6. Sonia's very detailed prompt gives a confused result. What trade-off?
Dilution, contradictions and saturation: cut to 5 precise elements then iterate.
1.3 · Application
7. Sonia's grid: no. 2 has the right composition but an approximate cup. What to do?
Execution problem = vary on the good lead, finish with subtle, upscale the keeper.
1.3 · Application
8. All 4 of Sonia's images miss (wrong framing, wrong mood). What to do?
Intent problem = the prompt is at fault: fix the description, do not vary into the void.
1.3 · Arbitration
9. After upscale, hands are deformed and text is unreadable. What trade-off?
Upscale adds details but does not fix structural issues — dedicated retouching required.
1.4 · Recall
10. Why learn on free tools before paying for Midjourney?
Structure, light and iteration learned for free apply as-is on Midjourney.
1.4 · Application
11. Sonia wants to train without paying. Which method?
Same prompt on two free tools: obedience, light and text compared, fundamentals trained.
1.4 · Application
12. Sonia wants to use a free-offer image for a client. What to check?
Free is often more restrictive: T&Cs + licence + content (people, brands) before invoicing.
What's next?
Score ≥ 80%: move to V3 Level 2 (parameters & image references — format, --no, stylize, seed, versions).
Score between 50% and 80%: revisit the missed sections (public by default in 1.1, template and --no in 1.2, execution/intent funnel in 1.3, quotas and T&Cs in 1.4), reread the explanations above, then retake the test.
Score < 50%: restart the level by running each template (/imagine 1.1-1.2, V/U cycle 1.3, free comparison 1.4) — the 1.3 V/U grid is the judge: iterate without rewriting before moving on.