Clients skim most of a cut. They follow the story and react to the music. Then a lower third comes up with their CEO’s name and they read every letter, because it’s the one part of the video they know better than anyone in the edit suite.
A missing accent, a title from before last month’s reshuffle or a surname split across two lines gets caught on the first viewing, and each one costs a round.
Lower thirds go wrong in the same few ways, so you can check them the same way every time. This list covers only the name and title graphic. For the rest of the video, see the guides on video quality control and the final QC checklist.
In short
- Copy each name from the person and check the accents in the rendered frame, not the text field.
- Get every job title approved in writing, and shorten one only with the client’s permission.
- Send the client a separate list of names and titles to confirm before the edit.
- Keep a lower third up long enough to read twice, inside title safe and readable on a phone.
Spell the name the way the person spells it
Copy it from the person, not from the interview audio or an old email. Accents and special letters (José, Zoë, Søren, Łukasz) stay when the graphic is set in capitals, so JOSÉ is still JOSÉ.
Particles and internal capitals are part of the name, as in van der Berg, de la Cruz or DeShawn. In a name like O’Neill, use the curly apostrophe, because the straight one looks like a stray tick in many display fonts.
And use the name the person goes by. The badge says Bob, the HR list says Robert, and only he knows which one he wants on screen.
Then check the rendered frame, not the text field. Plenty of display typefaces don’t include ø, ł or ș. When a letter is missing, the graphics software swaps in a fallback font for that one character or draws an empty box, and neither shows up while you’re typing.
Use the job title the client approved
Titles change faster than videos get made. The interview happens in March, the person is promoted in April, and the cut goes out in May with the old title. LinkedIn profiles go stale too.
Ask the client for each title, get a yes in writing and settle the house style at the same time: “VP of Sales” or “Vice President, Sales”, company name on every lower third or only for outside guests. Pick one rule and use it everywhere.
Shorten with permission. “Senior Director of Global Supply Chain Operations, EMEA” won’t fit on one line at a readable size. The client may accept “Supply Chain Director, EMEA”, but that’s their call, not the motion designer’s at 11pm.
Collect names and titles before the edit
The cheapest place to fix a name is before anyone types it. Put every lower third in the script, written exactly as it will appear. Our video script template shows where it goes.
Then send the client a separate list with only the names and titles, one per line, and ask them to confirm each one. People skim a six-page script, but they read a list of ten names carefully and can forward it to the people on it.
On the shoot, ask each interviewee to say and spell their name and title on camera before the first question. It leaves a record in the rushes for the day the list and the person disagree.
From then on the editor copies from that list, never from memory, and episode six of a series spells the name like episode one.
One or two lines, on screen long enough to read twice
Name on the first line, title on the second. Keep first name and surname together, the same rule subtitle guidelines follow for line breaks.
There’s no industry number for how long a lower third stays up, but you can work one out. Netflix’s English subtitle guide caps reading speed at 20 characters per second for adult programs.
“Siobhan O’Neill-Grant / Head of Supply Chain” has 41 characters, so one read takes about two seconds and two reads about four. Aim for two, because the viewer is also listening and looking at the person’s face, and start counting once the animation has settled.
Show it on the person’s first appearance, and again only if they come back after a long stretch in an event recording or a webinar people will skip through.
Bring it in after the shot starts and take it out before the cut, so it doesn’t ride across two shots. In a two-shot, each name goes on its person’s side.
Inside title safe, and readable on a phone
Title safe is where text is guaranteed to stay visible. Action safe is a slightly larger area for anything important that isn’t text.
For HD and 4K in 16:9, SMPTE ST 2046-1 and the EBU’s R 95 put title safe (the EBU calls it graphics safe) at 90% of the frame, a 5% margin on each edge, and action safe at 93%.
Flat screens crop far less than tube TVs did, but phones have rounded corners and social apps put buttons over the bottom of the picture, and both eat into the edges.
Check every format you deliver. A 9:16 version cut from the center of a 16:9 frame keeps less than a third of its width, and a lower third that starts at the left title-safe line disappears.
Burned-in subtitles live in the same strip at the bottom, so one of the two moves up while the lower third is on screen.
Contrast has to hold for the whole shot. White text over a white shirt or a bright sky disappears, and a handheld camera moves the background under the letters.
A plate or a soft shadow fixes most cases. The web accessibility guidelines (WCAG) ask for at least 4.5:1 contrast for normal text, a fair target here too.
Size is where graphics designed on a big monitor fall apart. A phone held upright shows a 16:9 video about 390 points wide, so a 1920-pixel frame shrinks to roughly a fifth.
Apple’s interface guidelines set the minimum text size on iPhone at 11 points, about 54 pixels in a 1080p comp. A title line set at 40 pixels looks elegant on a 27-inch monitor and reaches the phone at about 8 points. Watch the export on a phone before you send it.
The checklist before the cut goes out
- Every name matches the confirmed list, accents and capitals included.
- Every title is the approved one, in the client’s style.
- Accented letters render in the right typeface.
- One or two lines, first name and surname together.
- On screen long enough to read twice after the animation settles.
- First appearance only, never across a cut.
- Inside title safe in every format you deliver.
- Clear of burned-in subtitles.
- Readable over the actual footage, on a phone.
- Same template, position and timing across the series.
How this works in RecReview
In the RecReview player, the G key draws title safe at 90% and action safe at 93% over the picture.
The Guides button opens a panel with those areas, a rule of thirds grid and delivery crops for 9:16, 4:5, 1:1 and 16:9, one at a time, so you can see what’s left of a lower third in the vertical version before anyone exports it. Clients on the review link get the same guides.
When the client finds a wrong title, they click the moment, circle the lower third and type the note, which reaches the editor on that frame. Ask them to put the correct text in quotes.
AI Review adds one more layer. It reads the text on screen and flags clear spelling mistakes, names included, text cut off at the edge or too faint to read, and lower thirds of the same kind set in different fonts.
Once the client has commented, later versions are checked against their requests. Where they asked for a text change, AI Review reads the new version at that point, marks the request done, partial or not done and shows what it read.
Quoted text gives it something exact to look for. Every finding has a timecode, and the editor decides what to change.
AI Review doesn’t know a person’s correct job title, so a title that is spelled right but out of date still gets through. That one only the client’s confirmed list catches.
