NÁSTRÖND — Five Voices of Destruction
Original High-Gain Amplifier Collection | RavenForge × SpectralForge Chimera
Náströnd is an original five-channel high-gain amplifier, designed and coded by RavenForge for SpectralForge Chimera. It is a circuit model of our own DSP code, not a recreation of any commercial amplifier, hardware or plugin.
Its five channels take five different approaches to distortion, saturation, dynamics and tone, from immediate, aggressive attack to massive low-end weight and dense, heavily saturated textures. This pack brings all five into the Neural Amp Modeler ecosystem.
THE FIVE CHANNELS
FENRIR — The Relentless Hunter
Fast, aggressive and immediate: early pick attack, a sharp distortion response, lean lows and quick recovery. Built for articulate rhythm playing, fast alternate picking, tremolo-picked passages and tightly controlled riffs.
SURTR — The Burning Giant
Dense, forceful and compressed: the compression is spread across the gain stages, giving sustained distortion with a cohesive response and forward vocal mids. A strong starting point for heavy rhythm parts, thick power chords and saturated leads.
NÍÐHÖGGR — The Serpent Beneath
Raw, uneven and characterful: an asymmetric drive whose recovery is deliberately uneven, for a rougher, more irregular texture. For players who want a grinding, organic distortion rather than perfectly even saturation.
FIMBULVETR — The Endless Winter
Broad, full-bodied and expansive: body, bloom and a yielding supply, with open upper mids. Excellent for sustained chords, slower heavy riffs and atmospheric passages that want a fuller distortion.
RAGNARÖK — The Final Collapse
Deep, saturated and imposing: late-stage saturation, a powerful low-end hit and a stiff recovery. Built for dense high-gain textures, massive sustained notes and crushing rhythm sounds.
CAPTURE OVERVIEW
Source. The original Náströnd DSP, rendered offline through the plugin's own production signal path (4x oversampling and latency included) at 48 kHz.
Settings. Each channel at its factory default settings. Input and output trims at 0 dB.
Chain. Amp head only. No pedals, cabinet, microphone, room, effects or post-processing.
Training signal. A synthetic DI-style signal of plucked notes, noise bursts and decays from −30 to −3 dBFS. About 4.5 hours of it trained A2-Full and about 2 hours trained A2-Lite. Separate held-out signals were kept for validation and testing.
Model. NAM A2. Each file holds both A2-Full and A2-Lite; switch between them with the TONE3000 plugin's NAM A2 Size setting. A2-Lite uses less CPU but is less accurate.
Level. Each model's output is fixed so that its validation signal plays at −20 dBFS RMS through the channel EQ below. Keep Normalize off to preserve the balance between channels.
Accuracy. On the held-out test signal, A2-Full reaches a full-signal ESR of 0.008 to 0.022 per channel, with peak level within ±1 dB of the source. All five channels passed our A2 release qualification on both the validation and test signals.
These are neural approximations of the original DSP, not bit-exact replacements. Some nonlinear dynamics and signal-history behaviour can differ from the native Chimera engine, most noticeably on the very quiet tail after a note stops.
CHANNEL EQ (RELEASE VOICING)
Each channel has a six-band POST EQ, placed after the amp, that completes its intended voicing. In the TONE3000 plugin, enable the block's EQ, set it to POST, and enter the values below.
Band types: band 1 is a low shelf, bands 2 to 5 are bells, band 6 is a high shelf. Each entry is frequency / gain / Q.
FENRIR
1: 100 Hz / −1.0 dB / 0.7 · 2: 250 Hz / −1.5 dB / 1.0 · 3: 700 Hz / 0 dB / 1.0 · 4: 1.7 kHz / +1.5 dB / 0.9 · 5: 3.5 kHz / +0.5 dB / 1.2 · 6: 8 kHz / −0.5 dB / 0.7
SURTR
1: 100 Hz / 0 dB / 0.7 · 2: 280 Hz / −0.8 dB / 1.0 · 3: 850 Hz / +1.5 dB / 0.8 · 4: 1.7 kHz / +0.5 dB / 1.0 · 5: 3.5 kHz / −0.5 dB / 1.0 · 6: 8 kHz / −1.0 dB / 0.7
NÍÐHÖGGR
1: 100 Hz / −0.5 dB / 0.7 · 2: 280 Hz / −1.0 dB / 1.0 · 3: 750 Hz / −0.5 dB / 1.0 · 4: 1.8 kHz / +1.0 dB / 0.9 · 5: 3.6 kHz / +1.0 dB / 1.2 · 6: 8 kHz / −0.5 dB / 0.7
FIMBULVETR
1: 100 Hz / −0.5 dB / 0.7 · 2: 300 Hz / −2.5 dB / 0.8 · 3: 700 Hz / 0 dB / 1.0 · 4: 1.5 kHz / +2.0 dB / 0.75 · 5: 3.5 kHz / +1.0 dB / 1.0 · 6: 8 kHz / +0.5 dB / 0.7
RAGNARÖK
1: 90 Hz / +1.0 dB / 0.7 · 2: 250 Hz / −2.0 dB / 1.0 · 3: 650 Hz / −0.5 dB / 1.0 · 4: 2.0 kHz / +1.0 dB / 0.8 · 5: 3.5 kHz / +0.5 dB / 1.2 · 6: 8 kHz / −1.0 dB / 0.7
Without the EQ, each channel is the raw amp voice, and its level can differ from the calibrated −20 dBFS by about 0.5 to 1 dB.
RECOMMENDED USE
Load a channel into the TONE3000 plugin, or into any player that supports NAM A2.
Keep Normalize off, and set the channel's POST EQ from the table above.
Add your preferred guitar cabinet IR after the amp.
Adjust input gain to suit your guitar, pickups and playing.
For extended-range guitars and low tunings, choose a tight cabinet IR and trim low-frequency build-up to keep notes defined.
TECHNICAL SPECIFICATIONS
Format: Neural Amp Modeler (.nam), NAM A2 (A2-Full + A2-Lite in one file)
Sample rate: 48 kHz
Channels: 5 (one model each)
Gear type: amp head, circuit model of own DSP code
Capture settings: channel factory defaults
Cabinet, microphone, room: not included
Release voicing: channel-specific six-band POST EQ (values above)
Output calibration: −20 dBFS RMS on the validation signal, through the channel EQ
Normalize: off
Post-processing: none
ABOUT THE PROJECT
Náströnd is part of SpectralForge Chimera, an original amplifier and audio-processing project by RavenForge. This NAM collection is a portable way to explore five of Náströnd's voices outside the native Chimera signal chain.
Five channels. Five distinct distortion characters. One original amplifier architecture.
Developed by RavenForge · Powered by SpectralForge Chimera
