Waving at the TIDAL music service with Python. Runs on (at least) Windows, macOS, and GNU/Linux.
TIDAL is an artist-first, fan-centered music streaming platform that delivers over 100 million songs in HiFi sound quality to the global music community. © 2024 TIDAL Music AS
This project is inspired by qobuz-dl
, and, particularly, is a continuation of Tidal-Media-Downloader
. This project is intended for private use only: it is not intended for distribution of copyrighted content.
This software uses libraries from the FFmpeg project under the LGPLv2.1. FFmpeg is a trademark of Fabrice Bellard, originator of the FFmpeg project.
- Retrieve FLAC, Dolby Atmos, or AAC tracks; AVC/H.264 (up to 1920x1080) + AAC videos
- Note: as of 24 July 2024, TIDAL's catalog no longer includes tracks in MQA or Sony 360 Reality Audio format.
- Either a single track or an entire album can be retrieved
- Album covers are retrieved by default, and embedded into all tracks
- Highest-resolution, "original" album covers, which can be up to 6000x6000 pixels resolution, are retrieved if available
- Support for albums with multiple discs
- If available, lyrics are added as metadata to tracks
- If available, album reviews are retrieved as JSON
- If available, album credits are retrieved as JSON
- If available, artist bios are retrieved as JSON
- If available, artist images are retrieved as JPEG
- Playlist retrieval support (video or audio or both)
- Playlist .m3u8 file automatically created
- Mix retrieval support (video or audio)
- Artist's entire works retrieval support (video and audio; albums or albums and EPs and singles)
- Because of the use of the
requests
package, system proxies are respected (HTTP, HTTPs, Socks); or can be specified by typical environment variable - Also because of the use of
requests
, very simpleCache-Control
request caching occurs viaCacheControl
- If desired, all JSON responses from the TIDAL API can be saved for inspection or posterity or debugging
A current, valid TIDAL subscription is required in order to run tidal-wave
. Previously, TIDAL segmented the available audio qualities into HiFi and HiFi Plus plans: now,
All current TIDAL plans feature Max sound quality formats such as full lossless, HiRes FLAC, and Dolby Atmos (up to 24-bit, 192 kHz).
More information on sound quality at TIDAL's site here.
- As resources will be fetched from the World Wide Web, an Internet connection is required
- The venerable FFmpeg is necessary for audio and video data manipulation. This project's container image as well as its PyInstaller-created binaries build FFmpeg from source, so separate installation is unnecessary.
- Static builds of FFmpeg are available from John Van Sickle for GNU/Linux, or most package managers feature
ffmpeg
. - For macOS, the FFmpeg download page links to this download source; or there is always Homebrew
- For Windows, the FFmpeg download page lists 2 resources; or
chocolatey
is an option - If a minimal FFmpeg, compiled from source, is desired, take a look at this project's BUILDME.md file for decent instructions
- Static builds of FFmpeg are available from John Van Sickle for GNU/Linux, or most package managers feature
- This is a Python package, so to use it in the default manner you will need Python 3, version 3.8 or newer, on your system.
- However, as of December 2023, an OCI container image; and PyInstaller-created binaries for x86_64 GNU/Linux, Apple Silicon macOS, x86_64 macOS, and x86_64 Windows are provided for download and use that do not require Python to be installed
- Only a handful of Python libraries are dependencies:
Install this project with pip
: either with a virtual environment (preferred) or any other way you desire:
# GNU/Linux or macOS or Android (e.g. Termux)
$ python3 -m pip install tidal-wave
# Windows
PS > python.exe -m pip install tidal-wave
Alternatively, you can clone this repository; cd
into it; and install from there:
$ git clone --depth=1 https://github.com/ebb-earl-co/tidal-wave.git
$ cd tidal-wave
$ python3 -m venv .venv
$ source .venv/bin/activate
$ (.venv) pip install .
The release artifacts for this project are created with PyInstaller. It bundles Python 3.12.7, FFmpeg 7.0, and the tidal-wave
program into one binary, licensed under the terms of FFmpeg: with the GNU Lesser General Public License (LGPL) version 2.1. Installation is as simple as downloading the correct binary for your platform giving it execute permissions, and running it. Please make sure that the SHA256 checksum of the file that you have downloaded matches the corresponding .sha256
file on the releases page!
$ wget https://github.com/ebb-earl-co/tidal-wave/releases/latest/download/tidal-wave_ubuntu_24.04_amd64
$ wget https://github.com/ebb-earl-co/tidal-wave/releases/latest/download/tidal-wave_ubuntu_24.04_amd64.sha256
$ sha256sum --check tidal-wave_ubuntu_24.04_amd64.sha256
# ONLY CONTINUE IF THE OUTPUT IS THE FOLLOWING: 'tidal-wave_ubuntu_24.04_amd64.sha256: OK'
# Otherwise, delete the downloaded binary and try to download it again
$ chmod +x ./tidal-wave_ubuntu_24.04_amd64
$ ./tidal-wave_ubuntu_24.04_amd64 --help
# For just the lifetime of this PowerShell process, don't block the download from GitHub
PS > Set-ExecutionPolicy -ExecutionPolicy Bypass -Scope Process
PS > Invoke-WebRequest "https://github.com/ebb-earl-co/tidal-wave/releases/latest/download/tidal-wave_windows.exe" -OutFile "tidal-wave_windows.exe"
PS > Invoke-WebRequest "https://github.com/ebb-earl-co/tidal-wave/releases/latest/download/tidal-wave_windows.exe.sha256" -OutFile "tidal-wave_windows.exe.sha256"
# Get the checksum value from the tidal-wave_windows.exe.sha256 file and compare it to the just-downloaded EXE
# (Get-FileHash .\tidal-wave_windows.exe -Algorithm SHA256).Hash -eq (Get-Content .\tidal-wave_windows.exe.sha256)
PS > (Get-FileHash .\tidal-wave_windows.exe -Algorithm SHA256).Hash -eq "e02f69eb850a98e6e1df2bc033fd12566cf27305421a36ec5372fd432ccc8e70" # This checksum is from version 2024.4.3
# ONLY CONTINUE IF THE OUTPUT OF THE PREVIOUS COMMAND IS 'True'
PS > & .\tidal-wave_windows.exe --help
Pull the image from GitHub container repo:
$ docker pull ghcr.io/ebb-earl-co/tidal-wave:latest
# Or, the main branch of this repository, which will be ahead of `latest`:
$ docker pull ghcr.io/ebb-earl-co/tidal-wave:trunk
If your Python installation's location is available on path, run tidal-wave --help
to see the options available. Otherwise (including if you followed the repository cloning steps above), run python3 -m tidal_wave --help
from the repository root directory, tidal-wave
. In either case, you should see something like the following:
Usage: python -m tidal_wave [OPTIONS] TIDAL_URL [OUTPUT_DIRECTORY]
╭─ Arguments ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * tidal_url TEXT The Tidal album or artist or mix or playlist or track or video to download [default: None] [required] │
│ output_directory [OUTPUT_DIRECTORY] The parent directory under which directory(ies) of files will be written [default: ~/Music] │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --audio-format [Atmos|HiRes|Lossless|High|Low] [default: Lossless] │
│ --loglevel [DEBUG|INFO|WARNING|ERROR|CRITICAL] [default: INFO] │
│ --include-eps-singles No-op unless passing TIDAL artist. Whether to include artist's EPs and singles with albums │
│ --no-extra-files Whether to not even attempt to retrieve artist bio, artist image, album credits, album review, or playlist m3u8 │
│ --no-flatten Whether to treat playlists or mixes as a list of tracks/videos and, as such, retrieve them independently │
| --transparent Whether to dump JSON responses from TIDAL API; maximum verbosity |
│ --install-completion Install completion for the current shell. │
│ --show-completion Show completion for the current shell, to copy it or customize the installation. │
│ --help Show this message and exit. │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
Invocation of this tool will store credentials in a particular directory in the user's "home" directory: for Unix-like systems, this will be /home/${USER}/.config/tidal-wave
: for Windows, it varies: in either OS situation, the exact directory is determined by the user_config_path()
function of the platformdirs
package.
Similarly, by default, all media retrieved is placed in subdirectories of the user's default music directory: for Unix-like systems, this probably is /home/${USER}/Music
; for Windows it is probably C:\Users\<USER>\Music
. This directory is determined by platformdirs.user_music_path()
.
- If a different path is passed to the second CLI argument,
output_directory
, then all media is written to subdirectories of that directory.
Source: TIDAL
Low | High | Lossless | HiRes FLAC | Dolby Atmos | Video (H.264 + AAC) | |
---|---|---|---|---|---|---|
Android | ✔️ | ✔️ | ✔️ | ✔️ | ❌ | ✔️ |
Fire TV 🔷 | ✔️ | ✔️ | ✔️ | ❌ | ✔️ | ✔️ |
macOS | ✔️ | ✔️ | ✔️ | ✔️ | ❌ | ✔️ |
Windows | ✔️ | ✔️ | ✔️ | ✔️ | ❌ | ✔️ |
🔷 This is the default client for tidal-wave
, a spoofed Amazon Fire TV. It is the one invoked in all situations unless --audio-format hires
is passed as a command line flag:
$ tidal-wave https://listen.tidal.com/album/000000
$ # no --audio-format flag passed will instruct tidal-wave to use the Fire TV client, as it implies --audio-format lossless
$ tidal-wave https://listen.tidal.com/album/000000 --audio-format high
$ # specifying low, high, lossless, or atmos will instruct tidal-wave to use the Fire TV client
$ tidal-wave https://listen.tidal.com/album/000000 --audio-format hires
$ # the above forces tidal-wave to ask for an access token gleaned from an Android, macOS, or Windows device, as laid out in the above table
It is certainly useful for debugging, and, perhaps, using multiple versions of a software, to know which binary/package invoked is which version. From version 2024.7.1 of tidal-wave
, this is possible by adding the --version
flag to any command. This is an eager command, in typer
parlance, meaning that any other flag or argument passed to tidal-wave
will be ignored and the version will simply be returned. E.g.
$ tidal-wave --version
tidal-wave 2024.7.1
- First, find the URL of the album/artist/mix/playlist/track/video desired. Then, simply pass it as the first argument to
tidal-wave
with no other arguments in order to: retrieve the album/artist/mix/playlist/track in Lossless quality to a subdirectory of user's music directory and INFO-level logging in the case of audio; retrieve the video in 1080p, H.264+AAC quality to a subdirectory of user's music directory with INFO-level logging in the case of a video URL.
(.venv) $ tidal-wave https://tidal.com/browse/track/226092704
- By default, the track(s) and/or video(s) are retrieved, and other files are retrieved as well; such as the artist's bio JSON, the artist's TIDAL image, the playlist's .m3u8 file, the album's review JSON, and a few others. In order not to retrieve any of those, pass the
--no-extra-files
flag:
(.venv) $ tidal-wave https://tidal.com/browse/track/226092704 --no-extra-files
- To (attempt to) get a Dolby Atmos track, and you desire to see all of the log output, the following will do that
(.venv) $ tidal-wave https://tidal.com/browse/track/... --audio-format atmos --loglevel debug
Keep in mind that an access token from an Android (preferred), Windows, or macOS device will need to be extracted and passed to this tool in order to access HiRes FLAC tracks
- To (attempt to) get a HiRes FLAC version of an album, and you desire to see only warnings and errors, the following will do that:
$ tidal-wave https://tidal.com/browse/album/... --audio-format hires --loglevel warning
- To (attempt to) get a playlist, the following will do that. N.b. passing anything to
--audio-format
is a no-op when retrieving videos.
PS > C:\Users\User > & tidal-wave_windows.exe https://tidal.com/browse/playlist/...
- To (attempt to) get a mix, the following will do that. N.b. passing anything to
--audio-format
is a no-op when retrieving videos.
$ ./tidal-wave_ubuntu_24.04_amd64 https://tidal.com/browse/mix/...
- To (attempt to) get all of an artist's works (albums and videos, excluding EPs and singles) in AAC High format and verbose logging, the following will do that:
(.venv) $ python3 -m tidal_wave https://listen.tidal.com/artist/... --audio-format high --loglevel debug
- To (attempt to) get all of an artist's works (including EPs and singles), in HiRes format, the following will do that:
(.venv) $ tidal-wave https://listen.tidal.com/artist/... --audio-format hires --include-eps-singles
- As a throw-everything-at-the-wall-and-see-what-sticks option, there is the
--transparent
flag. In the directory wheretidal-wave
is called,--transparent
will write the responses from the TIDAL API to .json files
(.venv) $ tidal-wave https://listen.tidal.com/track/... --audio-format low --loglevel debug --transparent
By default, when passed a playlist or mix URL, tidal-wave
will retrieve all of the tracks and/or videos specified by that URL, and write them to a subdirectory of either Playlists
or Mixes
, which itself is a subdirectory of the specified output_directory
. E.g. ~/Music/Mixes/My Daily Discovery [016dccd302e9ac6132d8334cfbc022]
. In this directory, once all of the tracks and/or videos have been retrieved, they are renamed based on the order in which they appear in the playlist. E.g.
(.venv) $ tidal-wave https://listen.tidal.com/playlist/1b418bb8-90a7-4f87-901d-707993838346
$ ls ~/Music/Playlists/New Arrivals [1b418bb8-90a7-4f87-901d-707993838346]/
'001 - Dance Alone [CD].flac'
'002 - Kissing Strangers [CD].flac'
'003 - Sunday Service [CD].flac'
If this is not the desired behavior, pass the --no-flatten
flag. This flag instructs tidal-wave
to leave the tracks and/or videos in the directory where they would have been written if they had been passed to tidal-wave
independently. E.g.
(.venv) $ tidal-wave https://listen.tidal.com/playlist/1b418bb8-90a7-4f87-901d-707993838346 --no-flatten
$ ls ~/Music/
'Sia/Dance Alone [343225498] [2024]/01 - Dance Alone [CD].flac'
'USHER/COMING HOME [339249017] [2024]/05 - Kissing Strangers [CD].flac'
'Latto/Sunday Service [344275657] [2024]/01 - Sunday Service [CD].flac'
The command line options are the same for the Python invocation, but in order to save configuration and audio data, volumes need to be passed. If they are bind mounts to directories, they must be created before executing docker run
to avoid permissions issues! For example,
$ mkdir -p ./Music/ ./config/tidal-wave/
$ docker run \
--rm -it \
--name tidal-wave \
--volume ./Music:/home/debian/Music \
--volume ./config/tidal-wave:/home/debian/.config/tidal-wave \
ghcr.io/ebb-earl-co/tidal-wave:latest \
https://tidal.com/browse/track/...
Using Docker might be an attractive idea in the event that you want to retrieve all of the videos, albums, EPs, and singles in highest quality possible for a given artist. The following Docker invocation will do that for you:
$ mkdir -p ./Music/ ./config/tidal-wave/
$ docker run \
--name tidal-wave \
--rm -it \
--volume ./Music:/home/debian/Music \
--volume ./config/tidal-wave:/home/debian/.config/tidal-wave \
ghcr.io/ebb-earl-co/tidal-wave:latest \
https://listen.tidal.com/artist/... \
--audio-format hires \
--include-eps-singles
Perhaps you don't want a single-shot executable type of Docker invocation, but rather a long-lived container into which one can docker exec
in order to request media at one's leisure. This is one of the requested features from the GitHub Discussions, in particular for Unraid users. In order to do this, use the following, slightly-modified Docker command:
$ mkdir -p ./Music/ ./config/tidal-wave/
$ docker run \
--name tidal-wave \
-dit \ # is short for: --detach --interactive --tty
--volume ./Music:/home/debian/Music \
--volume ./config/tidal-wave:/home/debian/.config/tidal-wave \
--entrypoint "/bin/bash" \
ghcr.io/ebb-earl-co/tidal-wave:latest
$ docker exec -it tidal-wave tidal-wave https://tidal.com/browse/album/...
$ docker exec -it tidal-wave tidal-wave https://tidal.com/browse/mix/...
$ docker exec -it tidal-wave tidal-wave https://tidal.com/browse/playlist/...
$ docker exec -it tidal-wave tidal-wave https://tidal.com/browse/track/...
Note: the first tidal-wave
is whatever --name
you give the container, so that can be whatever your heart desires, but the second tidal-wave
is invoking the Python program inside the container and needs to exactly tidal-wave
.
The easiest way to start working on development is to fork this project on GitHub, or clone the repository to your local machine and do the pull requesting on GitHub later. In any case, there will need to be some getting from GitHub first, so, roughly, the process is:
- Get Python 3.8+ on your system
- Use a virtualenv or some other Python environment system (poetry, pipenv, etc.)
- Clone the repository:
$ git clone --depth=1 https://github.com/ebb-earl-co/tidal-wave/git
* Obviously replace the URL with your forked version if you've followed that strategy
- Activate the virtual environment and install the required packages (requirements.txt):
(some-virtual-env) $ python3 -m pip install -r requirements.txt
* optional packages to follow the coding style and build process; `pyinstaller`, `black`: `(some-virtual-env) $ python3 -m pip install black pyinstaller`
* optionally, Docker to build the OCI container artifacts
- From a Python REPL (or, my preferred method, an iPython session), import all the relevant modules, or the targeted ones for development:
from tidal_wave import album, artist, dash, hls, login, main, media, mix, models, oauth, playlist, requesting, track, utils, video
from tidal_wave.main import logging, user_music_path, Path