This the story of Tagorth, a magnificient cavern in which initiation rites took place in the ancient times.

STORY BEHIND
---------------------------------------------------------
This the story of Tagorth, a magnificient cavern in which initiation rites took place in the ancient
times. This haunting melody is describing the huge entrance with a soft waterfall dripping from the top, the main dark corridor, the hall of crystals and, finally, the Grand Dome, where an gigantic hall has a big hole in the ceiling, letting the light to pass through and lightening the center where is the altar. Wizards and the "Lords Of Darkling Light" meet here to perform the rite of Darkling Light. Hear reverberation bouncing everywhere, a dub-flute and a basoon following a 6/8 rhythm played by ancient drums. In this second edition you will note a different texture, better quality and plenty more air in between than in the previous version.

THE PRODUCTION
---------------------------------------------------------
Flutes are key instruments in “Tagorth, The Cavern” . How did I got such surrounding, thick sound?

The objective was to get a penetrating and thick flute sound with some well-marked tremolo to add some dramatic expression capabilities. To achieve that I created a three-layer sound using the factory “Flute” keymap in the Kurzweil K2500 and a Sine wave to add some thickness.

The Layer 1 was lowered two octaves and filtered with a 4-Pole Low-pass algorithm, with a cut frequency of 1568 Hz. Besides the cutting frequency was used to add the tremolo-like sound by connecting a LFO function controlled by the modulation wheel, this helped me to trigger the effect in the performance quite nicely. This LFO was coupled with key press so I gained extra-control in some passages of the tracks. I turned sound channel output of this layer full right and amplified overall level +6dB way up. For this layer amplitude envelope was set to “natural”. I rested 9 db of resonance in the filter that for this algorithm is quite steep.

The Layer 2 was born as a copy of layer 1, but I set a delay in triggering the sound of 85 milisec and turned sound channel output to fully left. This helped to, first, add thickness to the whole sound and second to de-phase two signals to help brain identify clearly both sources in the stereoscape. The level for this layer was set in +6 db also.

Finally, Layer 3 was made from a single, pure Sine Wave, with a pitch displacement of –12 semitones and no extra amplitude (just a humming sound in the background adding even more thickness but in a subtle way). For this layer I added no dynamic change of pitch so brain has a good “zero-reference which helps to identify clearly the tremolo effect in the other layers. Amplitude envelope was programmed to render a mid-soft attack and a quick release, most in the idea to avoid blurring the sound space when piped into the reverb unit. Again the sound was filtered with another 4-Pole Low Pass but this time I cut in 104 Hz to get such fat, deep sound. The sound output was set to the centre of panoramic so ears have a good (subtle) reference of where the centre is, both for the other flute layers and for the drums (which were also running in both positions, right and left separately).

The sound was pumped into the Mackie board, where I added and extra EQ boosting 12kHz+ freqs about 5 db,. mid frequencies (centre freq 2.5 kHz) were lowered about –3db and low freqs (<80 Hz) were lowered a little also. All this signal was the re-routed through a true-stereo (two Aux sends) reverb unit and the return was inserted in the general sound bus.

And, how I did the choral voices? Voices was of a great importance in Tagorth. Before I began to program any voice, I spent three weeks listening to Enya works, again and again. I thought this was a good reference to me, as her tracks are full of rich pads basically made of several layers of voices. Finally for Tagorth I decided to program two sets of programs, each one using different techniques an basic keymaps, and in each set several programs with subtle differences

For the first set of voices (I nicknamed “G Voices” from GNOMUSY to such programs, marking them from A to D and X depending on the different subtleties), I picked the factory keymap of Voices that came with Kurzweil K2500. In my opinion, this sample is a good starting raw material, in fact some of the most surprising programs I’ve found in the Web are using partially or completely this keymap. Nevertheless, the sound itself was not as thick as I was looking for. Besides the original program was pretty poor in terms of assignment of controls to be used dynamically in the performance.

The objective was to get a really rich and deep choirs with lots of parameters to play with in real time (assigning them to programmable controllers) so giving the opportunity to change sounds in the fly. The program finally used (G Voices X) was made in three layers, each one with its particularities.

Layer one, was a basic “Voices” with a 4Pole Low pass filter, with a base cut frequency of 622 Hz, but this could be changed through assignable slider controllers up to +3600 cents of tone. Together, the second slider controls the overall intensity of the layer. By moving the filter slider way up I get a brighter sound and amplitude slider gives more presence of voices. The output sound was perfectly centred in the pan position, as well as the remaining layers, as I was designing a “centred” choir sound, while the next set was the really stereo one.

Layer two was targeted to give a really “airy” mood to the whole choir, so I selected a Hi Freq Stimulator to enhance hi freqs of the same sample. To give a full palette of airy sound, I programmed a base frequency cut of 17 Hz, while the displacement of another slider (K2500 has 8 assignable sliders) augmented the airy effect (cut frequency) up tp +1700 cents of tone; besides I assigned another controller to balance the amount (amplitude) of this layer present in the complete program.

Layer three was filtered through a Bandpass Filter, quite useful in pads, to add colour to the airy effect, and then I routed it through a panner module in the same algorithm. In this case I was looking for a subtle sweep filter effect that gave the feeling of opening the choirs. I assigned the position of centre frequency of Bandpass to another controller (slider) and the same with the presence (amplitude) of the layer.

In this way I had a quite thick choir layer and 6 assigned controllers to vary sound quality in the fly, which gives interest to a choir pad. Nevertheless, this pad had finally relatively little presence in the overall arrangement of Tagorth. It was the second set of voices that should give the final touch ...

In programming the second set of choir voices I applied a bit more attention as such voices will be the most important and prominent sound in the whole track. The objective was to obtain deep, clear and thick voice sounds which represent the voices of Lords Of Darkling Light and Wizards, and giving the main feeling of darkness and haunting to the composition. Such voices might send an awesome feeling of mystery and spaciousness, as if listeners were walking inside the cave.

So I began the “quest for voices” among my pretty large library of samples, and finally I opted for one Mono voice sample of a sustained Aah from “Take 6” CD samples, a very good library I was gifted when I purchased K2500. Nevertheless, the sample was far distant to the sound I was seeking for, as it was targeted more to pop-rock background vocals, and I had to program stereo panning also. So again I opted to program a three layer sound, using the same sample. I followed a very similar process as for the set 1.

Layer 1 was delayed 0.4 secs respect to in-time note triggering, passed through a 4Pole Low pass filter with a dynamically varying cut frequency.(base 932 Hz up to +3500 cents of tone) and an extra amplitude boost of +6 dB. I had to work in detail the attack curve of amplitude envelope, as “natural” one was too quick to my purposes. Besides I displaced down overall layer pitch –10 cents of tone to get a thicker sound when combined with the rest of the layers. Output was programmed to fully right pan position.

Layer 2 was created as copy of Layer 1, displacing output to fully left and delaying 0 secs the note triggering. Again a Low-pass filter was applied, this time with a higher cut frequency and with an smaller dynamic variation respect to the layer 1. All these differences added thickness and enough cues to let brain differentiate among the two signals, every one going to a different ear.

Layer 3 was again a copy of layer one, but in this case I added by far more filtering and centred the output pan position , so giving a central reference to the whole choir program. Pitch was displaced 10 cents up to compensate previous layers and give opacity to the sound.

The resulting sound is quite convincing and fills sound space quite well. It is complementing the previous choir set nicely, as I took care of it analysing first the spectrum response of each set separately.

You will find first set of choir voices as background of main flute theme and the second one (deeper and with more presence) at the beginning (entrance of the cavern) and in the middle and final passages (the Grand Dome and the final rite). Given the quality of sample and thickness of sound, I let the second set to fly alone supporting the track in a single point, and it worked quite well!.

Maybe you’re listening to this program again in subsequent tracks (for sure you’ll have in “Lords of Darkling Light), I love it!