Watch the full unscripted 4-minute encounter against Voldrak the Troll:

https://www.youtube.com/watch?v=HJMoPBQ2Wx4

One Encounter, Two Divergent Paths

Facing Voldrak in the forest isn't limited to draining his 1200 HP bar:

• The Combat Route: Stamina management, directional guards, dodges, timed strikes, and active tolerance/temperament thresholds that escalate Voldrak's aggression when pushed past breaking points.
• The Negotiation Route: Using live text dialogue or letting your AI companion V.E.T.O.R. intervene directly with the boss mid-encounter, working toward a physical offering exchange at the negotiation table rather than a fight to the death.

How the Engine Handles Both Live

1. Custom HTML5 Canvas + Synchronized Video Layer: Idle, attack, guard, hurt, talk, and transition clips are pre-buffered and switched on frame/audio boundaries without breaking combat positioning or collision bounds.
2. State-Aware AI Dialogue & Guardrails: Spoken reactions are grounded in live combat state (remaining HP, guard/dodge state, temperament, item inventory, and prior utterances) so an unscripted conversation never invents fake items or bypasses encounter rules.
3. Full Voice Synthesis Across 7 Languages: Both the Traveler, V.E.T.O.R., and Voldrak synchronize lip/body animations directly to generated audio buffers before playback begins.

Project Status: Currently in active development for PC / Windows (no public playable demo yet).

Follow development updates and technical devlogs on the official site: vetorlogico.com

Here is a look at the encounter video and the architecture behind keeping spoken dialogue, subtitles, and combat state synchronized across languages.

https://www.youtube.com/watch?v=BP2v6d2mz44

1. Seven Working Languages Inside Real-Time Encounters

Right now, the game supports Brazilian Portuguese, English, Spanish, French, German, Italian, and Dutch. Rather than swapping pre-baked video clips for every single line in seven languages, the engine separates:

• Visual scene timing (1280×720 video/animation layers and hitbox windows running continuously in Canvas);
• Dynamic voice and dialogue playback synchronized per active language and character voice profile;
• Localized UI & tactical overlays so companion scans and enemy status badges remain readable mid-fight.

2. Why Real-Time Voices Matter When Dialogue Affects Gameplay

In V.E.T.O.R — A Ruptura, encounters with allies and bosses aren't locked behind paused menus. Player replies and reactions feed directly into encounter tolerance and behavior, meaning players need to understand character tone immediately while watching spacing, attacks, and guard timing.

See companion Albert and robot VETOR in action during a boss fight as well:

• Full combat dialogue gameplay: https://www.youtube.com/watch?v=CJakcjQfwrY

3. Project Status & Feedback

• V.E.T.O.R — A Ruptura is in active development for PC/Windows (no public demo yet).
• Official page and video archive: https://vetorlogico.com

For other solo/indie devs working on multi-language games: do you prioritize subtitles-only for early milestones, or invest early in multi-language voice pipelines to test pacing across regions?