The shift to 4K, longer retention mandates, and more cameras per site is pushing storage budgets past the breaking point. There is a way to cut the data — without cutting a single frame.
When surveillance programs were designed around 720p or 1080p cameras with 30-day retention, storage was manageable. Those assumptions are gone. Operators are upgrading to 4K for evidence quality. Regulations are extending retention to 90 days, one year, or longer. And camera counts keep climbing.
Each of those changes multiplies storage — independently of the others.
Storage infrastructure has to keep pace with all of it — or operators start making compromises: lower resolution, shorter retention, or accepting cost overruns as a line item. None of those are good outcomes.
Resolution, frame rate, codec, and retention period all interact to determine total storage requirement. Moving from 1080p to 4K alone roughly quadruples the data per camera. Adding a year of retention instead of 30 days multiplies the whole stack by 12.
| Configuration | Per Camera / Year | 100-Camera Site | With MForja (50% reduction) | Annual Savings (100 cams) |
|---|---|---|---|---|
| 1080p · H.264 · 30 days | 1.5TB | 150TB | 75TB | 75TB |
| 1080p · H.265 · 90 days | 2.2TB | 220TB | 110TB | 110TB |
| 4K · H.265 · 90 days | 8TB | 800TB | 400TB | 400TB |
| 4K · H.265 · 1 year | 32TB | 3,200TB | 1,600TB | 1,600TB |
| 4K · H.265 · 3 years | 96TB | 9,600TB | 4,800TB | 4,800TB |
These are conservative estimates. Motion-dense environments (retail, transit, crowd monitoring) generate significantly more data per camera. At cloud storage rates of $0.02–$0.08/GB, cutting 400TB equals $8,000–$32,000 per year — per site.
Surveillance operators are rightly cautious about anything that touches their video pipeline. Footage is evidence. Anything that alters it — even slightly — creates chain-of-custody risk and potential inadmissibility challenges.
Entropy conditioning does not modify pixel content. It does not resample, re-frame, re-time, or interpolate video data. It conditions the statistical properties of the video stream — reducing informational redundancy — before the signal reaches the encoder.
The encoder then produces a smaller file, because its input is more compressible. The output decodes identically to footage recorded without the technology. Every frame, every pixel, every timestamp is preserved.
What changes: file size. What stays the same: everything that matters for evidence.
Aggressive lossy recompression — common in some archival workflows — does alter pixel content. That is a legitimate admissibility concern. Entropy conditioning is not recompression. The distinction matters in any legal or regulatory context.
Mandatory retention windows are expanding across regulated industries. Longer retention means storage must grow every year just to stay compliant — even if camera counts stay flat.
| Sector | Typical Retention | Storage per 100 4K Cameras | With MForja | Annual Saving |
|---|---|---|---|---|
| Commercial / Retail | 30–90 days | 800TB–2.4PB | 400TB–1.2PB | 400TB–1.2PB |
| Financial Services | 90 days–1 year | 2.4PB–9.6PB | 1.2PB–4.8PB | 1.2PB–4.8PB |
| Casinos / Gaming | 30 days–6 months | 800TB–4.8PB | 400TB–2.4PB | 400TB–2.4PB |
| Law Enforcement | 1–7 years | 9.6PB–67PB | 4.8PB–33.5PB | 4.8PB–33.5PB |
| Critical Infrastructure | 1–3 years | 9.6PB–29PB | 4.8PB–14.5PB | 4.8PB–14.5PB |
| Healthcare | 90 days–3 years | 2.4PB–29PB | 1.2PB–14.5PB | 1.2PB–14.5PB |
At cloud storage rates, these savings convert directly to operating budget. At on-premises rates, they defer hardware procurement cycles — often by years.
Expand NAS/SAN capacity or add cloud tiers. Addresses the symptom, not the source. Capital expense grows in proportion to retention mandates and camera upgrades.
Move from H.264 to H.265 or AV1. Requires camera, encoder, or transcoding upgrades. Delivers 30–50% reduction, but requires infrastructure changes and still leaves the underlying data problem growing.
Applied upstream of the encoder in software. Works with existing cameras, NVR, and VMS. No hardware replacements. 50%+ reduction on any codec the encoder uses.
The codec upgrade path is appealing on paper, but the transition timeline is measured in years — especially across distributed multi-site deployments with mixed camera hardware. Entropy conditioning works with the infrastructure that exists today.
The most common concern from security system administrators is whether the technology disrupts their existing infrastructure. MForja integrates at the video pipeline layer — nothing downstream is affected.
These pages provide additional technical context on the underlying technology and deployment approach:
Share your camera count, resolution, and retention requirements. We'll model the storage reduction and cost impact for your specific environment.