SahrulRamadhani29/portfolio
Find a project.../
Projects/Jadi Gini Cuy

Jadi Gini Cuy

Private source

A 3D explainer video production pipeline, from research and storyboards to voice-over, animation, subtitles, and rendering.

SoloBlender pipeline, voice-over/Whisper processing, subtitles, caching, and quality gates
Private source code

My contribution highlights

Read all 8 contribution modules
Contents · 8 contribution modules

Main workflow

6 steps

The pipeline turns ideas into scenes, narration, and final video through inspectable production stages. Revisions target the affected stage while valid cached artefacts can be reused.

  1. 01Plan narration and scenes

    1. Research, script, and storyboard

      Creator & Python tools

    2. Build objects and scenes

      Blender / bpy builder

  2. 02Align audio and movement

    1. Process takes and timestamps

      VO recordings & faster-whisper

    2. Connect narration to the timeline

      Animation & subtitle tools

  3. 03Render, compose, and check

    1. Build video outputs

      Blender & FFmpeg

    2. Review results and revisions

      Quality checks & creator
Process details & branches

  1. Research, script, and storyboardCreator & Python tools

    Choose the topic, mechanism to explain, narration sequence, and visual decisions. Validate input before continuing production.

  2. Build objects and scenesBlender / bpy builder

    Prepare geometry, rigs, materials, and relationships between parts. Scene inspection helps check construction and readability.

  3. Process takes and timestampsVO recordings & faster-whisper

    Word-timestamp transcription helps select narration excerpts. Audio is normalised and mixed with SFX as needed.

  4. Connect narration to the timelineAnimation & subtitle tools

    Align animation beats with voice-over, group subtitle phrases, and prepare ASS/SRT outputs without breaking word timings.

  5. Build video outputsBlender & FFmpeg

    Render scenes, then combine frames, audio, and subtitles. Fingerprints help reuse artefacts that still match their inputs.

  6. Review results and revisionsQuality checks & creator

    Check relationships between scenes, audio, captions, frames, and the actual video against the quality contract.

    Ready to finalise?

    • Checks passed. Record artefacts and evidence for the revision that was actually reviewed.
    • Revision needed. Return to the affected stage; checkpoints and caches help limit repeated work.

Visual and audio reviews guide revisions. Fingerprints, caches, and checkpoints help reuse artefacts that still match their inputs.

CASE_STUDY.mdProduct & my contributions

About Jadi Gini Cuy

Jadi Gini Cuy is a 3D explainer video production pipeline connecting ideas, scripts, Blender scenes, voice-over, subtitles, and video outputs. Each job has preparation, preview, revision, and review stages so the relationship between narration and visuals can be followed throughout production.

I built the Python tooling, scene builders, audio processing integration, video composition, caching, and quality checks used in this pipeline.

My contributions

CLI and production job structure

I designed the CLI, video project structure, input templates, and script/scene preflight checks. I separated research/concepts, visual decisions, voice-over, previews, revisions, and finalisation so each stage’s inputs and artefacts can move clearly into the next process.

Scene builders, objects, rigs, and materials

I developed Blender/bpy builders for scenes, rig/material components, and example object mechanisms. I handled scale and relationships between parts, connecting them to animation requirements that follow the narration. Inspection tools help review construction, contact, and transitions between scene states.

Voice-over transcription with faster-whisper

I integrated faster-whisper for Indonesian transcription with word-level timestamps. Transcripts help assess takes, connect script units to recordings, select excerpts, and align animation beats. This turns timing information from audio into input for visual production and subtitles.

Audio processing, SFX, and edit versions

I worked on voice-over normalisation and mixing with SFX, original take management, and edit version records. I added timestamp checks and relationships between audio excerpts and script units so recording changes can be traced to the composition timeline.

ASS/SRT subtitles and phrase grouping

I built subtitle phrase grouping while preserving word timings and relevant pauses. I prepared ASS/SRT outputs for composition. This connects transcription results to the text displayed in the video without losing timing relationships.

Rendering, composition, and media inspection

I connected scenes and rendered frames to audio and subtitles through FFmpeg. ffprobe is used for media metadata inspection. Checks connect job specifications, scenes, audio, frames, and output video so artefacts used in the final result can be traced.

Fingerprint caching and checkpoints

I added input and artefact hashes/fingerprints so matching renders or compositions can be reused. Subtitle changes can be processed without always rebuilding scenes. Checkpoints and identical-failure tracking help resume work and limit repetition of the same stages.

Quality contracts and pipeline tests

I designed quality contracts, visual regression checks, and links between review evidence and the revision being worked on. I wrote tests for captions, audio, voice edits, scenes, preflight, artefact validation, and QA status. Pending/revise statuses identify stages requiring action before finalisation.

Fields & core technologies

Project fields

Automation & 3DAI EngineeringSoftware Quality

Core technologies

PythonBlenderFFmpegfaster-whisperNumPySoundFile

46 commits in the last 12 months

All branches
NovDecJanFebMarAprMayJunJulAugSeptOctNov
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Hover over a square or focus it to see the date.
LessMore
Current streak0 days
Longest streak in this period4 days
Active days in this period6 days
Verified data · updates scheduled hourly.
How it is counted · 1 repo · 1 branch

My authored commits reachable from all remote branches at sync time. Each SHA is counted once, even across branches or forks. Merge commits are excluded. Dates use the author timestamp in Western Indonesia Time (UTC+7), rather than push time. The current streak remains active when the last commit was yesterday and today is still in progress.

This calendar covers commits across all branches. Its totals may differ from GitHub’s contribution calendar, which includes other activity types and follows its own branch rules. This calendar covers only the project you are viewing.

Contribution history

46 of my non-merge commits · All branches · Data as of 3 November 2026.

Back to all projects