k2k audio Surface treatment for sound. Smooth it out, sharpen the edges, add grain, erode it, saturate it — like running sandpaper or varnish over audio.
Smooth
What it does — Blurs spectral detail using a Gaussian filter, softening transitions between frequencies and across time.
When you’d reach for it — Your recording has a brittle, grainy quality and you want something rounder and more pillowy. Also great for taming harsh digital artifacts before further processing.
Quick example
- Connect your source into Smooth.
- Set Sigma to around 3 and Kernel to 7 for a gentle polish.
- Listen back — edges soften, noise recedes.
- Try Phase on Blend if you hear phasing artifacts at higher sigma values.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Sigma | How wide the blur reaches — higher values spread smoothing further | 0.5 — 10 | 1.5 — 3 for subtle polish, 6+ for heavy diffusion |
| Kernel | Size of the smoothing window (odd numbers only) | 3 — 15 | 5 for most material, 9+ for aggressive smoothing |
| Phase | How phase is handled during blending: Preserve, Blend, Dominant, Complex | 4 modes | Preserve for transparency, Blend for creative smearing |
Sharpen
What it does — Enhances spectral edges and detail using Laplacian filtering, making transients and tonal boundaries more defined.
When you’d reach for it — A sound feels dull or buried in a mix and you want to bring out its internal detail and articulation without reaching for EQ.
Quick example
- Feed a pad or vocal into Sharpen.
- Set Amount to 1.0 and Kernel to 3x3.
- Transients and harmonic edges pop forward.
- Push Amount toward 3 — 4 for aggressive hyper-detail on sound design material.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Amount | Sharpening intensity — how much edge enhancement is applied | 0 — 5 | 0.5 — 1.5 for natural clarity, 3+ for aggressive detail |
| Kernel | Filter size: 3x3 (tighter, punchier) or 5x5 (broader, smoother enhancement) | 3x3 / 5x5 | 3x3 for transients, 5x5 for tonal detail |
| Phase | Phase handling during filtering: Preserve, Blend, Dominant, Complex | 4 modes | Preserve keeps the original character intact |
Spectral Saturation
What it does — Generates harmonic overtones above detected spectral peaks, adding richness and density to the sound.
When you’d reach for it — A thin synth patch or flat recording needs body and harmonic interest. You want the warmth of analog saturation but applied precisely to spectral content, not just waveform clipping.
Quick example
- Connect a digital synth sound into Spectral Saturation.
- Set Saturate around 50%, Warmth around 60%.
- Bring Orders to 3 for a natural harmonic tail.
- Adjust Decay to control how quickly upper harmonics fall off.
- Use Mix to blend the enriched signal with the dry original.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Saturate | How much harmonic content is generated above peaks | 0 — 100% | 30 — 60% for warmth, 80%+ for heavy coloring |
| Warmth | Biases saturation toward lower frequencies — higher values favor bass and mids | 0 — 100% | 40 — 70% for analog-style weight |
| Decay | How quickly overtone levels fall off with each harmonic order | 30 — 90% | 50 — 70% for natural roll-off |
| Orders | Number of harmonic overtones generated per peak | 1 — 5 | 2 — 3 for subtle richness, 4 — 5 for dense timbres |
| Mix | Dry/wet balance | 0 — 100% | 100% for full effect, dial back for parallel blending |
| Gain | Output level adjustment | -24 — +24 dB | 0 dB (compensate if saturation adds perceived loudness) |
Dynamic Brightness
What it does — Independently shifts spectral brightness for harmonic, percussive, and residual content, each on its own axis.
When you’d reach for it — You want to brighten the tonal body of a drum loop without making the snare harsher, or add air to a vocal’s noise floor without touching the pitch content. No EQ can do this — it separates what the sound is made of before adjusting brightness.
Quick example
- Feed a full mix or drum bus into Dynamic Brightness.
- Push Harmonic to +40% to open up tonal content.
- Pull Percussive to -30% to darken transient attacks.
- Leave Residual near zero or nudge it up for subtle air.
- Adjust Mix to taste.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Harmonic | Brightness shift for tonal/harmonic content | -100 — +100% | +20 to +50% for opening up, negative to darken |
| Percussive | Brightness shift for transient/percussive content | -100 — +100% | -20 to -40% to soften attacks |
| Residual | Brightness shift for noise/residual content | -100 — +100% | Small moves (+/-20%) for subtle air control |
| Mix | Dry/wet balance | 0 — 100% | 100% for full effect |
| Gain | Output level adjustment | -24 — +24 dB | 0 dB default |
Spectral Dither
What it does — Adds very subtle, frequency-shaped noise to break up the rigid perfection of digital audio.
When you’d reach for it — A sound feels too clean, too sterile — that unmistakable “digital” quality where every frame is mathematically identical. A touch of dither introduces just enough micro-randomness to feel natural again.
Quick example
- Place Spectral Dither after your processing chain.
- Set Level to -30 dB and Color to Pink.
- Narrow the frequency range with Freq Lo / Freq Hi to target only the mid-highs.
- Toggle Varying on so the noise pattern shifts over time.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Level | Noise floor amplitude relative to full scale | -60 — 0 dB | -30 to -20 dB for audible warmth, -40 dB for subliminal |
| Mix | Dry/wet balance | 0 — 100% | 100% to apply fully |
| Gain | Output level adjustment | -24 — +24 dB | 0 dB default |
| Color | Noise spectrum shape: White (flat), Pink (warm), Blue (bright), Shaped (psychoacoustic) | 4 modes | Pink for warmth, Shaped for most transparent results |
| Freq Lo | Low frequency boundary for the dither region | 20 — 2,000 Hz | 200 Hz to keep bass clean |
| Freq Hi | High frequency boundary for the dither region | 20 — 20,000 Hz | 16,000 Hz covers most useful range |
| Seed | Random seed for reproducible results | 0 — 999,999 | Change to get a different noise pattern |
| Varying | Whether the noise pattern evolves over time or stays static | On / Off | On for natural feel, Off for consistent mastering |
Spectral Roughness
What it does — Applies smooth, time-varying micro-modulation to spectral bins, creating organic movement and analog-like imperfection.
When you’d reach for it — A sound is lifeless and static — every moment sounds exactly the same. Roughness introduces the kind of subtle drift and wobble that makes tape machines, analog circuits, and acoustic instruments feel alive.
Quick example
- Connect a static pad or sustained tone into Spectral Roughness.
- Set Roughness to 25%, Rate to 1.5 Hz.
- Push Correlation toward 0.8 so neighboring frequencies drift together.
- Set Type to Smooth for gentle undulation.
- Nudge Bias negative to concentrate the movement in the low end.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Roughness | Depth of variation — how far bins deviate from their original level | 0 — 100% | 15 — 35% for subtle analog feel, 60%+ for obvious texture |
| Rate | Speed of the modulation in Hz | 0.1 — 20 Hz | 1 — 3 Hz for slow organic drift, 8+ Hz for shimmer |
| Correlation | How much neighboring frequency bins move together (0 = independent, 1 = locked) | 0 — 1 | 0.6 — 0.8 for natural spectral coherence |
| Bias | Which frequencies get more roughness: negative = lows, positive = highs | -1 — +1 | 0 for even distribution, -0.3 to weight the low end |
| Type | Noise character: Smooth, Filtered, S&H (stepped), Gaussian | 4 modes | Smooth for natural drift, S&H for rhythmic texture |
| Mix | Dry/wet balance | 0 — 100% | 100% for full effect |
| Gain | Output level adjustment | -24 — +24 dB | 0 dB default |
| Seed | Random seed for reproducible results | 0 — 999,999 | Change to get a different modulation pattern |
Spectral Erosion
What it does — Detects spectral edges and gradually eats them away, softening transitions and wearing down detail like weathering on stone.
When you’d reach for it — You want a sound to feel aged, decayed, or worn down. Useful for turning crisp recordings into something that sounds like it was played through old equipment, or for creative destruction where you want spectral detail to dissolve over time.
Quick example
- Feed a crisp recording into Spectral Erosion.
- Set Erosion to 40%, Sensitivity to 50%.
- Keep Depth at 1 for a single pass of erosion.
- Set Direction to Both so edges erode in frequency and time.
- Increase Depth to 3 — 4 for heavy weathering that eats deeper into the spectrum.
Parameters
| Parameter | What it controls | Range | Sweet spot hint |
|---|---|---|---|
| Erosion | How aggressively edges are attenuated | 0 — 100% | 30 — 50% for gentle aging, 70%+ for heavy decay |
| Sensitivity | What qualifies as an edge — lower values only catch sharp transitions | 0 — 100% | 40 — 60% for musical results |
| Depth | Number of erosion passes — higher values eat deeper into the spectrum | 0 — 5 | 1 for subtle, 3+ for dramatic wear |
| Direction | Which edges to erode: Frequency (spectral), Time (temporal), or Both | 3 modes | Both for natural decay, Frequency for timbral softening |
| Mix | Dry/wet balance | 0 — 100% | 100% for full effect, blend back for parallel processing |
| Gain | Output level adjustment | -24 — +24 dB | 0 dB default |