🎵 How Songs Are Organized

Deep .caustic project analysis and automated genre classification.

The Caustic 3 Archive contains a large collection of community songs recovered from surviving collections and historical sources. To make this material easier to explore, the Archive processes each .caustic project through a custom song-analysis system.

Rather than relying only on filenames, the sorter examines information stored inside the project itself, extracts useful musical and structural information, and uses multiple signals to organize songs by genre and subgenre.

The original .caustic projects remain preserved while the extracted information is documented alongside them.

🎼 Why Organize Songs?

🔍 In-Depth .caustic File Extraction

🎚️ Multi-Signal Classification Engine

📁 Handling Patch Demos & Patterns

📋 Metadata & Archive Tools

🧭 Transparent & Auditable

🔄 An Evolving Classification System

📦 From Individual Projects to an Accessible Archive

🔬 Part of the Technical Archive

🎼 Why Organize Songs?

A collection of hundreds or thousands of .caustic files quickly becomes difficult to navigate when filenames provide little indication of what each song contains.

A file named: Track01.caustic

doesn't tell you whether it is a trance track, a drum & bass project, a synthwave composition, a patch demonstration, or something else entirely.

The Archive therefore adds a layer of organization around the original projects.

Songs can be grouped into genres and subgenres such as:

Ambient · Drum & Bass · Dubstep · House · Synthwave · Techno · Trance · Trap · Chiptune · Harddance · Glitch Hop · Hip-Hop

Where the available evidence supports a more specific classification, subgenres can be used to refine the organization.

🔍 In-Depth .caustic File Extraction

Inside the Caustic Project File

Filenames rarely tell the full story.

Instead of treating a .caustic file as an opaque object, the Archive's tools examine its binary structure and extract information that can be used for documentation and classification.

Information currently extracted can include:

Session Information — Song name and BPM, where available.
Machine Roster — The Caustic machines present in the project.
Preset Information — Embedded preset names and other available preset information.
Sample Information — Embedded sample information where it can be identified.
Pattern Data — Note counts, pitch ranges, duration information and other pattern-level statistics.
Arrangement Data — Sequencer information describing pattern placement and timing where the structure can be reliably decoded.
Effects — Effects and Master Rack information where identifiable.

🔒 Preservation First: The original .caustic file is not modified or overwritten. Extracted information is written separately as companion metadata.

🎚️ Multi-Signal Classification Engine

How Genre Classification Works

Genre classification is rarely determined by one piece of information.

The sorter therefore combines several signals, with stronger evidence taking priority over weaker supporting evidence.

Signal 1 — Text & Keyword Markers

The strongest classification clues often come from information already present in the project.

The sorter can examine:

Song titles
Original filenames
Preset names
Machine names

Explicit terms such as dubstep, trance, DnB, house, or synthwave can provide strong evidence for a genre assignment.

Signal 2 — Musical Pattern Features

The sorter can also extract selected features from the project's pattern and note data.

Examples include:

Rapid percussion patterns that can provide supporting evidence for genres such as Trap or Drum & Bass.
Pitch movement and slide behaviour that can provide supporting evidence for bass-oriented styles.
Sustained notes and pad-like patterns that can provide supporting evidence for styles such as Ambient or Trance.

These are deliberately narrow rules rather than attempts to perform complete musical analysis.

They provide supporting evidence alongside the stronger textual and structural signals.

Signal 3 — Tempo & BPM

BPM is used as another supporting signal.

Tempo ranges overlap between genres, so BPM alone is not treated as proof of a particular genre. Instead, it can reinforce other evidence already found in the project.

Signal 4 — Artist Associations

Where a recognizable artist name appears in the available metadata or filename, known artist-to-genre associations can provide additional contextual evidence.

These associations carry less weight than direct information from the project itself.

📁 Handling Patch Demos & Patterns

Separating Non-Song Material

The historical Caustic community contained much more than finished songs.

Users regularly shared:

Synth patches
Sound-design experiments
Modular wiring tests
Machine demonstrations
Pattern experiments
Technical test projects

These files are valuable parts of the archive, but they shouldn't necessarily appear alongside finished musical compositions.

Pattern Only

The sorter uses structural information from the project to identify files that contain pattern material but do not appear to contain a conventional song arrangement.

These projects can be separated into a dedicated Pattern_Only collection rather than being incorrectly presented as finished songs.

Technical & Modular Demos

Additional filtering can identify projects containing technical terminology or modular experimentation that may otherwise resemble songs because they contain pattern or sequence data.

The aim is to preserve these projects while placing them in a more appropriate part of the archive.

📋 Metadata & Archive Tools

Companion Files & Collection Indexes

The extracted information is not kept only inside the sorting program.

Each processed song can be accompanied by a readable metadata report containing information such as:

Song name and BPM
Machine roster
Preset information
Pattern statistics
Arrangement information
Effects information where available
Classification information

The Archive can also generate collection-wide indexes:

manifest.csv — A master spreadsheet containing information about the processed collection, allowing songs to be searched, filtered and analysed in bulk.
treeview.txt — A readable map of the resulting folder structure.

These files make the collection searchable without requiring every .caustic project to be opened manually in Caustic.

🧭 Transparent & Auditable

Classification Is Evidence-Based

Automated classification is an evolving part of the Archive's research.

The sorter records the information used to make its classification decisions, allowing the reasoning behind an assignment to be examined rather than treating the result as unexplained automation.

A song may be classified using a combination of:

Title / Filename → Preset Names → Machine Roster → BPM → Pattern Features → Artist Context

Not every signal is available in every project, and not every genre can be determined with the same level of certainty.

Where the evidence is limited, the system is designed to avoid pretending that an uncertain classification is definitive.

🔄 An Evolving Classification System

The genre system is continually being tested against real Caustic projects.

As more songs are independently identified and more of Caustic's internal structures are understood, the classification rules can be refined.

Some parts of the .caustic format are already well understood, while other structures are still being researched.

This means the Archive's organization can improve over time without changing the original song files.

The goal is not to create an artificial certainty around historical material.

The goal is to make the collection more discoverable while remaining transparent about what is known, what is inferred, and what is still being researched.

📦 From Individual Projects to an Accessible Archive

The final result is more than a collection of .caustic files.

The preservation workflow turns:

Recovered Projects → Binary Analysis → Classification → Companion Metadata → Collection Indexes → Organized Archive

into a collection that can be searched, explored and eventually published as organized public archive material.

Explore Song Collections

🔬 Part of the Technical Archive

The song-organizing system depends on ongoing research into the internal structure of Caustic project files.

This includes investigating machine rosters, pattern data, sequencer records, embedded preset information, effects structures and other binary components.

The classification system is therefore both a practical archive tool and a product of the Archive's wider technical research.

Explore Technical Research

Page updated

Google Sites

Report abuse