The following audio files accompany the Mix Rescue of Phre The Eon in SOS August 2010. To get the most out of them, why not download them all and audition them in your DAW?
Undesirable levels of unnatural-sounding simulated hi-hat spill from FXpansion's BFD drums virtual instrument was one of the problems that I encountered on this mix, and I dealt with it in different ways for different tracks. I shelved 4dB from the kick track's high end above 1kHz using DDMF's LP10; I gated the under-snare mic channel with Cockos ReaGate, using the spill-free over-snare track as a gate trigger; and I ducked the overheads with Cockos ReaComp, using the hi-hat close mic to feed the compressor's side-chain input. In this example you can hear the full processed drum sound making use all of these spill-reduction tactics — I've muted the global ambience and stereo delay send effects to spotlight the effects of the processing. Compare this file with DrumsNoFX_MoreSpill.
The DrumsNoFX_LessSpill audio example showed the final drum sound complete with hi-hat spill-reduction measures. For this file I've defeated the spill reduction to show how this increases the hi-hat level and also adds an unwelcome softness and phasiness to the hi-hat tone.
Because the BFD overhead mics didn't include much snare room sound, this instrument's close mic felt very disconnected from the rest of the kit. The remedy to this involved supplementing the snare room sound with artificial reverb from SSL's Duende X-Verb plug-in, which is what you can hear in the DrumsNoFX_LessSpill audio file. This audio example, on the other hand, shows what the drums mix would have sounded like without this extra reverb support, for the sake of comparison.
Here's the final kick-drum sound from my remix, based on the raw output from the band's FXpansion BFD performance. It comprised an 'inside' mic with a small EQ notch at 97Hz to de-emphasise a dissonant pitched resonance and a 4dB shelving cut above 1kHz (both carried out in DDMF's LP10 EQ using minimum-phase processing); and an 'outside' mic with 20Hz high-pass filtering from Brainworx Bx_cleansweep, phase rotation from Betabugs Phasebug, and low-frequency transient processing from Voxengo's Transmodder. The added low-end punch that Transmodder is adding in the 55Hz and 390Hz regions isn't something that a static EQ can deliver, which is why I growing to like this plug-in a lot.
Voxengo's Transmodder is an unusual processor that can seem a bit forbidding, but it's worth investigating nonetheless because it's capable of fantasticly versatile transient processing. There's a good example of what it can do in this month's remix — it's used for one of the kick-drum layers. For this audio example, however, I've deliberately bypassed the Transmodder processing so that you can hear what effect it has on the final sound by comparing this file with the KickProc example.
The HF harshness of the raw DI track in the BassDIRaw example was easy to tone down with low-pass filtering from Flux's Epure II equaliser, while the resonant peak was traced to a frequency of 166Hz and cut out with a surgical peaking cut. For some additional warmth I applied a further broad peaking boost at 350Hz and then passed the signal to Universal Audio's software LA2A valve levelling amplifier emulation for a few decibels of dynamic control.
I'm not normally one for adding reverb to bass guitar, but in this particular remix the bass was balanced quite high, so a dusting of reverb on the DI signal gave a worthwhile blend improvement. Because the global ambience reverb send effect in question already had its low end rolled off below about 300Hz, there was no real danger of this reverb muddying up the LF region of the spectrum.
This audio example shows the sound of the band's original guitar cab recording. Although not exactly broken-sounding, the tone nonetheless comes across as woolly and lacklustre, despite being rather light on real low-frequencies. After some experimentation with EQ and phase matching, I decided that it didn't really have anything much to offer over the DI signal, so simply muted it.
Having decided to ditch the original bass cab recording (showcased in the BassCabRaw file), I reamped the DI signal through a virtual bass amplifier instead in search of more character and warmth. IK Multimedia's Ampeg SVX plug-in came up with the goods here, and its '60s Portaflex emulation delivering a good deal of character and rumble. In this file you can hear the sound processed (as in my final mix) with a 47Hz high-pass filter from Cockos ReaEQ. It's a bit flabby on its own, but that doesn't matter a great deal, because the DI track is provides ample note definition.
Here's the combination of the DI and reamped bass layers, as featured in the BassDIProcFX and BassAmpegProc audio examples — the reamped part is mixed about 3dB lower than the DI, as in the final remix. By the time I'd got this far with the sound it had a character that I really liked and also cut through the mix nicely, but there was no getting away from the fact that it simply didn't have enough low-end welly. EQ and other mix processing unfortunately didn't work as potential solutions, because all I got out of that was a flabby mess. (Import the file into your own system and have a go for yourself to see what I mean.) Therefore I opted for adding in a completely new sonic layer to beef up the low end.
This is a simple synth patch from the freeware Iblit synth, from developer Andreas Ersson, which I programmed from scratch in order to add apparent low end to the song's main bass guitar part. I only used one of the synth's oscillators, setting it up as a simple square-wave source and the using a static low-resonance low-pass filter to clear away most of the higher harmonics. After bouncing down its output to preserve CPU cycles I also dulled the sound further with another gentle low-pass filter from Cockos ReaEQ, while balancing its level in the mix against the bass DI and reamp tracks.
Adding in the new sub-synth part (the BassSub file) to the DI and reamp tracks (BassCombi_DIAmpeg) completed the puzzle as far as the bass sound was concerned, giving presence, attitude, and lots of controlled low end at once. Have a listen to the final bass sound in this example, and then listen back to the Remix to hear it in its proper context.
This example demonstrates the remix processing for the acoustic guitar part during the song's choruses. The three recorded layers were high-pass filtered with Brainworx Bx_cleansweep to remove some heinous low-frequency thuds and rumbles, but otherwise the only real processing was an instance of Sonnox Oxford EQ high-pass filtering a little further at 60Hz, dipping 1dB in a narrow band at 310Hz, and applying a broader 2dB boost centred around 1.1kHz. While this processing delivered a sound that worked in the mix, the tuning was adrift enough that the blend was really suffering, so that I didn't feel I could get the instrument to sit properly.
The treatment for the tuning ailments in the AcGtrUntuned audio file was a stiff dose of Melodyne Editor's DNA-powered polyphonic tuning correction, and you can hear the results of my efforts with that software here. I've learned to take care to refine the software's automatic pitch-detection results, but even with careful preparation in this department the side-effects of the processing turned out to be pretty severe, despite the small pitch-shifts: a chorus-like sound on a lot of strums and a general loss of transient definition.
In an attempt to compensate for the loss of transient information in the Melodyne-corrected acoustic guitar part (the AcGtrTuned_NoSPL file), for the final remix processing I stuck in an instance of SPL's Transient Designer, and this achieved some success, as you can hear in this audio example. Enough, in fact, that I was happy that the Melodyne cure was better than the original disease — sour tuning and bad blend. However, I didn't feel I could justify the same decision in the more exposed non-chorus sections of the song, where the processing artefacts didn't really stand up to the extra scrutiny.
Here's a section of the main lead guitar line in the song's chorus, without any channel processing. The muted, woolly sound was a real problem for the remix, because it really struggled to cut through at all unless I had it at a level which swamped everything else.
Processing the LdGtrRaw part with a 30Hz high-pass filter and 5dB of cut at 280Hz, both from DDMF's LP10, helped reduce the mushiness of the sound somewhat, but further EQ proved fairly fruitless — which served as my tip-off that I needed extra frequency information in the useful presence region, rather than just a higher level of the information that was already there.
In order to harden up the sound you can hear in LdGtrProc_NoSansamp, I tried adding further distortion to it from one of my favourite freeware distortion plug-ins: Mokafix's Noamp, which is a recreation of Tech 21's Sansamp. Bingo! The default switch and control settings hit the nail on the head, bar a small tweak of the plug-in's input Volume control to increase the distortion.
This track from Manchester-based band Phre The Eon is called 'Everybody's Falling Apart', and was recorded and mixed by their friendly engineer Warren Hilton.
My final remix of Phre The Eon's 'Everybody's Falling Apart', working from the same multitrack files — but with a little help from an additional sub-synth part at the low end.