How to Keep a Product the Same Size in Frame When Zooming in a Short AI Video

Quick answer

If your product grows or shrinks between shots, the problem is usually not the zoom setting — it is that you changed two variables at once. Fix the camera-to-product distance first, then change only the zoom level across your shots. When you want the product to look bigger, move the camera closer instead of zooming in; when you want it smaller, step back. Keep the product's outline in the same place on a grid overlay, and the clip will read as one continuous piece instead of three unrelated takes.

This guide uses a plain wooden block as a stand-in product so you can rehearse the method before applying it to a real item.

Why the product "breathes" between shots

Short vertical clips are judged in a second or two. If the block occupies 40% of the frame height in shot one and 55% in shot two, viewers read that as a jump cut or a different object, even if the lighting and background match.

Three things cause it:

• Distance drift. You moved the camera between takes without recording where it was.
• Zoom drift. You adjusted zoom to "fill the frame" and lost the reference.
• Framing drift. The product is the same size but sits higher, lower, or off-center, so the eye reads it as a size change.

The fix is a fixed reference: one object, one marked distance, one zoom value per shot.

Step 1: Plan the framing on paper before you touch the camera

Draw three rectangles in a vertical 9:16 proportion. Label them Shot A, Shot B, and Shot C. In each rectangle, sketch the block as a square and decide where it sits:

• Shot A — establishing. Block centered, occupying roughly the lower-middle third of the frame. Room above for a headline later.
• Shot B — closer. Block larger, still centered on the same vertical line.
• Shot C — detail. Block largest, cropped or nearly filling the width.

Mark a single horizontal line across all three rectangles where the block's top edge should land. That line is your anchor. If the top edge stays put while the block grows downward, the sequence reads as a push-in rather than a resize.

Write the planned zoom value for each shot next to its rectangle. Keep the numbers simple — for example 1.0, 1.3, 1.6 — and treat them as a plan, not a guarantee. What matters is that you record what you used so you can repeat it.

Printable shot-plan template

Copy this into a document and print it, or keep it open beside your timeline.

Fill in the distance and zoom columns after you shoot, not before. The plan tells you what to try; the record tells you what actually happened.

Step 2: Mark a fixed distance and change only one variable

Tape a mark on the table for each shot distance. Put the block's front edge on a second mark so it never creeps forward or back.

Then follow one rule: distance sets apparent size, zoom sets framing.

If you zoom in and the product gets bigger than planned, do not zoom back out to "fix" it. Step the camera back instead. That keeps the perspective consistent across the sequence, which is what makes the shots feel related.

A sample prompt for a planning pass

If you are using an AI video workspace to sketch the sequence before shooting, describe the framing constraints rather than the look. This is an example of how to phrase it, not a tested result:

You can build and compare these framing variations in the AI product video generator, then use the same plan when you shoot the real product. Treat any generated clip as a reference for composition, not as proof that your physical setup will match it.

Step 3: Review with a grid overlay

Do not judge size by memory. Put a grid on top of the footage.

1. Drop a 3×3 or 4×4 grid over the timeline so every shot shares the same lines.
2. Scrub to the first frame of each shot and note where the block's top, bottom, and side edges cross the grid.
3. Compare the numbers across shots. A difference of a few percent is usually invisible; a difference of a tenth of the frame height is not.
4. Check the anchor line first, then the block's height, then its horizontal center.
5. Watch the sequence at normal speed once. If your eye catches a jump, trust it over the numbers.

Output review checklist

• The block's top edge sits on the same grid line in all three shots.
• The block's horizontal center is on the same vertical grid line.
• Apparent size changes only in the direction you planned.
• Background elements have not shifted position between shots.
• Light direction and shadow length look consistent.
• No shot has the product touching the frame edge unless that was planned.
• The sequence reads as one clip at normal speed, not three takes.
• Any text or caption area is still clear of the product.

Common mistakes

Zooming to fill the frame on every shot. This is the most common cause of size drift. Fill the frame with distance, then use zoom only for small corrections.

Moving the product instead of the camera. Sliding the block forward changes both size and shadow. Move the camera and leave the block on its mark.

Changing camera height between shots. A few centimeters of height changes the apparent size of a flat object noticeably. Keep the tripod or phone mount at one height.

Judging on a phone screen at arm's length. Small size differences disappear. Review on the largest screen you have, or use the grid.

Re-lighting each shot. A different light angle changes how much of the block's face is visible, which reads as a size change. Lock the light before shot A.

Trusting a generated clip to match your physical setup. Generated footage can help you plan composition, but it does not tell you what your camera will capture. Always verify with a test frame.

FAQ

Should I zoom or move the camera to make the product bigger?

Move the camera if you want the perspective to change naturally. Zoom if you need the same perspective with a tighter crop. Pick one per shot and record which you used.

How much size difference is acceptable between shots?

There is no universal number. Compare the block's height against your grid: if it crosses a different grid line, most viewers will notice. If it stays within the same band, it usually reads as consistent.

Does this work for a real product instead of a wooden block?

Yes. The block is a rehearsal object because its edges are easy to measure. Once the method is familiar, swap in your product and keep the same distance marks and anchor line.

What if my product is reflective or transparent?

Reflections and see-through edges make size harder to judge. Add a small matte reference card beside the product in the same position in every shot, and compare the card's edges on the grid instead.

Can I fix size drift in editing?

You can scale a shot up or down, but that also changes sharpness and background scale. It is faster to reshoot with a marked distance than to correct three shots by hand.

Do I need special equipment?

A tape measure, masking tape, a fixed camera height, and a grid overlay are enough. A tripod helps, but a stack of books under a phone works for a test.

The takeaway

Consistency comes from constraints, not from better zooming. Mark the distance, lock the height, change one variable per shot, and check the result against a grid. Rehearse on a wooden block until the sequence holds together, then apply the same plan to the product you actually need to sell. If you want to compare framing options before you commit to a shoot, start with the AI product video generator; if you are weighing plans for a small team, the pricing page lists what is available.