ImpactMat: Continuous Material Estimation from Impact Sounds

Four bunnies of different materials (Ceramic, Plastic, Steel, Glass) dropped in a physics sim — the impact sounds are procedurally rendered from each material's estimated parameters at the simulated contacts (FEM modal synthesis, stereo-panned to match the screen). 🔊 Unmute to listen.

Abstract

Impact sound rendering synthesizes the sound produced when a 3D object is struck, reflecting its material properties. However, it often relies on fixed presets such as wood, plastic, and steel, which cannot capture the full range of material variations. Therefore, we study inverse sound rendering by estimating continuous material parameters from reference impact sounds. These estimates enable synthesis of impact sounds that acoustically resemble the reference without manual material-parameter tuning. To support this task, we introduce ImpactMat, a dataset and benchmark containing single- and blended-material impact sounds paired with physical parameters. We further propose a framework that supports continuous parameter prediction from one or more reference impact sounds. Our method outperforms conventional audio-feature regression and audio–language-model baselines and supports re-rendering from real-world impact recordings, providing a practical bridge from recorded sounds to controllable physics-based synthesis.

5continuous physical parameters (E, ρ, ν, α, β)

88kimpact clips · 1,100 objects · 8 base materials + blends

0.078Avg. NMAE (Ours) vs 0.17+ for the strongest baseline

Method

A set of K impact recordings from the same object is encoded, pooled as an unordered set, and mapped to continuous, renderer-compatible material parameters. Blend supervision regularizes both the prediction and the latent representation.

K impact soundssame object

→

Audio EncoderPANN CNN14

→

Set Poolingmean & std over hits

→

Regression Head+ aux. material class

→

Parametersθ = [logE, logρ, ν, logα, logβ]

E Young's modulus

ρ density

ν Poisson's ratio

α,β Rayleigh damping

Training combines a target-weighted regression loss, a soft-label material classification head, a blend regression term, and a latent blend consistency loss that aligns a blend's audio latent with the interpolation of its two base-material latents — teaching a continuous acoustic-to-physical mapping instead of a lookup over discrete classes.

Physical modal analysis

The parameters are renderer-compatible: every object is voxelized and solved with FEM modal analysis, and the estimated parameters drive the same modal synthesizer at re-render time — so the whole loop, from geometry to sound, stays physically grounded. Below: a voxelized mesh and several of its FEM vibration modes (displacement magnitude).

Listen · Synthetic re-rendering

For each object we play the input impact and the sound re-rendered from our predicted parameters (FEM modal synthesis, same object & hit). Close match = accurate parameters.

Ceramic

Bowl · pure

Input impact

Re-rendered (ours)

InputRe-render

Glass

Vessel · pure

Input impact

Re-rendered (ours)

InputRe-render

Plastic

Object · pure

Input impact

Re-rendered (ours)

InputRe-render

Log-mel spectrograms (onset-aligned): the re-render reproduces the input's modal structure and decay from the predicted parameters alone.

Continuous material blends

Because the targets are continuous, a material can be interpolated. Drag the slider to morph one object from Plastic to Ceramic: the impact sound and the underlying physical parameters move smoothly together (ρ, E, α, β by geometric mean, ν arithmetic) — the continuity our latent blend-consistency objective is built for.

Plastic Ceramic

—

Same object & strike throughout — only the material changes. Each step is a full FEM modal re-solve at the interpolated parameters.

Listen · Real-world objects

Real recordings from ObjectFolder-Real: each panel shows the log-mel of the real recording (left) next to the sound re-rendered from the material parameters our model predicted from that recording (right, playable). Hit point is arbitrary and hit time is estimated (no metadata used).

Wood

Cutting Board

Re-rendered (ours)

Real inputRe-render

Ceramic

Vase (green)

Re-rendered (ours)

Real inputRe-render

Glass

Salad Bowl

Re-rendered (ours)

Real inputRe-render

Iron

Kettlebell

Re-rendered (ours)

Real inputRe-render

🔁 Drag to rotate the 3D objects · scroll to zoom. Meshes & textures from ObjectFolder-Real.

Results

Physical-parameter estimation error (MAE, lower is better). ρ and E in log-space; ν linear; α, β in physical units. Our supervised regressor is far below conventional DSP regression and in-context learning with general audio–language models.

Method	Subset	ρ	E	ν	α	β	Avg. NMAE
Conventional regression
DSP + reg.	Single	0.190	0.443	0.047	9.93	3.26×10^-7	–
DSP + reg.	Blend	0.131	0.321	0.032	5.19	1.83×10^-7	–
In-context learning (audio–language models)
Qwen2.5-Omni-7B	Single	0.338	0.614	0.071	13.38	1.61×10^-1	–
Audio Flamingo 3	Single	14.1	2.14	0.130	21.7	1.8×10^-2	–
Supervised (ours)
Ours	Single	0.081	0.140	0.008	0.59	2.39×10^-8	0.078
Ours	Blend	0.119	0.206	0.012	1.08	1.32×10^-8	0.116

Averaging more impacts at inference (K = 1 → 8) further reduces error (all-sample NMAE 0.112 → 0.093), as set pooling stably combines multiple observations.

BibTeX

@article{cho2026impactmat,
  title   = {ImpactMat: Continuous Material Estimation from Impact Sounds for Inverse Sound Rendering},
  author  = {Cho, Hyebin and Kim, Bumsoo and Chung, Joon Son},
  journal = {arXiv preprint},
  year    = {2026}
}