Where Is That B-Roll from the Shoot Last Month

Raw footage piles up faster than anyone can organise it

There is a particular kind of frustration reserved for editors who know a shot exists. You were there when it was filmed. You remember the light, the framing, maybe even the lens. But the project wrapped, the cards got ingested, and now that perfect wide shot of the warehouse at golden hour is buried somewhere in 400 GB of footage across three folders named things like "Day2_B-Roll_FINAL" and "Card_07_misc."

This is one of the most common time sinks in post-production. Finding footage should take seconds. In practice, it takes minutes or longer, and those minutes compound across every project, every revision, every time a client changes direction and you need a shot you hadn't planned on using.

The root of the problem is rarely carelessness. It is timing. Between wrapping a shoot and starting the edit, there is almost never enough breathing room to sit down and organise every clip with thoughtful metadata, descriptive filenames, and a tidy folder hierarchy. The pressure to get into the timeline is immediate. So raw footage goes into a project folder with minimal sorting, and the editor relies on memory and scrubbing to find things later.

That approach works well enough for small projects. It falls apart at scale.

The real cost of disorganised footage

When footage lives across multiple drives, multiple projects, and multiple contributors, the lack of organisation creates problems that go well beyond a few wasted minutes. Editors start duplicating clips because they can't confirm whether a file already exists somewhere in the library. Storage costs creep up as the same media lives in three or four locations. Collaborative projects suffer because one editor's mental map of the footage doesn't transfer to another person stepping in.

There is also a creative cost. When finding a specific shot requires significant effort, editors tend to settle for whatever is close at hand. The b-roll selection narrows to what's already in the timeline or what happens to be visible in the most recent folder. That golden-hour warehouse shot, the one that would have been perfect, stays buried because nobody has time to go looking.

The traditional solution is to build better habits: stricter naming conventions, dedicated time for logging and tagging after every shoot, maybe a spreadsheet or database that catalogues key shots. These solutions work in theory. In reality, they demand discipline that production schedules rarely allow. Even well-run post houses find that their organisational systems decay under deadline pressure.

What a media asset management platform changes

A media asset management platform, often called a MAM, treats this as a systems problem rather than a discipline problem. Instead of relying on editors to manually tag and sort every clip, a MAM uses automation to handle the bulk of the organisational work. Media gets ingested, analysed, categorised, and made searchable without someone sitting down to do it by hand.

Fabric is a MAM built for creative teams, and its approach to footage organisation centres on two capabilities: semantic search and smart organisation.

Semantic search, available at https://fabric.so/features/search, lets you search by meaning rather than by filename or manually applied tags. Instead of trying to remember whether you tagged that warehouse shot as "exterior" or "building" or "warehouse_wide," you describe what you're looking for in plain language. A search for "wide angle of the warehouse exterior, golden hour" will surface clips that match that description, even if the filename is something unhelpful like "A024_C003_0214.braw."

This works because Fabric analyses the visual content of your media, not just its metadata. The search understands what's in the frame, so it can match your description to clips that were never manually tagged at all. For editors who have ever spent half an hour scrubbing through a bin of unsorted footage, this is a meaningful shift in how finding things works.

Smart organisation, found at https://fabric.so/features/smart-organization, handles the categorisation side. As media comes into Fabric, it gets auto-tagged and sorted based on its content. You still have full control over your folder structure and can add your own tags and labels, but the baseline level of organisation happens without manual effort. Footage from a shoot arrives already analysed and searchable, which means the gap between ingest and "I can find what I need" shrinks to almost nothing.

Searching across your entire library

One of the things that makes footage management particularly difficult is that useful b-roll doesn't expire. That warehouse shot from last month might be relevant again six months from now, for a completely different project. Aerial footage, establishing shots, texture clips, interview cutaways: these have a long shelf life, but only if you can find them.

Most editors' footage libraries are scattered across external drives, cloud storage accounts, and local project folders. Searching across all of that requires either a very good memory or a lot of patience. Fabric's cloud drive (https://fabric.so/features/your-cloud) brings media into a single searchable library without requiring you to consolidate everything onto one physical drive. Files stream on demand, so they don't consume local storage, but they're all accessible and searchable from one place.

Fabric also connects to existing storage through its connections feature (https://fabric.so/connections), integrating with services like Google Drive, Dropbox, and Frame.io. If your footage already lives in these places, you don't need to move it. You can search across connected sources alongside media stored directly in Fabric.

The practical effect is that your search isn't limited to the current project. When you need that warehouse shot, you're searching your entire media library, across every shoot and every project, in one query. The footage finds its way to you rather than the other way around.

Annotations and context

Finding the right clip is often only half the task. Once you've located it, you need to know where the usable moment is within the clip. Raw footage tends to include long takes, false starts, and plenty of material that didn't make the cut for a reason.

Fabric's annotation tools (https://fabric.so/features/annotations) let you add timestamped notes directly on video and audio files. If a director or producer has already reviewed footage and marked selects, those annotations travel with the file. When you find the clip through search, the notes are already there, pointing you to the section that matters. This turns a found clip into a usable clip much faster.

For teams, annotations also serve as a form of institutional memory. When multiple people work with the same footage library, the notes and markers left by one person become a resource for everyone else. A producer's feedback on selects, a colourist's notes about exposure issues, an editor's marker for a favourite reaction shot: all of this context accumulates and makes the library more useful over time.

Building a footage library that holds its value

There is a broader point here about how creative teams think about their media. Raw footage is an asset. Every shoot produces material that has potential value beyond the immediate project. But that value is only realisable if the footage is findable and accessible. A hard drive sitting in a drawer, full of beautifully shot b-roll that nobody can search, is a wasted resource.

The shift from thinking about footage as "files on a drive" to "searchable assets in a library" changes the economics of production. Shoots become investments in a growing library rather than one-off expenses. B-roll from a location shoot in March becomes available for a pitch deck in September, a social media cut in November, and a showreel the following year. The footage pays for itself repeatedly, provided someone can find it when it's needed.

Fabric's AI assistant (https://fabric.so/features/ai-assistant) adds another layer to this. Beyond searching, you can ask questions about your media library and get answers informed by the content itself. It's a way of interacting with your footage that goes beyond traditional file browsing.

The publish feature (https://fabric.so/features/publish) also plays a role here. When you do find that perfect shot and want to share a selection with a client or colleague, you can create polished shareable pages with link analytics, so you know who viewed what and when.

The time arithmetic

Consider a conservative estimate. An editor spends ten minutes per session looking for footage that better organisation would surface immediately. Over the course of a week with multiple projects, that might add up to an hour. Over a month, four or five hours. Over a year, several full working days spent doing nothing but looking for files that already exist.

That arithmetic gets worse with larger teams. Three editors searching for footage independently, each maintaining their own mental map of what's where, will spend collectively far more time than one editor working from a well-organised, searchable library.

The investment in proper footage management is small relative to the time it recovers. The difficulty has always been that "proper footage management" traditionally meant manual labour that nobody had time for. When the organisation happens automatically and search works by description rather than by filename, the cost of keeping things tidy drops close to zero.

That is the case for treating footage management as infrastructure rather than as a chore. It is what a MAM does at the footage level, and it is the problem Fabric was built to solve.

Frequently asked questions

How does Fabric's semantic search differ from searching by filename?

Traditional file search matches the exact text in a filename or folder name. Fabric's semantic search (https://fabric.so/features/search) analyses the visual and audio content of your media, so you can describe what you're looking for in natural language. A search for "person walking through a field at sunset" will return matching clips regardless of what the files are named.

Does Fabric replace my existing storage drives?

Fabric's cloud drive (https://fabric.so/features/your-cloud) can serve as your primary media storage, streaming files on demand without consuming local disk space. You can also use Fabric alongside existing storage by connecting services like Google Drive, Dropbox, and Frame.io through Fabric's connections (https://fabric.so/connections).

Can Fabric auto-tag footage from cameras that use generic filenames?

Yes. Smart organisation (https://fabric.so/features/smart-organization) analyses the content of your media, so it works regardless of the original filename. Clips from cameras that produce names like "A001_C004_0312" get categorised based on what's in the frame, not what the file is called.

Is Fabric suitable for large volumes of raw footage?

Fabric is built as a media asset management platform designed for creative teams working with substantial media libraries. The cloud drive streams files on demand, which means your local machine doesn't need to store everything at once, and semantic search remains effective across large collections.

Can multiple editors search the same footage library?

Yes. Fabric is designed for team use. All connected media is searchable by everyone with access, and annotations (https://fabric.so/features/annotations) left by one team member are visible to others. This eliminates the problem of each editor maintaining a separate mental map of where footage lives.

Does Fabric work offline?

Fabric requires an internet connection. Since files stream from the cloud and search relies on Fabric's servers, an active connection is necessary to access and search your media library.

How do annotations help with footage organisation?

Annotations (https://fabric.so/features/annotations) let you add timestamped notes to video and audio files. When you or a colleague mark selects, flag issues, or note favourite moments, those markers travel with the file. When someone later finds the clip through search, the annotations point them directly to the relevant section.

Can I share specific clips or selections with clients?

Fabric's publish feature (https://fabric.so/features/publish) lets you create polished shareable pages from your media. These pages include link analytics so you can see who viewed the content and when, which is useful for client review workflows.

What types of media does Fabric's semantic search work with?

Semantic search works across video, audio, images, and documents. For video and images, it analyses visual content. This makes it effective for finding b-roll, stills, and other visual assets by describing what's in them.

How quickly is new footage searchable after uploading?

Media becomes searchable shortly after upload as Fabric processes and analyses the content. The exact time depends on the volume and type of media, but the goal is to close the gap between ingest and searchability so that footage is findable well before you sit down to edit.

Related pages

• Your Cloud

• Search

• Smart Organisation

• Annotations

• Connections

• AI Assistant

• Publish