Why did my extended product-ad ending turn black instead of holding the last frame?

Decide between a picture hold and a separate card

A black tail is not necessarily a failed video generation. First identify the padding operation in the finishing chain. In add mode, tpad creates solid-color frames. In clone mode, it copies the last frame of the stream for end padding. A deliberately designed blank background and a frozen product picture therefore require different choices.

The filter offers both frame counts and duration options. Explicit start_duration or stop_duration overrides the corresponding count option. Record the chosen mode, length and color instead of assuming a colleague used the same defaults. A stop count of -1 means indefinite padding, so do not substitute it for a bounded delivery duration.

Executed one-second-tail comparison

We generated a two-second, 10-fps testsrc2 source containing 20 frames, then compared two FFmpeg 7.1 filter outputs. tpad=stop_duration=1:stop_mode=clone produced 30 frames and 20 distinct checksums; the additional tail matched the final source-frame checksum. The add-mode comparison also produced 30 frames, but its added black tail had a new checksum.

The first output timestamp was zero and the last was 2.9 seconds in both cases. These are filter measurements before muxing, not a claim about every encoded container or player. No real product ad, soundtrack extension or subjective-quality test was performed.

Choose a trustworthy final frame first

Illustrative exercise: input A is an approved product hero shot ending with a stable, readable label. Input B ends during a hand movement and includes a distorted package edge. A may support a deliberate hold after frame review. B should not gain a longer display merely because clone mode is available; choose an earlier approved ending or a separate approved card.

If the intended ending is a designed offer card, construct and approve that card separately. Cloning the last frame of an unrelated demonstration does not create the card’s text, layout or offer conditions. If a solid-color ending was intended, check its transition and typography rather than treating every black frame as an error.

Check the finished ending as a sequence

Play the transition into the hold and the complete ending. Check that its stillness is intentional and that the product remains accurately represented. Review sound and captions independently; the documented filter pads video, so the soundtrack must have its own explicit ending policy.

Keep the approved source and finishing settings. If the picture and sound end at different moments, inspect the editing chain before regenerating the source clip. A longer file is not necessarily a more complete advertisement.

Inspect the actual measurement data

These data describe the synthetic checks scoped in the article, not real-product, publishing-platform or subjective-quality acceptance.

• synthetic-video-filter-measurements.csv · SHA256 2d8619cda22dc3ba1219061ade31f1132d6f67a9b6cf81efbe7986396221220a

Illustrative exercise — intentional ad ending

These are illustrative exercise rows. Save a copy and record your own observations; the example is not an executed acceptance test.

Download review worksheet (CSV)