sound. Although this example is particularly extreme, the same kind of problem is evident to a lesser extent in most amateur recordings, and doesn't help you achieve a natural-sounding snare presentation.


The most promising snare sound in this instance actually came courtesy of the band's undersnare mic. Although still harbouring some unpleasant resonances and a great deal of spill from all the other kit instruments, the basic sound is much more usable, albeit lacking some attack.


Here's a section of the under-snare microphone recording which has been EQ'd as in the final remix using Cubase's built-in channel EQ processing. A high-pass filter at 116Hz is removing the worst of the kick-drum spill; an 8dB notch is removing a nasty high-frequency whistle (as well as softening the hat and cymbal spill); and a series of notches (3dB at 310Hz, 6dB at 470Hz, 15dB at 599Hz, and 3dB at 2359Hz) are targeting unwanted pitched resonances. While this results in a more appealing snare timbre, there are still dynamics issues to content with: the level of cymbal spill, a lack of snare attack, and some unevenness in the snare performance.


This is what the under-snare microphone sounded like once I'd added dynamics processing to the EQ demonstrated in the SnareUnderEQ file. The processing comprised gating (Attack: 0.3ms; Hold: 41ms; Release 137ms; Side-chain band-pass filtering at 445Hz) to reduce the cymbal spill, limiting (Release: 100ms) to even out the snare hits, and transient enhancement (Attack: 3dB; Attack Length: 26ms) to improve the drum's punch. Notice how the gating threshold is being adjusted under automation control to allow snare ghost notes through for some fills (at 0:17 for instance) and to extend the snare hits after 0:19.


This file showcases the two snare effects that I used to enhance the EQ/dynamics-processed under-snare mic for the final remix. The first is a simple single-tap delay, with some of its attack shaved off using Cubase's Envelope Shaper plug-in so that it added sustain to the snare hits within the mix. Although it sounds clearly like a distinct echo with the snare track soloed, it isn't audible in this way once everything else in the mix is hammering away. The second effect is a convolution-based studio reverb patch from the Reverence plug-in, with 10ms of pre-delay and high-pass filtering at 230Hz, designed to blend the drier close-mic sound into the kit as a whole.


Although all the drummer's ghost notes made it tricky to replace the whole snare part with a more suitable sound (as I've done for a number of previous Mix Rescue remixes), I still sneaked this snare-drum sample in underneath the under-snare mic to try to bolster the tone for the main hits. I first selected a stereo snare sample within Cubase's Halion One sample-based virtual instrument (to help widen the snare image a little), triggered it from Koen Tanghe's KTDrumTrigger plug-in (in a fairly slap-dash manner, it has to be said!), and then removed all its attack so that it wouldn't risk flamming with the main snare in context.


Here's what the under-snare mic sounded like with the triggered sample layered behind it. Although the samples aren't always triggered reliably, it doesn't actually matter too much, as the extraneous triggers just sound like additional ghost notes and any vestigial unnaturalness is easily concealed once the rest of the instruments and vocals are playing.


At the last moment, I decided to add in a little of the over-snare close-mic to supplement the attack phase of the combined snare sound, and you can hear this element isolated and processed as in the remix here. To isolate and control the added attack, I high-pass filtered at 730Hz; gated with a high threshold, 20ms hold time, and 29ms release time; and then limited.


The full composite snare sound, comprising the processed under-snare mic, a triggered sample from Halion One, the processed over-snare mic, and delay and reverb send effects. It sounds a little ragged in solo, but within the mix it nonetheless does a fairly passable job.


Here's the full drum-kit sound to show how the composite snare sound worked in context. It's still not about to win any Grammies, but it's good enough to serve its purpose fairly solidly in this particular mix.


This is one of the three microphones used to record the main acoustic guitar part. It's been placed close to the instrument's frets, and while it does manage to capture some useful high-frequency string sound, there's also a lot of less desirable information too: string-damping noise (for example at 0:05), fret squeaks (for example at 0:21), and lots of pick noise.


This is the second acoustic-guitar mic, placed close to the instrument's soundhole — which is usually one of the worst locations to put a mic for studio work. The first problem is that a mic in this location typically picks up masses of pick noise, and also the booming air resonance of the sound hole whenever the lowest string comes into play (as you can hear quite dramatically at 0:19). In this case, there were also fret squeaks to worry about (for example at 0:21). Despite all of these problems, there was still some body to the timbre which I was able to use for the remix sound.


The third acoustic-guitar mic didn't offer anything that the other mics didn't, and exhibited bad pick/fret noise and soundhole boom into the bargain, so I decided to discard this one. There's no sense in using a microphone simply for the sake of form if it doesn't contribute anything useful to the mix!


This audio example isolates the final acoustic-guitar sound as it appears in the remix. EQ, transient processing, and multiband compression were all employed to tackle the pick/fret-noise and soundhole boom problems in the original recordings, and a parallel distortion effect thickened the midrange tone. Filling out and blending the sound further are generous helpings of ambience and hall reverbs, and tempo-synchronised delay — the latter under the control of mixer automation.


Distortion isn't something you might normally think of adding to acoustic guitar parts, but in this instance it really helped to fill out the sound so that it could hold its own against the rest of the instrumentation. To illustrate the impact of this processing, compare this file, where the distortion processing is bypassed, to the AcGtrFinal file.


In order to get the bass guitar to support the guitars better during the chorus sections, I doubled it at the octave above using a clavinet sound from Halion One. The added part was reamped through Cubase's Amp Simulator to thicken the tone, and then EQ'd and flanged to distribute it suitably across the frequence spectrum and stereo field. In this example file you can hear it within the context of the remix backing track — I've left the vocal out to make it easier to hear. Compare this file with BassFinalNoHalion.


This file is identical to BassFinal, with the exception that I've muted the extra Halion One clavinet layer I added to fill out the bass sound. Compare this with the BassFinal example to hear the difference this component of the mix makes.


The original mix of the song '54', by the band Flags. The band had completed the mix on their own Cubase system using some Waves third-party plug-ins.


This is my remix based on the band's original multitracks, which was carried out in Steinberg Cubase 5 using only its built-in processing and effects.