SCTE‑35 and SCTE‑104 are complementary signalling standards that indicate when a video stream should switch to another type of content – typically replacing programme content with an ad for a defined period.

This signalling can be done in two ways: splice insert or time signal. In this blog post, we’ll explore these two modes and when to use each one.

What do SCTE-35 and SCTE-104 do?

SCTE-104 and SCTE-35 both provide cues that say: “This is the time to switch content.”

SCTE-104 does this in the studio, with SDI carrying the SCTE-104 cues inside facility equipment, typically embedded in SDI ancillary data packets as defined by SMPTE.

SCTE-35 does the same further downstream, indicating to edge systems that it’s time to show an ad. The signal is typically carried in MPEG-TS, HLS, DASH, or other IP transport formats to encoders, packagers, CDNs, ad decision systems, and players.

What are splice insert and time signal?

Splice insert and time signal are both ad-break switching modes. But they differ in how the cue is represented and what can be signalled in parallel.

Splice insert

With splice insert, the signalling typically targets an explicit splice action that is executed deterministically by downstream systems. Time can be conveyed in one of two ways:

• ASAP: splice should happen now.
• Presentation Time Stamp (PTS): splice should happen at a specific moment in the timeline.

In other words, splice insert can tell the system “do it right now” or “do it at this exact time”, and downstream workflows are expected to act accordingly.

Time signal

With time signal, the cue is about timing, but it can also signal different types of elements in parallel, for example:

• start/end of programme,
• start/end of an ad break (also called “ad screen”),
• start/end of a specific ad.

This matters because it enables more granular use cases than “replace everything for the ad break”. For example, time signal can be used to:

• replace only a single ad rather than an entire ad screen,
• support behaviours that are relevant for time-shifting / start-over / STB and box UX (e.g., preventing viewers from skipping ads during timeshift, depending on the workflow),
• drive “start over” experiences by enabling servers that can restart content from the beginning with consistent ad replacement logic.

The differences between splice insert and time signal

Given that the cue is expressed differently in these two modes, there are differences in how downstream systems interpret it and, consequently, what they can do.

• Time signal is timing-centric and, as we have seen above, can carry element-type meaning in parallel (programme boundaries, ad-screen boundaries, and even specific ad boundaries). This enables more granular and richer replacement/scheduling use cases.
• Splice insert is action-centric and (typically) deterministic: it expresses “splice now” (ASAP) or “splice at this exact PTS”, with clearer splice command semantics for systems that rely on explicit splice instructions.

Should you use Splice Insert or Time Signal?

As often with technology – it depends. There is no one system that is better in all use cases, so it depends what you are trying to achieve and what equipment you have. Here is a practical guide of when to use which mode.

Use Splice Insert when:

• Your downstream chain does not support time signal (many existing systems only support splice insert)
• The downstream systems expect explicit splice information to ensure seamless transitions.

Use Time Signal when your downstream systems (including encoders, packagers, or set-top boxes) support time signal end to end, and:

• Regional or operator policies (e.g., national standards) require it,
• You want a cue that can support more granular replacement semantics (e.g., ad-screen vs single-ad replacement), using parallel signalling of element types, or
• You need to address timeshift and “start-over” behaviours where the cue semantics need to be consistent with how the platform rewinds/replays content.

You can even use a hybrid approach. Many workflows generate SCTE-104 in the studio, convert to SCTE-35 for transport, and choose either splice insert or time signal depending on the granularity they require and what downstream components support.

In some markets, local standards favour time signal.

AF2M Guidance in France

AF2M stands for Association Française pour le développement des services et usages Multimédias Multi-opérateurs (French Association for the Development of Multi-operator Multimedia Services and Uses). This association aims to develop and promote B2B2C multi-operator services and uses.

Together with France’s national television advertising syndicate, SNPTV (now called AMDTV), it published guidelines for ad replacement in 2020, the year that targeted advertising became legal in France.

These guidelines stipulate that time signal should be used for ad-replacement and any kind of targeted TV workflow.

Video Distribution with EasyTools

The EasyTools Video Distribution solution can convert, insert, and manipulate cues to fit many distribution and monetisation needs. In particular, it enables broadcasters and operators to:

• Convert SCTE-104 to SCTE-35 markers, and vice versa, to bridge the studio and transport domains;
• Insert SCTE-35 markers based on sequence recognition or via API triggers for automated, programmatic workflows;
• Convert time signal into splice insert when downstream systems require explicit splice commands;
• Manipulate metadata (custom tags, ad IDs, duration overrides) while preserving provenance for reporting and verification.

Learn more about the EasyTools Video Distribution solution.