Skip to main content
Participant
September 12, 2026

Wrong spelling output - jumbled letters. Generation was perfect except the spelling was wrong. Veo 3.1, 16:9. It used 400 credits, not usable video,

  • September 12, 2026
  • 1 reply
  • 7 views

this was the prompt.

Use the reference images to generate a 14 seconds video. Begin with a slow, subtle push-in on the central cross-section image of skin and hair follicle, rendered in soft clinical lighting against a dark teal, softly blurred medical background with faint floating light particles. Hold on the center for 1–2 seconds to establish it as the anchor.

Then, in clockwise order starting from the top-left, animate each surrounding circular panel into view one at a time with a short delay between each (roughly 0.8–1 second apart):

Thyroid disorders (top-left) — fades and scales in from a soft glow, with its connecting line drawing outward from the center follicle to the panel, label box sliding in beneath it.
Iron deficiency (top-right) — same style: panel scales up from a pulse of light, connector line animates from center outward, label appears.
Hormonal changes (right) — panel materializes, connecting line traces from the follicle, label box fades in.
Medications (bottom-right) — panel appears with a gentle glow burst, line draws to center, label slides in.
Scalp infection (bottom-left) — panel scales in, connector animates, label appears.
Physical or emotional stress (left) — final panel appears, completing the circle, connector line draws in, label fades in.

Each connecting line should animate as a glowing tracer moving from the central follicle outward to its panel (not the reverse), synced with the panel's appearance. Add a subtle pulsing highlight at each of the six anchor points on the follicle cross-section as its corresponding line begins to draw, to visually tie cause back to the hair structure.

End with a 2-second hold on the fully assembled diagram, all six panels and labels visible, before a slow fade-out or gentle overall zoom-out.

 

BTW, the images did not follow the reference image. The movement was spot on. Use 400 credits and this is not usable output. Please help! Attached is the original image and output with garbled spelling

 

    1 reply

    N8.
    Community Expert
    September 14, 2026

    Hey ​@Lorenzo33770401irte 

    this is definitely frustrating. AI image generators simply suck at spelling. AI Image Generators are designed for images and videos.  LLMs are designed to create text. 

    For consistent spelling, especially with complex images and text, I would use:
    Photoshop, Illustrator, or Adobe Express

    Adobe Express is a free app, with paid options, you can animate, change the font size, colour, and many more options.

    How to Add Text In Adobe Express

    Cheers

    Nate