polish README, viewer header, minor cleanups

.gitignore

@@ -84,5 +84,6 @@ tensorboard/
 
 # ── misc ────────────────────────────────────────────────────────────────
 CLAUDE.md
+.claude
 docs/*
 docs/

README.md

@@ -2,7 +2,7 @@
 
 Reconstruct a continuous 3D sound pressure field from sparse microphone measurements, with the Helmholtz wave equation enforced as a hard constraint during training.
 
-**Apple M4 Pro · No cloud GPU**
+On **Apple M4 Pro (48GB VRam) · No cloud GPU**
 
 ---
 
@@ -123,11 +123,11 @@ where β = 0.4 is the specific admittance (absorption coefficient).
 
 ## Key Implementation Decisions
 
-**SIREN activations (ω₀=1.0, not 30):** Sine activations give smooth second derivatives — critical for the Laplacian computation. ω₀=30 was tested but caused the physics loss to blow up (second derivatives O(1e9) with 8 stacked layers). Lowered to 1.0 since positional encoding already handles high frequencies.
+**SIREN activations (ω₀=1.0, not 30):** Sine activations give smooth second derivatives, critical for the Laplacian computation. ω₀=30 was tested but caused the physics loss to blow up (second derivatives O(1e9) with 8 stacked layers). Lowered to 1.0 since positional encoding already handles high frequencies.
 
 **Finite difference Laplacian:** PyTorch autograd `create_graph=True` is unstable on MPS for second derivatives. Central differences with h=5e-3 give residual error ~0.05 on plane waves (acceptable). Coordinate scaling: multiply each dimension's contribution by (2/L_dim)² to convert from normalized to physical metres.
 
-**2000 microphones:** The model with 128 mics produced visually noisy fields despite good physics metrics — 99.9% Helmholtz residual reduction but no visible wave structure. Root cause: ~800:1 param-to-data ratio. pyroomacoustics computes all mic RIRs in a single call regardless of count (128 mics → 2000 mics costs 1.2s extra). 2000 mics × 4 frequencies = 8000 samples drops the ratio to ~62:1 and produces clear wave rings.
+**2000 microphones:** The model with 128 mics produced visually noisy fields despite good physics metrics. *99.9%* Helmholtz residual reduction but no visible wave structure. Root cause: ~800:1 param-to-data ratio. pyroomacoustics computes all mic RIRs in a single call regardless of count (128 mics → 2000 mics costs 1.2s extra). 2000 mics × 4 frequencies = 8000 samples drops the ratio to ~62:1 and produces clear wave rings.
 
 ---
 
@@ -158,4 +158,4 @@ docs/          raw_log.md, article_draft.md
 
 ---
 
-*Pooya Esfahani · M4 Pro · April 2026*
+*Pooya Esfahani - April 2026*

src/api/main.py

@@ -23,7 +23,7 @@ from src.data.generate_synthetic import generate_room_data
 from src.visualization.export import export_field, DEFAULT_CFG, ROOM_DIMS, SOURCE_POS, FREQUENCIES
 from src.physics.helmholtz import helmholtz_residual_fd
 
-# ── Global state ──────────────────────────────────────────────────────────────
+# Global state
 
 STATE = {}
 
@@ -76,7 +76,7 @@ app.add_middleware(
 app.mount("/viewer", StaticFiles(directory="viewer", html=True), name="viewer")
 
 
-# ── Routes ────────────────────────────────────────────────────────────────────
+#Routes
 
 @app.get("/")
 def root():

src/data/generate_synthetic.py

@@ -1,7 +1,5 @@
 """
 Synthetic room acoustic data generation using pyroomacoustics.
-Produces complex pressure measurements at sparse microphone positions
-and dense ground-truth grids for validation.
 """
 
 import numpy as np

viewer/index.html

@@ -150,7 +150,8 @@
 
 <script>
 (async () => {
-  const isLocal = !window.location.hostname || window.location.hostname === '';
+  const h = window.location.hostname;
+  const isLocal = !h || h === '' || h === 'localhost' || h === '127.0.0.1';
   const fieldUrl = isLocal ? 'field_data.json' : '/viewer/field_data.json';
 
   let pct = 0;