Vet clinic Halloween clip: chocolate, xylitol and costume checks

The ASPCA's Halloween guidance names three risks a clinic can show in a few seconds each: chocolate in all forms, sugar-free candy with xylitol, and costumes that limit movement, sight or breathing. One 24-second reel with three caption cards covers them. On Sume it costs $0.22: one trim at $0.02 and one caption job at $0.20, or $0.36 if you cut three ranges and join them with a render.

What the ASPCA page says

The page gives no dose tables or toxicity thresholds, so the reel should add none; a clinic's own vet should write advice of that kind.

Shoot it as silent b-roll

A clinic does not need an actor. Film three shots on a phone: a bowl of wrapped candy on a table with a dog's nose out of focus behind it, a costume laid flat with a hand pressing the neck opening, and a collar with a tag on a hook. Each is 6 to 8 seconds.

Because the reel is silent, speech-to-text is the wrong tool; it would fail with caption_no_speech. Use video captions with cues instead, which burns the exact text you wrote at the exact seconds you set.

• Card 1: "Chocolate and sugar-free candy are not for pets."
• Card 2: "A costume should never limit movement, sight or breathing."
• Card 3: "Save the poison control number in your phone."

Build and price it

If the three shots come from one recording, use video trim to cut a 24-second range, or cut three ranges and join them in Timeline 1.0. The join bills one output minute at $0.10. Then run one caption job with all three cards, since a caption job covers a video up to 60 seconds.

Keep the medical lines short

A clip can remind people of a hazard. It cannot assess a case. End the reel with your clinic's number and the ASPCA's, and leave dosing, symptoms and what to do in a specific emergency to a veterinarian on the phone.

The page also advises a separate quiet room for pets during trick-or-treating. If you want a fourth card, that is the one, and it needs no extra clip: reuse the table shot.

Before you publish

Sume tools read media from your workspace, so import the recording first with POST /v1/media-imports and use the returned media.sume.com URL as the trim source. A trim needs start plus exactly one of end or duration, and precision defaults to exact, which lands on the frame you wrote down.

Keep the clinic's phone number as the last card, on screen for at least two seconds longer than the others, since it is the one line a viewer may want to copy. Captions with cues take your text with start and end seconds and skip speech-to-text, so a clip with no speech never hits caption_no_speech. A caption job covers a video up to 60 seconds, which is why each clip here stays well under that.