Presentation intelligence: synchronised assets, not continuous video
cRUD Research TeamThe cRUD Company
A presentation is not naturally a continuous video. It is a set of visual assets, an order, a current position and a series of authorised changes. Treating every slide as camera pixels consumes delivery capacity, reduces readability and makes recovery depend on a stream that was never the true source. This whitepaper defines the presentation-intelligence model behind Present in cRUD Morph.
Executive summary
- Eligible presentation material should be converted into protected static assets before delivery.
- The current slide and presentation state need one server-authoritative timeline.
- A viewer can recover the correct slide from state without replaying missed video frames.
- Dynamic demonstrations remain on a moving-media path when motion carries meaning.
- Conversion, access, synchronisation and output evidence should be reviewable without exposing private implementation.
Why continuous video is the wrong default
Screen sharing is a useful general tool because it can carry almost anything visible on a presenter's device. That flexibility has a cost. Static slides are repeatedly encoded as changing video frames. Text and diagrams can become soft at a low bitrate. A momentary network problem can make the audience miss the current slide even though the underlying asset has not changed.
The general tool also joins unrelated responsibilities. The presenter's screen becomes the source, the capture path becomes the access control and the video stream becomes the timeline. A notification, pointer movement or accidental window can enter the programme. Presentation intelligence separates the approved material from the presenter's desktop and separates slide state from video delivery.
The presentation as assets and state
An eligible presentation is ingested, validated and converted into a set of delivery assets. The system retains the relationship between the source, its approved version, the generated assets and their order. The live state identifies the current presentation and position. A viewer receives the authorised asset and the state required to display it at the correct moment.
This model does not assume every page is a simple image. Conversion may preserve dimensions, orientation and other approved properties, while the customer experience decides how navigation, presenter notes or accessibility information appear. The public contract exposes the result and state, not the private conversion pipeline.
Ingestion and conversion boundaries
The system should establish what it accepts before processing starts. File type, size, page count, ownership, malware controls and environment all belong to the governed request. An accepted upload receives a durable identity. Conversion produces versioned assets and an explicit outcome. A failed conversion does not silently fall back to presenting an unreviewed source.
Source lineage matters because a presentation can change shortly before a session. Replacing the file should create a new approved version, not overwrite the evidence of what earlier viewers received. The programme can then identify which source and asset version belonged to the live output, recording and replay.
Synchronisation needs one authority
A slide change is a state transition. The authorised presenter or producer requests the change, the server validates it and the resulting position becomes authoritative. Viewers observe that state. A reconnecting viewer receives a current snapshot plus any ordered changes needed to catch up. The viewer does not guess the slide from a timestamp embedded in a video stream.
This approach also keeps multiple outputs aligned. An interactive speaker view, a large audience path, Program Output and recording can all reference the same presentation state while rendering it appropriately for their purpose. A delayed route can still identify the correct state for its own programme timeline.
Static and dynamic material need different paths
Presentation intelligence should not force moving material into static assets. A live product demonstration, animated model or continuously changing dashboard may need a moving-media path. The decision follows purpose. If change carries meaning, motion should be preserved. If the material is fundamentally static, repeating it as video is unnecessary work.
A programme can use both forms. Slides may be delivered as assets while a demonstration switches to approved screen content. The transition belongs to the session and Program Output, so recording and replay preserve what the audience was meant to see rather than a collection of unrelated files.
Access and protection
Presentation assets are part of the authorised product experience. A public-looking asset address must not become permanent access. Delivery should use tenant and environment boundaries, time-limited authority and the customer's entitlement decision. Revocation, expiry and replay access remain explicit.
The customer product owns registration, admission and business permission. cRUD Morph uses the resulting capability to deliver the approved asset. Provider storage paths and private credentials stay behind the boundary. The operating record can show which version was authorised and delivered without revealing the storage implementation.
Recovery and degraded conditions
Because the asset and state are separate, recovery does not require the viewer to reconstruct missed frames. The client can obtain the current authorised position and asset. If an asset cannot be delivered, the interface should show a clear state rather than leaving an old slide visible as though it were current.
Prefetching can improve transitions, but it should not change access policy or state authority. A prefetched asset is not permission to display it. The current state and customer entitlement still control what appears. This distinction keeps performance optimisation from becoming a security shortcut.
Readability, accessibility and output
Static assets can preserve text and diagrams more clearly than a constrained screen-share video. That does not by itself make a presentation accessible. The product still needs an accessibility plan for captions, spoken explanation, contrast, keyboard behaviour and any alternative information associated with the source.
Program Output may compose the current slide with speaker video, branding, captions and overlays. The composition should reference the same authoritative presentation state as the audience experience. Recording then preserves the approved programme, while replay can reuse the output and source lineage instead of rebuilding the relationship afterwards.
Operating roles and review
Presentation delivery crosses several responsibilities. A content owner approves the source and version. A producer controls when the approved material enters the programme. A presenter can request authorised navigation. The media foundation converts, protects and distributes assets while maintaining state. The customer product decides who holds each business role and whether a viewer may enter the experience.
These responsibilities should be visible in operating procedures. A last-minute replacement needs a named approval and a version transition. A conversion warning needs a review path before the session starts. A producer needs to know whether an unavailable asset should hold the current state, switch to a safe alternative or pause the programme. Support needs the asset identity and state history without gaining unnecessary access to the source file.
A rehearsal should test these boundaries. Upload the actual presentation, replace it, navigate non-linearly, join late, reconnect and move between presentation and dynamic demonstration content. Review the audience view, Program Output, recording and replay. The goal is not only to confirm that slides appear. It is to prove that authorised people can control them and that every output can identify what was shown.
Evidence and evaluation
Evaluation should compare more than picture quality. Test upload and replacement, page ordering, slide changes, late joins, reconnects, two audience routes, Program Output, recording and replay. Inspect whether every output identifies the approved source and version. Confirm that a failed or stale state is visible rather than silently masked.
Measure legibility, transition time, recovery accuracy and delivered work against a matched screen-sharing path. Do not claim a data reduction or quality improvement without that evidence. The purpose of the model is to remove unnecessary repetition and preserve meaning, but the measured result depends on content, providers, devices and audience conditions.
Conclusion
Presentation intelligence treats a presentation according to what it is. Protected assets preserve the material. Authoritative state preserves timing and position. A moving path remains available when motion matters. By separating those responsibilities, the product can improve readability, recovery, output lineage and operating clarity without turning the customer application into a presentation-processing system.
Basis and methodology
This paper explains the public role of Present within cRUD Morph. It reports no benchmark or universal saving. Conversion support, asset handling, accessibility responsibilities, providers and measurable outcomes are confirmed for the approved engagement.
For the public product boundary, review the cRUD Morph approach, capabilities and documentation.
Frequently asked questions
How does cRUD Morph keep slides synchronized?
What happens to animated or live demonstration content?
Are presentation assets public?
Does asset delivery guarantee lower cost than screen sharing?
About this whitepaper. It explains the public product and operating principles of cRUD Morph. It does not disclose source code, private implementation, provider credentials or customer data.