Update app.py

app.py

@@ -1,5 +1,6 @@
 import gradio as gr
 from pydub import AudioSegment
+from pydub.silence import detect_nonsilent
 import numpy as np
 import tempfile
 import os
@@ -102,9 +103,9 @@ def apply_phaser(audio, rate=0.5, depth=0.7, feedback=0.2, mix=0.5):
     return audio._spawn(audio.raw_data, overrides={"frame_rate": int(audio.frame_rate * rate)})
 
 def apply_bitcrush(audio, bit_depth=8):
-    samples = np.array(audio.get_array_of_samples())
+    samples = np.array(audio.get_array_of_samples()).astype(np.float32)
     max_val = np.iinfo(np.int16).max
-    crushed = (samples // (max_val // (2 ** bit_depth))).astype(np.int16)
+    crushed = ((samples / max_val) * (2 ** bit_depth)).astype(np.int16)
     return array_to_audiosegment(crushed, audio.frame_rate, channels=audio.channels)
 
 def apply_auto_gain(audio, target_dB=-20):
@@ -116,6 +117,16 @@ def apply_vocal_distortion(audio, intensity=0.3):
     distorted = samples + intensity * np.sin(samples * 2 * np.pi / 32768)
     return array_to_audiosegment(distorted.astype(np.int16), audio.frame_rate, channels=audio.channels)
 
+def apply_harmony(audio, shift_semitones=4):
+    shifted_up = apply_pitch_shift(audio, shift_semitones)
+    shifted_down = apply_pitch_shift(audio, -shift_semitones)
+    return audio.overlay(shifted_up).overlay(shifted_down)
+
+def apply_stage_mode(audio):
+    processed = apply_reverb(audio)
+    processed = apply_bass_boost(processed, gain=6)
+    return apply_limiter(processed, limit_dB=-2)
+
 # === Vocal Isolation Helpers ===
 def load_track_local(path, sample_rate, channels=2):
     sig, rate = torchaudio.load(path)
@@ -188,7 +199,15 @@ if not preset_choices:
         "Music Production": ["Reverb", "Stereo Widening", "Pitch Shift"],
         "ASMR Creator": ["Noise Gate", "Auto Gain", "Low-Pass Filter"],
         "Voiceover Pro": ["Vocal Isolation", "TTS", "EQ Match"],
-        "8-bit Retro": ["Bitcrusher", "Echo", "Mono Downmix"]
+        "8-bit Retro": ["Bitcrusher", "Echo", "Mono Downmix"],
+
+        # 🎤 Vocalist Presets
+        "🎙 Clean Vocal": ["Noise Reduction", "Normalize", "High Pass Filter (80Hz)"],
+        "🧪 Vocal Distortion": ["Vocal Distortion", "Reverb", "Compress Dynamic Range"],
+        "🎶 Singer's Harmony": ["Harmony", "Stereo Widening", "Pitch Shift"],
+        "🌫 ASMR Vocal": ["Auto Gain", "Low-Pass Filter (3000Hz)", "Noise Gate"],
+        "🎼 Stage Mode": ["Reverb", "Bass Boost", "Limiter"],
+        "🎵 Auto-Tune Style": ["Pitch Shift (+1 semitone)", "Normalize", "Treble Boost"]
     }
 
 preset_names = list(preset_choices.keys())
@@ -252,7 +271,9 @@ def process_audio(audio_file, selected_effects, isolate_vocals, preset_name, exp
             "Flanger": lambda x: apply_phaser(x, rate=1.2, depth=0.9, mix=0.7),
             "Bitcrusher": lambda x: apply_bitcrush(x, bit_depth=8),
             "Auto Gain": lambda x: apply_auto_gain(x, target_dB=-20),
-            "Vocal Distortion": lambda x: apply_vocal_distortion(x)
+            "Vocal Distortion": lambda x: apply_vocal_distortion(x),
+            "Harmony": lambda x: apply_harmony(x),
+            "Stage Mode": apply_stage_mode
         }
 
         effects_to_apply = preset_choices.get(preset_name, selected_effects)
@@ -317,7 +338,7 @@ def transcribe_audio(audio_path):
     text = " ".join([seg.text for seg in segments])
     return text
 
-# === TTS Tab ===
+# === TTS Voice Generator ===
 tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC", progress_bar=False)
 
 def generate_tts(text):
@@ -355,7 +376,7 @@ def detect_silence(audio_file, silence_threshold=-50.0, min_silence_len=1000):
     if not nonsilent_ranges:
         return audio.export(os.path.join(tempfile.gettempdir(), "trimmed.wav"), format="wav")
 
-    trimmed = audio[nonsilent_ranges[0][0]:nonsilent_tracks[-1][1]]
+    trimmed = audio[nonsilent_ranges[0][0]:nonsilent_ranges[-1][1]]
     out_path = os.path.join(tempfile.gettempdir(), "trimmed.wav")
     trimmed.export(out_path, format="wav")
     return out_path
@@ -369,7 +390,7 @@ def mix_tracks(track1, track2, volume_offset=0):
     mixed.export(out_path, format="wav")
     return out_path
 
-# === Dummy Voice Cloning Tab – Works on Local Only ===
+# === Dummy Voice Cloning Tab – Works Locally Only ===
 def clone_voice(*args):
     return "⚠️ Voice cloning requires local install – use Python 3.9 or below"
 
@@ -432,7 +453,9 @@ effect_options = [
     "Flanger",
     "Bitcrusher",
     "Auto Gain",
-    "Vocal Distortion"
+    "Vocal Distortion",
+    "Harmony",
+    "Stage Mode"
 ]
 
 with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
@@ -512,6 +535,47 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
             description="Convert voice to text and edit it before exporting again."
         )
 
+    # --- Vocal Presets for Singers ===
+    with gr.Tab("🎤 Vocal Presets for Singers"):
+        gr.Interface(
+            fn=process_audio,
+            inputs=[
+                gr.Audio(label="Upload Vocal Track", type="filepath"),
+                gr.CheckboxGroup(choices=[
+                    "Noise Reduction",
+                    "Normalize",
+                    "Compress Dynamic Range",
+                    "Bass Boost",
+                    "Treble Boost",
+                    "Reverb",
+                    "Auto Gain",
+                    "Vocal Distortion",
+                    "Harmony",
+                    "Stage Mode"
+                ]),
+                gr.Checkbox(label="Isolate Vocals After Effects"),
+                gr.Dropdown(choices=[
+                    "🎙 Clean Vocal",
+                    "🧪 Vocal Distortion",
+                    "🎶 Singer's Harmony",
+                    "🌫 ASMR Vocal",
+                    "🎼 Stage Mode",
+                    "🎵 Auto-Tune Style"
+                ], label="Select Vocal Preset", value="Default"),
+                gr.Dropdown(choices=["MP3", "WAV"], label="Export Format", value="MP3")
+            ],
+            outputs=[
+                gr.Audio(label="Processed Vocal", type="filepath"),
+                gr.Image(label="Waveform Preview"),
+                gr.Textbox(label="Session Log (JSON)", lines=5),
+                gr.Textbox(label="Detected Genre", lines=1),
+                gr.Textbox(label="Status", value="✅ Ready", lines=1)
+            ],
+            title="Create Studio-Quality Vocal Tracks",
+            description="Apply singer-friendly presets and effects to enhance vocals.",
+            allow_flagging="never"
+        )
+
     # --- Voice Cloning (Local Only) ===
     with gr.Tab("🎭 Voice Cloning (Local Only)"):
         gr.Interface(
@@ -563,7 +627,7 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
         return None, None, None, None
 
     with gr.Tab("🧾 Auto-Save & Resume"):
-        gr.Markdown("Save your current state and resume editing later.")
+        gr.Markdown("Save your current state and resume later.")
 
         action_radio = gr.Radio(["save", "load"], label="Action", value="save")
         audio_input = gr.Audio(label="Upload or Load Audio", type="filepath")
@@ -622,7 +686,7 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
         )
 
     # --- Mix Two Tracks ===
-    with gr.Tab("🔀 Mix Two Tracks"):
+    with gr.Tab(" remix mode"),
         gr.Interface(
             fn=mix_tracks,
             inputs=[