Shaker bottles: the shake is Vidu; the oz marks must not crawl

Ice rattle is diegetic. Printed graduations are overlay or a locked still, never Veo type.

A shaker clip that holds is two objects. The bottle in the still is the SKU: lid, sleeve print, oz marks a buyer will match on the shelf. The shake is physics plus sound — ice, powder, the click of a lid that already exists in the photo. Motion may rattle what is in the cup. It may not redraw the scale.

Lock the bottle. Buy the rattle. Typeset the marks if the still cannot hold them.

The still is the bottle

The cover is the wrap, the lid colour, the blender ball if it is in frame. Text-to-video will invent a prettier cup with melted type and a cousin logo. That is a different product.

Vidu Q3 Image to Video is the row when the still is already the bottle. Image required. Audio on. Durations only 5s, 10s, and 15s. Max output 1080p. Sixteen credits listed; the row is marked discounted. Crop the still to the delivery ratio before upload. This slug has no aspect-ratio list. 9:16 in the prompt does nothing.

Image-to-video is the launcher. Frame one is the file. Prompt the shake — a short rattle, a wrist flick, condensation that is already on the wall — not a second bottle. Re-describing the print in the paragraph is how the model starts drawing a second sleeve.

If packing tape covers the oz line, reshoot. Do not ask motion to sharpen graduations that were never there. A last frame is allowed when you have a second approved still of the same cup — lid on, then lid cracked — not a generated interior of ice the photo never contained.

Ice rattle is diegetic

Native audio means soundtrack and frames from the same pass. Ice against plastic is the textbook case: the hit is visible, so the hit has to be audible in the same denoise. Retrofitting a "shake" sample onto invented motion is how the crash lands on the backswing.

Write the mix into the 5s prompt. Name two or three sounds: ice rattle, lid click, room tone. State "no speech, no music" if you want clean foley. Leave audio unspecified and Q3 invents a mix — including a gym murmur that does not belong on a pack shot.

Do not generate Q3 silent-on-purpose. Audio is on. A "silent" brief still gets a band. Mute plates belong on a silent image-to-video row. Five seconds of native rattle is not a subtitle track.

A 15-second miss of a wrong bottle with a soundtrack attached is a long miss. A shake-and-settle is often five seconds. Do not pad because the slider has two more stops.

Graduations are overlay or they crawl

Oz, ml, the fill line, a "20 oz" callout — those strings have to match the mold and the listing. A video model drawing them on the wall, or sliding them as the cup shakes, will miss a tick. That is a false spec.

If the marks are already sharp in the photograph, keep the camera off the 6-point line. The still is the typesetter. If they are not readable, do not generate a sharper bottle. Burn the spec with a text overlay: one line you wrote, whole clip, top / center / bottom. It does not transcribe. It does not listen. Position is required.

Several beats — size, then BPA-free, then a CTA — are timed overlays on the same locked take. Do not regenerate the cup to "include the oz this time."

Veo, or any cinematic talking row, is the wrong typesetter. "Shaker bottle, ice, 20 oz marks crisp" is how you pay for crawling glyphs and a cousin sleeve. Overlay is composited, so the number does not crawl when the ice does.

This is not a gym floor or a powder comparison

Gyms sell member energy: the room, the PR, the class. Equipment photography is the miss on that page. A shaker brand that copies the gym playbook — generated members, a packed floor, a cup that is almost yours — has used the wrong still. The bottle is the listing. The floor is their listing.

Supplement brands are the other neighbour. That page is comparison by ingredient and dose, real testimonials, a label close-up of the powder. A shaker OEM that ships a "this vs that protein" talking head has left the cup. Overlay a size line on the locked bottle. Do not Omni a supplement ad and call it the shaker.

Boutique studios and personal trainers are further out: class feel, form checks. They may hold your cup in a real clip. They do not own the graduation mold. If they need the oz to read, they overlay or they film the real marks. They do not ask Veo to typeset them.

Three plates:

• The cup. Photograph it. Q3 for the rattle.
• The gym. Film the room. That is /for/gyms.
• The powder. Label still, comparison copy. That is /for/supplement-brands.

The shake is Vidu. The oz marks must not crawl.

FAQ

Can I generate a 3-second shake on Vidu Q3?

Not on this ladder. Durations are 5s, 10s, and 15s. The shortest legal clip is five seconds. If you need a three-second hit, pick a model that lists 3s, then cut.

Why not prompt Veo to "keep the oz marks exact"?

You can lock a still and buy motion, and the small type will still drift. If the words are the listing, overlay them. If the wrap print is the listing, I2V the bottle and keep the camera off the scale.

Does Q3 require a photo of our actual bottle?

Yes, as image-to-video. The still is the contract. A text-only "protein shaker with ice" is a different family. If the lid, the sleeve, or the marks must survive, lock that frame first.

Should I overlay the oz line on every take?

Only when the still does not already carry it. Readable mold marks do not need a caption. Unreadable marks need overlay, not a second generate. Crop first, then print.