README.md

metadata

license: cc-by-nc-4.0
pretty_name: Time-Lapse Artifacts
annotations_creators:
  - expert-generated
source_datasets:
  - original
language:
  - en
size_categories:
  - n<1K
task_categories:
  - video-classification
  - feature-extraction
tags:
  - drawing-process
  - longitudinal-study
  - traditional-art
  - single-subject
  - video
  - computer-vision
  - video-understanding
  - creative-process
  - human-motion
dataset_info:
  features:
    - name: video
      dtype: video
    - name: record_id
      dtype: large_string
    - name: collection_tier
      dtype: large_string
    - name: is_canonical
      dtype: bool
    - name: acquisition_protocol_version
      dtype: large_string
    - name: date
      dtype: large_string
    - name: time
      dtype: large_string
    - name: tool
      dtype: large_string
    - name: medium
      dtype: large_string
    - name: support
      dtype: large_string
    - name: dimensions
      dtype: large_string
    - name: duration_seconds
      dtype: float64
    - name: width_px
      dtype: int64
    - name: height_px
      dtype: int64
    - name: orientation
      dtype: large_string
    - name: encoded_frame_rate
      dtype: int64
    - name: video_codec
      dtype: large_string
    - name: pixel_format
      dtype: large_string
    - name: container
      dtype: large_string
    - name: has_audio
      dtype: bool
    - name: hub_xet_hash
      dtype: large_string
    - name: original_filename
      dtype: large_string
    - name: provenance_status
      dtype: large_string
    - name: legacy_provenance_path
      dtype: large_string
    - name: legacy_related_references
      dtype: large_string
    - name: evidence_method
      dtype: large_string
    - name: provenance_reference_duration_delta_seconds
      dtype: float64
    - name: size_bytes
      dtype: int64
    - name: notes
      dtype: large_string

time-lapse-artifacts

467 time-lapse recordings of one artist's traditional drawing practice across nearly two years (September 2024–August 10, 2026), with standardized acquisition since July 15, 2025. Across approximately 1.89 TB of video, the archive follows the same artist working in physical media—including pencil, ballpoint pen, fountain pen, graphite, and ink—on fixed-format paper and illustration board. The release combines 338 canonical Standard videos with per-record content hashes and validated technical metadata, plus 129 earlier Pre-Standard recordings with lighter descriptive metadata.

The underlying practice also includes a 2012–2016 livestream layer, but those videos are not yet published. This is an active construction release.

Snapshot and Coverage

These counts describe the current main branch. The 467-row v0.2 media membership was established at revision 3967721ce363316cb7475be648af34a66bc83c27; later card and schema revisions preserve the same 338 Standard / 129 Pre-Standard membership.

Collection tier	Published MP4s	Indexed rows	Date coverage	Metadata state
standard	338 direct files	338	2025-07-15–2026-08-10	Complete core technical metadata for every direct Standard file
pre_standard	129	129	2024-09-17–2025-07-13, plus 9 unknown dates	Filename-level descriptive metadata; technical ingestion incomplete
legacy_livestream	0 public videos	0	Historical scope 2012–2016	Placeholder layer; ingestion pending

Current published index (v0.2): the live Hub Viewer exposes 467 rows: 338 Standard and 129 Pre-Standard. The preceding v0.1 state at revision 339e6149de130cbe2d001f7a0e5247872bc788ec exposed 465 rows: 335 Standard and 130 Pre-Standard. v0.2 added three previously unindexed Standard files and removed one Pre-Standard row whose media did not meet the ownership and provenance threshold. The rebuild is complete; 467 is the current state, not a pending estimate.

The builder-generated split name train is an archive-loading default, not a recommended machine-learning train split. No official train/validation/test partition has been defined.

Viewer and Metadata Discovery

The current Hub build automatically discovers and loads both metadata files into one default/train Viewer split:

• root metadata.csv indexes the 338 direct files in Standard_Time_Lapses/
• Pre_Standard_Time_Lapses/metadata.csv indexes the 129 direct files in that folder

The Viewer therefore has 467 rows and a 29-column union schema. That schema is declared explicitly in the card's dataset_info block so file-discovery order cannot change field availability or infer incompatible types. Both CSV files use the same 29 columns; undocumented Pre-Standard values are empty and load as null rather than as zero, false, or inferred numeric data. Both indexes declare tier membership directly: the root canonical CSV sets collection_tier = standard and is_canonical = true; the Pre-Standard CSV sets collection_tier = pre_standard and is_canonical = false.

Published-index completeness is defined by the 467 rows in default/train and the matching generated Parquet export. Hugging Face's separate column Statistics job is an auxiliary, asynchronous surface; its availability is not used to claim or reject row-level completeness.

There is no explicit configs: block in this release. A broad canonical folder glob could accidentally include the nested Standard_Time_Lapses/Provenance/ subtree. Separate canonical and Pre-Standard configurations therefore remain a fail-closed migration task described in VIEWER_MIGRATION.md, rather than an untested card-only change.

Overview

time-lapse-artifacts documents analog drawing processes across different recording periods and acquisition conditions. Raw video is the primary record. Metadata identifies, organizes, compares, and documents recordings without imposing aesthetic scores or stylistic labels.

The collection is a single-creator longitudinal archive. This makes it useful for studying within-practice change, but it does not represent artists, materials, studios, or drawing practices broadly.

Repository Structure

Standard_Time_Lapses/

The directory contains 338 direct MP4 files. metadata.csv is their complete machine-readable canonical index: every direct file has one validated row with core technical metadata. Direct placement alone is not sufficient for future additions; metadata and release validation are required.

Each canonical metadata row now includes:

• a persistent record_id
• collection_tier = standard
• is_canonical = true
• acquisition_protocol_version = standard-2025-07-15
• descriptive session fields
• encoded video properties
• repository content hash and size
• provenance fields where documented

Files use a structured filename convention described in SCHEMA.md. A filename is descriptive, but it is not the stable identity of a record; external references should use record_id.

Standard_Time_Lapses/Provenance/

This subtree contains supporting historical records and retained exceptions. It is not part of the canonical record set.

2026-07-31/ preserves the completed provenance release, including mappings between historical Series_9x12 filenames and canonical records. The former full duplicate video snapshot is not retained. Relationships are preserved through mappings, hashes, metadata, and repository history.

Unresolved_Series_9x12/ contains distinct historical recordings that do not yet have enough documentation for promotion into the canonical collection.

Pre_Standard_Time_Lapses/

These 129 recordings predate the standard acquisition protocol. They are published for archival continuity but are not interchangeable with canonical standard records. Their technical metadata remains incomplete. Missing values must not be interpreted as zero, false, or evidence that a property was absent.

The folder-local metadata.csv is the active index for these 129 videos and is loaded by the current Viewer alongside the root canonical index. It follows the same declared 29-field schema as the canonical index while leaving undocumented extended values empty. The explicit dataset_info types prevent those missing values from changing field types during loading.

Material that does not meet the archive's ownership and provenance threshold is excluded from both the published media and the active metadata index.

Legacy_Livestreams_2012_2016/

This directory reserves the historical livestream layer. At the audited snapshot it contains no public video data. References to 2012–2016 therefore describe archive scope, not current downloadable coverage.

Metadata and Provenance

The metadata distinguishes documented facts, established historical relationships, and unresolved information. Evidence may include:

• exact file/blob identity
• original filenames and repository paths
• file sizes and hashes
• duration comparison
• visual continuity across recordings
• documented physical references
• repository commit history

Uncertain relationships are not promoted to verified relationships solely because dates or filenames look similar.

The active CSV files and live Viewer use semicolons for multiple tool and medium values, for example Pencil;FountainPen and Graphite;Ink. A pipe (|) is not a valid separator in v0.2. The two fields are independent observed-value lists; equal list lengths and positional pairing are not guaranteed. See SCHEMA.md.

encoded_frame_rate is encoded playback rate. It is not the original temporal sampling rate of the drawing process.

Acquisition and Stored-Video Orientation

The acquisition protocol effective July 15, 2025 is documented in Archive_Specifications.md.

The protocol's capture orientation and a file's stored pixel orientation are different concepts. The orientation, width_px, and height_px metadata describe the published encoded file. Historical canonicalization may include a rotation, crop, or transcode; those transformations are documented only where evidence currently exists and must not be inferred from dimensions alone.

The July 15, 2025 protocol boundary is explicitly encoded in canonical metadata. It supports era-aware comparison of repeated physical-media practice, but it is an acquisition change, not a skill-change label. Differences across that boundary may reflect capture conditions as well as practice development.

Annotation State

Session-level descriptive and provenance metadata are curated by the creator from the physical practice, filenames, file properties, and retained historical evidence. This is the scope of annotations_creators: expert-generated; it does not mean that frame-level task labels exist. Likewise, language: en describes the filename vocabulary, metadata, and documentation, not incidental audio.

The video-classification task facet refers to possible session-level targets such as documented tool, medium, and support. The release does not define label encodings, an official benchmark, or train/validation/test partitions.

The videos themselves are primarily raw and unannotated. The release does not provide:

• frame-level bounding boxes or segmentation masks
• hand, pen-tip, pose, or stroke labels
• aesthetic or quality scores
• imposed stylistic classifications
• calibrated physical trajectories
• original capture timestamps for individual time-lapse frames

Researchers may derive annotations, but derived labels should identify the source record_id and pinned repository revision.

Intended and Unsupported Uses

Reasonable uses in the current release

• archival and drawing-process research
• qualitative within-creator analysis of practice and skill development across repeated physical-media sessions
• self-supervised video representation experiments
• frame-change, scene, and workspace analysis
• development of annotation and tracking methods

Uses requiring additional validation or calibration

• hand or pen-tip tracking
• optical-flow comparison across acquisition eras
• temporal behavior analysis
• real-world trajectory, velocity, acceleration, or motor-control measurement

The release does not currently provide capture intervals, acceleration factors, source-frame timestamps, camera calibration, or pixel-to-physical coordinate mapping. Encoded time-lapse playback must not be treated as real drawing time.

Not validated

• population-level claims about artists or drawing behavior
• clinical, biometric, authorship, or identity inference
• a leakage-safe training or evaluation benchmark
• commercial use outside the CC BY-NC 4.0 terms

Limitations and Sources of Bias

• The archive documents one creator and one evolving practice.
• Current public video coverage begins in 2024, not 2012.
• The standard and pre-standard tiers have substantially different metadata completeness and acquisition conditions.
• The workspace can contain hands, tools, phones, tablets, reference material, and other objects that may create visual shortcuts or occlusion.
• Stored resolution and frame rate do not establish physical or temporal calibration.
• Large files make exhaustive inspection and replication expensive.
• No official leakage-safe split exists.

Rights, Privacy, and Audio

The creator confirms ownership of the published video files and depicted creator-owned drawings. Material that cannot pass the ownership and provenance review is not part of the active public index and is not distributed by this release. This statement does not independently grant rights in incidental third-party material that may appear on screens, in references, or in captured audio.

Most canonical masters retain an audio stream. Users are responsible for reviewing audio, visible screens, references, reflections, tattoos, and other potentially identifying or third-party content before redistribution or model release. A future proxy release should remove audio by default while preserving the archival masters.

Access and Reproducibility

The active 467-video release contains 1,885,402,632,655 bytes of video (approximately 1.89 TB, or 1.71 TiB). Canonical files are often several gigabytes each, so users should select records through metadata.csv before downloading video.

Bandwidth-safe Quick Start

Install Hugging Face Datasets, stream the archive, and disable decoding while selecting records:

from datasets import Video, load_dataset

dataset = load_dataset(
    "maxwellinked/time-lapse-artifacts",
    split="train",
    streaming=True,
)
metadata = dataset.cast_column("video", Video(decode=False))
first = next(iter(metadata))
print(first["video"]["path"], first["tool"], first["medium"])

This exposes the complete metadata schema without transferring video bytes. Default decoded row access additionally requires torchcodec and a compatible video-decoding stack. Use streaming=True unless a full archive download is intentional.

For reproducible work:

1. Pin a repository commit or release tag.
2. Record the selected record_id values.
3. Preserve the published hub_xet_hash values.
4. State whether audio was retained.
5. Document all rotation, crop, sampling, decoding, and proxy-generation steps.

Run the included validator before publishing a metadata change:

python tools/validate_metadata.py metadata.csv

The v0.2 index can be regenerated from the pinned 335-row source index plus the three audited additions:

python tools/repair_metadata.py source_metadata.csv metadata.csv \
  --additions tools/metadata_additions_v0.2.csv

Versioning

The repository is under active construction. Record identity is carried by record_id; a filename may be corrected without creating a new conceptual record. Content replacement must retain provenance and should receive a new content hash.

See CHANGELOG.md for release-level changes.

Documentation

• metadata.csv — canonical machine-readable index
• Pre_Standard_Time_Lapses/metadata.csv — active pre-standard public index
• SCHEMA.md — filename and metadata contracts
• Archive_Specifications.md — acquisition specifications and known gaps
• VALIDATION.md — validation scope and current results
• Standard_Time_Lapses/Provenance/ — historical evidence and mappings

Citation

Until a DOI-backed release is available, cite the dataset and pinned revision:

@dataset{maxwellinked_time_lapse_artifacts_2026,
  author    = {maxwellinked},
  title     = {time-lapse-artifacts},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/maxwellinked/time-lapse-artifacts},
  note      = {Active construction release; include the repository revision used}
}

License

The dataset is released under CC BY-NC 4.0. The license applies only to material for which the licensor holds the necessary rights. Users must conduct their own review of incidental third-party content and downstream use.