Applies to: Mac or Apple TV output to a display, with independent documented player/computer and display sizing controls.
The useful distinctions
- Apple directs users to both source and TV size settings.
- An available Mac underscan adjustment is display-dependent.
- Diagnostic inference to test: Compensation at two sizing stages can conceal which stage adds a margin.
Map the two boundaries
Distinguish the reference image boundary from the player’s output rectangle and the display’s physical edge. A border may be part of the file, added by a source fit choice or introduced by display sizing. Begin with an owned reference whose boundary is known. A film already containing bars makes this initial map ambiguous.
Avoid compensating for compensation
Changing both stages at once can produce a visually full screen while preserving a hidden crop. Test source sizing with display sizing fixed, restore it, then reverse the roles. Record which margins move in each run. The pattern shows the contribution of each control even when neither adjustment alone produces the final desired framing.
Return to the actual programme
After the boundaries are understood, choose a combination that preserves intended picture area and proportions. Remaining bars in the original programme require a content-aspect check. The experiment cannot label every border a fault or promise that one source control fits every display. Keep the working reference state and output mode written down so later changes can be reversed without guesswork.
A focused way to proceed
- Use an owned full-frame border reference and record all current sizing states.
- Keep display sizing fixed while comparing the source’s documented size control.
- Restore the source state.
- Keep source sizing fixed while comparing the display’s documented fit control.
- Record which border changes in each comparison.
Worked example · synthetic
Hypothetical example: a source underscan adjustment changes all four margins, while a TV zoom choice removes only the outer margins and crops reference corners. Testing them independently exposes two sizing contributions that a combined adjustment had hidden.
When to stop or ask for help
Use documented controls and preserve original values. A visually full screen is insufficient evidence that all original picture area remains.