My contribution to the MOSQITO repository

docs/sound-level-meter.md

@@ -0,0 +1,28 @@
+# MOSQITO Documentation
+## Sound level meter functions
+
+### Introduction
+
+There are operations that are important. We need to be able to perform them to have a better understanding of the data with which we work and to be able to use them to calculate data of interest. For this are the sonometer functions.
+
+A step by step description of how to use MOSQITO to use the sonometer functions is given in [this tutorial](../tutorials/tuto_sound_level_meter.ipynb)
+
+### Validation of the implementation
+
+This section of the MoSQITo repository are mathematical accounts. For this reason, the validations have consisted of comparing two results: the results of the software and the simulated results in Excel.
+
+Accounts are backed by regulations. Among them we have ISO 1996 part 1, ISO 1996 part 2 and the international standard IEC 61672: 2014 (This is for the A weighting and the C weighting).
+
+We have taken three signals to carry out the calculations. These signals are located at ../tests/input and These are a white noise (349315__newagesoup__white-noise-10s.wav), a pink noise (349312__newagesoup__pink-noise-10s.wav) and a pure tone at 1000 Hz (554329__patobottos__beep-sound-of-1000-hz).
+
+For the functions that work with Thirds of an Octave, the three signals (input array) were introduced to the program. In the same way, the data of the signals are entered in an Excel in which the desired operation of the function has been replicated. If the function and Excel return the same result, it is shown that the function works well and returns the correct result.
+
+In the case of functions that work with temporary samples, only one of these three signals is entered and the procedure is the same, results are compared between the function and Excel.
+
+### References
+
+ISO 1996 - 1: Description, measurement and evaluation of environmental noise Part 1: Basic magnitudes and evaluation methods.
+
+ISO 1996 - 2: Description, measurement and evaluation of environmental noise Part 2: determination of sound pressure levels.
+
+IEC 61672-1:2014, Electroacoustics - Sound level meters - Part 1: Specifications.

mosqito/functions/sound_level_meter/LN.py

@@ -0,0 +1,46 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Thu Jan 12 19:45:20 2023
+
+@author: Igarciac117 
+"""
+
+# Third party imports
+import numpy as np
+
+# Local imports
+from mosqito.utils.conversion import amp2db
+
+def LN(signal):
+    """Calculate the percentile value of the series of levels (dB) of the one signal.  
+
+    Parameters
+    ----------
+    signal : numpy.array
+        time signal values [Pa]
+
+    Outputs
+    -------
+    percentiles : numpy.ndarray
+        The values in dB of L90, L50 and L25 of the signal.
+    """
+    # Empty array to store the values in dB of the signal.
+    dB_values = np.zeros(signal.shape[0])
+    # Performs the conversion to dB with all the values of the signal.
+    for i in range(signal.shape[0]):
+        #If the value is negative value.
+        if signal[i] <= 0:
+            # we convert it to positive.
+            signal[i] = np.sqrt(np.mean(signal[i] ** 2))
+        # Conversion Pa to dB.
+        dB = amp2db(np.array(signal[i]))
+        # Save all values in dB of the third octave in another array.
+        dB_values[i] = dB
+    # Calculate the percentiles with the values. "q" of np.percentile = 100 - N (N of LN).
+    L90 = np.percentile(dB_values, 10,interpolation='linear') 
+    L50 = np.percentile(dB_values, 50,interpolation='linear') 
+    L25 = np.percentile(dB_values, 75,interpolation='linear')
+    # Save the calculated percentile values.
+    percentiles = np.array([L90,L50,L25])
+
+    return percentiles

mosqito/functions/sound_level_meter/max_level.py

@@ -0,0 +1,42 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Wen Dic 1 18:08:08 2021
+
+@author: Igarciac117 
+"""
+
+# Third party imports
+import numpy as np
+
+# Local imports
+from mosqito.utils.conversion import amp2db
+
+def max_level(signal):
+    """Calculate the maximum value of the series of levels (dB) of the one signal.
+
+    Parameters
+    ----------
+    signal : numpy.array
+        time signal values
+
+    Outputs
+    -------
+    max_level : numpy.ndarray
+        return the maximum value of the signal values.
+    """
+    # Empty array to store the values in dB of the signal.
+    dB_values = np.zeros(signal.shape[0])
+    # Performs the conversion to dB with all the values of the signal.
+    for i in range(signal.shape[0]):
+        #If the value is negative value.
+        if signal[i] <= 0:
+            # we convert it to positive.
+            signal[i] = np.sqrt(np.mean(signal[i] ** 2))
+        # Conversion Pa to dB.
+        dB = amp2db(np.array(signal[i]))
+        # Save all values in dB of the third octave in another array.
+        dB_values[i] = dB
+    # Save the maximum level.
+    max_level = np.array(max(dB_values))
+
+    return max_level

mosqito/functions/sound_level_meter/max_level_3oct.py

@@ -0,0 +1,66 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Thu Jan 12 19:40:20 2023
+
+@author: Igarciac117 
+"""
+
+# Third party imports
+import numpy as np
+
+# Local imports
+from mosqito.sound_level_meter.noct_spectrum.noct_spectrum import noct_spectrum
+from mosqito.utils.conversion import amp2db
+
+def max_level_3oct(data_all_signals, fs, f_min, f_max):
+    """Return the maximum value of the frequency bands you choose. Each one is calculated with the levels (dB)
+    of its band in the different signals.
+
+    Parameters
+    ----------
+    data_all_signals : numpy.ndarray
+        Array which each row corresponds to the data of a signal [Pa].
+    fs : float
+        Sampling frequency [Hz].
+    fmax : float
+        Max frequency band [Hz].
+    fmin : float
+        Min frequency band [Hz].
+
+    Outputs
+    -------
+    max_level_3oct : numpy.ndarray
+        The maximum values of each frequency band.
+    """
+    # We initialize the array that stores the third octave values (in Pa) of the all signals ​​with the first signal.
+    spectrum_all_signals_Pa = noct_spectrum(data_all_signals[0],fs,f_min,f_max)[0]
+    # We initialize the center frequencies of the third octaves with the first signal.
+    freq = noct_spectrum(data_all_signals[0],fs,f_min,f_max)[1]
+    # We initialize the number of the signals.
+    num_signals = data_all_signals.shape[0]
+    # We initialize the number of frequency bands.
+    num_bands = freq.shape[0]
+
+    # Calculate the value of the third octave in Pa of each signal.
+    for i in range(num_signals):
+        # We skip the first signal because we have initialized with it.
+        if i != 0:
+            # We calculate and save the values ​​of the third octaves of the signals
+            spectrum_all_signals_Pa = np.append(spectrum_all_signals_Pa,noct_spectrum(data_all_signals[i],fs,f_min,f_max)[0],axis=1)
+
+    # Creating a list of zeros of the size of the frequency bands (to keep the maximum level values).
+    max_level_3oct = np.zeros(num_bands)
+    # Empty array to store the values in dB of the third octave whose maximum value is going to be calculated.
+    band_value_all_signals = np.zeros(num_signals)
+    # For each frequency band you perform the operation.
+    for i in range(num_bands):
+        # Performs the conversion to dB with all the values of the frequency band in the different signals.
+        for j in range(num_signals): 
+            # Conversion Pa to dB.
+            dB = amp2db(np.array(spectrum_all_signals_Pa[i][j]))
+            # Save all values in dB of the third octave in another array.
+            band_value_all_signals[j] = dB
+        # Calculate and keep the maximum value found in the array. That value will be the maximum of the third of an octave.
+        max_level_3oct[i] = max(band_value_all_signals)
+
+    return max_level_3oct

mosqito/functions/sound_level_meter/min_level.py

@@ -0,0 +1,42 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Thu Jan 12 19:44:20 2023
+
+@author: Igarciac117 
+"""
+
+# Third party imports
+import numpy as np
+
+# Local imports
+from mosqito.utils.conversion import amp2db
+
+def min_level(signal):
+    """Calculate the minimum value of the series of levels (dB) of the one signal.
+
+    Parameters
+    ----------
+    signal : numpy.array
+        time signal values
+
+    Outputs
+    -------
+    min_level : numpy.ndarray
+        return the minimum value of the signal values.
+    """
+    # Empty array to store the values in dB of the signal.
+    dB_values = np.zeros(signal.shape[0])
+    # Performs the conversion to dB with all the values of the signal.
+    for i in range(signal.shape[0]):
+        #If the value is negative value.
+        if signal[i] <= 0:
+            # we convert it to positive.
+            signal[i] = np.sqrt(np.mean(signal[i] ** 2))
+        # Conversion Pa to dB.
+        dB = amp2db(np.array(signal[i]))
+        # Save all values in dB of the third octave in another array.
+        dB_values[i] = dB
+    # Save the minimum level.
+    min_level = np.array(min(dB_values))
+
+    return min_level

mosqito/functions/sound_level_meter/min_level_3oct.py

@@ -0,0 +1,66 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Thu Jan 12 19:43:20 2023
+
+@author: Igarciac117 
+"""
+
+# Third party imports
+import numpy as np
+
+# Local imports
+from mosqito.sound_level_meter.noct_spectrum.noct_spectrum import noct_spectrum
+from mosqito.utils.conversion import amp2db
+
+def min_level_3oct(data_all_signals, fs, f_min, f_max):
+    """Return the minimum value of the frequency bands you choose. Each one is calculated with the levels (dB)
+    of its band in the different signals.
+
+    Parameters
+    ----------
+    data_all_signals : numpy.ndarray
+        Array which each row corresponds to the data of a signal [Pa].
+    fs : float
+        Sampling frequency [Hz].
+    fmax : float
+        Max frequency band [Hz].
+    fmin : float
+        Min frequency band [Hz].
+
+    Outputs
+    -------
+    min_level_3oct : numpy.ndarray
+        The minimum values of each frequency band.
+    """
+    # We initialize the array that stores the third octave values (in Pa) of the all signals ​​with the first signal.
+    spectrum_all_signals_Pa = noct_spectrum(data_all_signals[0],fs,f_min,f_max)[0]
+    # We initialize the center frequencies of the third octaves with the first signal.
+    freq = noct_spectrum(data_all_signals[0],fs,f_min,f_max)[1]
+    # We initialize the number of the signals.
+    num_signals = data_all_signals.shape[0]
+    # We initialize the number of frequency bands.
+    num_bands = freq.shape[0]
+
+    # Calculate the value of the third octave in Pa of each signal.
+    for i in range(num_signals):
+        # We skip the first signal because we have initialized with it.
+        if i != 0:
+            # We calculate and save the values ​​of the third octaves of the signals
+            spectrum_all_signals_Pa = np.append(spectrum_all_signals_Pa,noct_spectrum(data_all_signals[i],fs,f_min,f_max)[0],axis=1)
+
+    # Creating a list of zeros of the size of the frequency bands (to keep the maximum level values).
+    min_level_3oct = np.zeros(num_bands)
+    # Empty array to store the values in dB of the third octave whose maximum value is going to be calculated.
+    band_value_all_signals = np.zeros(num_signals)
+    # For each frequency band you perform the operation.
+    for i in range(num_bands):
+        # Performs the conversion to dB with all the values of the frequency band in the different signals.
+        for j in range(num_signals): 
+            # Conversion Pa to dB.
+            dB = amp2db(np.array(spectrum_all_signals_Pa[i][j]))
+            # Save all values in dB of the third octave in another array.
+            band_value_all_signals[j] = dB
+        # Calculate and keep the maximum value found in the array. That value will be the maximum of the third of an octave.
+        min_level_3oct[i] = min(band_value_all_signals)
+
+    return min_level_3oct