import os import sys import wave import subprocess import tempfile import numpy as np # Required: point to espeak-ng data bundled with Piper os.environ.setdefault("ESPEAK_DATA_PATH", r"espeak-ng") # ── Config ──────────────────────────────────────────────────────────────────── # Indian female voices from the L2-ARCTIC corpus (Hindi L1 speakers): # SVBI → speaker_id=2 (Indian female) # TNI → speaker_id=9 (Indian female) MODEL_PATH = "/home/nityansco/public_html/nscorm_app/script/en_US-hfc_female-medium.onnx" #MODEL_PATH = "/home/nityansco/public_html/nscorm_app/script/en_US-hfc_male-medium.onnx" SPEAKER_ID = 9 # 2 = SVBI (Indian female) | 9 = TNI (Indian female) OUTPUT_FILE = "female_artict_tts_audio.wav" PIPER_EXE = r"piper" TARGET_RATE = 44100 # upsample output to 44.1 kHz text = ( "The presentation outlines a global corrective action plan initiated by " "Sun Pharmaceutical in response to a U S F D A observation at its Dadra " "site, where multiple unaddressed deviations related to missing GMP " "documents were found." ) #text = sys.argv[1] #OUTPUT_FILE = sys.argv[2] # ── Synthesis ───────────────────────────────────────────────────────────────── def synthesize_with_python_api(text, model_path, syn_params, speaker_id=None): """Use piper-tts Python package directly (preferred).""" from piper.voice import PiperVoice, SynthesisConfig voice = PiperVoice.load(model_path) cfg = SynthesisConfig( speaker_id=speaker_id, length_scale=syn_params["length_scale"], noise_scale=syn_params["noise_scale"], noise_w_scale=syn_params["noise_w_scale"], ) chunks = [] src_rate = 22050 for chunk in voice.synthesize(text, syn_config=cfg): pcm = np.frombuffer(chunk.audio_int16_bytes, dtype=np.int16).astype(np.float32) chunks.append(pcm) src_rate = chunk.sample_rate audio = np.concatenate(chunks) / 32768.0 return audio, src_rate def synthesize_with_subprocess(text, model_path, piper_exe, syn_params, speaker_id=None): """Fall back to piper binary via subprocess.""" tmp = tempfile.NamedTemporaryFile(suffix=".wav", delete=False) tmp.close() cmd = [ piper_exe, "--model", model_path, "--length_scale", str(syn_params["length_scale"]), "--noise_scale", str(syn_params["noise_scale"]), "--noise_w", str(syn_params["noise_w_scale"]), "--output_file", tmp.name, ] if speaker_id is not None: cmd += ["--speaker", str(speaker_id)] proc = subprocess.Popen(cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True) _, err = proc.communicate(text) if proc.returncode != 0: raise RuntimeError(f"piper failed: {err}") with wave.open(tmp.name, "rb") as wf: src_rate = wf.getframerate() raw = wf.readframes(wf.getnframes()) os.unlink(tmp.name) audio = np.frombuffer(raw, dtype=np.int16).astype(np.float32) / 32768.0 return audio, src_rate # Lower noise_scale → crisper consonants; length_scale 1.0 → natural pace syn_params = dict(length_scale=1.3, noise_scale=0.25, noise_w_scale=0.4) try: audio, src_rate = synthesize_with_python_api(text, MODEL_PATH, syn_params, speaker_id=SPEAKER_ID) print("Synthesis: piper-tts Python API") except ImportError: audio, src_rate = synthesize_with_subprocess(text, MODEL_PATH, PIPER_EXE, syn_params, speaker_id=SPEAKER_ID) print("Synthesis: piper subprocess (piper-tts not found in this Python)") # ── Audio enhancement ───────────────────────────────────────────────────────── def soft_compress(audio, threshold=0.30, ratio=3.5, makeup=1.35): """ Soft-knee downward compressor. Brings up quiet consonants and reduces peak-to-average gap so every syllable is equally intelligible — the single biggest clarity booster for non-native TTS voices. """ abs_a = np.abs(audio) gain = np.ones_like(audio) above = abs_a > threshold if above.any(): excess = abs_a[above] - threshold target = threshold + excess / ratio gain[above] = target / abs_a[above] return audio * gain * makeup def enhance_audio(audio, src_rate, target_rate): try: from scipy import signal as sg # 1. High-quality polyphase upsample (22 050 → 44 100 Hz) audio = sg.resample_poly(audio, target_rate, src_rate) nyq = target_rate / 2.0 # 2. Remove DC offset audio -= audio.mean() # 3. High-pass at 80 Hz — removes low-frequency rumble sos_hp = sg.butter(6, 80.0 / nyq, btype="high", output="sos") audio = sg.sosfilt(sos_hp, audio) # 4. Cut muddiness (300–600 Hz) — removes boxiness that blurs words sos_mud = sg.butter(2, [300.0 / nyq, 600.0 / nyq], btype="bandpass", output="sos") audio -= sg.sosfilt(sos_mud, audio) * 0.10 # 5. Strong presence boost (1–3.5 kHz) — primary speech intelligibility band; # consonants (t, d, s, p, k) live here and define sharpness sos_pres = sg.butter(3, [1000.0 / nyq, 3500.0 / nyq], btype="bandpass", output="sos") audio += sg.sosfilt(sos_pres, audio) * 0.30 # 6. Brilliance boost (5–9 kHz) — adds air, attack and sharpness to sibilants sos_brill = sg.butter(2, [5000.0 / nyq, 9000.0 / nyq], btype="bandpass", output="sos") audio += sg.sosfilt(sos_brill, audio) * 0.18 # 7. Gentle low-pass at 13 kHz — removes aliasing without killing brilliance sos_lp = sg.butter(4, 13000.0 / nyq, btype="low", output="sos") audio = sg.sosfilt(sos_lp, audio) except ImportError: # scipy not available — numpy-only FFT upsample print("scipy not found — using numpy-only processing") factor = target_rate // src_rate n = len(audio) fft = np.fft.rfft(audio) fft_pad = np.zeros(n * factor // 2 + 1, dtype=complex) fft_pad[:len(fft)] = fft * factor audio = np.fft.irfft(fft_pad, n * factor) audio -= audio.mean() # 8. Soft-knee compression — evens out dynamics, raises quiet syllables audio = soft_compress(audio, threshold=0.30, ratio=3.5, makeup=1.35) # 9. Safety soft-clip (tanh) to handle any post-compression peaks cleanly audio = np.tanh(audio * 0.95) / 0.95 # 10. Peak-normalise to −0.3 dB for maximum loudness without clipping peak = np.max(np.abs(audio)) if peak > 0: audio *= 0.966 / peak return audio audio = enhance_audio(audio, src_rate, TARGET_RATE) # ── Save ────────────────────────────────────────────────────────────────────── audio_int16 = (audio * 32767).astype(np.int16) with wave.open(OUTPUT_FILE, "wb") as wf: wf.setnchannels(1) wf.setsampwidth(2) wf.setframerate(TARGET_RATE) wf.writeframes(audio_int16.tobytes()) duration = len(audio_int16) / TARGET_RATE print(f"Done: {OUTPUT_FILE} ({duration:.1f}s @{TARGET_RATE} Hz)")