Stop Paying Hot Storage Prices for Cold Media — A Real Archive Strategy for Post in 2026

By PostForward

The single most common email I'm getting right now starts with some version of: "Our SAN is full again." Then a pause, then: "We don't really have an archive strategy."

Welcome to the club. Almost every studio I walk into has the same problem. There's a turn-and-burn rhythm to modern production — shoot, edit, deliver, move on — and finished projects just keep piling up on the most expensive storage in the building. Nobody touches them. Nobody wants to delete them. The bill keeps growing.

A real archive strategy isn't "back up the finished cut to a hard drive in the closet." It's a tiered system that moves media from hot to warm to cold as it ages, with an asset manager that keeps it findable, and a folder discipline that doesn't rely on someone's memory. Here's what that actually looks like in 2026.

Why this got harder than it used to be

Three things have changed in the last few years:

Volumes exploded. 4K is table stakes, 8K and high-frame-rate originals are landing on more shoots, and AI/VFX workflows now generate proxy and rendered intermediates that didn't exist as a category five years ago.

Hot storage hasn't gotten meaningfully cheaper. Per-TB pricing on SAN, fast NAS, and LucidLink-style cloud working storage has barely budged while your library has 3x'd. Studios that don't tier are paying premium prices to store finished projects nobody is touching.

Compliance and client retention grew teeth. "Keep this project for seven years" used to mean a stack of LTO tapes in a closet. Now it often means provable redundancy across geographies, auditable access logs, and recoverability SLAs.

The hot / warm / cold framework

Think of every project on a timeline from active to dormant. The right storage tier follows that timeline.

Hot tier — active projects (0–60 days)

This is editorial, color, VFX, finishing. Sub-millisecond latency, sustained 4K/8K throughput, real-time collaboration. SAN, fast NAS, LucidLink, EditShare, Studio Network Solutions, Hammerspace. Highest cost per TB by a wide margin, and that's fine because the work needs it.

Warm tier — recent or probably-coming-back (2 months – 2 years)

Wrapped projects waiting on client QC, revisions, sequels, anything you can imagine a client emailing about in the next year or two. Minutes-to-first-byte is fine; you don't want to wait hours. Good options here: on-prem object storage (Cloudian, MinIO, Scality), S3 Standard-IA, S3 Intelligent-Tiering, or PostForward's S3 cloud storage for video teams — S3-compatible, no egress surprises, and built for this exact workload.

Cost on warm is usually 30–60% of hot per TB, and the math gets really good when you realize most of your library actually belongs here, not on the SAN.

Cold tier — long-term retention (2+ years, or compliance)

Master deliverables, project archives, camera originals you may never touch again. Hours-to-day retrieval is acceptable. LTO-9 tape, S3 Glacier Deep Archive, Azure Archive.

Cost on cold is 5–15% of hot. The catch is retrieval — if you're going to be pulling stuff back a lot, cold is a trap.

The mistake I see most: studios that treat archive as binary. Either it's on the SAN or it's "gone." The whole point of three tiers is that nothing is ever fully gone — but you stop paying hot prices to store cold data.

Picking the actual tools

LTO

Still the gold standard for cold if you have the scale to justify the hardware. LTO-9 holds 18TB native (45TB compressed) per cartridge, sips power at rest, lasts 30+ years in proper environmental conditions, and once you own the library the marginal cost per TB is brutally low. LTO-10 is right around the corner with 30TB (!) of native storage and shipping in 2026!

Downsides: capital expense up front, requires physical handling and rotation discipline, slow random access, and you need someone who actually knows how to drive it. Best for: studios over ~500TB of cold media, or anyone with strict offsite/airgap requirements.

S3 Glacier Deep Archive

LTO's main cloud competitor. Around $1/TB/month at rest, no hardware to maintain, geographically redundant by default. The catch is retrieval — pricing punishes you hard if you pull a lot back, and the slowest tier takes up to 12 hours to start returning data.

Best for: archives you're almost certainly not going to touch. Compliance retention, "we have to keep it but the chance we ever ask for it is near zero."

PostForward S3 cloud storage

This is what we built for exactly this problem. S3-compatible warm storage tuned for post-production workflows — predictable per-TB pricing, no surprise egress fees when you actually need to pull a project back, and direct integration with the MAM and editorial tools you already use. For most studios this is the warm tier we drop in to replace whatever expensive hot storage is currently holding finished work. More on our S3 cloud storage for video teams.

On-prem object storage

Lots of small to medium storage platforms offer S3-compatible APIs without the cloud bill - even some great open source solutions that are easy to set up. Pairs well with a hybrid setup — warm tier on-prem, cold tier in actual cloud cold. Best for studios that already have a data center footprint and IT staff who don't flinch at "object storage."

Hybrid (this is what most working studios actually end up at)

A common shape that works: LucidLink/SuiteStudios/Shade for hot, PostForward S3 or on-prem object for warm, LTO plus S3 Glacier for cold (two cold copies, one local, one cloud, which satisfies most disaster-recovery requirements without overthinking it).

The MAM is the glue

Here's the part most teams get wrong: an archive without an asset manager isn't an archive — it's a graveyard. Storage nobody can search is just expensive disk.

A good MAM (like Iconik) does three things that make tiered archiving actually work:

1. It keeps proxies hot even when the masters go cold. When a client emails "do you still have that footage from the 2023 campaign?" you can show them the actual content from a low-res proxy in seconds — without restoring an LTO tape or paying Glacier retrieval fees.
2. It tracks metadata that lives forever. Project name, client, shoot date, talent, locations, keywords, custom fields. None of which the filesystem captures.
3. It manages the restore workflow. Good MAMs integrate directly with LTO libraries and cloud cold storage. Your editor clicks "restore" in the MAM, the right file gets pulled back from cold, and the editor gets a notification when it lands in their NLE bin.

If you only do one thing from this whole post: get a MAM running with proxies before you start aggressively archiving. Otherwise you'll spend the next decade re-ingesting tapes to find b-roll you forgot you had.

Project and media organization (before you archive)

The best archive strategy in the world fails if your projects are organized like a freelance editor's external drive. Standardize this stuff before media leaves the hot tier:

• Top-level folder schema. [YEAR] / [CLIENT] / [PROJECT] / [DATE]_[SHOOT-NAME]. Same shape every time. No exceptions, even for "quick" projects.
• Camera card naming at ingest. [DATE]_[CAM]_[CARD#]. Don't trust default camera names — they collide, and you find out three years later when you're restoring from Glacier.
• Project closeout checklist. Finals exported, master file naming standard followed, project file consolidated, working files purged, archive metadata filled in. Run it the same way every time.
• Required MAM metadata at archive time. Client, project, shoot date, deliverable type, retention years, legal holds. If a required field is missing, the archive job doesn't run.

This isn't glamorous work. But studios that skip it end up paying for archive storage and paying staff to manually hunt for files when clients come back. The labor cost of bad organization usually dwarfs the storage cost over a few years.

A pre-archive checklist

When a project moves off active storage, walk through this every single time:

• Deliverables shipped and accepted in writing
• Master files exported to deliverable spec (typically ProRes 422 HQ or DNxHR HQX, plus the actual delivered format)
• Project file saved with renders consolidated and dependencies inside the project folder
• Camera originals verified with checksums (xxhash or md5 — and yes, this matters; LTO and Glacier won't tell you a corrupted file is corrupted)
• MAM metadata complete (client, project, retention class, restore workflow noted)
• Two cold copies created before you delete the hot copy. Two. Always two.
• Proxies confirmed present and playing in the MAM. If the MAM can't preview it, you haven't archived it — you've just hidden it.

A 2026 default stack that works

For studios in the 100–500 TB range without a current archive strategy, here's what a clean tiered setup typically looks like:

• Hot: LucidLink, Suite Studios, Shade, or an on-prem fast NAS for active projects
• Warm: PostForward S3 cloud storage — S3-compatible, no retrieval fees, billing you can predict
• Cold: LTO-9 in-house if you're north of 500TB cold; S3 Glacier Deep Archive if you're below
• Asset management: iconik, with proxies stored in warm tier and masters in cold
• Discipline: standardized folder schema, project closeout checklist, MAM metadata gates

For studios above 1 PB, the math shifts toward owning more LTO capacity and treating cloud as the secondary cold copy rather than the primary.

The bottom line

Archiving used to be a once-a-year cleanup project. In a turn-and-burn world it has to be a continuous workflow — projects move tier by tier from the day they finish, the MAM holds the institutional memory, and the folder discipline means anyone on the team can find what they need years later.

If you're getting calls from clients asking about old projects and the answer is anything other than "yes, here's the proxy in your inbox in five minutes," your archive strategy is leaking. Usually the fastest win is just identifying what's actually cold on your hot tier — even one good tier-down pass typically pays for the next year of a real warm/cold setup.

Want help building yours?

We help studios do three things on archive: figure out what should actually be on which tier, design and roll out the storage and MAM workflow to make it work day-to-day, and write the operational discipline (checklists, closeout procedures, metadata gates) that keeps it from drifting back into chaos six months later.

[Book a free 30-minute archive consultation →](https://postforward.co/contact/)

You can also read more about our post-production storage consulting and custom iconik solutions for the asset-management layer.

PostForward is a post-production consultancy based in Washington, DC. We work with creative agencies, corporate marketing teams, and independent filmmakers on editorial workflows, storage strategy, MAM rollouts, and cloud migration. [Get in touch.](https://postforward.co/contact/)