user_mobilelogo

Outlining The Fundamental Equalisation Aesthetics Of Extreme Heavy Metal Music Utilising Empirical Analysis

Dan Turner

University of Glamorgan

 

Introduction

This paper will examine some of the exaggerated approaches that are often utilised in the equalisation of audio in extreme metal music. It will mainly focus on the extraction of empirical data from the component parts of a master mix session recorded and mixed by the Grammy-award winning producer Andy Sneap. This empirical data was derived using FFT analysis, and using graphical representation, I will explain how the component parts interact and why some of these techniques employed by Sneap work effectively.

A Question of Genre

In 1981, Franco Fabbri proposed that:
"A musical genre is a set of musical events, real or possible, whose course is governed by a set of socially accepted rules".
This is certainly true of the various sub-categories that fall under the umbrella term of Heavy Metal. However, the governance of this acceptance and consensus among the fan community is the source of constant debate. How can one define the music that deserves this classification? Walser (1993) correctly states that this accepted term is heavily contested among both fans and musicians.
"That's not heavy metal!" is the most damning music criticism a fan can inflict, for that genre name has great prestige among fans"
He argues that the boundaries of genre are often difficult to define and that debate is often furious over the stylistic rules of these genres and sub-genres. However, the more extreme a sub-genre, the more definitive are the rules commanding that categorisation.
Although now slightly out of date, Weinstein states in her book Heavy Metal that "Thrash" (considered one of the most extreme categories at the time of its publication) is Heavy Metal's "fundamentalist strain" (Weinstein 1991). By this she refers to the notion that fundamentalism arises as a reaction to happenings of a decadent nature, in this case a reaction to the chart dominance of "pop metal". This led to a purer, more distilled form of the music that was certainly, at its core, more aggressive.
Kahn-Harris (2006) goes some way to defining some of the facets typical of extreme Metal. He defines these as sonically transgressive elements (a term also used with regularity in Walser's work), and goes some way towards explaining the difference between extreme metal and its more commercialised cousins. Although no accepted published lexicon currently exists for Metal music, or indeed its production, there are terms quoted within Kahn-Harris' definitions that are commonplace within the industry and fan scene. These definitions are as follows:
TEMPO: Often incredibly fast (or conversely, extremely slow), and pieces are commonly between 150-250 bpm. Extreme Metal pioneered the use of the "blast beat", an illusion of high speed performed on the drums, whereby the kicks and ride/hi-hat play a repeated eighth-note pattern while the snare is struck in repeated sixteenth-note subdivisions. Also common is the performance of repeated sixteenth-note patterns on the kicks, which require the use of either two kick drums or a double pedal. Often, the drums perform at these frantic speeds while the guitars play rhythmically simpler patterns, although many extreme motifs employ the use of "tremolo picking", referring to sixteenth-note patterns performed on guitar and usually in tandem with drum parts of the same rhythmic subdivisions.
TUNING: Instruments are almost always tuned a step (or more) down from standard A440 pitch, consequently evoking sounds with a "darker" overtone.
KEYBOARDS: The music of the extreme sub-genre generally known as "black metal" often makes use of keyboard parts of a symphonic nature. Many of these arrangements are somewhat reminiscent of the grand operas of the late Romantic period, such as those by Wagner.
VOCALS: Vocalisation in extreme Metal is usually a feature that differentiates it from the more commercialised sub-genres. Traditional Metal often relies on pseudo-operatic stylings that ostensibly exist to suggest "intensity and power" (Walser 1993). Extreme metal vocalisation takes this fundamental concept further, by showing little regard for melody, vibrato and musical pitch. Often the lyrics are deliberately completely indistinguishable from those printed on the lyric sheets (indeed, John Tardy from the popular death metal act Obituary was renowned for inventing completely new words during his performances). Death metal vocalists tend to employ low grunts, while black metal vocalists rely on higher screeches.
GUITARS: All genres of Heavy Metal generally feature extremely distorted electric guitars as their key defining ingredient. Walser describes it as the "most important aural sign". He also suggests that it "functions as a sign of extreme power and intense expression by overflowing its channels and materialising the exceptional effort that produces it".
These definitions suggest that extreme Metal music could not be successfully reproduced using the same production and mixing techniques utilised for the recording of other genres of music i.e. Country and Western or Dixieland Jazz. The busy nature of the drums and guitars, coupled with the use of heavy distortion, signify that the production of extreme Metal often requires extreme production measures of its own in order to achieve maximum sonic clarity. The resultant effect, when applied correctly, I define here as "heaviosity".
An excellent example of a producer currently succeeding in this field is Andy Sneap. Sneap is a Grammy-award winning record producer that works exclusively in the field of metal music. With a discography that reads like a veritable "who's-who" of the industry, Sneap has produced or mixed many of the major practitioners in the field.
To demonstrate the application of this extreme approach to production, Sneaps work exists as an excellent vehicle for this research. For the purposes of this paper, the multitrack master used for analysis is Cradle of Filth's "Tragic Kingdom", from the 2008 album "Godspeed on the Devil's Thunder" (kindly donated by Roadrunner/Warner Bros. records). The Pro Tools master session file is perfect for research purposes, as Sneap mixes exclusively "in the box", or rather, completely in the Pro Tools environment without the use of any outboard mixing equipment. The further ramifications of this are that it is possible to analyse and bypass any plug-in settings as well as panoramic information and amplitude data. It is even possible with this multitrack to analyse the original drum takes before they were compiled and edited to make a final master compilation.
As this paper is a precursor to further research using this multitrack, I will focus here solely on the basic rhythm section: drums, bass and rhythm guitars.

The Issues with Mixing Metal Music

Walser stated that guitars are a Metal pieces most important aural signature. This is definitely a truth in this case study. Guitar sounds that heavily employ the use of harmonic distortion tend to have heightened activity over a considerable portion of the frequency spectrum, as displayed in Figure 1:

Image
Figure 1: Sonogram of Equalised Sustained Guitar Chord from Middle Section of "Tragic Kingdom"
Figure 1 represents the frequency makeup of a low sustained chord from the middle section of the song. It is taken from the left channel guitar (the guitars are double-tracked throughout the piece) and lasts for 6.267 seconds. It has all the equalization plug-ins applied.
 As with all the examples in this paper, the analysis was performed in both Adobe's Audition 3 and Sonic Visualiser. The resolution for the sonogram was 16384 bands, for more precise results, and a Hanning windowing function was chosen due to its good frequency resolution, handling of spectral leakage and fair amplitude accuracy (there is an element of compromise with all types of FFT windowing functions). All audio was recorded in 24-bit at a sampling rate of 44.1 kHz.
There can be seen clearly significant energy in the lower end of the spectrum, up to 400Hz (as outlined by the black box) and also noise formants from 2-2.9 kHz and 3 kHz (highlighted by the white box). There is a sharp notch termination from approximately 400 Hz to 2 kHz, signifying what is often referred to as a "scooped" midrange (Guitar Techniques, March 1994).
However, the two most salient points are that the guitar has reasonably high energy throughout the entire frequency spectrum, up to a cut-off point at 7 kHz. It also has unnaturally long sustain with high energy in the noise formants until 5.5 seconds, where the transient attack of the next chord can be observed. Both of these points are common to all well-recorded extreme metal guitars, and Berger reaffirms these two points:
"In sum, distortion simulates the conversion of the guitar from an impulsive to a sustained or driven instrument, and this transformation may be part of the acoustic correlate to the perceptual experience of heaviness"
"Distortion strengthens the upper harmonics of the signal" (Berger 2005)
Sneap confirmed in a telephone interview (Sept. 2009) that little had been applied to the original source. The guitar was played through a Randall RM100 amplifier (including the XL modification) with a Peavey 6505 cabinet loaded with Peavey Sheffield speakers. An Ibanez Tube Screamer pedal was included in the chain, to induce more harmonic distortion. A single Shure SM57 microphone was used, positioned slightly off-axis but approximately an inch from the speaker grill. No outboard equalization or compression was utilised during the tracking process, and at mixdown only a light URS A-10 EQ and GRM Tools Filterbank were applied. Figure 2 displays these settings, and highlights the minimalistic approach to equalisation required here. There is some filtering in the lower and higher ends of the spectrum, but that is to reduce the level of higher and lower harmonics in order to make some space for the bass and cymbals respectively.

Image
Figure 2: Plugin Settings for Rhythm Guitars in Mixdown

All of these factors highlight that extreme metal guitar timbres naturally take up a large portion of the frequency spectrum and have an unnaturally long sustain as a consequence of severe harmonic distortion. The guitars make the mixdown process difficult, as creating a sensation of space is immediately hampered. They could be described as a "sonic wall" that somehow the busy drums, keyboards, bass and vocals have to scale in order to occupy their appropriate positions in the mix.

Scaling the Sonic Wall

The first example of how Sneap negotiates this problem is the kick drum. Kick drums have a role of paramount importance in extreme metal, as they often play intricate parts that emphasise the rhythmic movement of the guitar riffs. An excellent example of this is the 90's thrash act Pantera's "Slaughtered" from the 1994 Atco album "Far Beyond Driven".
Audio Example 1: Pantera's "Slaughtered" from "Far Beyond Driven"
Audio Example 1 amply demonstrates the basic core of good metal arrangement, in that sonic "heaviosity" is the combined perceptual impact of kick/snare drum, bass and guitar all occurring simultaneously to create constant, consistent accents. The kick drum particularly has to scale the aforementioned sonic wall. However, this was not always the case, especially when extreme metal production was at its nascent stage. For example, a comparison can be made that underlines this point, using Metallica's breakthrough 1986 record "Master of Puppets", and Cradle of Filth's "Tragic Kingdom"
Audio Example 2: Metallica's "Battery" from "Master of Puppets"
Audio Example 3: Cradle of Filth's "Tragic Kingdom" from "Godspeed on the Devil's Thunder"
Both examples 2 and 3 feature fast, frantic, repeated sixteenth-note patterns on the kick drums. However, it is immediately obvious in the latter example that the kicks are certainly sharper sounding in their attack and slice through the sonic wall more adequately when compared to the Metallica recording.

Image
Figure 3: Sonogram of Double Kick Section of Metallica's "Battery"
Figure 3 is a sonogram that represents frequency makeup over time for "Battery". Again, a Hanning windowing function was used, but the FFT size was 512 bands because visually it is far more comprehensible than a higher resolution. The sonogram shows high energy in the lower end of the spectrum (outlined by the white boxes), but the transient attack of each kick drum is not clearly visible beyond approximately 500 Hz. However, the Cradle of Filth example is completely different. Figure 4 is a sonogram that represents the audio demonstrated in Example 3, and immediately it is obvious that the kick drum transient attacks can be seen all the way up to beyond 10 kHz.

Image
Figure 4: Sonogram of Double Kick Section from Cradle of Filth's "Tragic Kingdom"
There are considerable differences between these different kick drum timbres. Sneap adopts two techniques in the production of his drum sounds: Extreme equalisation and sample augmentation (colloquially referred to as "triggering").
Sneap discussed the concept of sample augmentation in an online interview in 2006, and in doing so provided a great argument that goes some way to helping define the previous comparison:
"I won't mention any names, because I'll get myself in trouble, but I mixed a lot of the old school guys, as well; you go back and listen to some of their old recordings, especially the earlier thrash albums where everyone was thinking, "these guys are killer!", but you can't actually hear the bass drums that clear... it's more of a low rumble going on. I guarantee you; if you put a sample there you'd be quite surprised to find out what's going on with people's feet."
When questioned on the validity of sample augmentation in other styles of music production, he added:
"I won't use triggers on the slower, rock stuff I produce - it's really just the faster metal stuff, where you've got to get the clarity to the playing. And the way guitar sounds have gone now, where you'll have four guitars that are really thick - and you're trying to get a good low end for everything - and if you try to use a natural kit on that, you'll never get the clarity - the smack of the high end from the kick drums. So I don't know if triggers would really apply to other styles..."
(Maelstrom Webzine, October 2009)
Sneap seems to be referring to the consistency of kick drum strikes. Although a kick drum always acoustically excites the same modes on the batter head, due to the beater being at a fixed position, the force of attack is variable. This is a consequence of the limitations of human motor control.  For example, in 1938 C.E Seashore found that, when instructed to do so, musicians could not play completely without expression. They could not play any sequence of notes without any variable inconsistencies (that is to say, like a machine generated sequence could on contemporary MIDI equipment).
The use of frantic, sixteenth-note kick drum patterns is commonplace in extreme metal. In the case of Audio Example 3, the time between each kick drum strike is 67 ms. At this speed, it would be humanly impossible to hit each kick drum with precisely the same degree of force when the pedals are depressed in such fast succession. Therefore, augmenting the original source kick drum recording with a pre-recorded sample would help settle some of these inconsistencies. This approach gained popularity in metal production during the early 1990's, and is omnipresent in the Pantera example demonstrated in Audio example 1.
The second technique is that of extreme equalization. Sneap used three separate sources that were mixed together in order to create the final kick drum timbre. These are:
1. The original recorded kick drum (captured with a close microphone slightly inside the shell)
2. A Yamaha "Sub-Kick" microphone positioned outside the drum to capture sub-harmonic frequencies.
3. A pre-recorded kick sample to augment the original source recording.
Using Auditon 3's frequency analysis tool, set to an FFT size of 65536 bands and still using a Hanning windowing function, the separate sources were analysed and the data exported to Microsoft's Excel 2007. In order to simplify the appearance of the resultant graphs, an average reading was taken at 100 Hz intervals. Line graphs were then produced that show the frequency relationships between the respective sources. Figure 5 is a line graph that represents the three kick sources with no equalisation applied, taken from one kick drum strike with a length of 100 ms. The X axis represents frequency averages per every 100 Hz interval, and the Y axis represents frequency amplitude in dB.

Image
Figure 5: Line Graph of Unequalised Kick Drum Frequencies
Figure 5 shows that each source has a strong fundamental frequency, as highlighted by the black box. The kick sample shows less activity in the lower-midrange (purple box), but more presence in the higher-midrange, between 2-6 kHz (blue box). It then gradually decreases in frequency amplitude, with a notable drop beyond 13 kHz (green box).
The real kick (labeled on the legend as Kick 1 Flat) gradually decreases in frequency amplitude beyond the fundamental, but has more presence in the lower-midrange than the sampled kick. Again, it shows more activity in the high-mids, as the blue box outlines. However, it has more energy in the higher frequencies than the sample, as the grey box highlights.
The Sub-Kick again has a strong fundamental, but decreases rapidly in frequency amplitude. However, the orange box denotes that there is still some activity in the higher-midrange.
Overall, this graph shows lower amplitudes in the high-end of the spectrum on all sources.
Figure 6 shows the same analysis process applied to each source post-equalisation. It shows some major changes in frequency makeup.

Image
Figure 6: Line Graph of Equalised Kick Drum Frequencies
The fundamentals look similar to their unequalised counterparts, as the black box highlights. However, both sampled and real kicks have considerably less activity in the lower-mids (purple box). The blue box shows that both have considerably more energy in the higher midrange, particularly between 4-6 kHz and both have higher frequency amplitudes in the higher end of the spectrum- considerably more in the case of the sampled kick.
The Sub-Kick has much less activity in the higher-midrange, but still maintains its strong presence in the low-end (orange box).
These dramatic boosts and cuts suggest the use of extreme equalisation. Figure 7 shows Sneap's plugin settings for these 3 sources. The images were cut from his Metric-Halo Channelstrip GUI displays.

Image
Figure 7: Channelstrip EQ Settings for the Three Kick Drum Sources
Sneap's settings, both for the real kick and sample, display midrange troughs represented by the red box in Figure 6. The real kick has a cut of -5.1 dB at 438 Hz on a lower bandwidth, while the sample has a major -18.2 cut at 289 Hz, again on a lower bandwidth. Both feature extreme boosts in the higher-midrange and top end, while the Sub-Kick has a low-pass filter implemented to heavily reduce midrange activity.
The fact that the sample has the most severe equalisation applied compounds the argument for augmentation further. A sample, by its very nature, has no leakage from other components of the kit, as it is taken from a single clean kick drum hit. Consequently, it is possible to apply extreme cuts and boosts without affecting the overall timbre of other parts of the kit. Figure 8 displays peak frequency sonograms of the real kick drum, both pre and post equalisation. The areas outlined in white and yellow are snare leakage on the kick drum track. As the area in yellow shows, the high-end of the snare spill is dramatically increased when the kick drum is equalised to such a high degree.

Image
Figure 8: Peak Frequency Sonograms of the Source Kick Drum, Pre and Post Equalisation
To help circumvent this issue, Sneap heavily gates the real kick track by using a duplicate of the sampled kick track to trigger the gates key input. However, this fails to remedy the issue when the kick drum and snare drum occur simultaneously, as Figure 9 demonstrates:

Image
Figure 9: Peak Frequency Sonogram of the Source Kick Drum, with Simultaneous Snare Strikes

The problem with this scenario, even with gates applied, is that this extra high-end activity on the snare spill would have an adverse effect on the overall timbre of the snare. Consequently, Sneap mixes the sampled kick at the forefront of the balance of the individual components:

Image
Figure 10: Amplitude Measurements for the Three Kick Drum Sources
Figure 10 shows that the sample is the most prominent source in the overall balance. The real kick appears to be added to inject an element of realism, and the Sub-Kick to supplement the overall low-end. The total average frequency range makeup of the kick drum and its relationship to the equalized guitars is represented in Figure 11.

Image
Figure 11: Line Graph of Equalised Kick Drum and Guitar Frequencies
This graph clearly shows the kick drum in the region from approximately 6.4-14.5 kHz clearly transcending the frequency amplitudes of the guitar. This proves that the kicks have more high-end activity, partly as a consequence of the low-pass filtering on the guitar tracks, thus scaling the sonic wall.
The same approach to equalisation is apparent in the snare. This time there are four sources:
1. The original snare, captured with a microphone on the batter head.
2. The original snare, captured with a microphone on the resonant head.
3. A sample snare, in a similar tuning to the source.
4. A second sample, of a higher overall pitch (labeled on the multitrack as "Piccolo Snare")


Image
FIGURE 12: LINE GRAPH OF UNEQUALISED SNARE DRUM FREQUENCIES
Figure 12 is a line graph displaying average frequency ranges for the four sources in question. The sample taken for analysis was of a length of 98 ms, and all samples were captured between the same start and end point in the multitrack. The area outlined by the black box shows the snare top has a strong fundamental and a markedly more prominent lower-midrange than the other sources. However, it naturally has less activity in the higher-midrange and high-end than the three other components (purple box). There is a significant valley at approximately 1.2 kHz within the frequency range of the first sample snare (green box), but the piccolo sample shows the greatest activity in high-midrange and high-end of the spectrum, as outlined by the two orange boxes.
Analysis of the equalised snares shows that Sneap has subjected these to extreme frequency cuts and boosts, in keeping with the methodology described for the kick drums:

Image
Figure 13: Line Graph of Equalised Snare Drum Frequencies
The black box in Figure 13 displays that the low-midrange prominence of the snare top is significantly reduced. But the fundamentals on all sources are still high in amplitude. Overall, the higher-midrange is boosted on all sources, with the piccolo sample still strongest (orange box). The activity in the high-end is boosted overall, with the piccolo sample again showing the highest frequency amplitudes here (green box).
Sneap's Channelstrip settings again confirm these findings:

Image
Figure 14: Channelstrip EQ Settings for the Four Snare Drum Sources
Figure 14 shows that the snare top has heavy midrange trough of -13.6 dB at 722 Hz on a low bandwidth, hence the aforementioned reduction in lower-midrange activity. It appears that all sources are boosted in the higher-midrange and more dramatically in the high-end. Again, Sneap has boosted one of the samples to a more extreme degree than the other sources, in this case the piccolo snare, as already highlighted by Figure 13. This would appear to be for the same reason as the kicks; the snare top and bottom are heavily gated in order to minimize spill from other components of the drum kit.
In the case of a membranophone with sharp transients, such the snare in question, it would seem that this kind of extreme boosting also increases the perception of attack on the drum- possibly to an unpleasant degree. Sneap counteracts this with the use of compression.
Sneap uses a relatively light amount of compression on each source post-equalisation. The Channelstrip setting for the snare top shows that this is relatively light:

Image
Figure 15: Compression Settings for Snare Top
From this it can be seen that the threshold and ratio settings are not extreme, and that the attack and release are set to reduce the transient attack by a slight degree. There is no make-up gain applied and the compression is applied post-equalisation.
The compression on all sources is of a similar nature. Overall, it makes quite a difference in frequency makeup. Figure 16 demonstrates this by plotting the frequency ranges of the snare pre and post-compression against each other:

Image
Figure 16: Line Graph of Snare (All Sources Combined), Pre and Post-Compression
The black boxes outline the main areas where the frequency makeup appears to be different. These differences appear to be most dissimilar in the higher-midrange, from about 4.4 kHz onwards. As a consequence of this, the attack of the snare is considerably less harsh. As Alex U. Case explains, this is due to the reshaping of the amplitude envelope caused by compression "leading to complex and dynamic upper and lower sidebands associated with the gain riding by the compressor" (Case 2007).
In a similar fashion to the kick, Sneap clearly favours the samples over the real snare in the balance, as Figure 17 clearly displays:

Image
Figure 17: Amplitude Measurements for the Four Snare Drum Sources
The direct conclusion from this graph is that snare sample 1 leads, with the piccolo snare slightly behind. Both component sources of the real snare are far quieter in the balance.
All of these factors add together to make the snare cut through the guitars, and thus scale the sonic wall. Figure 18 features the combined snare sources against the guitar frequency range:

Image
Figure 18: Line Graph of Equalised Snare Drum and Guitar Frequencies
In a similar nature to the kick, the snare has consistently higher frequency amplitudes than the guitar from 4.4 kHz onwards, as the black boxes outline.    
Sneap not only employs liberal frequency cuts and boosts, he also makes extreme use of filtering. The overheads are an excellent example of this. His Channelstrip settings for the overheads show a high-pass filter applied from 693 Hz- clearly an extraordinary filtering of frequencies.

Image
Figure 19: Channelstrip Equalisation Setting for Left Overhead
This is to reduce the low frequencies of the kicks and snare captured in the overheads, thus allowing Sneap more control over levels of the cymbals. With the filter applied, the overhead tracks can be boosted in overall level without affecting the timbre of the kick nor snare, and vice-versa.
Figure 20 displays the equalized kick, with a sample taken from the unequalised left overhead track at the same reference point as the kick in the multitrack. The sample length is 100 ms.

Image
Figure 20: Line Graph of Equalised Kick Drum and Unfiltered Overhead

This graph shows that the lower frequencies of the overheads, particularly around 100-200 Hz area, have high activity. This is also true around the lower-midrange, from 200-500 Hz.

Image
Figure 21: Line Graph of Equalised Kick Drum and Filtered Overhead
When the filter is applied, the huge drop in frequency amplitude on the overheads can be observed, thus keeping it away from the kick drum. The snare is affected also in a similar way by this filtering.
There is no extra equalisation present on any of the overheads, as the cymbals naturally cut through the guitars spectrum, and consequently need no boosting. Figure 20 demonstrates the relationship between the equalized guitar and filtered overheads.

Image
Figure 22: Line Graph of Equalised Overhead and Guitar Frequencies

The final component to analyse is the bass guitar. Sneap uses two sources in the creation of the final timbre- An effected direct-injected signal, and a bass amplifier.
The direct-injected signal goes through a number of processes. Sneap employed the use of a C4 multiband compressor, set to a high threshold on all bands. This was followed by a Sansamp PSA-1 amplifier modelling plugin, with higher settings on the high-end and crunch, but little lo-end boost. Next, he applied a second C4 with a similar setting to the first, and finally an L1 Ultramaximiser, set on a less extreme threshold. Figure 23 displays these plugins, and their settings.

Image
Figure 23: Plugin Settings for Direct-Injected Bass signal
The amplified bass, however, is completely flat. There were no effects applied. It would seem the intention of this extreme multiband compression on the direct-injected signal would have been to level out the bass attack and applying a distorted signal to complement the cleaner amplified tone. Figure 24 demonstrates the relationship of frequencies between the two sources.

Image
Figure 24: Line Graph of Direct-Injected (All Plugins Applied) and Amplified Bass Signals
The black box outlines the fact that the bass amplifier has a more significant trough in the lower-midrange than the direct-injected signal. However, the green box shows that there is more activity overall in the higher-midrange on the direct-injected bass than the amplified signal. It also has a more gradual decrease in frequency amplitude. Finally, it seems amplified signal has more presence in the higher-end.
When analysed with the rhythm guitar sample, it can be seen that the bass does not scale the sonic wall as it needs to occupy the lower frequencies. Figure 25 clearly demonstrates that the guitar has significantly more energy in the higher-midrange than the bass (purple box).


Image
Figure 25: Line Graph of Equalised Guitar and Effected Bass Frequencies

However, when the low frequencies are closely analysed it can be seen that the high-pass filter applied to the guitar has been utilised in order to allow room for the bass guitar, as the fundamental frequency can clearly be seen to have more presence. Figure 26 demonstrates this:

Image
Figure 26: Line Graph of Equalised Guitar and Effected Bass Frequencies- 0-2 kHz

Conclusion

This paper has examined the audio equalisation process undertaken by Sneap for the basic rhythm section. Throughout the paper, the same concept has been referred to repeatedly- that of extremity, both in regard to boosting and cutting of frequencies, and also of filtering.
The evidence from the multitrack points towards the notion that extreme music can tolerate extreme mixing methods. Sneap's liberal approach to sample augmentation also adds credence to this argument. However, the approach of moderation in equalisation to the guitars and bass highlights that Sneap does not "EQ for EQ's sake". The application of extreme EQ in this instance is not an arbitrary process, but one based in a clear production methodology and an insightful musical rationale. Further research would have to be undertaken to quantify the extremity of Sneaps approach to compression, utilisation of effects, and also the effect these have on the equalisation of individual components of the mix.
Further research is also required to determine if these exaggerated approaches would only function correctly within the confines of an extreme metal mix. The lack of study into extreme mixing methodologies in other genres hampers the establishment of any "moderatation vs. extremity" argument at this time. For example, would these approaches to equalisation work effectively in the mixing of a Jazz recording? In the Journal of the Art of record Production Vol. 3, Dr. Toulson argues that heavy equalisation would not effectively repair poorly recorded audio, and that the engineer should always endeavour to record at source with the optimum choice and placement of microphones. However, he does not examine the effect of extreme equalisation on well-recorded audio.
In order to examine this further, it is clear that the academic community requires access to more multitracks from the industry. Although acquiring these from Artist and Repertoire departments is a process fraught with difficulty, this paper has demonstrated that it is possible to have the requisite permissions granted, and also proves to the industry at large that the analysis process does not damage sales or even examine the musicological process, let alone contribute to the piracy problem currently ravaging the established music industry model. However, the optimum sources for analysis are clearly from digital audio software applications, like the Pro Tools session in question. Without the ability to bypass plug-in settings, it is almost impossible to analyse any audio pre-and-post any processing added at the mix stage, and to therefore examine these processes further.
It is hoped, ultimately, that these processes may yield a production methodology for extreme Metal, in addition to progressing the field of record production as a bone fide scholarly discipline.

REFERENCES

Berger. Harris M. 2005. "Heaviness" in the Perception of Heavy Metal Guitar Timbres: The Match of Perceptual and Acoustic Features over Time. Wired for Sound: Engineering and Technologies in Sonic Cultures. Wesleyan University Press
Case. Alex U. 2007. Sound FX: Unlocking the Creative Potential of Recording Studio Effects. Focal Press
Fabbri, Franco. 1981. A Theory of Musical Genres: Two Applications.  Paper presented at the First International Conference on Popular Music Studies
Kahn-Harris, Keith. 2006. Extreme Metal: Music and Culture on the Edge. Berg Publishers
Kilminster, David. 1994. "Walk" Transcription. Guitar Techniques. Future Publishing
Martinelli, Roberto. Andy Sneap Interview. http://www.maelstrom.nu/ezine/interview_iss45_208.php. Accessed Sept. 2009
Seashore. C.E.1967. Psychology of Music. New York: Dover (Original work published in 1938)
Walser,Robert. 1993. Running With The Devil: Power, Gender and Madness in Heavy Metal Music. Wesleyan University Press
Weinstein, Deena. 2000. Heavy Metal: The Music and Its Culture. Da Capo Press
Audio Examples
Audio Example 1: Pantera. 1994. Slaughtered from Far Beyond Driven. Atco (01:08-01:20 minutes)
Audio Example 2: Metallica. 1986. Battery from Master of Puppets. Elektra (03:46-04:00 minutes)
Audio Example 3: Cradle of Filth. 2008. Tragic Kingdom from Godspeed on the Devil's Thunder. RoadRunner. (05:17-05:33 minutes)