Tamale pre-order clip in English and Spanish: one still, one clip

For a bilingual tamale pre-order clip, make one still of the product, animate it into a 6-second clip with Wan 3.0, record the pre-order line once in English and once in Spanish, and render the same picture under each voice. The total on Sume is $1.19: $0.075 for the still, $0.75 for the clip, $0.0442 for both voices, $0.125 for one music bed and $0.20 for two renders.

Holiday tamales are a pre-order business: the kitchen needs counts before it cooks, and the customer needs a last order day. The clip should carry two facts only, what you sell and the date you stop taking orders.

How do you make the picture and the clip?

Generate the still with Ideogram 4.5 at medium quality ($0.075) and describe the product as it will arrive: a steamer basket, a dozen wrapped tamales, a ribbon. Then send that still as the first frame to a Wan 3.0 request at 720p for six seconds. Sume's video docs list Wan 3.0 at 2 to 30 seconds, and the 720p rate is $0.10 a second at list, or $0.125 with Sume's 1.25 margin, which is $0.75 for six seconds.

Ask for gentle motion only, such as steam rising. A small movement is more likely to keep the product recognizable from the first frame to the last.

Can the same clip speak two languages?

Yes. Sume's text-to-speech catalog lists Spanish (es) next to English for its Sonic models, so the English and the Spanish read are two separate voice jobs against the same clip. Write the Spanish line by hand or with a translator you trust, not by word-for-word translation, since a pre-order line is a promise about dates and prices.

Then run two Timeline renders: the same clip, a different audio spine each time. Each render is $0.10 for up to a minute of output, so the second language adds the cost of one voice and one render, about $0.12.

What goes in the script?

Name the product, the box size, the price and the last day to order, then one way to order. Say the date as a weekday and a number, for example the Friday before Christmas, so it does not clash with the caption on screen. Keep each read under 30 seconds so one music bed fits both versions. Ask a Spanish speaker on your team to listen to the Spanish read before it posts, since a number or a street name that sounds wrong will cost you more than the voice job did.