PacificSoundClassifyBlueA
Copyright (c) 2021, MBARI
- Distributed under the terms of the GPL License
- Maintainer: dcline@mbari.org
- Authors: Danelle Cline dcline@mbari.org, John Ryan ryjo@mbari.org
Kernel Selection¶
If running in SageMaker, the Python 3 (Data Science) kernel is sufficient. The Python 3 (TensorFlow 2.3 Python 3.7 GPU Optimized) will run the inference code faster for a higher cost. For more advanced users, SageMaker Batch Transform is recommended to process data in bulk.
Applying Machine Learning to classify blue whale A calls¶
Essential to detection and classification of marine mammal vocalizations are the distinct acoustic attributes of those vocalizations. Machine learning (ML) is an effective way to recognize acoustic attributes and reliably classify such vocalizations.
In this brief tutorial, we will:
- tap into an extensive (6+ years and growing) archive of sound recordings from a deep-sea location along the eastern margin of the North Pacific Ocean,
- illustrate the beautiful songs produced by baleen whales, and
- demonstrate the application of ML to classify one of the three types of calls produced by blue whales in their songs.
If you use this data set or tutorial, please cite our project.
Install dependencies¶
First, let's install dependencies and include all packages used in this tutorial. This only needs to be done once for the duration of this notebook.
In [1]:
Copied!
!pip install oceansoundscape==1.0.1
!pip install boto3
!pip install soundfile
!pip install tensorflow==2.4.1
import boto3
from botocore import UNSIGNED
from botocore.client import Config
from oceansoundscape.spectrogram.signal import psd_1sec
from oceansoundscape.raven import BLEDParser
from oceansoundscape.spectrogram import conf, denoise, utils
import os
from PIL import Image
from pathlib import Path
import soundfile as sf
import json
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
from matplotlib.patches import Rectangle
!pip install oceansoundscape==1.0.1
!pip install boto3
!pip install soundfile
!pip install tensorflow==2.4.1
import boto3
from botocore import UNSIGNED
from botocore.client import Config
from oceansoundscape.spectrogram.signal import psd_1sec
from oceansoundscape.raven import BLEDParser
from oceansoundscape.spectrogram import conf, denoise, utils
import os
from PIL import Image
from pathlib import Path
import soundfile as sf
import json
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
from matplotlib.patches import Rectangle
Requirement already satisfied: oceansoundscape==1.0.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (1.0.1)
Requirement already satisfied: librosa==0.8.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (0.8.1)
Requirement already satisfied: opencv-python==4.5.3.56 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (4.5.3.56)
Collecting numpy==1.20
Using cached numpy-1.20.0-cp37-cp37m-manylinux2010_x86_64.whl (15.3 MB)
Requirement already satisfied: pandas==1.3.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (1.3.2)
Collecting h5py==3.4.0
Using cached h5py-3.4.0-cp37-cp37m-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (4.1 MB)
Requirement already satisfied: SoundFile==0.10.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (0.10.2)
Requirement already satisfied: scipy==1.6.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (1.6.2)
Requirement already satisfied: Pillow==8.3.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (8.3.1)
Requirement already satisfied: matplotlib==3.4.3 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from oceansoundscape==1.0.1) (3.4.3)
Requirement already satisfied: cached-property in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from h5py==3.4.0->oceansoundscape==1.0.1) (1.5.2)
Requirement already satisfied: numba>=0.43.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (0.54.0)
Requirement already satisfied: scikit-learn!=0.19.0,>=0.14.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (1.0)
Requirement already satisfied: audioread>=2.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (2.1.9)
Requirement already satisfied: pooch>=1.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (1.5.1)
Requirement already satisfied: resampy>=0.2.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (0.2.2)
Requirement already satisfied: decorator>=3.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (5.1.0)
Requirement already satisfied: packaging>=20.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (21.0)
Requirement already satisfied: joblib>=0.14 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from librosa==0.8.1->oceansoundscape==1.0.1) (1.0.1)
Requirement already satisfied: pyparsing>=2.2.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from matplotlib==3.4.3->oceansoundscape==1.0.1) (2.4.7)
Requirement already satisfied: python-dateutil>=2.7 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from matplotlib==3.4.3->oceansoundscape==1.0.1) (2.8.2)
Requirement already satisfied: kiwisolver>=1.0.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from matplotlib==3.4.3->oceansoundscape==1.0.1) (1.3.2)
Requirement already satisfied: cycler>=0.10 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from matplotlib==3.4.3->oceansoundscape==1.0.1) (0.10.0)
Requirement already satisfied: pytz>=2017.3 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from pandas==1.3.2->oceansoundscape==1.0.1) (2021.1)
Requirement already satisfied: cffi>=1.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from SoundFile==0.10.2->oceansoundscape==1.0.1) (1.14.6)
Requirement already satisfied: pycparser in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from cffi>=1.0->SoundFile==0.10.2->oceansoundscape==1.0.1) (2.20)
Requirement already satisfied: six in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from cycler>=0.10->matplotlib==3.4.3->oceansoundscape==1.0.1) (1.15.0)
Requirement already satisfied: llvmlite<0.38,>=0.37.0rc1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from numba>=0.43.0->librosa==0.8.1->oceansoundscape==1.0.1) (0.37.0)
Requirement already satisfied: setuptools in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from numba>=0.43.0->librosa==0.8.1->oceansoundscape==1.0.1) (58.1.0)
Requirement already satisfied: requests in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from pooch>=1.0->librosa==0.8.1->oceansoundscape==1.0.1) (2.26.0)
Requirement already satisfied: appdirs in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from pooch>=1.0->librosa==0.8.1->oceansoundscape==1.0.1) (1.4.4)
Requirement already satisfied: threadpoolctl>=2.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from scikit-learn!=0.19.0,>=0.14.0->librosa==0.8.1->oceansoundscape==1.0.1) (2.2.0)
Requirement already satisfied: charset-normalizer~=2.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests->pooch>=1.0->librosa==0.8.1->oceansoundscape==1.0.1) (2.0.6)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests->pooch>=1.0->librosa==0.8.1->oceansoundscape==1.0.1) (1.26.7)
Requirement already satisfied: idna<4,>=2.5 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests->pooch>=1.0->librosa==0.8.1->oceansoundscape==1.0.1) (3.2)
Requirement already satisfied: certifi>=2017.4.17 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests->pooch>=1.0->librosa==0.8.1->oceansoundscape==1.0.1) (2021.5.30)
Installing collected packages: numpy, h5py
Attempting uninstall: numpy
Found existing installation: numpy 1.19.5
Uninstalling numpy-1.19.5:
Successfully uninstalled numpy-1.19.5
Attempting uninstall: h5py
Found existing installation: h5py 2.10.0
Uninstalling h5py-2.10.0:
Successfully uninstalled h5py-2.10.0
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
tensorflow 2.4.1 requires h5py~=2.10.0, but you have h5py 3.4.0 which is incompatible.
tensorflow 2.4.1 requires numpy~=1.19.2, but you have numpy 1.20.0 which is incompatible.
Successfully installed h5py-3.4.0 numpy-1.20.0
Requirement already satisfied: boto3 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (1.18.48)
Requirement already satisfied: jmespath<1.0.0,>=0.7.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from boto3) (0.10.0)
Requirement already satisfied: s3transfer<0.6.0,>=0.5.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from boto3) (0.5.0)
Requirement already satisfied: botocore<1.22.0,>=1.21.48 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from boto3) (1.21.48)
Requirement already satisfied: python-dateutil<3.0.0,>=2.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from botocore<1.22.0,>=1.21.48->boto3) (2.8.2)
Requirement already satisfied: urllib3<1.27,>=1.25.4 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from botocore<1.22.0,>=1.21.48->boto3) (1.26.7)
Requirement already satisfied: six>=1.5 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from python-dateutil<3.0.0,>=2.1->botocore<1.22.0,>=1.21.48->boto3) (1.15.0)
Requirement already satisfied: soundfile in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (0.10.2)
Requirement already satisfied: cffi>=1.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from soundfile) (1.14.6)
Requirement already satisfied: pycparser in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from cffi>=1.0->soundfile) (2.20)
Requirement already satisfied: tensorflow==2.4.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (2.4.1)
Requirement already satisfied: tensorflow-estimator<2.5.0,>=2.4.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (2.4.0)
Requirement already satisfied: protobuf>=3.9.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (3.18.0)
Collecting numpy~=1.19.2
Using cached numpy-1.19.5-cp37-cp37m-manylinux2010_x86_64.whl (14.8 MB)
Requirement already satisfied: astunparse~=1.6.3 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.6.3)
Requirement already satisfied: keras-preprocessing~=1.1.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.1.2)
Requirement already satisfied: opt-einsum~=3.3.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (3.3.0)
Collecting h5py~=2.10.0
Using cached h5py-2.10.0-cp37-cp37m-manylinux1_x86_64.whl (2.9 MB)
Requirement already satisfied: google-pasta~=0.2 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (0.2.0)
Requirement already satisfied: flatbuffers~=1.12.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.12)
Requirement already satisfied: termcolor~=1.1.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.1.0)
Requirement already satisfied: wheel~=0.35 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (0.37.0)
Requirement already satisfied: grpcio~=1.32.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.32.0)
Requirement already satisfied: absl-py~=0.10 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (0.14.0)
Requirement already satisfied: typing-extensions~=3.7.4 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (3.7.4.3)
Requirement already satisfied: wrapt~=1.12.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.12.1)
Requirement already satisfied: six~=1.15.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (1.15.0)
Requirement already satisfied: gast==0.3.3 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (0.3.3)
Requirement already satisfied: tensorboard~=2.4 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorflow==2.4.1) (2.6.0)
Requirement already satisfied: werkzeug>=0.11.15 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (2.0.1)
Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (0.6.1)
Requirement already satisfied: google-auth<2,>=1.6.3 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (1.35.0)
Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (1.8.0)
Requirement already satisfied: markdown>=2.6.8 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (3.3.4)
Requirement already satisfied: setuptools>=41.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (58.1.0)
Requirement already satisfied: requests<3,>=2.21.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (2.26.0)
Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from tensorboard~=2.4->tensorflow==2.4.1) (0.4.6)
Requirement already satisfied: rsa<5,>=3.1.4 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.4->tensorflow==2.4.1) (4.7.2)
Requirement already satisfied: cachetools<5.0,>=2.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.4->tensorflow==2.4.1) (4.2.2)
Requirement already satisfied: pyasn1-modules>=0.2.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.4->tensorflow==2.4.1) (0.2.8)
Requirement already satisfied: requests-oauthlib>=0.7.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.4->tensorflow==2.4.1) (1.3.0)
Requirement already satisfied: importlib-metadata in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from markdown>=2.6.8->tensorboard~=2.4->tensorflow==2.4.1) (4.8.1)
Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from pyasn1-modules>=0.2.1->google-auth<2,>=1.6.3->tensorboard~=2.4->tensorflow==2.4.1) (0.4.8)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.4->tensorflow==2.4.1) (1.26.7)
Requirement already satisfied: certifi>=2017.4.17 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.4->tensorflow==2.4.1) (2021.5.30)
Requirement already satisfied: idna<4,>=2.5 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.4->tensorflow==2.4.1) (3.2)
Requirement already satisfied: charset-normalizer~=2.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.4->tensorflow==2.4.1) (2.0.6)
Requirement already satisfied: oauthlib>=3.0.0 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.4->tensorflow==2.4.1) (3.1.1)
Requirement already satisfied: zipp>=0.5 in /raid/dcline-admin/pacific-sound/docs/.venv/lib/python3.7/site-packages (from importlib-metadata->markdown>=2.6.8->tensorboard~=2.4->tensorflow==2.4.1) (3.5.0)
Installing collected packages: numpy, h5py
Attempting uninstall: numpy
Found existing installation: numpy 1.20.0
Uninstalling numpy-1.20.0:
Successfully uninstalled numpy-1.20.0
Attempting uninstall: h5py
Found existing installation: h5py 3.4.0
Uninstalling h5py-3.4.0:
Successfully uninstalled h5py-3.4.0
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
oceansoundscape 1.0.1 requires h5py==3.4.0, but you have h5py 2.10.0 which is incompatible.
oceansoundscape 1.0.1 requires numpy==1.20, but you have numpy 1.19.5 which is incompatible.
Successfully installed h5py-2.10.0 numpy-1.19.5
2021-09-25 22:54:59.808252: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory 2021-09-25 22:54:59.808276: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
Why use 2 kHz data?¶
Because we are studying the low-frequency calls of blue whales, we don't need the original recordings sampled at 256 kHz. Instead, we will use the 2 kHz decimated data in WAV format; these are stored in an s3 bucket named pacific-sound-2khz. For more information on the storage and organization of the 2kHz data, please see the 2kHz example.
List the contents of a monthly directory¶
Between August 2015 and July 2021 (a 6-year period), the highest levels of blue whale song activity off central California were detected during November 2017. Let's start by listing the daily 2 kHz files for that month.
In [2]:
Copied!
s3 = boto3.client('s3',
aws_access_key_id='',
aws_secret_access_key='',
config=Config(signature_version=UNSIGNED))
year = "2017"
month = "11"
bucket = 'pacific-sound-2khz'
for obj in s3.list_objects_v2(Bucket=bucket, Prefix=f'{year}/{month}')['Contents']:
print(obj['Key'])
s3 = boto3.client('s3',
aws_access_key_id='',
aws_secret_access_key='',
config=Config(signature_version=UNSIGNED))
year = "2017"
month = "11"
bucket = 'pacific-sound-2khz'
for obj in s3.list_objects_v2(Bucket=bucket, Prefix=f'{year}/{month}')['Contents']:
print(obj['Key'])
2017/11/MARS-20171101T000000Z-2kHz.wav 2017/11/MARS-20171102T000000Z-2kHz.wav 2017/11/MARS-20171103T000000Z-2kHz.wav 2017/11/MARS-20171104T000000Z-2kHz.wav 2017/11/MARS-20171105T000000Z-2kHz.wav 2017/11/MARS-20171106T000000Z-2kHz.wav 2017/11/MARS-20171107T000000Z-2kHz.wav 2017/11/MARS-20171108T000000Z-2kHz.wav 2017/11/MARS-20171109T000000Z-2kHz.wav 2017/11/MARS-20171110T000000Z-2kHz.wav 2017/11/MARS-20171111T000000Z-2kHz.wav 2017/11/MARS-20171112T000000Z-2kHz.wav 2017/11/MARS-20171113T000000Z-2kHz.wav 2017/11/MARS-20171114T000000Z-2kHz.wav 2017/11/MARS-20171115T000000Z-2kHz.wav 2017/11/MARS-20171116T000000Z-2kHz.wav 2017/11/MARS-20171117T000000Z-2kHz.wav 2017/11/MARS-20171118T000000Z-2kHz.wav 2017/11/MARS-20171119T000000Z-2kHz.wav 2017/11/MARS-20171120T000000Z-2kHz.wav 2017/11/MARS-20171121T000000Z-2kHz.wav 2017/11/MARS-20171122T000000Z-2kHz.wav 2017/11/MARS-20171123T000000Z-2kHz.wav 2017/11/MARS-20171124T000000Z-2kHz.wav 2017/11/MARS-20171125T000000Z-2kHz.wav 2017/11/MARS-20171126T000000Z-2kHz.wav 2017/11/MARS-20171127T000000Z-2kHz.wav 2017/11/MARS-20171128T000000Z-2kHz.wav 2017/11/MARS-20171129T000000Z-2kHz.wav 2017/11/MARS-20171130T000000Z-2kHz.wav
A view of baleen whale song¶
Let's produce a spectrogram with sufficient resolution in time and frequency to see the blue whale song with enough resolution to visually identify. We'll limit the exercise to a single hour from a day with calls of variable received intensity (signal strength).
Download a single 2 kHz file¶
In [3]:
Copied!
year = "2017"
month = "11"
filename = 'MARS-20171101T000000Z-2kHz.wav'
bucket = 'pacific-sound-2khz'
key = f'{year}/{month}/{filename}'
print('Downloading')
session = boto3.Session()
s3_session = session.resource('s3')
s3_session.Bucket(bucket).download_file(key, filename)
print('Done')
year = "2017"
month = "11"
filename = 'MARS-20171101T000000Z-2kHz.wav'
bucket = 'pacific-sound-2khz'
key = f'{year}/{month}/{filename}'
print('Downloading')
session = boto3.Session()
s3_session = session.resource('s3')
s3_session.Bucket(bucket).download_file(key, filename)
print('Done')
Downloading Done
Subset to the 5th hour of the day¶
In [4]:
Copied!
sample_rate = int(2e3)
start_frame = int(sample_rate * 5 * 3600)
duration_frames = int(sample_rate* 3600)
pacsound_file = sf.SoundFile(filename)
pacsound_file.seek(start_frame)
x = pacsound_file.read(duration_frames, dtype='float32')
sample_rate = int(2e3)
start_frame = int(sample_rate * 5 * 3600)
duration_frames = int(sample_rate* 3600)
pacsound_file = sf.SoundFile(filename)
pacsound_file.seek(start_frame)
x = pacsound_file.read(duration_frames, dtype='float32')
Plot the full 2 kHz spectrogram¶
Lots of biophony (sounds of ocean life) are represented in this spectrogram. Humpback whale songs are dominant above ~ 100 Hz, while blue and fin whale songs are dominant below ~ 100 Hz. The energy of blue whale A calls is largely between ~70 and 90 Hz (between the white dashed lines).
For more details about creating a calibrated spectrogram, see the 2 kHz tutorial.
In [5]:
Copied!
sg, f = psd_1sec(x, sample_rate, 177.9) # create calibrated psd
plt.figure(dpi=300)
plt.axhline(73,linestyle='--', color='white')
plt.axhline(91,linestyle='--', color='white')
plt.imshow(sg,extent=[0, 3600, min(f), max(f)],aspect='auto',origin='lower',vmin=30,vmax=100)
plt.yscale('log')
plt.ylim(10,1000)
plt.colorbar()
plt.xlabel('November 01, 2017 Hour 5')
plt.ylabel('Frequency (Hz)')
plt.title('Calibrated spectrum levels')
sg, f = psd_1sec(x, sample_rate, 177.9) # create calibrated psd
plt.figure(dpi=300)
plt.axhline(73,linestyle='--', color='white')
plt.axhline(91,linestyle='--', color='white')
plt.imshow(sg,extent=[0, 3600, min(f), max(f)],aspect='auto',origin='lower',vmin=30,vmax=100)
plt.yscale('log')
plt.ylim(10,1000)
plt.colorbar()
plt.xlabel('November 01, 2017 Hour 5')
plt.ylabel('Frequency (Hz)')
plt.title('Calibrated spectrum levels')
Out[5]:
Text(0.5, 1.0, 'Calibrated spectrum levels')