Song: Halcyon Days
Key: A Major
Timo Kotipelto – vocals
Matias Kupiainen – guitar, mixing, production
Jens Johansson – keyboard
Rolf Pilve – drums
Lauri Porra – bass
Stratovarius formed in Finland in 1984, and released their first album, Fright Night in 1989. They are a power metal band, who originally played a very melodic, very technical variant of traditional power metal. With Nemesis and their previous album, Elysium, Stratovarius embraced the use of sampling (with voiceovers and foley in Elysium) and synthesizers (a major part of the compositions on Nemesis), which gave their music a whole new dimension, often replacing what would be a lead guitar or keyboard part.
Halcyon Days is their fourth track on Nemesis and really emphasises the change in direction from their previous albums. It is a powerful sounding song, using punchy bass drum to emphasise beats, and utilises wide panning on the rhythm guitars and a sizeable reverb on the vocals to generate it’s sense of space. The production sets up two very distinct moods in Halcyon Days, constantly going from “heavy” to “soft” segments, which contrasts to many power metal songs that often have a very fluid or progressive feel. It features 21 song-changes, which I will use to refer to parts of the song.
The song features an intro (1), heavy intro (2), four different bridges (3, 8, 14, 17), four soft verses (4, 6, 11, 13), three heavy verses (5, 7, 12), three choruses (9, 15, 19), and one heavy bridge (10) that resembles the heavy intro. It also includes a synthesizer-focused interlude (16) that includes several different synth sounds at different intervals. Shortly after the bridge following the interlude there is a bass drop (18) before going into the chorus. The song ends with a heavy outro (20) that then leads to a soft outro (21).
The list of instruments in this song includes (in order of analysis):
- Drum Kit (Kick, Snare, Toms, Cymbals)
- Church Bell
- Rhythm Guitar
- Bass Guitar
- Vocals (lead, whisper, choir, vocal processing)
- Synthesizer (9 different sounds, given a subjective description)
The kick drum in Halcyon Days sounds quite punchy without becoming overbearing and sits at the centre of the sound field. The frequencies boosted to emphasise the kick drum are a 90hz shelf on the bass end of the frequency spectrum, while the click of the kick pedal hitting the drum skin is brought up in the mix through a band at the 3.5khz-5khz range, as well as approximately 12khz. This produces a cushiony-sounding tail end to the drum hit. In some of the heavier sections of the song such as the heavy intro and chorus, it sounds as though the 90hz shelf is lowered to allow for the bass and rhythm guitars’ lower frequencies. The kick drum does not sustain for long after the initial hit, without sounding unnaturally cut off. It sounds to me like this has been achieved through a compressor with a fast attack and well-defined release of about 300-500ms.
The snare sounds very snappy, without much bottom end (likely due to a high-pass filter), although it doesn’t need it to stand out. It sits at the centre of the mix alongside the kick, as these are the two parts of the drum kit that listeners are likely to use to determine the rhythm of the song. It sounds like the main frequencies that have been used to bring out the snare are at approximately 2khz, with a slight boost at about 150hz to help it emphasise the rhythm alongside the kick. It is important that the snare has a different focal point in the low range to the kick, as otherwise they would both lose clarity due to so much of the same frequency being played at the same points in time. Compression has also been used on the snare, however it has a much faster release than the kick, as you can hear the recorded reverb of the snare being cut off by the release of the compression, which leads me to believe the release was set to about 100ms or less.
The toms have quite a wide stereo image, heard particularly in the last two bars of the heavy intro (2), where you can hear the higher-pitch tom to the far right (about a 70%-80% pan), and it shifts over to the left at about a 30-40% pan as the drum roll changes toms. The preciseness of the EQ and compression suggests that the toms were individually mic’d, although the overheads help make the drum roll a more gradual change in panning than a sudden shift between each.
The toms have a warm, resonant sound, although their resonance is not particularly long. This seems to be a recurring theme in Kupiainen’s way of designing ADSR envelopes. He lets the toms sustain in his mix until the point at which their envelopes would start to release naturally, at which point he has a faster than natural release as I discovered upon analysis of the snare. It is a good way to keep the mix sounding clean, letting other instruments become more present in the other-wise occupied frequency range earlier, especially in the case of 16th-beat drum rolls where many of them are sharing certain frequencies within the low-mid range. The toms also have a very clear transient that sits at about 9khz-12khz which usually needs to be boosted by about 4-5db to be clearly audible. Compressing the individual recordings of the toms would have helped in making this transient more audible as well.
The cymbals, bar the hi-hat, are used sparingly in the mix. This is understandable as cymbals aesthetically are mainly used as decoration to emphasise song or riff changes. The hi-hat, however, is another instrumental part of the listener’s perception of the rhythm and is more easily digestible at a louder volume than a ride cymbal. With this in mind, it sounds as though Kupiainen has recorded the hi-hat, ride cymbal, and overheads to cover his cymbals.
The hi-hats are mainly utilised by the drummer in the verses and bridges, while the ride is reserved for the chorus. The hi-hats sit with the kick and snare in the centre of the mix and sound quite crisp. In the verses you can hear the looseness of the hats, suggesting that an additional microphone was used to the side of the hi-hat at the same height as where the hi-hats would clash with each other when going from an open to closed positioning, pointing towards that gap. The hi-hat’s focal frequency range is between 8khz-12khz which is a relatively quiet frequency range in the mix otherwise, with only other cymbals, some harmonics from guitars and transients from drums that occupy the same range.
The ride is quite quiet in the mix except for the resonance from harder hits, which gives the ride a nice dynamic range. It is panned to the left by about 70% and occupies a fairly small frequency range around 7khz. It’s reverb sounds like it was achieved mainly through recording rather than processing, of which there may be a small amount just to give it a little more sustain.
The overheads capture the sound of the kit as a whole, however they are particularly important in capturing the crash cymbals, as it does not sound like the crash cymbals were microphoned individually. The crash cymbals sound relatively quiet in the mix, only becoming louder at the end of the chorus where the drummer plays a 16th-beat roll on his crash. This was probably achieved by automating the volume of the right channel of the overheads. Similar to the ride, the majority of the reverb from the crash cymbals is likely from the recording process, excluding the harder hits which seem to have some processing to provide some additional sustain. The cymbals captured from the overheads all sit at between 7khz and 12khz in the mix with some harmonics at higher than 12khz, which have been preserved by a high shelf.
In the intro and outro of Halcyon Days, you can hear a church bell in the background at the end of climactic passages. It sits at the centre of the mix but doesn’t get lost in the mix too much because it is only played when most of the other instruments in the centre aren’t playing. It occupies from 1khz through to 1.3khz and has quite a lengthy reverb (I would guess one and a half to two beats). This could be due to bell’s size and the location of the recording, but I would guess that some processing was involved. You can also hear that the tail drops off rather fast (about 100ms-200ms) at the end of the sustain. This relates back to the technique that I suggested Kupiainen was using for some of the drums.
The rhythm guitar is arguably the most important element of a metal song, and as such is the loudest instrument in this mix, which is quite evident in the master track’s waveform when the song goes from a soft verse into a heavy verse, as seen below.
The rhythm guitar comes in for the heavy verse, and the song immediately goes from being peaceful and intimate to the glorious, grandiose feel that is very familiar to fans of the subgenre. It sounds like there are two identically recorded rhythm guitar tracks on either side of the mix, with the left being slightly louder than the right. You can hear this particularly in the chorus. They both seem present at about 1khz and 3.5khz-5khz in the mid to high-mid frequency range, with what sounds like one channel being more focused at 600hz (giving a more sharp, precise tone) while the other is focused on 150hz (achieving a thicker, crunchier tone).
The rhythm guitar has been gated through Kupiainen’s TC Electronic Nova System, which also provides him with his distortion, overdrive, and delay among many other functions to shape his guitar tone. This is done in combination with his ENGL amplifier and cabinet.
The bass guitar in Halcyon days is not immediately noticeable, but loud enough to hold the two rhythm guitar channels together with its centre-stage pulse at 60hz-140hz, with some of the transients being present at about 400hz. The dynamic range of the bass seems relatively flexible. Some of the quieter frequencies have been brought up through compression, however you can still make out the velocity of the finger picking.
The violins come in after the first two bars of the soft outro, which can almost only be heard at 7.5khz and above. After the guitars fade out in this section, they are a lot more present around 1.7khz to 2.2khz. I am guessing that the reason these frequencies aren’t as noticeable on the violin track prior to this is that the guitar transients and violin transients share the same frequencies. This means that due to the guitars being so loud in the mix, the violins are drowned out. This is not necessarily a bad thing, as it gradually introduces the violin as a new focal point for the melody in this final section of the song.
Halcyon Days does an amazing job of showcasing the talent of lead vocalist Timo Kotipelto through equalisation of the high and high-mid frequencies that had been captured by a ribbon microphone, as seen at 4:10 in the video below.
The body of the vocal sounds like it mostly comes from around 870hz, which translates to A5 in musical terms. This matches the song so well because the song is in the key of A Major. This space in the frequency spectrum is also fairly bare compared to the 1khz-2khz and 450hz range, meaning that the vocalist Kotipelto has his own space to show off in. The harmonics from Kotipelto’s voice are also highlighted by a high shelf from approximately 2khz onwards. Kotipelto’s voice rarely crashes with the low and low-mid frequency ranges, suggesting a high pass filter from about 300hz and below. In combination with a plate reverb and the quiet delay that follows the beat of the song in the majority of the song, this creates an airy vocal sound that is a feature in many power metal productions. The only sections to exclude this combination of equalisation, reverb, and delay are the soft verses which bring the vocal right forward in the mix, lessening the high shelf and instead focusing on 850-900hz.
In the soft verses, Kotipelto records a second version of the verse in a whisper, which sits exclusively in the high frequency range (from 3khz onwards), with a small room reverb. It occupies the back left for most of the soft verses, occasionally shifting to about a 20% pan to the right.
The choir sits quite far back in the mix, but the stereo width achieved by a left and right track bring a sense of enormity to the song. The choir occupies a wide band at around 970hz, with a high shelf to highlight the harmonics like the lead vocals. The choir that was recorded was called the “Shark Finns” who don’t appear on many other works than Nemesis, so I’m assuming it is just a group of trusted associates of the band. It is unclear as to whether the reverb on the choir is natural and from the recording process, or whether a hall reverb has been added in the mixing stage.
In the interlude, there are some non-word, down-pitched vocals that have been used as a layer in an additive synthesizer, combined with a low-frequency saw wave to create a strange-sounding groaning that is quite far forward in the mix, and occupies a band around 150hz and another at 900hz in the frequency spectrum. This kind of reminds me of the sound they use when Neo is entering the matrix for the first time at 0:54 into this video.
There are a total of 9 different synth sounds in Halcyon Days, 10 if you include the vocal processing mentioned previously. Due to the amount of synths, I have described them in the sound field by way of my own subjective descriptors. The sythesizers add to the duplicity in tones on this song, where all of the synthesizers either use a very bass-centric body or a very high-frequency body, with none of them having a main focus on the mid frequency range. This adds to the dramatic atmosphere of the song quite nicely and takes advantage of the two-faced composition.
I will attempt to analyse the synth sounds in order of appearance. Analyzing synthesizers is not exactly my strong point as I have a very rock and heavy metal musical upbringing which typically doesn’t use synth, so I will do my best to describe the synthesizers utilised.
What I refer to as the “Blip” synthesizer is a mix of square and sawtooth oscillators that have a very fast, if not instantaneous attack and release. This synth follows the chord progression of the song and pans from left to right and back again at about a 50% pan in either direction as an underlying secondary melody. It is heard for the first time in the fourth bar of the intro, and persists in sections 3-5, 11-13, and 16-18. It sits in the middle of the mix, being quite noticeable but not in-your-face in the quiet sections, and a nice decorative addition to the heavy sections.
The “Slide” synthesizer kind of sounds like if you took a pick sliding down a guitar string and reversed the sample. The only way I can think that they would have achieved this sound is through a combination of a sine, sawtooth, and a square wave that is the most prominent part to the synth. It sounds as though it has a slow attack and fast release. I think that playing just one midi note of that sort of setup and automating the equalisation to gradually boost higher frequencies instead of the low ones would achieve this sort of sound. This synth is centre-stage but fairly far back in the mix. This synth appears at the end of the intro.
I named this the killswitch synth because of how fast the signal cuts out after it plays, as though someone just completely kills the volume at the tail end of the signal. It can be heard as a bridge before the chorus. I’m fairly sure that this is mostly just a sawtooth wave, possibly with a short spring reverb and a little bit of a sine wave mixed in. It is panned slightly to the left in preparation for all of the instruments to come in at the centre of the mix at the start of the chorus.
This is a square-wave-focused version of the original blip, which sounds a lot cleaner and slightly more percussive, with a longer reverb and a more exaggerated decay in it’s ADSR envelope. It also has a delay at about the length of a 16th beat. This synthesizer is exclusive to the chorus sections and sits behind the lead vocal in the mix. It pans from left to right similarly to the original blip, with a slightly narrower max pan of about 50%.
The bass synthesizer appears in the interlude section and bass drop section exclusively. It sounds like a fairly standard bass synth, being a low-volume sawtooth combined with a low-frequency oscillator that sits at about 150hz. It does not sound like it has much high frequency content, so my guess is that a low-pass filter has been added at around 600hz. It is slightly quieter than glassy blip from the chorus and doesn’t seem to deviate from the centre of the mix.
This synthesizer sound comes in about four bars into the interlude, sitting at the centre of the mix and its focal frequencies are at 1.2khz-3.2khz, with some undertones at around 800hz. This provides some of the edgy, crunchy tone that is not as present without the rhythm guitar. It seems to be another combination of a sawtooth wave and a low-frequency oscillator. It is about as loud as the glassy blip in the chorus, but due to the rhythm guitars not playing in the interlude, it seems louder to the listener as there is more room for it to breathe than the glassy blip.
This is another low-frequency oscillator combined with a sawtooth that has been heavily distorted and it appears directly after the digital sounding synthesizer in the interlude. The frequencies emphasised through equalisation are around 130hz for the body of this sound, with the distortion being boosted at 2.5khz as well to provide the crunch from the harmonics.
The slimey sounding synth I’m referring to is in the latter half of the interlude and its sawtooth, low frequency oscillator body sits at 200hz-500hz while the high frequency sawtooth that gives it the slimey, squelchy aesthetic is lower in volume at about 3khz. It sits between about 20% pan to the right and centre of the mix.