# Asset credits

## Engine

Three.js 0.170.0 and RGBELoader: Three.js contributors, MIT. See THREE-LICENSE.txt. https://threejs.org/

## Photographic surface materials and sky

Poly Haven, CC0. Source images converted to 1K JPEG and used as color, normal and roughness maps. Color multipliers, scale and normal intensity are adjusted in the scene.

- https://polyhaven.com/a/white_sandstone_blocks_02
- https://polyhaven.com/a/sandstone_blocks_08
- https://polyhaven.com/a/stone_pathway
- https://polyhaven.com/a/weathered_planks
- https://polyhaven.com/a/hessian_380
- https://polyhaven.com/a/marble_01
- https://polyhaven.com/a/kloofendal_43d_clear_puresky
- License: https://polyhaven.com/license

These are reusable surface scans, not scans of the reconstructed historical location.

Static indirect-light occlusion is precomputed by casting rays against the scene geometry. It is an approximation used alongside the sun and HDR environment; this is not real-time path tracing. Character foot-contact shadows are soft rendering overlays.

## People and fitted clothing

MakeHuman base mesh, morph targets and system skin textures: CC0. The full base was morphed, garments fitted through MHClO barycentric bindings, visible body surfaces extracted, and 17-joint skinning weights added. Adult male and female and child anatomy use separate macro targets. Skin textures are resized and encoded as JPEG; colors and pose are adapted. These are anatomical models, not scans of historical people.

- https://github.com/makehumancommunity/makehuman
- https://static.makehumancommunity.org/assets/assetpacks/makehuman_system_assets.html
- System textures: young_lightskinned_female_diffuse.png and young_lightskinned_male_diffuse.png.
- See MakeHuman-CC0.txt.

Mycenaean Tunic by WDG: current official catalog CC0. Fitted to male and child anatomy, skinned, recolored, textures converted to JPEG. Original source: http://www.makehumancommunity.org/node/1763 ; catalog https://static.makehumancommunity.org/assets/assetpacks/dress01.html . The old MHClO exporter header says AGPL3; the current official catalog explicitly lists CC0. Both source records are retained for review.

Long robe: adapted upper robe by MaciekG/MGgame, CC BY 2.5 Sweden per the current official catalog; original source http://www.makehumancommunity.org/node/1482 ; catalog https://static.makehumancommunity.org/assets/assetpacks/suits03.html . Combined with folded lower-skirt geometry from Donitz’s Monk’s Robe (CC0), http://www.makehumancommunity.org/node/3238 ; catalog https://static.makehumancommunity.org/assets/assetpacks/suits02.html . WDG textile, CC0 as credited above. Cape, hood and rope belt removed, sleeves narrowed, collar lowered, weights adjusted, material recolored and woven belt added. This is a visual approximation, not an authenticated priestly costume.

Hair: MakeHuman short01 and short02 system assets, CC0. Full beard and moustache: grinsegold, CC BY 2.5 Sweden; original sources http://www.makehumancommunity.org/node/188 and http://www.makehumancommunity.org/node/187 . Scruffy Beard 1: Elvaerwyn, CC BY 2.5 Sweden; original source http://www.makehumancommunity.org/node/2611 . Catalog https://static.makehumancommunity.org/assets/assetpacks/bodyparts06.html . Fitted to the male and child anatomy, assigned to the head joint, recolored; alpha textures resized to 1K and converted to WebP, beard normal map to JPEG. License: https://creativecommons.org/licenses/by/2.5/se/ .

The older grinsegold and upper-robe exporter headers say AGPL3; the current official catalogs explicitly list CC-BY. The official FAQ https://static.makehumancommunity.org/assets/assetpacks/faq.html links CC-BY to Attribution 2.5 Sweden. Source metadata and this discrepancy are preserved with the project.

Elvs Goddess Dress 6 by Elvaerwyn, via MakeHuman Community, CC BY 2.5 Sweden; fitted, skinned and adapted for this experience. Source http://www.makehumancommunity.org/node/2563 ; catalog https://static.makehumancommunity.org/assets/assetpacks/dress02.html ; license https://creativecommons.org/licenses/by/2.5/se/ . The official FAQ https://static.makehumancommunity.org/assets/assetpacks/faq.html links its CC-BY label to that version. Textures converted to JPEG (dress normal map reduced to 1K), opacity made solid, colors adjusted and head covering added. Garment geometry is not a documented period reconstruction.

## Audio

“Niggun Habaal Shem Tov”: a melody traditionally attributed to the Baal Shem Tov, performed on alto recorder by Nizzan Zvi Cohen, 15 October 2010, CC BY 4.0. Plays near the street musicians. This is a traditional Hasidic nigun, not a Sukkot-specific song or Temple-era instrument recording. Converted to mono MP3, filtered, loudness-normalized, faded and spatialized.

- Source: https://commons.wikimedia.org/wiki/File:Niggun_Habaal_Shem_Tov.ogg
- License: https://creativecommons.org/licenses/by/4.0/

“Hoshanas Group”: Rabbi Hoffman leading Hoshana Rabbah at Kohelet in 2009, published by Rafi Simcha in the collection “Yomim Norayim: Days of Awe” (2010), CC BY 3.0.

- Item and recording license: https://archive.org/details/YomimNorayimDaysOfAwe
- Original file: https://archive.org/download/YomimNorayimDaysOfAwe/Hoshanas_Group.mp3
- License: https://creativecommons.org/licenses/by/3.0/
- Modification: excerpt 05:50–07:42, filtered, loudness-normalized, faded and spatialized. Used at the pilgrims’ table and Temple. This is modern Sukkot liturgical singing, not a Temple-era recording or a newly licensed commercial song.

FluidR3_GM instrument samples: Fluid Soundfont, CC BY 3.0, obtained from MIDI.js Soundfonts. Instruments: harp, flute, nylon-string guitar, woodblock and taiko drum. Sequenced, mixed and spatialized in original arrangements. These remain in the family and courtyard sukkot.

- https://github.com/gleitz/midi-js-soundfonts
- https://creativecommons.org/licenses/by/3.0/

Three Yiddish speech clips and a loop of indistinct crowd/footstep ambience were generated for this project using the connected audio service. All intelligible dialogue is Yiddish; the historical recording has Hebrew singing. The old Hebrew greeting is not used in this build. Generated voices are not named historical or living people. Native-speaker pronunciation review has not been completed.

Greeting: אַ גוטן יום־טוב! שלום־עליכם! קומט אַרײַן אין אונדזער סוכה. זעצט זיך און פֿרייט זיך מיט אונדז!

Courtyard: אַ גוטן יום־טוב! קומט אַרײַן אין דער סוכה. הערט זיך צו צום ניגון, און זינגט מיט!

Temple: מיר זענען בײַם בית המקדש אין ירושלים. ס׳איז סוכות, און פֿון אַלע זײַטן קומען די עולי־רגל. קוקט אויף די גרויסע טויערן, די שטיגן און דעם הויף. די כּהנים גרייטן זיך צו דער עבודה, און די לויים שפּילן און זינגען. אַ פֿריילעכן יום־טוב!

## Temple references

Museum reconstruction photographs informed the geometry but are not embedded in the scene. Source authors and exact reference URLs are documented in HISTORICAL-NOTES.md. Their photographs show a modern model, not the original Temple. Precise ornaments and movement are interpretive.

## Donkeys and traveling baggage

Authored donkey and pack meshes/textures from 0 A.D., Wildfire Games, copyright 2009 Wildfire Games, CC BY-SA 3.0 Unported: https://creativecommons.org/licenses/by-sa/3.0/ . Source https://github.com/0ad/0ad/tree/master/binaries/data/mods/public/art . Original files: meshes/skeletal/donkey.dae, textures/skins/skeletal/animal_donkey.dds, meshes/props/donkey_pack_01.dae, textures/skins/props/celt_prop_1.png. Converted to indexed JSON, normalized in metres with +Z forward, pack aligned, DDS converted to PNG and pack texture to JPEG. A simplified walking deformation uses the supplied weights and pivots. The modified animal and pack assets remain CC BY-SA 3.0. These terms apply to those assets; other assets retain their own licenses. See DONKEY-CREDITS.txt and TRAVEL-ART-LICENSE.txt.

## New Yiddish introductions

Three additional generated speech clips introduce the nighttime celebration, mikvah and arriving pilgrims. Voices are fictional performances; native-speaker pronunciation review remains pending.

night_yi: ס׳איז שמחת בית השואבה אין בית המקדש. קוקט אויף די גרויסע לייכטער, די פֿאַקלען און די טענצער. די לויים שפּילן בײַ די פֿופֿצן שטיגן. אַ פֿריילעכן יום־טוב!

mikvah_yi: דאָס איז אַ מקוה פֿון דער צײַט פֿונעם צווייטן בית המקדש. די עולי־רגל האָבן זיך געגרייט אין טהרה. איר קענט אַראָפּגיין די שטיגן און באַטראַכטן די מקוה.

arrivals_yi: שלום־עליכם! מיר קומען קיין ירושלים אויף סוכות. די אייזלען טראָגן אונדזערע פּעקלעך. נאָך דער לאַנגער נסיעה זענען מיר זייער פֿריילעך אָנצוקומען.


## Expanded household and four-minim assets
- `food.js`: newly authored shared 3D meshes for bread, figs, dates, grapes, olives, cups and shallow ceramic dishes. No external food images or scans.
- `four-minim.js`: newly authored shared meshes for the closed lulav, myrtle, willow and citron; a small generated rind bump texture. Botanical and ritual proportions are interpretive.
- `neighborhood.js`: newly authored residential quarter, house interior and public approach additions; reuses the credited material textures and human meshes above.
- `house_yi.mp3`: generated male Yiddish welcome, 6.4 seconds. Script: “שלום־עליכם! קומט אַרײַן אין אונדזער הויז. אַ פֿריילעכן יום־טוב!” Pronunciation has not been independently reviewed.


### Expanded Temple and crowd

Temple interiors, vessel approximations, circulation, furniture, interaction markers and environment additions were authored for this reconstruction. The new crowd retains the existing MakeHuman, garment, hair and beard credits above. Background garments are reduced mesh adaptations with transferred UV coordinates after baking the carrying pose, prepared using fast-simplification (MIT; build dependency: https://github.com/pyvista/fast-simplification). Texture and garment licenses remain unchanged. Core human assets and detailed foreground meshes remain in the source project. The mesh-reduction helper, code and static assets are retained for reproducibility.


### Water, surface finishes and distant crowd update

`rendering-finishes.js` adds original material shader work, water response, wet margins, small household details and sukkah lashings. The existing photographic textures and licenses remain as credited above. The second crowd mesh level uses the same credited character assets, seam welding, quadric reduction and surface attribute transfer; `tools/build_crowd_distance.py` is retained for reproduction. No additional music or voice recordings were introduced in this update.


### v12 living activities and Yiddish guide

`living-scene.js` contains new authored interaction, staging, prop and animation work using the previously credited materials, people, food and travel assets. No new external music recording was added. The existing Hoshanas, nigun and instrumental arrangement credits remain unchanged.

Twenty-two additional fictional Yiddish speech performances were generated for this scene using the Elias preset via Runway, model eleven_v3, speed 0.95. The first generation with an explicit language code failed; the successful clips used text-language detection. Exact scripts and generation task identifiers are retained in `audio/life-transcripts.json` in the source. Pronunciation remains unreviewed by a native speaker. The voices do not impersonate a named person.

The geometry combiner now transfers vertex color data, preserving authored bread and fruit colors. Cloth movement, local Foley, hand placement, skin response and foot-contact shadows are authored visual/audio approximations.

## v17 music selections

Three additional excerpts of **Hoshanas Group**, Rabbi Hoffman and congregation at Kohelet, recorded/published by Rafi Simcha, 2009/2010, CC BY 3.0:
- `hoshanot_response.mp3`: 07:50–10:50.
- `hoshanot_procession.mp3`: 12:00–15:00.
- `hoshanot_festival.mp3`: 18:00–21:00.

Source: https://archive.org/details/YomimNorayimDaysOfAwe ; original: https://archive.org/download/YomimNorayimDaysOfAwe/Hoshanas_Group.mp3 ; license: https://creativecommons.org/licenses/by/3.0/ . Each was excerpted, high/low-pass filtered, loudness-normalized, faded, converted to mono MP3 and spatialized. These are selections from one performance, not separate songs. Collection metadata is retained under `audio/sources-v17/metadata.json`. The collection's spoken study recordings were researched but are not included.

The Royal Stoa, underground rooms and added architectural details are new procedural geometry. Skin/cloth relief and blink morphs use the existing credited MakeHuman meshes and texture assets. No new photogrammetry is claimed.