bandcrash 0.4.3

Tools for publishing albums to the web and digital stores

Bandcrash (formerly pyBlamscamp)

Bandcrash is a standalone program that automatically encodes an album of songs into a bunch of different formats for distribution on various platforms, such as itch.io, or for hosting on your own website. The embedded player is originally based on the one from blamscamp by @blackle, although it has been pretty thoroughly modified at this point.

See it in action!

Features

• Output as mp3, ogg, FLAC, and web preview (HTML5+mp3 at a lower bitrate)
• Optionally upload everything to your page on itch using butler
• High-quality encoding and metadata, with support for cover songs, per-track artwork, embedded lyrics, and more
• Web player also supports per-track artwork

Usage

Bandcrash comes with everything it needs to operate out of the box.

GUI

The easiest way to install the GUI is to download it from my itch.io page. Better yet, use the itch app to keep it updated!

If you want to build the GUI locally, install Python, poetry, and make (from your favorite F/OSS package repository). After that you can run:

poetry run bandcrash-gui

to run the app without building it, or

make app

to build it into an executable, which will be stored in the dist/ directory. However, there are some extra steps you have to do for this on macOS (due to how pyInstaller works with universal binaries). I hope to automate this process eventually.

The GUI is implemented in the bandcrash.gui module.

Command-line

For the command-line version, the suggested approach is to install Python, after which you can install Bandcrash with:

pip install bandcrash

bandcrash --help will provide a lot more detailed information. For the most basic usage you can run:

bandcrash --init input_dir output_dir

which will try to guess the track order, titles, and artwork from the files in that directory, and store the resulting album configuration as input_dir/album.json. It will then process the album and put its results into output_dir.

You can also specify the input as a path to the JSON file, e.g.:

bandcrash input_dir/my_album.json output_dir

or if you want to keep the JSON file in a different directory or use a name other than album.json, you can use a relative path like:

bandcrash input_dir -j ../data/my_album.json

Note that the -j path is relative to input_dir, not the current working directory.

The entry point for the CLI is in bandcrash/cli.py.

Use as a library

Bandcrash can also be used as a Python library in order to embed it into a web service or application. Here is a minimal example of how to use it:

import bandcrash
import bandcrash.options
import collections.defaultdict
import concurrent.futures

config = bandcrash.options.Options()
album = {
    'artist': 'My Band',
    'title': 'My album',
    'tracks': [
        {'filename': 'track 1.wav', 'title': 'The First Track'},
    ],
}

pool = concurrent.futures.ThreadPoolExecutor()
futures = collections.defaultdict(list)

bandcrash.process(config, album, pool, futures)
for task in concurrent.futures.as_completed(list(itertools.chain(*futures.values()))):
    try:
        task.result()
    except Exception:
        # handle the exception in some helpful way

1. Load your album metadata into a Python dict structure (e.g. by doing json.loads on a JSON file or the like)
2. Create a concurrent.futures.ThreadPoolExecutor and a collections.defaultdict(list) to store its futures
3. Create a bandcrash.options.Options and set its options accordingly
4. Call bandcrash.process(options, album_data, threadpool, futures_dict
5. Wait for all of the futures in the futures_dict to complete

The futures_dict is a mapping from process step → list of futures, and is fully populated after process() returns. You can use something like list(itertools.chain(*futures_dict.values())) to collapse it into a single list of futures if you don't care about tracking individual stages in the pipeline.

It is also fine (and, in fact, preferable) to share a ThreadPoolExecutor across multiple concurrent batches.

Album format

For the CLI and GUI, album information is stored in a JSON file (normally named album.json) that looks something like this:

{
    "artist": "The artist of the album",
    "title": "The title of the album",
    "bg_color": "black",
    "fg_color": "white",
    "highlight_color": "#cc00ff",
    "artwork": "album_cover.jpg",
    "tracks": [{
        "filename": "the first track.wav",
        "title": "The First Track",
        "artwork": "track1_cover.jpg",
        "lyrics": ["This is the first line",
            "This is the second line",
            "This is the third line",
            "",
            "This is the second verse",
            "This song just keeps getting worse"],
        "hidden": false,
        "preview": true
    }]
}

You can also automatically generate a stub album.json file with the --init option. Here are a few examples of its use:

# stored as input_dir/album.json
bandcrash --init input_dir output_dir

# stored as input_dir/filename.json
bandcrash --init input_dir/filename.json output_dir

# stored as input_dir/data/filename.json
bandcrash --init input_dir -j data/filename.json output_dir

which will try to guess the track order and titles from the audio files in input_dir.

Basically, the top-level album contains the following properties (all optional):

• artist: The artist for the album as a whole
• title: The album's title
• year: The release year
• composer: The album's composer
• artwork: an image file to use for the album's cover art (relative or absolute paths okay)
• bg_color, fg_color, highlight_color: The color theme for the player
• genre: The default genre for all tracks
• tracks: an array of track descriptions, in album order
• do_mp3: Whether to build the album in MP3 format
• do_ogg: Whether to build the album in Ogg Vorbis format
• do_flac: Whether to build the album in FLAC format
• do_preview: Whether to build the web preview
• do_zip: Whether to build a zip file of each output format
• do_butler: Whether to upload the builds to itch.io
• butler_target: The itch.io Butler target (e.g. "fluffy/songs-of-substance")
• butler_prefix: Any prefix to add to the Butler channel name (e.g. "bonus-" will upload it as "fluffy/songs-of-substance:bonus-mp3")

And each track contains (all optional except filename):

• filename: The audio file to encode into the final output track (ideally wav or aif)
• group: The title of the track's grouping (i.e. a work with multiple movements)
• title: The title of the track
• artist: The performing artist of this track
• composer: The composer of this track
• cover_of: The original artist that this track is a cover of, if any
• genre: The genre of this track
• artwork: Track-specific cover art
• lyrics: An array of strings, one line of lyrics per string; alternately, the name of a text file to read the lyrics from (relative or absolute paths okay)
• hidden: A boolean for whether to hide this track from the web player entirely (e.g. a purchase bonus); defaults to false
• preview: A boolen for whether to generate a preview for the web player; defaults to true
• about: Detailed commentary about the track

See the sample album JSON file for a rough example.

Publishing to Itch

Here's the process for publishing an album to itch.io:

1. Install butler and log in with butler login

2. Create a new project

3. Set it as a "soundtrack," with the kind of project being "HTML"

4. Set your pricing, add preview artwork, etc., and save. Don't do any uploading from this interface.

5. Run bandcrash with a -b user/project flag; for example:

   bandcrash novembeat-2021/wav novembeat-2021/out -b fluffy/novembeat-2021
   

6. Wait a moment for itch to finish processing; you can use butler status user/project (e.g. butler status fluffy/novembeat-2021)

7. Reload your project edit page; you should now have a few targets, such as preview, mp3, etc.

8. Set the preview target to "This file will be played in the browser". Set all the other targets to "soundtrack" and, optionally, change the display name.

9. View the project page, and when you're ready to publish, publish!

Recommended "embed options"

• Set it to "embed in page" with "manually set size"
• Enable "mobile friendly" (with orientation "default") and "automatically start on page load"
• Disable "fullscreen button" and "enable scrollbars"

Publishing to other sites

If you have a website of your own, upload the preview directory to your site somewhere, and then embed it using an iframe, e.g.

<iframe src="/path/to/album/" width=640 height=360></iframe>

Contributing

Pull requests are welcome! But please note the following:

The generated web player must not receive any added dependencies. The generator must stay as a single, self-contained file that is as small as reasonably possible. The point is for the generated file to be lightweight. Stick to Vanilla JS.

Roadmap

See the github issues for details, but roughly:

• Local GUI and/or web UI to make setting up the album.json easier (and easier installation, especially on Windows!)
• Various player improvements
• Easy embedding into websites (opengraph et al)

FAQ

How is this different from blamscamp, scritch, etc.?

Blamscamp and scritch are both great programs for publishing album previews on itch.io and other websites! However, their functionality is only to bundle already-encoded audio files into a web-based player. They don't do the difficult work of encoding and tagging your files, which can be an extremely tedious and error-prone process. Bandcrash's intention is to make the process of encoding and uploading your albums easier to as many stores as possible.

Why was it renamed?

Back when this project started, it was named pyBlamscamp as the intention was to be basically a Python version of the blamscamp GUI which would also handle encoding steps for you, but it very quickly drifted away from that and became something else. Unfortunately, the similarity of the names was incredibly confusing.

Currently the only connection between Bandcrash and blamscamp is that Bandcrash uses a highly-modified version of blamscamp's web player. They serve different goals.

Why run it locally instead of as a web app?

You already have your large .wav files on your local hard drive. Your local drive is also a good place to keep your previous encoding results. Your local computer also has a lot more space available than a typical cloud server, doesn't have to juggle cloud storage credentials, doesn't have to worry about the security of the server running the encoder app, the cost of running servers or paying for cloud storage, and so on.

Basically, it's easier for everyone.

Sometimes local apps are just Better™.

Credits

• Main code: @fluffy-critter
• Original player code and this project's name: @blackle