HN 日本語サマリー

← 一覧へ戻る
科学・技術

Frieve Vinyl Explained – Microscopic stylus/groove physics simulation

Frieve Vinyl Explained – Microscopic stylus/groove physics simulation (frieve-a.github.io)

71 pointsby XzetaU89 コメント

要約

この記事は、アナログレコードのスタイラス(針)と溝の物理的な相互作用を微視的なスケールでシミュレーションするウェブアプリケーション「Frieve Vinyl」について解説しています。Hertz/Winkler接触理論、粘弾性減衰、剛体動力学といった物理モデルに基づき、レコードの再生プロセスを詳細に再現し、ノイズ源や歪み、トラッキングエラーなどの要因を定量的に分析できる機能を提供しています。

全文翻訳

Frieve Vinyl Explained Microscopic stylus/groove physics simulation Drag=orbit / wheel/pinch=zoom View Language English Japanese Zoom Speed Pause Contact markers (red=no contact) Part labels (by zoom) PVC chain-scale reference (high zoom) Ideal-tracking ghost stylus (offset=tracking error) Stylus transparency Travel-axis ruler (0=stylus) Time Distance Bit Ruler (Digital Reference) L-wall ruler (modulation axis) R-wall ruler (modulation axis) Bits Auto-scale with zoom Signal (Mastering) Signal Pink noise (music-like) Sine wave (test) Silence (noise view) Frequency 1003151000 31501000015000 Peak level HF cutoff Bass mono below Record Speed 33⅓ rpm 45 rpm 78 rpm Groove radius Roughness σ Dust deposition Static Scratches Stylus / Cartridge Stylus shape Elliptical (0.3×0.7 mil) Spherical (0.6 mil) Side radius Scanning radius Tracking force Tip mass Compliance Measurement / Analysis Run Measurement Calibrate to SNR 60dB Measurement time 0.25 s (fast) 0.5 s 1.0 s (more accurate) Frequency response (after phono EQ, 1/24-oct smoothing) Noise spectrum (silent groove) Run Measurement to measure frequency response, SNR, and skip rate for the current settings. It takes a few seconds. Guide (Physics Model) What is this? This app simulates an analog stylus tracing a vinyl groove in slow motion, using microscopic physics rather than a visual approximation: Hertz/Winkler contact, viscoelastic damping, and rigid-body dynamics. All core quantities use SI units, and the observable SNR, distortion, and resonance frequencies are checked against real-world orders of magnitude. Cutting Signal Chain Pink noise (a music-like long-term spectrum) passes through a 20Hz rumble filter, an 80Hz music LF shaper, bass mono filtering, HF cutoff, RIAA recording pre-emphasis including the Neumann 3.18µs pole, and velocity-to-displacement integration. The result drives 45/45 left and right wall normal displacement. The 0dB reference is a 5cm/s peak wall velocity. Contact Physics At 2g VTF, each wall carries roughly 14mN of normal force and the contact pressure reaches about 0.4GPa, well above the yield stress of PVC. Microscopic asperities inside the ~5×7µm contact patch are therefore crushed and averaged. The simulation separates visible roughness from the roughness actually felt by the stylus with an analytic patch-averaging factor. Tracing Distortion At each step, the simulator evaluates the stylus scanning-radius parabola against the wall gap, so inner-groove treble loss and pinch-effect distortion emerge from geometry rather than from an added distortion formula. Use sine mode to inspect it directly. Bit Ruler The ruler divides a ±15µm wall displacement full scale along each 45° modulation axis. The default roughness σ=13.17nm is calibrated to SNR 60.0dB, or 9.7 effective bits; at the default +12dB peak level this corresponds to DR 72dB. White markers show the instantaneous wall displacement. Noise Sources The main noise source is three-scale AR(1) groove roughness representing stamper precision and molecular-scale placement error. Dust follows a physical deposition model with flakes, fibers, and grit; only particles actually pressed by the stylus deform. Static is injected as an electrical pulse, as in a real playback chain, not as a mechanical vibration. Main Simplifications Two degrees of freedom in the groove cross-section; groove-direction vibration and stick-slip are omitted. Roughness is one-dimensional along the groove, with lateral patch averaging handled analytically. Groove walls are rendered as non-overhanging height fields. Wow, flutter, eccentricity, warp, and cartridge electrical loading are outside the model. Validated Orders of Magnitude 1kHz, 5cm/s sine: output error <0.03dB, THD≈0.6%. Contact resonance ~40kHz, cantilever static deflection ~0.3mm, arm resonance ~12Hz. SNR 60.0dB at σ=13.17nm, yielding DR 72dB at the default +12dB peak level.