Subway Sounds

This repository contains code for the real-time story Subway Sounds, completed for the class Storytelling with Streaming Data at Columbia University in Spring 2016.

Dependencies

System dependencies

This project has two non-standard dependencies:

• Redis
• websocketd

Please install these before attempting to run this locally. This project was developed using Redis version 3.0.5 and websocketd 0.2.11. We cannot guarantee compatibility with other versions of these software packages, although we use relatively basic funcionality from both, so it should be fairly tolerant to different versions.

Python dependencies

The Python dependencies are listed in requirements.txt. You can install them in one go using the following command:

pip install -r requirements.txt

Folder Structure

├── 0_streaming_server
│   ├── analysis
│   │   ├── A_weighting.py
│   │   ├── calibration.py
│   │   ├── day1
│   │   │   └── processed-data
│   │   │       ├── 1_street.WAV.json
│   │   │       ├── 2_concourse.WAV.json
│   │   │       ├── 3_1-2-3.WAV.json
│   │   │       └── 4_7.WAV.json
│   │   ├── day2
│   │   │   └── processed-data
│   │   │       ├── 1_street.WAV.json
│   │   │       ├── 2_concourse.WAV.json
│   │   │       ├── 3_1-2-3.WAV.json
│   │   │       └── 4_7.WAV.json
│   │   ├── generate_data.py
│   │   ├── spl.py
│   │   └── wavio.py
│   ├── graphs
│   │   ├── recalibration\ with\ A-Weights.png
│   │   ├── x.png
│   │   ├── y.png
│   │   └── z.png
│   ├── playback.py
│   └── server.py
├── 1_ingestor
│   ├── ingest.py
│   └── store.py
├── 2_view_server
│   ├── 90
│   ├── trains.py
│   └── viewserver.py
├── 3_frontend
│   ├── chart-test.html
│   ├── css
│   │   └── main.css
│   ├── index.html
│   ├── js
│   │   ├── chart-maker.js
│   │   ├── highcharts.js
│   │   ├── jquery-1.9.1.min.js
│   │   ├── main.js
│   │   └── update.js
│   ├── media
│   │   └── cover.jpg
│   ├── references.txt
│   ├── run_server
│   └── thoughts.txt
├── README.md
└── requirements.txt

13 directories, 38 files

System structure

Here is a sketch of the system's structure:

To Run

Run the static data files generator

This is strictly optional. The static data are already this repository in the location. This is to be used if you have want the full experience.

Download files from: https://drive.google.com/a/columbia.edu/folderview?id=0ByTEfsZoEb1-YkdzNm5OaDB3cnc&usp=sharing

From the root directory:

// to get the static data files 
cd 0_streaming_server/analysis && python get_data.py {{dirname}}
// to get the calibrated data
cd 0_streaming_server/analysis && python calibration.py

where {{dirname}} is the directory of audio files you would like to process into static datasets.

Start the backend

To start the simulated MTA stream of sound intensities, use the following command.

cd 0_streaming_server && python playback.py | python server.py

Start the ingestor

Note: before you start the ingestor, make sure you have started the Redis server and that it is serving over the default port. The below commands must be run from the root directory.

cd 1_ingestor && python ingest.py | python store.py

Start the view server

Run the following command to start the view server, which is a websocket pumping out updates to the stream.

websocketd --port 6999 python 2_view_server/viewserver.py

Start the frontend

Finally, you can start the front-end server using the following command.

cd 3_frontend && ./run_server

Navigate your browser to localhost:8000 to view the story, updating in (simulated) real time.