bumped to version 1.3

changed defaults for matching threshold, alternations, tone_a_min and tone_b min

refactored frequency_extraction.py to get even more precise time for tones.

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "icad_tone_detection"
-version = "1.2"
+version = "1.3"
 authors = [
   {name = "TheGreatCodeholio", email = "[email protected]"},
 ]

src/icad_tone_detection/audio_loader.py

@@ -38,7 +38,8 @@ def load_audio(audio_input):
         audio = audio.set_channels(1)  # Ensure the audio is mono
         audio = audio.set_frame_rate(22050)  # Set the frame rate to 22050 Hz
         samples = np.array(audio.get_array_of_samples()).astype(np.float32)
-        samples /= np.iinfo(audio.sample_width * 8).max  # Adjusted for correct normalization
+        max_val = float(2 ** (audio.sample_width * 8 - 1))
+        samples /= max_val
     except Exception as e:
         raise RuntimeError(f"Error processing audio: {e}")
 

src/icad_tone_detection/frequency_extraction.py

@@ -78,7 +78,7 @@ def calculate_times(self, start_index, end_index, time_samples):
         Calculates accurate start and end times for frequency matches.
         """
         start_time = round(time_samples[start_index], 3)
-        end_time = round(time_samples[end_index - 1], 3)
+        end_time = round(time_samples[end_index], 3)  # Use end_index directly
         return start_time, end_time
 
     @staticmethod
@@ -99,15 +99,15 @@ def amplitude_to_decibels(amplitude, reference_value):
 
     def match_frequencies(self, detected_frequencies, time_samples):
         """
-            Identifies and groups matching frequencies from a list of detected frequencies based on the matching threshold.
-            Each group's start time, end time, and the matching frequencies are returned.
+        Identifies and groups matching frequencies from a list of detected frequencies based on the matching threshold.
+        Each group's start time, end time, and the matching frequencies are returned.
 
-            Parameters:
-                detected_frequencies (list of float): The detected frequencies from the audio sample.
-                time_samples (np.array): Array of times corresponding to each frequency sample.
+        Parameters:
+            detected_frequencies (list of float): The detected frequencies from the audio sample.
+            time_samples (np.array): Array of times corresponding to each frequency sample.
 
-            Returns:
-                list of tuples: Each tuple contains the start time, end time, and a list of matching frequencies.
+        Returns:
+            list of tuples: Each tuple contains the start time, end time, and a list of matching frequencies.
         """
 
         if not detected_frequencies:
@@ -125,15 +125,15 @@ def match_frequencies(self, detected_frequencies, time_samples):
                     current_match.append(frequencies[i])
                 else:
                     if len(current_match) >= 2:
-                        start_time, end_time = self.calculate_times(start_index, i, time_samples)
-                        freq_length = round((end_time - start_time) + .1, 2)
+                        start_time, end_time = self.calculate_times(start_index, i - 1, time_samples)
+                        freq_length = round(end_time - start_time, 3)
                         matching_frequencies.append((start_time, end_time, freq_length, current_match))
                     current_match = [frequencies[i]]
                     start_index = i
 
             if len(current_match) >= 2:
-                start_time, end_time = self.calculate_times(start_index, len(frequencies), time_samples)
-                freq_length = round((end_time - start_time) + .1, 2)
+                start_time, end_time = self.calculate_times(start_index, len(frequencies) - 1, time_samples)
+                freq_length = round(end_time - start_time, 3)
                 matching_frequencies.append((start_time, end_time, freq_length, current_match))
 
             return matching_frequencies

src/icad_tone_detection/main.py

@@ -10,22 +10,22 @@ def __init__(self, two_tone_result, long_result, hi_low_result):
         self.hi_low_result = hi_low_result
 
 
-def tone_detect(audio_path, matching_threshold=2, time_resolution_ms=100, tone_a_min_length=0.8, tone_b_min_length=2.8, hi_low_interval=0.2,
-                hi_low_min_alternations=4, long_tone_min_length=3.8, debug=False):
+def tone_detect(audio_path, matching_threshold=2.5, time_resolution_ms=25, tone_a_min_length=0.7, tone_b_min_length=2.7, hi_low_interval=0.2,
+                hi_low_min_alternations=6, long_tone_min_length=3.8, debug=False):
     """
         Loads audio from various sources including local path, URL, BytesIO object, or a PyDub AudioSegment.
 
         Parameters:
            - audio_input: Can be a string (path or URL), bytes like object, or AudioSegment.
            - matching_threshold (float): The percentage threshold used to determine if two frequencies
                 are considered a match. For example, a threshold of 2 means that two frequencies are considered matching
-                if they are within 2% of each other.
-           - time_resolution_ms (int): The time resolution in milliseconds for the STFT. Default is 100ms.
+                if they are within x% of each other. Default 2.5%
+           - time_resolution_ms (int): The time resolution in milliseconds for the STFT. Default is 25ms.
            - tone_a_min_length (float): The minimum length in seconds of an A tone for two tone detections. Default 0.8 Seconds
            - tone_b_min_length (float): The minimum length in seconds of a B tone for two tone detections. Default 2.8 Seconds
            - long_tone_min_length (float): The minimum length a long tone needs to be to consider it a match. Default 3.8 Seconds
            - hi_low_interval (float): The maximum allowed interval in seconds between two consecutive alternating tones. Default is 0.2 Seconds
-           - hi_low_min_alternations (int): The minimum number of alternations for a hi-low warble tone sequence to be considered valid. Default 4
+           - hi_low_min_alternations (int): The minimum number of alternations for a hi-low warble tone sequence to be considered valid. Default 6
            - debug (bool): If debug is enabled, print all tones found in audio file. Default is False
 
         Returns: