A split screen should compare like with like. If one side is a real test and the other is a generated ideal, it does not demonstrate product performance. Label illustrative concepts and use comparable evidence for claims viewers may act on.
Define the comparison claim
Write what the viewer is meant to conclude: a color option, a workflow difference or an actual measured outcome. For a factual test, document the conditions and use real evidence. Generated footage can explain a concept, but cannot establish measured performance by looking convincing.
In Imagild's Film Studio, plan the story or supporting visuals. If the current interface does not offer split-screen assembly, use an external editor. Do not present an editing operation as a built-in feature merely because the final video contains it.
Make the two sides comparable
A fictional furniture video compares two approved fabric options on the same chair. Use consistent framing and lighting so the viewer can judge the color difference. Do not make one side dimmer, blurrier or more flattering to force a preference.
For a workflow demonstration, record the actual operations and keep elapsed-time claims separate from edited runtime. A shortened recording is not proof that a task completed in that many seconds.
Add labels that explain the evidence
Label each side with what it actually shows. If either side is a generated concept, make that clear in the visible video, not only in a distant description. Avoid calling a generated cleanup “before and after” of a real tested product effect.
Review the complete export at mobile size. Each panel must remain readable, labels should not obscure the subject, and narration should not overstate the conclusion. Keep sources and editing notes so the comparison can be checked later.