From 08983d7e726853f8a3312b87dcf176bd14303de4 Mon Sep 17 00:00:00 2001
From: Joel Mathew Thomas <90510078+joelmathewthomas@users.noreply.github.com>
Date: Mon, 6 Jan 2025 22:11:49 +0530
Subject: [PATCH] add code to generate spectogram array and plot_data

---
 src/spectogram/generator.py | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
 create mode 100644 src/spectogram/generator.py

diff --git a/src/spectogram/generator.py b/src/spectogram/generator.py
new file mode 100644
index 0000000..8fa230d
--- /dev/null
+++ b/src/spectogram/generator.py
@@ -0,0 +1,34 @@
+import librosa
+import numpy as np
+
+def generate_spectogram(audio_file: str, spectogram_type: str = 'mel', sr: int = 22050):
+    """
+    Generates a spectrogram array from an audio file.
+
+    Args:
+        audio_file (str): Path to the input audio file.
+        spectrogram_type (str): Type of spectrogram ('stft', 'mel'). Default is 'mel'.
+        sr (int): Sampling rate for audio. Default is 22050.
+
+    Returns:
+        tuple: A tuple containing:
+            - np.ndarray: Spectrogram array (in decibels).
+            - dict: Metadata for plotting (sr, x_axis, y_axis).
+    """
+
+    # Load the audio file
+    waveform, sr = librosa.load(audio_file, sr=sr)
+
+    # Create the spectogram
+    if spectogram_type == 'mel':
+        spec = librosa.feature.melspectrogram(y=waveform, sr=sr)
+        spec_db = librosa.power_to_db(spec, ref=np.max) # Convert to decibels
+        plot_data = {'sr': sr, 'x_axis': 'time', 'y_axis': 'mel'}
+    elif spectogram_type == 'stft':
+        spec = np.abs(librosa.stft(waveform))
+        spec_db = librosa.amplitude_to_db(spec, ref=np.max)
+        plot_data = {'sr': sr, 'x_axis': 'time', 'y_axis': 'log'}
+    else:
+        raise ValueError(f"Unsupported spectogram type: {spectogram_type}")
+
+    return spec_db, plot_data