Create app.py

app.py

@@ -0,0 +1,70 @@
+import streamlit as st
+import numpy as np
+import librosa
+import soundfile as sf
+from PIL import Image
+from io import BytesIO
+
+# Process the color information to create audio frequencies with extended parameters
+def process_color_to_audio(array, sample_rate=44100, duration=5, **kwargs):
+    amplitude_modulation = kwargs.get('amplitude_modulation', 1.0)
+    frequency_modulation = kwargs.get('frequency_modulation', 0.0)
+    harmonic_content = kwargs.get('harmonic_content', 1.0)
+    attack_time = kwargs.get('attack_time', 0.005)
+    decay_time = kwargs.get('decay_time', 0.1)
+    sustain_level = kwargs.get('sustain_level', 0.7)
+    release_time = kwargs.get('release_time', 0.3)
+    vibrato_freq = kwargs.get('vibrato_freq', 5.0)
+    vibrato_depth = kwargs.get('vibrato_depth', 0.005)
+
+    normalized_colors = array / 255.0
+    brightness = np.mean(normalized_colors, axis=(0, 1))
+    brightness = brightness ** harmonic_content
+    base_freq = 220.0
+    frequencies = base_freq * (1 + frequency_modulation * brightness)
+    t = np.linspace(0, duration, int(duration * sample_rate), endpoint=False)
+    vibrato = vibrato_depth * np.sin(2 * np.pi * vibrato_freq * t)
+    t_env = np.linspace(0, duration, int(duration * sample_rate), endpoint=False)
+    env = np.ones_like(t_env)
+    env[:int(attack_time * sample_rate)] = np.linspace(0, 1, int(attack_time * sample_rate))
+    env[int(attack_time * sample_rate):int((attack_time + decay_time) * sample_rate)] = np.linspace(1, sustain_level, int(decay_time * sample_rate))
+    env[int((attack_time + decay_time) * sample_rate):int((duration - release_time) * sample_rate)] = sustain_level
+    env[int((duration - release_time) * sample_rate):] = np.linspace(sustain_level, 0, int(release_time * sample_rate))
+    audio_data = env * amplitude_modulation * np.sin(2 * np.pi * (frequencies + vibrato) * t)
+
+    return audio_data, sample_rate
+
+def main():
+    st.title('PNG to Audio Streamlit App')
+    st.write("Upload a PNG image and get the corresponding audio!")
+
+    uploaded_file = st.file_uploader("Choose a PNG image", type="png")
+    if uploaded_file is not None:
+        image = Image.open(uploaded_file)
+        st.image(image, caption='Uploaded PNG image', use_column_width=True)
+
+        with st.expander("Audio Generation Parameters", expanded=True):
+            amplitude_modulation = st.slider("Amplitude Modulation", 0.0, 2.0, 1.0, 0.1)
+            frequency_modulation = st.slider("Frequency Modulation", -1.0, 1.0, 0.0, 0.1)
+            harmonic_content = st.slider("Harmonic Content", 0.1, 2.0, 1.0, 0.1)
+            attack_time = st.slider("Attack Time", 0.001, 0.1, 0.005, 0.001)
+            decay_time = st.slider("Decay Time", 0.01, 0.5, 0.1, 0.01)
+            sustain_level = st.slider("Sustain Level", 0.1, 1.0, 0.7, 0.1)
+            release_time = st.slider("Release Time", 0.1, 1.0, 0.3, 0.1)
+            vibrato_freq = st.slider("Vibrato Frequency", 0.1, 10.0, 5.0, 0.1)
+            vibrato_depth = st.slider("Vibrato Depth", 0.001, 0.01, 0.005, 0.001)
+
+        if st.button("Generate Audio"):
+            array = np.array(image)
+            audio_data, sample_rate = process_color_to_audio(array, amplitude_modulation=amplitude_modulation, frequency_modulation=frequency_modulation,
+                                                             harmonic_content=harmonic_content, attack_time=attack_time,
+                                                             decay_time=decay_time, sustain_level=sustain_level,
+                                                             release_time=release_time, vibrato_freq=vibrato_freq,
+                                                             vibrato_depth=vibrato_depth)
+
+            with BytesIO() as output:
+                sf.write(output, audio_data, sample_rate, format='wav')
+                st.audio(output.getvalue(), format='audio/wav')
+
+if __name__ == '__main__':
+    main()