Original films by LeoGuoLabs / Codex Film Studio. Complete 1920 × 1080, 24 FPS web editions of the final 180-second English films. Full quality production masters are retained separately.
# Source and rights ledger - Reference: **The Fields Under the Forest / 林下的田**, LemoLab, 40 seconds, 1920×1080 at 24 fps. Original film downloaded from https://github.com/lemomo-ai/lemo-opuscar/releases/download/films/lidar-survey.mp4. Film/stills CC BY 4.0; source engine MIT. Adapted here: rendering, fonts and effect grammar. New survey locations, narration, measurements, camera route and score. Original source and MIT license retained under `reference/`. - Engine reference: https://github.com/lemomo-ai/lemo-opuscar/tree/main/styles/lidar-survey. Attribution retained in both film end cards. No claim of endorsement by LemoLab. - Laser data: USGS 3DEP, US Government Public Domain, cloud-hosted EPT registry https://registry.opendata.aws/usgs-lidar/. Forest: `USGS_LPC_CT_Statewide_B8_2016`; quarry: `CT_Statewide_B3_2016`. EPT metadata, node selection, raw LAZ returns and derived grids are retained in each film's `assets/raw/`. - Survey season: Connecticut statewide 2016 data acquired March 11–April 16, 2016; classified point cloud and 1 m DEM described by UConn/CT DEEP at https://maps.cteco.uconn.edu/data/lidar/ct_lidar/. Spring point cloud and summer NAIP imagery are separate acquisitions; they are not represented as simultaneous capture. - Aerial photographs: USDA NAIP, public domain, delivered by Microsoft Planetary Computer. Forest item `ct_m_4207264_sw_18_.6_20160706_20160913` (2016-07-06); quarry item `ct_m_4107233_se_18_.6_20160727_20160913` (2016-07-27). STAC search responses retained in each film. The exact crop is 744 × 1000 local metres, north up. - LiDAR explanation: USGS https://www.usgs.gov/faqs/what-lidar-data-and-where-can-i-download-it and https://pubs.usgs.gov/circ/1399/pdf/circ1399.pdf. Ground classification supports a bare-earth surface; height-above-ground is calculated relative to that surface. No invented tree count, species identification or perfect detection claim. - Forest location: Nipmuck forest region / east Connecticut; https://ctparks.com/parks/nipmuck-state-forest and https://portal.ct.gov/deep/state-parks/forests/nipmuck-state-forest-bigelow-hollow-state-park/overview. No forest-wide height claim: all reported statistics refer to this downloaded window. - Quarry history: Connecticut State Parks https://ctparks.com/parks/sleeping-giant-state-park and Sleeping Giant Park Association https://sgpa.org/quarry-history/. Only the documented fact of former quarrying is narrated. No excavation-volume estimate or invented date is inferred from the single lidar survey. - Fonts: Barlow Condensed (light/regular/medium), IBM Plex Mono (regular/medium), Noto Sans SC, downloaded from Google Fonts; OFL license texts copied from the source demo. - Narration: English adaptations of the original scripts, synthesized with Microsoft Edge TTS, stock `en-US-GuyNeural` voice at -4% requested synthesis rate. No clone of any identified person. Source text hashes and unmodified generated takes retained. - Score and effects: original procedural synthesis authored for these timelines. No reference soundtrack copied. ## Measurement rules Raw measured positions remain fixed. East/north use a local tangent approximation with the film's stated origin; EPT vertical coordinates are retained as delivered. Their numerical scale is consistent with metres; the EPT EPSG:3857 WKT does not independently specify the vertical datum, so this production does not assert a transformed vertical reference. Class 2 is provider-classified ground; classes 7 and 18 are excluded. Other classes are not assumed to identify individual trees. Heights are the raw return elevation minus the interpolated ground model. P95 is over non-ground returns more than 2 m above that model, **not** the 95th percentile of individual tree heights. All-return counts and ratios refer to the full rectangular data window. The analysis terrain is a 1 m grid triangulated from real class-2 points; the display mesh uses 0.5 m sampling. Ground surfaces fade where nearest measured ground is 5–10 m away, and are absent past 10 m. Only display micro-relief is doubled, with an on-screen note; numbers use the undistorted ground model. Three east–west transects each span 420 m, at northing -160, 0, +160 m. Each profile uses returns in a 4 m strip and median ground elevations in 2 m bins. The chart reports its own vertical display exaggeration. Amber transects are measurement guides, not archaeological identifications. No original-film farm-wall measurements are reused.