maxwellinked commited on
Commit
ed96dab
·
verified ·
1 Parent(s): 1cead61

Ingest October 23 recordings and byte-bound clock observations

Browse files

Add 51 Standard MP4s across five creator-confirmed works and five separately indexed real-time MOVs. Preserve original source-to-canonical SHA-256 mappings, unknown completeness, file-specific displayed-clock scope, and existing selected photographs. Immutable staging identity checks, all validators, and 222 repository tests passed. Generated views bind to exact index revision dfcddd0409ce8a0a5846f74b6ca3891cc0135ea1.

This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +5 -0
  2. CHANGELOG.md +8 -0
  3. Clock_Evidence/README.md +56 -0
  4. Clock_Evidence/SCHEMA.md +25 -0
  5. Clock_Evidence/clock_crops/20_2_corrected.jpg +3 -0
  6. Clock_Evidence/clock_crops/21_2_corrected.jpg +3 -0
  7. Clock_Evidence/clock_crops/30_2.jpg +3 -0
  8. Clock_Evidence/clock_crops/40_2_corrected.jpg +3 -0
  9. Clock_Evidence/clock_crops/42_2.jpg +3 -0
  10. Clock_Evidence/clock_crops/43_0.jpg +3 -0
  11. Clock_Evidence/clock_crops/43_2.jpg +3 -0
  12. Clock_Evidence/clock_crops/44_0.jpg +3 -0
  13. Clock_Evidence/clock_crops/45_0.jpg +3 -0
  14. Clock_Evidence/clock_crops/45_2.jpg +3 -0
  15. Clock_Evidence/clock_crops/51_2.jpg +3 -0
  16. Clock_Evidence/observations.json +0 -0
  17. README.md +36 -22
  18. Real_Time_Recordings/2026-10/IMG_3222.MOV +3 -0
  19. Real_Time_Recordings/2026-10/IMG_3223.MOV +3 -0
  20. Real_Time_Recordings/2026-10/IMG_3225.MOV +3 -0
  21. Real_Time_Recordings/2026-10/IMG_E3206.MOV +3 -0
  22. Real_Time_Recordings/2026-10/IMG_E3221.MOV +3 -0
  23. Real_Time_Recordings/README.md +62 -0
  24. Real_Time_Recordings/SCHEMA.md +22 -0
  25. Real_Time_Recordings/SHA256SUMS +5 -0
  26. Real_Time_Recordings/metadata.csv +6 -0
  27. SCHEMA.md +1 -1
  28. SHA256SUMS +0 -0
  29. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec01.mp4 +3 -0
  30. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec02.mp4 +3 -0
  31. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec03.mp4 +3 -0
  32. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec04.mp4 +3 -0
  33. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec05.mp4 +3 -0
  34. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec06.mp4 +3 -0
  35. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec07.mp4 +3 -0
  36. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec08.mp4 +3 -0
  37. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec09.mp4 +3 -0
  38. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec10.mp4 +3 -0
  39. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec11.mp4 +3 -0
  40. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec12.mp4 +3 -0
  41. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec13.mp4 +3 -0
  42. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec14.mp4 +3 -0
  43. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec15.mp4 +3 -0
  44. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec16.mp4 +3 -0
  45. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec17.mp4 +3 -0
  46. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec18.mp4 +3 -0
  47. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec19.mp4 +3 -0
  48. Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec20.mp4 +3 -0
  49. Standard_Time_Lapses/2026-10-20.1057.Pencil.Graphite.FountainPen.Ink.Paper.9x12.Rec01.mp4 +3 -0
  50. Standard_Time_Lapses/2026-10-20.1057.Pencil.Graphite.FountainPen.Ink.Paper.9x12.Rec02.mp4 +3 -0
.gitattributes CHANGED
@@ -65,3 +65,8 @@ Migration_Staging_2026/2026-06-18.FountainPenInk.Paper.11x14in.TL.Source.Rec01 f
65
  Migration_Staging_2026/2026-06-29.FountainPenInk.Paper.11x14in.TL.Source.Rec03 filter=lfs diff=lfs merge=lfs -text
66
  Migration_Staging_2026/2026-06-29.FountainPenInk.Paper.11x14in.TL.Source.Rec02 filter=lfs diff=lfs merge=lfs -text
67
  Migration_Staging_2026/2026-06-29.FountainPenInk.Paper.11x14in.TL.Source.Rec01 filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
65
  Migration_Staging_2026/2026-06-29.FountainPenInk.Paper.11x14in.TL.Source.Rec03 filter=lfs diff=lfs merge=lfs -text
66
  Migration_Staging_2026/2026-06-29.FountainPenInk.Paper.11x14in.TL.Source.Rec02 filter=lfs diff=lfs merge=lfs -text
67
  Migration_Staging_2026/2026-06-29.FountainPenInk.Paper.11x14in.TL.Source.Rec01 filter=lfs diff=lfs merge=lfs -text
68
+ Real_Time_Recordings/2026-10/IMG_E3206.MOV filter=lfs diff=lfs merge=lfs -text
69
+ Real_Time_Recordings/2026-10/IMG_E3221.MOV filter=lfs diff=lfs merge=lfs -text
70
+ Real_Time_Recordings/2026-10/IMG_3222.MOV filter=lfs diff=lfs merge=lfs -text
71
+ Real_Time_Recordings/2026-10/IMG_3223.MOV filter=lfs diff=lfs merge=lfs -text
72
+ Real_Time_Recordings/2026-10/IMG_3225.MOV filter=lfs diff=lfs merge=lfs -text
CHANGELOG.md CHANGED
@@ -1,5 +1,13 @@
1
  # Changelog
2
 
 
 
 
 
 
 
 
 
3
  ## Unreleased — work-level high-resolution primary layer
4
 
5
  - Add the work-centric `Work_Images` layer with one active research-grade
 
1
  # Changelog
2
 
3
+ ## October 23 coordinated ingest
4
+
5
+ - Add 51 unchanged MP4s and five confirmed work identities, with completeness unknown.
6
+ - Preserve five real-time MOV representations in a separate index, including reader timestamp qualifications and candidate pairings.
7
+ - Attach 333 displayed-clock readings to 125 checksum-bound time-lapse identities; no external calibration or archive-wide conversion.
8
+ - Refresh recording views while preserving prior photograph selections and null images for new works.
9
+ - Retain the established audit clarifications and historical evidence scopes.
10
+
11
  ## Unreleased — work-level high-resolution primary layer
12
 
13
  - Add the work-centric `Work_Images` layer with one active research-grade
Clock_Evidence/README.md ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Displayed-clock evidence
2
+
3
+ This sidecar records **333 accepted display readings at 330 distinct media
4
+ positions across 125 recordings**. It combines 116 positive records from the
5
+ earlier archive review with nine positive records in the October 23 ingest.
6
+ These are file counts, not independent sessions, creators, or calibrated clocks.
7
+
8
+ The supported claim is **file-specific displayed-clock evidence**: a readable
9
+ device displayed the transcribed value at the observed frame. External wall-clock
10
+ accuracy, uninterrupted capture, active drawing time, physical pen velocity, and
11
+ an archive-wide timing conversion have not been established.
12
+
13
+ ## Evidence and inspection scope
14
+
15
+ `observations.json` joins each positive record to its persistent record ID,
16
+ published path, ordinary SHA-256 and byte size. The earlier 116 positive records
17
+ were reconciled using 12 retained full-file hash receipts and 104 fresh hashes,
18
+ with no identity mismatch. This is separate from the visual reading of clocks.
19
+
20
+ The earlier review screened five frames from each of 1,333 indexed recordings
21
+ and made documented follow-up inspections. The new batch was reviewed at three
22
+ positions per MP4, with native crops for the eleven accepted new readings.
23
+ This is sampled review, not continuous inspection of every file. A missing row
24
+ means no accepted reading in this sidecar; it does not establish absence of a
25
+ clock elsewhere in that video. Seven earlier detailed-review records had no
26
+ accepted readings and are not positive entries here.
27
+
28
+ Historical observations retain the decimal decoded PTS recorded by that review;
29
+ their integer PTS/time base is unknown in this export. New observations retain
30
+ integer PTS and time base. Do not invent greater temporal precision. New crops
31
+ are included under `clock_crops/`. Historical native crops remain in the retained
32
+ local review; their basenames and report/sample locators are provided. Researchers
33
+ can re-extract frames from the hash-bound published recordings. This package does
34
+ not claim that every historical crop is publicly included.
35
+
36
+ ## Interpretation
37
+
38
+ - A displayed civil-time value is an observation of the display, not independent
39
+ validation of UTC accuracy or device synchronization.
40
+ - Multiple readings may support conditional elapsed-display-time arithmetic.
41
+ Date, AM/PM, clock adjustment, continuity and display resolution still matter.
42
+ No automatic date rollover or archive-wide conversion is applied here.
43
+ - Unrecorded intervals have unknown activity. File gaps cannot establish what
44
+ happened during them.
45
+ - One creator limits population generalization. It does not invalidate an
46
+ observed clock display; clock calibration is a separate measurement question.
47
+ - Recording-app counters, media-player progress, photograph metadata and
48
+ handwritten finish labels are not included as device-clock readings.
49
+
50
+ The [real-time cohort](../Real_Time_Recordings/README.md) preserves additional
51
+ recorded observations, including intervals compressed by time-lapse. Its
52
+ candidate links do not supply a validated conversion to other files.
53
+
54
+ Run `python tools/validate_observational_evidence.py --root .` to check joins,
55
+ position bounds, duplicate entries and the limited claim scope. Passing this
56
+ validator is not a new visual review or external clock calibration.
Clock_Evidence/SCHEMA.md ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Clock observation sidecar, version 1
2
+
3
+ `observations.json` uses schema `displayed-clock-evidence/v1`. This additive
4
+ sidecar does not alter the established 35-field time-lapse indexes.
5
+
6
+ | Field | Meaning |
7
+ | --- | --- |
8
+ | `claim_scope` | Fixed to `file_specific_displayed_clock_evidence`. |
9
+ | `externally_calibrated` | `false` for this evidence packet. |
10
+ | `archive_conversion_established` | `false`; no transfer model is published. |
11
+ | `records[].record_id`, `file_path`, `sha256`, `size_bytes` | Exact canonical identity binding. |
12
+ | `identity_verification` | Full-file hash reconciliation method, separate from visual interpretation. |
13
+ | `observation_source_revision` | Revision used by the review, or null for observations made before initial publication. |
14
+ | `observations[].video_pts_seconds` | Decimal string for an observed media position, not elapsed drawing time. |
15
+ | `pts`, `time_base` | Exact integer PTS and rational time base where retained; both null otherwise. |
16
+ | `position_basis` | `integer_pts_and_time_base` or `observed_decoded_pts_seconds`. |
17
+ | `display_literal` | Human-readable transcription; no date, AM/PM or timezone is inferred. |
18
+ | `display_kind`, `displayed_date_literal` | Display type and date text if recorded; null date remains unknown. |
19
+ | `display_precision` | Visible display resolution; currently minute-level. |
20
+ | `am_pm_visibility`, `year_visibility` | Recorded visibility, including unknown/not-recorded states. |
21
+ | `evidence_locator` | Public crop path or explicitly local report/sample locator. |
22
+ | `retained_crop_basenames`, `note` | Optional retained evidence locators and reading qualifications. |
23
+
24
+ Different displays at the same media position are separate observations. Count
25
+ distinct positions within each file before reporting temporal coverage.
Clock_Evidence/clock_crops/20_2_corrected.jpg ADDED

Git LFS Details

  • SHA256: e55dd963aa18b8475bd921dd2c998443ca3c2613dabfc610d0f2fc78c051ce08
  • Pointer size: 131 Bytes
  • Size of remote file: 106 kB
Clock_Evidence/clock_crops/21_2_corrected.jpg ADDED

Git LFS Details

  • SHA256: 34486833f30f4e83d522138384e619973c2c93d66a67a80ac923e0aba10cf731
  • Pointer size: 131 Bytes
  • Size of remote file: 141 kB
Clock_Evidence/clock_crops/30_2.jpg ADDED

Git LFS Details

  • SHA256: d6fd256118c311005487de54ed037b4ab8ccf45742ae1bada399ca2ba92f40f1
  • Pointer size: 130 Bytes
  • Size of remote file: 88.4 kB
Clock_Evidence/clock_crops/40_2_corrected.jpg ADDED

Git LFS Details

  • SHA256: 2e1c7bf408058988966a84859b6c6daadfdbf538c8bb05bc0c1cd14b6a6f66b6
  • Pointer size: 131 Bytes
  • Size of remote file: 135 kB
Clock_Evidence/clock_crops/42_2.jpg ADDED

Git LFS Details

  • SHA256: 495c0e5911c7b0fd14886400188134564114dd139014f700cd32850d2841638b
  • Pointer size: 131 Bytes
  • Size of remote file: 124 kB
Clock_Evidence/clock_crops/43_0.jpg ADDED

Git LFS Details

  • SHA256: 5a0e59c4fbac56b3450e860bcfd001f15c46cdc5ed90a35a1f48241dd2b0e2a4
  • Pointer size: 131 Bytes
  • Size of remote file: 129 kB
Clock_Evidence/clock_crops/43_2.jpg ADDED

Git LFS Details

  • SHA256: 04e28701c60752eedaace09e739b70909b15473002fce32526a8ef8acf88eff1
  • Pointer size: 131 Bytes
  • Size of remote file: 140 kB
Clock_Evidence/clock_crops/44_0.jpg ADDED

Git LFS Details

  • SHA256: 0e8dcec56d71e3f0d9e9014158455853f528e18ac9f7eaff89b45dde3ba6eeab
  • Pointer size: 131 Bytes
  • Size of remote file: 146 kB
Clock_Evidence/clock_crops/45_0.jpg ADDED

Git LFS Details

  • SHA256: 2711f83dc45888b98551d4d98d1ee2ac62de77ef38117c84c3af945844bd6088
  • Pointer size: 131 Bytes
  • Size of remote file: 160 kB
Clock_Evidence/clock_crops/45_2.jpg ADDED

Git LFS Details

  • SHA256: 8f80677af08b0fdcf7d0812461498073d3951950515764b3e8ba025d54e92871
  • Pointer size: 131 Bytes
  • Size of remote file: 155 kB
Clock_Evidence/clock_crops/51_2.jpg ADDED

Git LFS Details

  • SHA256: 94f1bae19e32da9260a90148004569793a68c9552f418bd0969067c4a533bd41
  • Pointer size: 131 Bytes
  • Size of remote file: 119 kB
Clock_Evidence/observations.json ADDED
The diff for this file is too large to render. See raw diff
 
README.md CHANGED
@@ -92,7 +92,7 @@ dataset_info:
92
  dtype: large_string
93
  splits:
94
  - name: archive
95
- num_examples: 1205
96
  - config_name: pre_standard
97
  features:
98
  - name: video
@@ -208,7 +208,7 @@ dataset_info:
208
  dtype: large_string
209
  splits:
210
  - name: archive
211
- num_examples: 1333
212
  # BEGIN generated workflow-subsets dataset_info
213
  - config_name: standard_earlier
214
  features:
@@ -359,7 +359,7 @@ dataset_info:
359
  dtype: large_string
360
  splits:
361
  - name: archive
362
- num_examples: 841
363
  # END generated workflow-subsets dataset_info
364
  configs:
365
  - config_name: canonical
@@ -711,11 +711,12 @@ configs:
711
  # Time-Lapse Artifacts
712
 
713
  Time-Lapse Artifacts is a longitudinal dataset and archive. Its
714
- **1,333 indexed video files** document one artist's traditional drawing practice.
715
- The recorded finish dates span September 17, 2024 through October 16, 2026;
 
716
  nine Pre-Standard dates remain unknown. Standardized acquisition began July 13,
717
- 2025. The current indexes contain 2,365,664,498,906 indexed video bytes
718
- (approximately 2.37 TB).
719
 
720
  The recordings began as personal practice documentation and a durable record of
721
  manual work. The archive was initially organized as a dataset retrospectively;
@@ -738,7 +739,7 @@ The archive documents recorded drawing processes and photographed drawing states
738
 
739
  | Configuration | Indexed video files | Scope |
740
  | --- | ---: | --- |
741
- | `canonical` (default) | 1,205 | All Standard files in the root [index](metadata.csv). |
742
  | `pre_standard` | 128 | Earlier acquisition/documentation era; [index](Pre_Standard_Time_Lapses/metadata.csv). |
743
 
744
  Both use the `archive` split and the same 35-field schema. Historical descriptive
@@ -746,13 +747,23 @@ fields may be unknown even when technical metadata is available. Blank/null
746
  values are not zero or evidence of absence. No official train, validation, or
747
  test partition is defined.
748
 
749
- The four recording configurations have 2,538 aggregate rows, counting Standard
750
  through both the combined view
751
- and its two subdivisions below. There are 1,333 unique indexed video files.
752
  For archive-wide selection, use
753
  `canonical` + `pre_standard`, or the two Standard views + `pre_standard`.
754
  Do not combine all four configurations as independent records.
755
 
 
 
 
 
 
 
 
 
 
 
756
  The separate [2012–2016 archival holding](Earlier_Archive_2012_2016/README.md)
757
  has a public file manifest, but its media is held locally and excluded from these
758
  totals. Its historical scope label does not date every listed file.
@@ -782,16 +793,17 @@ confirmed photograph. Historical receipts retain their original claims.
782
 
783
  ### Recording and photograph projection
784
 
785
- `recording_physical_references/archive` contains the same 1,333 indexed
786
- recordings in Standard then Pre-Standard order. Its single `image` column embeds
787
- the selected display photograph for each of the 1,205 Standard recording rows,
788
- joined through the positive work registry. The 128 Pre-Standard rows have no
789
- registered work link and retain true null images. This view adds no recordings
 
790
  or work identities. Neither a photograph nor its selection establishes a
791
  recording's endpoint, timing, completeness, finish state, or frame alignment.
792
 
793
- The following counts describe **images and metadata included in the published
794
- projection**, using distinct active Standard `work_id` values at revision
795
  `6fed0d073bdc307cdb81ea1dd8d4db6b1ecb04c7`:
796
 
797
  | Property | Work count | Scope |
@@ -840,11 +852,11 @@ Standard again. Both retain the `archive` split.
840
  | Configuration | Existing protocol value | Indexed files | Distinct registered works |
841
  | --- | --- | ---: | ---: |
842
  | `standard_earlier` | `standard-2025-07-13` | 364 | 342 |
843
- | `standard_individual_recordings` | `standard-2026-08-27` | 841 | 89 |
844
 
845
  Counts use the video indexes and [work registry](Work_Identity/README.md) in the
846
  same containing repository revision. The combined Standard total is
847
- **431 distinct registered works**, calculated from the union of work IDs;
848
  **0 works occur in both workflows** at this revision. Work counts must not
849
  be assumed additive in later revisions. File totals depend on recording
850
  segmentation: multiple files can belong to one drawing. They do not measure
@@ -930,13 +942,14 @@ identify earlier releases, not the 1,333-file revision used above; see
930
 
931
  ## Recording and drawing identities
932
 
933
- A **`record_id`** identifies one published MP4 file. Historical files may combine
 
934
  multiple source recordings; the Standard individual-recording workflow preserves
935
  one unchanged source recording per file. Neither a metadata row nor a file
936
  boundary establishes a drawing, session, or complete process.
937
 
938
  A **`work_id`** identifies a creator-confirmed drawing/composition. Currently,
939
- 1,205 files link to 431 active work IDs. The 128 Pre-Standard files have blank
940
  `work_id` values, but each has a separate, scope-limited creator-supplied
941
  physical-reference link. These links do not assign work IDs or establish
942
  whole-recording membership. Blank does not mean a different drawing. The
@@ -1073,7 +1086,8 @@ rotation, cropping, sampling, proxy generation, and whether audio was retained.
1073
  The [verified downloader guide](Workflow_Subsets/README.md#retain-a-selection-before-downloading-videos)
1074
  shows a metadata-only check before intentional media transfer.
1075
 
1076
- Audio streams occur in 486 of 1,333, or 36.5% of indexed files. `has_audio` reports
 
1077
  stream presence, not meaningful signal. Review incidental audio, visible
1078
  references, and identifying or third-party content before redistribution or
1079
  model release. Derived proxies should remove audio by default while preserving
 
92
  dtype: large_string
93
  splits:
94
  - name: archive
95
+ num_examples: 1256
96
  - config_name: pre_standard
97
  features:
98
  - name: video
 
208
  dtype: large_string
209
  splits:
210
  - name: archive
211
+ num_examples: 1384
212
  # BEGIN generated workflow-subsets dataset_info
213
  - config_name: standard_earlier
214
  features:
 
359
  dtype: large_string
360
  splits:
361
  - name: archive
362
+ num_examples: 892
363
  # END generated workflow-subsets dataset_info
364
  configs:
365
  - config_name: canonical
 
711
  # Time-Lapse Artifacts
712
 
713
  Time-Lapse Artifacts is a longitudinal dataset and archive. Its
714
+ **1,384 indexed time-lapse files and five separately indexed real-time recordings**
715
+ document one artist's traditional drawing practice.
716
+ The time-lapse recorded finish dates span September 17, 2024 through October 22, 2026;
717
  nine Pre-Standard dates remain unknown. Standardized acquisition began July 13,
718
+ 2025. The time-lapse indexes contain 2,400,224,173,549 indexed video bytes
719
+ (approximately 2.40 TB). The real-time cohort adds 62,641,549,258 bytes.
720
 
721
  The recordings began as personal practice documentation and a durable record of
722
  manual work. The archive was initially organized as a dataset retrospectively;
 
739
 
740
  | Configuration | Indexed video files | Scope |
741
  | --- | ---: | --- |
742
+ | `canonical` (default) | 1,256 | All Standard files in the root [index](metadata.csv). |
743
  | `pre_standard` | 128 | Earlier acquisition/documentation era; [index](Pre_Standard_Time_Lapses/metadata.csv). |
744
 
745
  Both use the `archive` split and the same 35-field schema. Historical descriptive
 
747
  values are not zero or evidence of absence. No official train, validation, or
748
  test partition is defined.
749
 
750
+ The four time-lapse recording configurations have 2,640 aggregate rows, counting Standard
751
  through both the combined view
752
+ and its two subdivisions below. There are 1,384 unique indexed time-lapse files.
753
  For archive-wide selection, use
754
  `canonical` + `pre_standard`, or the two Standard views + `pre_standard`.
755
  Do not combine all four configurations as independent records.
756
 
757
+ The [real-time collection](Real_Time_Recordings/README.md) has its own five-row
758
+ CSV index and MOV checksum manifest. It is excluded from the five native
759
+ Parquet configurations and all time-lapse workflow totals above. Its candidate
760
+ pairings do not establish calibrated timing or work membership.
761
+
762
+ [Displayed-clock observations](Clock_Evidence/README.md) attach 333 readings
763
+ at 330 media positions to 125 checksum-bound time-lapse recordings. They record
764
+ what devices displayed in those files. External clock accuracy and an archive-wide
765
+ conversion remain unestablished.
766
+
767
  The separate [2012–2016 archival holding](Earlier_Archive_2012_2016/README.md)
768
  has a public file manifest, but its media is held locally and excluded from these
769
  totals. Its historical scope label does not date every listed file.
 
793
 
794
  ### Recording and photograph projection
795
 
796
+ `recording_physical_references/archive` contains the same 1,384 indexed
797
+ time-lapse recordings in Standard then Pre-Standard order. Its single `image`
798
+ column preserves the selected display photograph for 1,205 Standard rows. The
799
+ 51 new Standard rows have registered work links but no selected photograph;
800
+ they retain true null images. The 128 Pre-Standard rows also retain true null
801
+ images because they have no registered work link. This view adds no recordings
802
  or work identities. Neither a photograph nor its selection establishes a
803
  recording's endpoint, timing, completeness, finish state, or frame alignment.
804
 
805
+ The following counts describe the **prior photograph selection retained in the
806
+ current projection**, using distinct Standard `work_id` values at revision
807
  `6fed0d073bdc307cdb81ea1dd8d4db6b1ecb04c7`:
808
 
809
  | Property | Work count | Scope |
 
852
  | Configuration | Existing protocol value | Indexed files | Distinct registered works |
853
  | --- | --- | ---: | ---: |
854
  | `standard_earlier` | `standard-2025-07-13` | 364 | 342 |
855
+ | `standard_individual_recordings` | `standard-2026-08-27` | 892 | 94 |
856
 
857
  Counts use the video indexes and [work registry](Work_Identity/README.md) in the
858
  same containing repository revision. The combined Standard total is
859
+ **436 distinct registered works**, calculated from the union of work IDs;
860
  **0 works occur in both workflows** at this revision. Work counts must not
861
  be assumed additive in later revisions. File totals depend on recording
862
  segmentation: multiple files can belong to one drawing. They do not measure
 
942
 
943
  ## Recording and drawing identities
944
 
945
+ A **`record_id`** identifies one published video object. The time-lapse indexes
946
+ contain MP4 files; real-time MOVs have a separate index. Historical files may combine
947
  multiple source recordings; the Standard individual-recording workflow preserves
948
  one unchanged source recording per file. Neither a metadata row nor a file
949
  boundary establishes a drawing, session, or complete process.
950
 
951
  A **`work_id`** identifies a creator-confirmed drawing/composition. Currently,
952
+ 1,256 time-lapse files link to 436 active work IDs. The 128 Pre-Standard files have blank
953
  `work_id` values, but each has a separate, scope-limited creator-supplied
954
  physical-reference link. These links do not assign work IDs or establish
955
  whole-recording membership. Blank does not mean a different drawing. The
 
1086
  The [verified downloader guide](Workflow_Subsets/README.md#retain-a-selection-before-downloading-videos)
1087
  shows a metadata-only check before intentional media transfer.
1088
 
1089
+ Audio streams occur in 486 of 1,384, or 35.1% of indexed time-lapse files.
1090
+ All five separately indexed real-time MOVs have audio streams. `has_audio` reports
1091
  stream presence, not meaningful signal. Review incidental audio, visible
1092
  references, and identifying or third-party content before redistribution or
1093
  model release. Derived proxies should remove audio by default while preserving
Real_Time_Recordings/2026-10/IMG_3222.MOV ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6e7a551d54d203fce283d1c0e4e08a08fbc2023cc4098de1b41b714e30423c5
3
+ size 16082615335
Real_Time_Recordings/2026-10/IMG_3223.MOV ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2973d1ab13cf114d738e2f1bd918717dbba049f7c326db30527c83196ef8d1a4
3
+ size 12032805079
Real_Time_Recordings/2026-10/IMG_3225.MOV ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d4cd6668ab3b5d0fa0b8e91bca9b45c73a6cf3a92ae7e27095419c8df577739
3
+ size 13270216513
Real_Time_Recordings/2026-10/IMG_E3206.MOV ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c695fe3fc778f94f52e1d59939d4c870ad0f17989b1deeba280067f08dd07322
3
+ size 9257697642
Real_Time_Recordings/2026-10/IMG_E3221.MOV ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:281ac10e7815dae62e1f57a7b34b42fede760f849111ad1e48ebcd29fdb45427
3
+ size 11998214689
Real_Time_Recordings/README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Real-time observational recordings
2
+
3
+ This separately indexed cohort preserves five MOV files, totaling
4
+ **62,641,549,258 bytes**, as observational source evidence. It is separate from
5
+ the Standard and Pre-Standard time-lapse indexes and their acquisition-protocol
6
+ partitions. Inclusion does not require a claim of externally calibrated time.
7
+
8
+ | Included representation | Candidate time-lapse source | Demux result |
9
+ | --- | --- | --- |
10
+ | `IMG_E3206.MOV` | `VID_20261020_003433.mp4` | Completed with timestamp errors |
11
+ | `IMG_E3221.MOV` | `VID_20261022_005729.mp4` | Completed with timestamp errors |
12
+ | `IMG_3222.MOV` | `VID_20261022_021031.mp4` | No logged errors |
13
+ | `IMG_3223.MOV` | `VID_20261022_024937.mp4` | No logged errors |
14
+ | `IMG_3225.MOV` | October 22, 04:30 finish-labeled MP4 | No logged errors |
15
+
16
+ `metadata.csv` provides exact paths, record IDs, sizes, ordinary SHA-256, Hub Xet
17
+ identities, encoded durations and candidate paired record IDs. `SHA256SUMS`
18
+ covers this MOV collection; the root checksum manifest retains its established
19
+ MP4/ZIP scope. Files are preserved byte-for-byte, without remuxing or re-encoding.
20
+
21
+ ## Selection and provenance
22
+
23
+ Seven MOVs were reviewed. Two pairs (`IMG_3222` / `IMG_E3222` and `IMG_3223` /
24
+ `IMG_E3223`) have matching encoded durations/frame counts, capture metadata and
25
+ reviewed visual samples. They are not byte-identical, and full-frame equivalence
26
+ was not established. One representation per candidate event is included: the
27
+ unprefixed files where both exist. The two alternates remain locally retained.
28
+ The source folder contains no unprefixed counterpart for `IMG_E3206` or
29
+ `IMG_E3221` at the exact expected paths checked.
30
+
31
+ The creator stated that no export edits were made. This is creator-reported
32
+ provenance, not independent verification of each file's export lineage. An `E`
33
+ prefix does not establish intentional editing. Source-role status remains
34
+ `not_independently_verified_as_camera_original`.
35
+
36
+ Candidate pairings are supported by reviewed imagery and capture metadata.
37
+ They are not calibrated alignments or creator-confirmed MOV-to-work assignments.
38
+ The `work_id` field therefore remains empty. The five-work confirmation applies
39
+ to the listed MP4 memberships only.
40
+
41
+ ## Timestamp behavior
42
+
43
+ FFmpeg 9.0 default packet reading showed 26 non-increasing DTS events in each of
44
+ `IMG_E3206` and `IMG_E3221`. The retained alternate `IMG_E3222` showed one. Reading
45
+ with `-ignore_editlist 1` removed these events and returned packet counts matching
46
+ the declared video frame counts. Default decoding of six-second windows around
47
+ the first affected regions produced 361 frames each, no logged errors and
48
+ strictly increasing decoded PTS.
49
+
50
+ This implicates MOV edit-list/timeline reader handling. It does not establish
51
+ corrupted imagery or intentional user editing. The source files were not
52
+ repaired. The review did not fully decode every MOV or establish every edit-list
53
+ semantic. Preserve reader settings when attempting replication.
54
+
55
+ ## Temporal scope
56
+
57
+ These files can ground claims in recorded observations and provide the origin
58
+ for future file-specific studies. Encoded media duration is not independently
59
+ validated wall-clock duration. Visible clock values have their own
60
+ [observation evidence](../Clock_Evidence/README.md). No uninterrupted drawing
61
+ session, whole-session coverage, activity during gaps, common time-lapse factor,
62
+ cross-artist prediction, or historical priority is claimed.
Real_Time_Recordings/SCHEMA.md ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Real-time index
2
+
3
+ This collection has its own CSV and checksum manifest. It is not assigned a
4
+ time-lapse acquisition protocol or included in time-lapse workflow counts.
5
+
6
+ | Field group | Contract |
7
+ | --- | --- |
8
+ | `record_id`, `file_name` | Permanent object ID and repository-relative MOV path. IDs use the existing `tla-` namespace and are unique across collections. |
9
+ | `original_filename`, `sha256`, `hub_xet_hash`, `size_bytes` | Source name and exact byte identities. |
10
+ | `duration_seconds`, `width_px`, `height_px`, `encoded_frame_rate`, `video_codec`, `pixel_format`, `has_audio`, `time_base`, `nb_frames` | Probe observations of the encoded representation; no physical-time calibration follows. |
11
+ | `work_id` | Empty for this initial cohort: candidate pairing is not positive work-identity authority. |
12
+ | `paired_record_id`, `pairing_status` | Candidate Standard recording and `candidate_not_calibrated`. |
13
+ | `temporal_basis` | `encoded_media_timeline`. |
14
+ | `clock_accuracy_status` | `not_externally_calibrated`. |
15
+ | `source_role_status` | `not_independently_verified_as_camera_original`. |
16
+ | `demux_status` | `passed_no_logged_errors` or `completed_with_timestamp_errors`; read the diagnostic before interpreting a warning. |
17
+ | `notes` | Representation-specific qualifications. |
18
+
19
+ Record IDs initially use the existing namespaced path-hash method. They must
20
+ survive later descriptive renames. Never substitute the Xet identity for ordinary
21
+ file SHA-256. The accompanying validator checks the initial cohort's bounded
22
+ claims, not general reproducibility of drawing behavior.
Real_Time_Recordings/SHA256SUMS ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ d6e7a551d54d203fce283d1c0e4e08a08fbc2023cc4098de1b41b714e30423c5 Real_Time_Recordings/2026-10/IMG_3222.MOV
2
+ 2973d1ab13cf114d738e2f1bd918717dbba049f7c326db30527c83196ef8d1a4 Real_Time_Recordings/2026-10/IMG_3223.MOV
3
+ 7d4cd6668ab3b5d0fa0b8e91bca9b45c73a6cf3a92ae7e27095419c8df577739 Real_Time_Recordings/2026-10/IMG_3225.MOV
4
+ c695fe3fc778f94f52e1d59939d4c870ad0f17989b1deeba280067f08dd07322 Real_Time_Recordings/2026-10/IMG_E3206.MOV
5
+ 281ac10e7815dae62e1f57a7b34b42fede760f849111ad1e48ebcd29fdb45427 Real_Time_Recordings/2026-10/IMG_E3221.MOV
Real_Time_Recordings/metadata.csv ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ record_id,file_name,original_filename,sha256,hub_xet_hash,size_bytes,duration_seconds,width_px,height_px,encoded_frame_rate,video_codec,pixel_format,has_audio,time_base,nb_frames,work_id,paired_record_id,pairing_status,temporal_basis,clock_accuracy_status,source_role_status,demux_status,notes
2
+ tla-086bbc187b9b862f,Real_Time_Recordings/2026-10/IMG_E3206.MOV,IMG_E3206.MOV,c695fe3fc778f94f52e1d59939d4c870ad0f17989b1deeba280067f08dd07322,085b2a09d2ccef9cd655c8fffa37f7080528ccf23759a40db47203d8315bcb83,9257697642,2669.145000,3826,2150,59.972388,hevc,yuv420p10le,true,1/600,160075,,tla-e7dd855a7896d0c9,candidate_not_calibrated,encoded_media_timeline,not_externally_calibrated,not_independently_verified_as_camera_original,completed_with_timestamp_errors,Unchanged supplied representation. Pairing supported by imagery and capture metadata; no calibrated alignment or independent source-lineage claim.
3
+ tla-5cc8bff9d170577f,Real_Time_Recordings/2026-10/IMG_E3221.MOV,IMG_E3221.MOV,281ac10e7815dae62e1f57a7b34b42fede760f849111ad1e48ebcd29fdb45427,fcbe68a1bdb52e4a75da6396e6d13e58b0c944dca370cc24404e8e559f566995,11998214689,3490.723333,3826,2150,59.972384,hevc,yuv420p10le,true,1/600,209347,,tla-9d0913586a134a39,candidate_not_calibrated,encoded_media_timeline,not_externally_calibrated,not_independently_verified_as_camera_original,completed_with_timestamp_errors,Unchanged supplied representation. Pairing supported by imagery and capture metadata; no calibrated alignment or independent source-lineage claim.
4
+ tla-910c68c14d743e13,Real_Time_Recordings/2026-10/IMG_3222.MOV,IMG_3222.MOV,d6e7a551d54d203fce283d1c0e4e08a08fbc2023cc4098de1b41b714e30423c5,1e47521ac308c7950c4029d9dacacf85e8686421e8ec8a003466cde8942d1457,16082615335,1555.000000,3840,2160,59.972347,hevc,yuv420p10le,true,1/600,93257,,tla-9b648b2c608eb523,candidate_not_calibrated,encoded_media_timeline,not_externally_calibrated,not_independently_verified_as_camera_original,passed_no_logged_errors,Unchanged supplied representation. Pairing supported by imagery and capture metadata; no calibrated alignment or independent source-lineage claim.
5
+ tla-1173f78b2c88e6ca,Real_Time_Recordings/2026-10/IMG_3223.MOV,IMG_3223.MOV,2973d1ab13cf114d738e2f1bd918717dbba049f7c326db30527c83196ef8d1a4,3ad5fd7d446d146dc404effc43c1fc1cde5c935804f6bd45b7d9802cd5ef1e52,12032805079,1154.030000,3840,2160,59.972358,hevc,yuv420p10le,true,1/600,69210,,tla-6e997c9a39698d6b,candidate_not_calibrated,encoded_media_timeline,not_externally_calibrated,not_independently_verified_as_camera_original,passed_no_logged_errors,Unchanged supplied representation. Pairing supported by imagery and capture metadata; no calibrated alignment or independent source-lineage claim.
6
+ tla-32e62eed05b62568,Real_Time_Recordings/2026-10/IMG_3225.MOV,IMG_3225.MOV,7d4cd6668ab3b5d0fa0b8e91bca9b45c73a6cf3a92ae7e27095419c8df577739,8c556d2b5c7881574880494f84d6968493e3dc459912326e9294e81ab4c23fda,13270216513,1301.018333,3840,2160,59.97525,hevc,yuv420p10le,true,1/600,78029,,tla-12fb3f2c65d95902,candidate_not_calibrated,encoded_media_timeline,not_externally_calibrated,not_independently_verified_as_camera_original,passed_no_logged_errors,Unchanged supplied representation. Pairing supported by imagery and capture metadata; no calibrated alignment or independent source-lineage claim.
SCHEMA.md CHANGED
@@ -139,7 +139,7 @@ the authoritative relationship source; the index column is a generated copy.
139
  Join by exact `record_id` within the same repository revision.
140
 
141
  Only positive, owner-confirmed links to active work IDs are projected. The
142
- current 1,205 linked recordings share 431 active work IDs; 128 Pre-Standard recordings remain blank.
143
  No IDs are minted, reassigned or inferred from dates, filenames, sequence
144
  numbers, shared materials or appearance by this projection. A work ID does not
145
  assert recording-set completeness, gapless capture, session boundaries or
 
139
  Join by exact `record_id` within the same repository revision.
140
 
141
  Only positive, owner-confirmed links to active work IDs are projected. The
142
+ current 1,256 linked recordings share 436 active work IDs; 128 Pre-Standard recordings remain blank.
143
  No IDs are minted, reassigned or inferred from dates, filenames, sequence
144
  numbers, shared materials or appearance by this projection. A work ID does not
145
  assert recording-set completeness, gapless capture, session boundaries or
SHA256SUMS CHANGED
The diff for this file is too large to render. See raw diff
 
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec01.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c237cbe1bfeddc18d9d8bf48fc947bd17636ca9646b4d76c601a6f934f585c0
3
+ size 342133709
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec02.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c37a344b3b3b5e06d5b77ba53b7ac12a103d8aae975eb37c45640d74a3724bb
3
+ size 533208509
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec03.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ba31f9df90bf6733ab3e79efea8183bc878433de319fba2f710c3b7f4157456
3
+ size 331085831
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec04.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8dbed15e9c822c47599c7e44b5dd5b04859ed1823d47b6ea71f6add90e83c13
3
+ size 444169981
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec05.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:169845442078fb7d383d0900b7bf965f93587b0617f1604938d5c771fa371eb6
3
+ size 423852407
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec06.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53eeb95061b54f881205c2c272d8cfd105da261054fd67209018a8410708983d
3
+ size 436541821
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec07.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ecaa2bd999671a1d59709ca383bc0ae35b05f8fc4cc59fe897ff7d71b477d2b8
3
+ size 201996876
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec08.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4336e7a7d0ae38b37f646aaa54ab28588e069c3d5cb4cb006d4b5549b837b83
3
+ size 608714437
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec09.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67fc0ff288f4910b73134bd6b6816b588edfbae1e2fdd528c89f82c02cab7a43
3
+ size 235472117
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec10.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c9c93c9bbdef9c792c754b6cbdf06473b93a6d4f7876139569a0f8f24e7e9c6
3
+ size 1198331455
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec11.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1dee31e62b3115076f68a732d64f926438fd4535156fb9cfb8f5096bb38cc810
3
+ size 328106647
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec12.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef5ccb048a792c5c6a7e79b479a41922cfd64f1928516a4922bd22311bf851fa
3
+ size 682578580
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec13.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:488ce9ea62c5a66498153c4bbbede98534c654740e02982a04d58c149fe5fc56
3
+ size 634167039
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec14.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7bffa59491e3716dc35197e5445120ac7da8da15bacf29f8a9afb38416163395
3
+ size 1595436182
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec15.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d44eb579ad6b1a208a53c433199f53df788c6de63bac551b2eaac3533b11f278
3
+ size 1270234616
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec16.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3277de0b47302650fc08792113e5a735b5c213119db4baa15a2cc50e6c46313a
3
+ size 1155639330
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec17.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:007d5469520b657ff920b1fa7cf9bb5f2403d4fc15dd0810779e7b8e1e7dca8d
3
+ size 581843929
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec18.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:363acc5baab2d6c559a4dbf28c20921c1060e13918ba92971236af0a7ce8786d
3
+ size 962463112
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec19.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:693d20a3cc49f6804bc27113401e4f14a8b59b4693b592db64160b6cefdaf15d
3
+ size 407078891
Standard_Time_Lapses/2026-10-19.2245.Pencil.Graphite.BallpointPen.Ink.Paper.9x12.Rec20.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9a80c2f7935cd6d30b20210f928c214383b72d89ebe137f97ca2b30c8e27eb2
3
+ size 343365316
Standard_Time_Lapses/2026-10-20.1057.Pencil.Graphite.FountainPen.Ink.Paper.9x12.Rec01.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f22390a48a95beabd2c7922b313c74ee50fc3d3e62d229ac89aeadcd57ebf514
3
+ size 1943767328
Standard_Time_Lapses/2026-10-20.1057.Pencil.Graphite.FountainPen.Ink.Paper.9x12.Rec02.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d7c0a43e66bd6b177e81e1092e5511667b6f3f60d5c7211f1bd58516ec3b46e
3
+ size 313604747