Short answer
Video asset management is the part of DAM that handles video ingest, transcoding, frame-accurate review, subtitles, chapters and streaming delivery. Video multiplies storage and processing cost: one hour of 4K ProRes 422 HQ at 25 frames per second takes about 332 GB. Buyers with large video libraries should test transcoding, review tools and egress pricing before choosing a vendor.
What video asset management covers
Video asset management covers the full video lifecycle inside a DAM, from master file to streamed rendition. Video needs capabilities that image libraries never use: time-based review, proxies, subtitles, chapters and adaptive streaming.
| Stage | What happens | DAM capability |
|---|---|---|
| Ingest | Masters and camera files arrive, often hundreds of GB per project | Accelerated or resumable upload, checksum verification |
| Proxy | A light preview version is created for browsing and review | Automatic proxy generation |
| Review | Stakeholders comment on exact frames and approve versions | Time-coded comments, version comparison, approval workflow |
| Enrichment | Transcripts, subtitles, chapters and tags are added | Speech transcription, subtitle files per language, scene detection |
| Transcoding | Masters convert into channel formats | Rendition presets per channel |
| Delivery | Streams play on websites, apps and social | Adaptive streaming, player, delivery URLs |
| Archive | Masters move to cheaper storage after the campaign | Storage tiers with restore times |
Transcoding is the conversion of a video or audio file from one format or bitrate to another. A rendition is a derived version of an asset in a specific format, size or crop for a particular channel. The Learn pillar introduces both as core DAM concepts.
Capabilities to test
A video DAM test uses the buyer's own footage, at full size, with the buyer's reviewers. Short demo clips hide upload, processing and review problems that appear with real masters.
| Capability | What good looks like | Test |
|---|---|---|
| Large file upload | Resumable upload that survives a dropped connection | Upload a 50 GB master over an office connection and interrupt it once |
| Proxy and playback | Playback starts quickly from the proxy, with frame-by-frame stepping | Scrub through a 20-minute file and step frame by frame |
| Frame-accurate review | Comments pinned to timecode and on-screen area | Three reviewers comment on the same scene; compare versions |
| Transcoding presets | Channel formats generated automatically on approval | Generate web, social vertical and broadcast renditions from one master |
| Subtitles | Subtitle files per language linked to the video | Import, edit and export subtitle files in three languages |
| Transcription | Speech converted to searchable text | Transcribe a Dutch and a German clip; check accuracy |
| Chapters and scenes | Scenes detected and searchable | Search for a scene and jump to its timecode |
| Streaming delivery | Adaptive streaming through a player or delivery URL | Embed on a test page and throttle the connection |
Record processing time for each test. Transcoding a one-hour master into five renditions shows whether the platform keeps up with a campaign launch.
Formats and standards that matter
Video asset management depends on a small set of formats for masters, renditions, streaming and captions. The DAM should accept the buyer's master formats and produce every rendition format the channels need.
- Masters: production codecs such as Apple ProRes keep quality for editing. Fact: Apple's ProRes white paper lists 184 Mbps for ProRes 422 HQ at 1920 by 1080 and 25 frames per second. That rate equals about 83 GB per hour.
- Renditions: compressed formats such as H.264 or H.265 in MP4 containers serve web and social channels.
- Streaming: Fact: RFC 8216 describes HTTP Live Streaming (HLS), which delivers video as a series of short segments at several bitrates. MPEG-DASH is a comparable adaptive streaming format.
- Captions: Fact: WebVTT is a W3C format for timed text tracks such as subtitles and captions.
- Timecode: frame-accurate review depends on timecode carried from the master into proxies and comments.
Cost drivers: storage, processing and delivery
Video cost in a DAM has three drivers: storage of masters, processing minutes and delivery volume. DAMSelection position: storage, egress and AI credit consumption drive five-year DAM cost more than licence fees, and video makes that visible fastest.
| Driver | What scales it | How vendors usually meter it | Control |
|---|---|---|---|
| Storage | Hours of masters, codec, resolution, retention | Storage tiers in TB | Archive tier for old masters; retention rules |
| Processing | Hours transcoded, renditions per master, AI transcription | Processing or output minutes, credits | Presets limited to channels in use |
| Delivery | Views, average bitrate, view duration | Bandwidth or egress in TB | Adaptive streaming; CDN caching; external video platform for high-volume public video |
Worked example with stated assumptions: a brand holds 2,000 hours of HD masters and 200 hours of 4K masters, all in ProRes 422 HQ at 25 frames per second. Using the ProRes white paper rates of 83 GB and 332 GB per hour, masters occupy about 166 TB plus 66 TB, or 232 TB. Public videos receive an assumed 500,000 views per month of 3 minutes each, at an assumed average stream of 5 Mbps. Each view then transfers 112.5 MB, and monthly delivery reaches about 56 TB.
Ask each vendor to price these exact volumes over five years. The TCO Calculator and the five-year TCO guide separate storage, processing and delivery lines.
Accessibility: captions and transcripts
Captions and transcripts are accessibility requirements as well as search assets, so treat them as mandatory metadata for published video. Fact: WCAG 2.2 success criterion 1.2.2, Captions (Prerecorded), is a Level A requirement. It requires captions for all prerecorded audio content in synchronised media, except clearly labelled media alternatives for text.
Three practices keep captions manageable at scale:
- Store one subtitle file per language as a linked asset, with status and owner.
- Use AI transcription for the first draft and human review before publication, because names, product terms and brand language need correction.
- Pass caption files to every channel through the integration, so editors never upload them by hand.
Transcripts also make every spoken word searchable inside the DAM. The accessibility and alt text guide covers the image side of the same obligation.
AI for video
AI for video turns hours of footage into searchable, reusable segments. DAMSelection assessment: transcription, scene detection and logo detection deliver reliable search value today, while automatic clipping and generated edits need review before publication.
| AI capability | Value | Review need |
|---|---|---|
| Speech transcription | Every spoken word becomes searchable; caption drafts | Review before captions go live |
| Scene and shot detection | Search jumps to the right moment in long footage | Spot checks |
| Logo and object detection | Find footage showing a product or sponsor | Spot checks; confirm before rights decisions |
| Face recognition | Find footage of a named person | Legal basis check under the GDPR before enabling |
| Automatic clipping and reframing | Short vertical cuts from long masters | Brand and rights approval per clip |
Ask whether AI video processing is metered per minute, and include it in the processing line of the cost model. Test transcription in every language the library holds, because accuracy can differ by language and accent. The AI in DAM guide gives the test protocol and the credit cost method.
Integration points and system of record
Video often flows through several systems, so the buyer must decide which system holds the master and which serves the stream. Unclear ownership produces duplicate libraries and broken links.
| System | Typical role | Decision to make |
|---|---|---|
| Editing tools | Create and edit the master | Whether editors check masters in and out of the DAM |
| DAM | System of record for masters, metadata, rights and approvals | Whether the DAM also streams to public channels |
| Video platform | Streaming, player, analytics for public video | Whether the platform pulls from the DAM or receives pushed renditions |
| CDN | Delivery close to viewers | Who pays egress and on which contract |
| CMS and e-commerce | Embed players on pages | Whether pages reference DAM or video platform URLs |
| Social platforms | Publish native uploads | How rights and expiry reach native copies |
DAMSelection assessment: keep the DAM as the system of record for masters, rights and approvals in every setup. Public streaming can run from the DAM delivery layer or a specialised video platform, depending on volume. The headless DAM guide explains the delivery patterns.
Rights in video
Video carries more layered rights than images, because one file combines footage, music, voices, talent and locations. Each layer can expire at a different date.
- Music: library or label licences limited by channel, territory and duration.
- Talent and voice-over: contracts limited by period, channel and buyout terms.
- Stock footage: licences per clip, often with end-use limits.
- Locations and property: releases for recognisable places and objects.
Record each layer as a linked rights record, and let the earliest expiry drive the video's expiry date. The rights management guide describes the fields and enforcement levels.
Video selection checklist
A video selection checklist turns the capabilities, costs and integrations into scored requirements. Use it in the RFP and in the scripted demo scenarios.
- Upload and processing tests passed with the buyer's own masters.
- Time-coded review and version comparison tested with real reviewers.
- Subtitle and transcription workflow tested in the buyer's languages.
- Rendition presets cover every channel specification.
- Storage, processing and delivery priced for the buyer's five-year volumes.
- System of record and streaming role agreed with the video platform owner.
- Rights layers recorded with the earliest expiry driving the video status.
- Masters exportable with metadata and subtitles at contract end.
Frequently asked
What is the difference between a DAM and a video platform?
A DAM manages masters, metadata, rights, review and approval across all asset types. A video platform focuses on streaming, players and viewer analytics for published video. Organisations can run both, with the DAM as the system of record and the video platform serving high-volume public streams.
How much storage does video need in a DAM?
Storage depends on codec, resolution and hours. Apple's ProRes white paper lists about 83 GB per hour for ProRes 422 HQ at 1920 by 1080 and 25 frames per second, and about 332 GB per hour at 3840 by 2160. Compressed renditions and proxies add a smaller share.
Can a DAM stream video to a website?
A DAM with a streaming delivery layer serves video through adaptive streaming and an embeddable player or delivery URL. Test playback on a throttled connection and check how delivery volume is metered. High-volume public video often runs through a specialised video platform fed by the DAM.
Does a DAM need to support subtitles?
A DAM that manages published video needs subtitle support, because captions are an accessibility requirement under WCAG 2.2 success criterion 1.2.2. Store subtitle files per language as linked assets, review AI drafts and pass the files to every channel through integrations.
How do you test video features in a DAM demo?
Send the vendor two weeks ahead a full-size master, a set of reviewers and subtitle files in your languages. Test upload with an interruption, frame-accurate comments, rendition presets, transcription accuracy and streaming on a throttled connection. Record processing times for each step.