v4.0 — CPU-Only, Fully Offline, No GPU Required

Your Footage,
Optimized Per Shot by AI

BAVE analyzes each scene individually using machine learning to predict the optimal x264 encoding settings. Deliver theatrical-quality output, streaming-ready files, and festival-compliant masters — all from one tool, all running entirely on your CPU. No subscriptions, no phone-home.

500+
Active Licenses
30+
Countries
1 Year
Updates Included
500+
Active Licenses Issued
896
Training Clips Analyzed
178K
Parameter Candidates Evaluated
30+
Countries Reached

Built for Professional Encoding

Every feature serves one goal: optimal quality per shot with predictable, repeatable results.

AI

AI Per-Shot Encoding

The forward-surrogate ML model evaluates ~500 parameter candidates per shot and selects optimal x264 settings (CRF, preset, tune, AQ, psy-rd, bframes, subme, deblock, me_range, rc_lookahead, AQ-mode, b-adapt, trellis, direct, hqdn3d, CAS) for that specific scene.

SC

Dual-Detector Scene Detection

Fused AdaptiveDetector + ContentDetector cascade with adaptive thresholds. Catches hard cuts and gradual transitions. Precision guard drops boundaries where frame-MAD ≤ 2× baseline. No magic numbers — thresholds derived from content's own motion energy.

DP

Three Delivery Presets

Theatrical: CRF unconstrained, VMAF 96 target, veryslow, grain rewarded.
Streaming: CRF+VBV, ~2500kbps target, VMAF 93, slow preset.
Festival: Two-pass bitrate, exact 1GB file size, slow preset.

VT

ViT-B/32 Semantic Analysis

INT8-quantized Vision Transformer computes whole-video visual diversity (cosine similarity of frame embeddings). Sampled 30–300 frames adaptively. Cached once per video, reused across all shots. Runs fully on CPU.

AU

Smart Audio Routing

Auto-detects stereo/surround (5.1/7.1), downmix/upmix with libfdk_aac VBR5, 48kHz SOXR resampling. Single continuous audio track — no per-shot cuts, no clicks. Loudness measured via EBU R128.

FX

Format Conversion Pipeline

HDR→SDR tonemap, deinterlace, scale to FHD 8-bit BT.709 in one FFmpeg filter chain. Lossless intermediate (qp=0, ultrafast). Blackbar detection via cropdetect at multiple sample points. All reuse cached media_info — zero redundant ffprobe.

CK

Crash-Resilient Checkpointing

Shot-level state machine: pending → processing → encoded → concatenated → finalized. SQLite checkpoint DB survives power loss/OOM. Resume from any shot. Automatic 72-hour TTL cleanup.

NM

Professional Output Naming

Project folder: FilmName Year Director
Video: FilmName Year Director FHD STEREO THEATRICAL BAVE.mp4
Stills: FilmName Year Director/Stills/FilmName Year Director Stills 00h00m18s.png
Zero underscores. Metadata & copyright embedded in output and stills.

UI

Custom Frameless UI

Premium dark theme with glassmorphism. Horizontal filmstrip with shot thumbnails, palette strips, per-shot parameter preview. Real-time resource gauges (CPU/RAM/disk I/O). Drag-drop import. Batch queue with multiple preset selection.

How the AI Pipeline Works

BAVE's encoding pipeline reuses every heavy computation exactly once. No redundant ffprobe, no re-analysis, no silent fallbacks.

Phase 1: Preview (Runs Once Per File)

  1. ffprobe → media_info (duration, fps, resolution, codec, color space, HDR)
  2. inspect_media → deep container analysis
  3. detect_audio_config → audio topology
  4. measure_audio_loudness → LUFS/peak
  5. detect_blackbars_full_video → crop info
  6. analyze_format → conversion decision (HDR→SDR, deinterlace, scale)
  7. convert_video_for_model → FHD 8-bit intermediate (shared with encode)
  8. detect_scenes_fast → shot_starts (dual detector fusion)
  9. compute_semantic_similarity → ViT-B/32 INT8 (cached LRU=20)
  10. analyze_shots_parallel → features + thumbnails + palettes (ThreadPoolExecutor)

Phase 2: Encode (Per-Shot AI Selection)

  1. VideoProfile from pre-provided media_info — zero new ffprobe
  2. Reuse: converted path, shot_starts, features, semantic similarity (cached)
  3. frame_align_cut_points → exact frame boundaries (0-frame drift)
  4. Per-shot: forward surrogate predicts (bitrate, VMAF, time) for each candidate
  5. Scoring: quality_reward − bitrate_penalty − time_penalty − ceiling_penalty ± tune_reconciliation
  6. Best candidate → FFmpeg x264 command (crf, preset, tune, aq, psy-rd, bframes, subme, deblock, me_range, rc_lookahead, aq-mode, b-adapt, trellis, direct, hqdn3d, cas)
  7. Parallel encode pool (mp.Pool, workers = encoding_workers())
  8. Retry: up to 2 inner retries, preset escalation on timeout
  9. Concat via concat demuxer with exact frame counts (-frames:v)
  10. mux_final_output → single continuous audio track + verify_output
For Engineers: Model Performance Metrics →

Forward Surrogate Models (v5.0)

Three CatBoost GBDT models exported to ONNX for CPU inference. Trained on 178,208 rows from 896 unique clips.

📊

Bitrate Model

R² = 0.633 | MAE = 3591 kbps
Kendall τ = 0.635 | Spearman ρ = 0.823

VMAF Model

R² = 0.530 | MAE = 1.79
Kendall τ = 0.592 | Spearman ρ = 0.775

⏱️

Time Model

R² = 0.757 | MAE = 7.00s
Kendall τ = 0.709 | Spearman ρ = 0.888

Available on Windows, macOS & Linux

Native installers for all major desktop platforms. CPU-only — no GPU, CUDA, Metal, or drivers required.

WIN

Windows 10/11

NSIS installer (~2 GB)
Intel Quick Sync, AMD VCE, NVIDIA NVENC passthrough

Download Windows
MAC

macOS 12+

DMG bundle (~2 GB)
Apple Silicon native, Metal Performance Shaders ready

Download macOS
LNX

Linux (Beta)

AppImage (~2 GB)
Ubuntu 20.04+, VAAPI, NVENC, AMF

Download Linux

Product Tour

A 90-second walkthrough of BAVE's per-shot AI encoding pipeline and delivery presets.

🎬

1. Import Your Footage

Drag and drop any video file. BAVE analyzes resolution, codec, audio channels, and HDR metadata automatically.

🧠

2. AI Per-Shot Analysis

Each scene is evaluated individually. The forward-surrogate model tests ~500 candidate parameter sets per shot and selects the optimal one.

🎯

3. Choose Delivery Preset

Theatrical (CRF unconstrained, VMAF 96), Streaming (CRF+VBV, ~2500kbps), or Festival (two-pass, exact file size).

4. Export & Deliver

Single continuous audio, zero-frame-drift cuts, metadata embedded. Ready for distribution, DCP, or festival submission.

Proven Quality at Real Bitrates

Independent benchmark on a 30.7s ProRes HQ showreel (612 MB, 160 Mbps). VMAF v0.6.1 and PSNR measured on full-resolution frames. BAVE uses CRF18 veryslow + custom AI presets as its theatrical baseline.

MethodVMAFPSNROutput SizeBitrate
Source (ProRes HQ)612 MB160 Mbps
BAVE Cinematic (CRF18 veryslow)96.0145.1047.4 MB12 Mbps
BAVE Streaming (CRF20 slow)95.1944.3030.7 MB8 Mbps
x264 CRF 18 veryslow94.243.80~49 MB~12 Mbps
2-pass 2500kbps CBR91.9142.519.5 MB2 Mbps

Test corpus: 1 clip (20–30s, 1080p, ProRes HQ → H.264). VMAF computed with Netflix VMAF v0.6.1 model on full-resolution frames. PSNR computed on YUV444.

License Tiers

Per-seat perpetual license with 1 year updates. Offline activation — no subscription, no phone home.

Personal
$195/seat

1 workstation • Under 45 min videos

  • 1 encoding workstation
  • Theatrical + Festival presets
  • AI per-shot encoding
  • Ed25519 offline license
  • 1 year updates
  • Email support
  • Device transfer
  • Grant eligibility
  • All 3 presets (no Streaming)
  • DCP module
Buy Personal
Enterprise
$4,999/site

Unlimited workstations, one site license

  • Unlimited workstations
  • All 3 delivery presets
  • AI per-shot encoding
  • Ed25519 offline license
  • 1 year updates
  • 24/7 dedicated support
  • Device transfer + bulk mgmt
  • Air-gapped certificates
  • API access for automation
  • Custom SLAs
Contact Sales

Try Before You Buy

Free 7-day trial — full features, no credit card. Offline activation via machine config exchange.

Download Trial

Independent Filmmaker Grant

We provide 100% free Personal licenses for student directors, indie filmmakers, and creators facing regional or financial hardship. Reviewed directly by director Benyamin Khodabandeh.

Apply for Grant