Every question about MForja, answered.

Entropy conditioning is a signal processing technique that reduces the informational complexity of a video stream before it reaches a video encoder. By conditioning the statistical structure of the input, it makes the signal more compressible — so the encoder produces a smaller output file while preserving all visual content. The result is 50%+ file size reduction with no pixel-level changes and no downstream infrastructure modifications required. Full technical explanation →

Standard video compression (H.264, H.265, AV1, etc.) operates on raw video to produce a compressed file. Entropy conditioning operates on the video stream before it reaches the encoder — conditioning its statistical properties so the encoder's existing algorithms can compress it more effectively. They are complementary: entropy conditioning improves the result of whatever encoder is already in the pipeline, rather than replacing it.

MForja is codec-agnostic. It works with H.264, H.265/HEVC, AV1, VP9, and the top 20 video codecs in production use. Because entropy conditioning conditions the stream before encoding, the encoder itself does not need to change. See the codec compatibility details →

No. Entropy conditioning does not alter pixel content. Video decoded from a MForja-conditioned file is pixel-for-pixel identical to the same video encoded without the technology. There is no perceptible quality difference because there is no actual difference — the visual data is unchanged.

MForja delivers 50%+ file size reduction across tested codecs and content types. The exact reduction varies by content complexity, existing encoding settings, and codec. High-motion, high-complexity content may yield somewhat lower reductions; static or low-motion content may yield higher reductions. The 50% figure represents a conservative baseline across production workloads.

Codec upgrades (e.g., H.264 → H.265, or H.265 → AV1) deliver 30–50% compression improvements but require encoder upgrades, decoder compatibility updates, and transition timelines measured in years. MForja delivers 50%+ reduction with no transition timeline — it works with the codec already in production. The two approaches are also complementary: MForja applied to AV1 delivers larger gains than either alone. H.265 vs AV1 comparison →

No. MForja's output is a standard codec file — H.264, H.265, AV1, or whichever codec the encoder uses. Any player, device, or system that could decode the video before MForja was deployed can decode it afterward. No decoder updates, firmware changes, or player modifications are required.

MForja is a pure software deployment that inserts upstream of the encoder. For VOD workflows, it processes video before encoding and storage. For live streaming, it operates in the encoding pipeline. For surveillance, it sits between camera output and the NVR encoder. Integration is additive — existing infrastructure continues to function exactly as before. Full integration guide →

None. MForja is a pure software solution. It requires no dedicated hardware, no FPGA cards, no custom silicon, and no hardware replacements. It runs on standard compute infrastructure — on-premises servers, cloud instances, or hybrid deployments.

Yes. MForja can be deployed in cloud environments including AWS, Azure, and GCP. It integrates with cloud encoding services and pipelines without requiring changes to the cloud environment itself.

Yes. MForja is designed to operate in real-time encoding pipelines, including live streaming workflows. It introduces minimal latency — well within the tolerances of standard live streaming delivery chains.

Yes. MForja is resolution- and format-agnostic. It works with SD, HD, 4K, and 8K content, and is compatible with HDR formats including HDR10, HLG, and Dolby Vision. Higher-resolution content benefits proportionally — more data means more compressibility gains.

CDN egress is billed per gigabyte of data delivered. A 50% reduction in video file size produces a 50% reduction in egress cost — directly. For platforms delivering 500TB per month at $0.01–$0.085/GB, this represents $2,500–$42,500 in monthly savings per 500TB of delivery volume. Full CDN cost analysis →

A 100-camera 4K surveillance site with 90-day retention generates approximately 800TB annually. With MForja, that becomes approximately 400TB — a 400TB annual saving. At cloud storage rates of $0.02–$0.08/GB, that equals $8,000–$32,000 per site per year, compounding as camera counts and retention periods grow. Surveillance storage deep dive →

Yes. AI and ML pipelines that store large video datasets benefit directly from 50%+ file size reduction. At 1PB of training data, standard cloud storage costs $20,000–$50,000/month. Entropy conditioning cuts that to $10,000–$25,000/month — plus proportional reductions in egress costs on every training run. AI/ML storage cost analysis →

MForja uses a usage-based pricing model tied to data volume processed. This aligns cost directly with the storage and delivery savings generated — the more data processed, the larger the savings, and the pricing scales proportionally. Contact MForja for specific pricing based on your volume and use case.

No. Entropy conditioning does not alter pixel content — the decoded video is identical to footage recorded without the technology. Admissibility depends on chain-of-custody procedures, not file size. MForja does not introduce any new admissibility variables. It is categorically distinct from lossy recompression, which does alter pixel data and may raise legitimate admissibility questions. Evidence integrity explanation →

No. AI and ML models train on decoded pixel data — not on compressed bitstreams. Since entropy conditioning preserves pixel-level content exactly, the training signal is unchanged. Models trained on MForja-conditioned video produce identical results to models trained on unconditioned video. Smaller files additionally reduce I/O bottlenecks during training. AI/ML use case →

Yes. Entropy conditioning can be applied as a batch re-encoding process on existing video archives. Files are re-encoded with conditioning applied, producing smaller files with identical decoded content. This is particularly valuable for large archives where storage costs are ongoing and compounding.

Deployment timelines depend on pipeline complexity. Because MForja is a software-only solution that inserts at a defined point in the encoding pipeline without changing anything downstream, integrations are typically faster than codec migrations or infrastructure upgrades. Initial deployments in straightforward VOD pipelines can be completed in days.

The starting point is a conversation about your pipeline: what codec you use, where you encode, your current file sizes, and your storage or delivery volume. From that, MForja can model the specific savings and outline an integration path. Contact the team or read the white paper for full technical details.