Lower Thirds Explained: What a Lower Third Is and How to Design One
The anatomy, the three states, the safe areas, the readability rules, and the difference between a lower third somebody types and one that builds itself from match data.
A lower third is an on-screen graphic that appears in the lower portion of the video frame to identify who or what is on camera. It normally carries a name, a role or title, and sometimes one line of extra context. It animates in, holds long enough to be read, then animates out. After the score bug, it is the most-used graphic in live sport.
Everything else follows from that job: identify, quickly, without getting in the way.
Why it is called a lower third
Split the frame into three horizontal bands. The graphic lives in the bottom one. That is the whole derivation.
The name describes position, not size. Almost no modern lower third fills a third of the frame height. Most sit in a strip between a tenth and a fifth of it, anchored low and to one side.
Other names, depending on where you learned the job:
- Chyron. American, from the Chyron Corporation, whose character generators were standard in US newsrooms. Genericised, like hoover.
- Aston. British, from Aston Broadcast Systems, used the same way.
- Super, name super or CG, after the character generator that used to make them.
- Name bar, player bar or ident. Common in sport production.
All the same graphic. If one person asks for an aston and another for a name super, build the same thing.
The anatomy of a lower third
A lower third is a set of slots, not one object. What separates an amateur one from a broadcast one is that every slot has a defined behaviour when its content is missing, long or accented.
| Element | What it carries | What goes wrong |
|---|---|---|
| Primary line | The name. Largest type, highest contrast, the reason the graphic exists. | Sized to a short name, then broken by a long one or by diacritics. |
| Secondary line | Role, position, club or title. Smaller and usually tinted rather than white. | Treated as somewhere to put a sentence. |
| Brand mark | Competition, club or broadcaster logo. | Repositioned per instance, so it jitters between graphics. |
| Accent | A colour bar or edge keyed to the team or the competition. | Baked into the artwork, so every club needs its own file. |
| Backing plate | The shape behind the text. It is what makes text legible over grass, crowd or sky. | Set to a light tint that looks refined on a black edit background and vanishes over a bright pitch. |
| Sponsor slot | A defined region with its own clear space and minimum size. | Added late, on top of a layout with no room for it. |
Design against the worst case in your competition, not the example in front of you: the longest player name, the club with two sponsors, the substitute with no listed position.
The three states: in, hold, out
Every lower third has three states, and they are not equally important.
In. The reveal, typically a few tenths of a second. Direction matters more than duration: a package where the score bug wipes from the left and the lower third flies in from the right reads as two designs stapled together.
Hold. The read, and the only state with a real decision in it. It must be long enough to read the longest line twice at a comfortable pace, because the first read is often interrupted by the action. For a name plus a role that usually lands between four and seven seconds. Shorter, and anyone who glanced away has lost it. Longer, and it risks still being on screen after the shot has cut away.
Out. Usually faster than the entrance, and rarely a straight reverse of it. Nobody needs to watch a graphic leave.
Two rules save more broadcasts than any amount of motion polish. A lower third must never be interrupted by another one, so queue them a second apart. And the content must not change mid-hold: a name that swaps under the viewer looks like a mistake even when it is a correction.
Title safe, action safe, and the modern version of the problem
Safe areas exist because the edge of the frame was never reliably visible. The older SMPTE convention places action safe about 5 per cent in from each edge and title safe about 10 per cent. EBU R95 for 16:9 HD is tighter: 3.5 per cent action safe, 5 per cent graphics safe.
The original reason has largely gone, since flat panels do not overscan the way tubes did. The modern reason is worse: the platform draws its own interface over your picture. Player controls, video titles, chat and automatic captions all live in the bottom of the frame.
- Keep the whole graphic within 5 per cent of every frame edge.
- Expect the bottom 8 to 12 per cent of frame height to be covered intermittently by player chrome and captions. Sit above that band, not in it.
- Anchor to a percentage of frame height, never a pixel value, so the layout survives rescaling.
- Build the vertical variant, do not crop the horizontal one. Cropping 16:9 to 9:16 removes most of the width and takes the ends of the lower third with it.
Readability at streaming bitrates and on a phone
A lower third that looks immaculate in the design tool can be unreadable by the time it reaches a viewer. Two causes: compression and screen size.
Compression. Encoders spend their bit budget on motion. A grass pitch, a moving camera and a crowd is close to the worst case, and the fine detail in your graphic is what gets discarded first: hairline strokes, tight letter spacing, light serifs, soft drop shadows. Use a medium or semibold weight, give the text a backing plate with real opacity rather than a 20 per cent tint, and be careful with saturated red on dark, because 4:2:0 chroma subsampling stores colour at a quarter of the luminance resolution and strong reds fringe first.
Screen size. Most live sport outside the top tier is watched on a phone. Size type as a percentage of frame height, not in pixels, because the pixel value stops meaning anything once the stream is rescaled. A working floor is around 3 per cent of frame height for the primary line and 2 per cent for the secondary, roughly 32 and 22 pixels at 1080. Check it on a phone, not on the edit monitor.
Contrast. Aim for a luminance difference between text and plate, not only a hue difference. Compression erodes colour before brightness, and near-white text on a dark plate survives more abuse than any clever tint pairing.
When to use a lower third, and when not to
Use one when a person appears for the first time, when somebody speaks who has not been identified in this segment, to attribute a statistic, to name a location, or to credit a contributor. Do not use one when:
- The same person was identified thirty seconds ago. Once per segment is enough.
- Play is live and the viewer needs the whole frame. Identify in breaks and after replays.
- The information needs more than two lines. That is a full-frame graphic.
- The shot has cut away from the subject. Naming somebody no longer on screen is worse than no graphic.
- The score bug already carries the same information.
Lower thirds for sports
Sport uses lower thirds harder than most genres: dozens of people to identify, and a constant supply of events worth attributing to one of them.
| Use | What it shows | When it fires |
|---|---|---|
| Player introduction | Name, squad number, position, club. | Pre-match in line-up order, or on a first close shot. |
| Scorer | Scorer name and event time, sometimes a season tally. | On the scoring event, after the replay rather than over live play. |
| Substitution | Two names with on and off indicators and shirt numbers. | On the substitution event. |
| Card, foul or penalty | Name, the offence where known, and the sanction. | Once the official has signalled it. |
| Interview | Name and role of the speaker. | A second or two into the first answer, not over the question. |
| Commentator or presenter ident | Name and role. | Once, early, and again after a long break. |
| Statistic attribution | One player, one number, one label. Never a table. | In a stoppage or an interval, never during play. |
| Presented-by variant | The same graphic with a sponsor lock-up in its slot. | Wherever the rights deal says, which is why the slot must exist from the start. |
Data-bound lower thirds versus typed ones
This is the structural distinction, and it matters more than any design choice above.
A typed lower third is a template with empty text fields. A person types the name in before it goes to air. Fine for one match. Across a season it fails three ways. Spelling and accented characters drift, because they depend on whoever is at the keyboard. The graphic is always late, because the operator has to work out who scored before they can start typing. And coverage is bounded by staffing, so any fixture without an operator gets nothing.
A data-bound lower third has no fields to type into. Every element is bound to a field in the match data: the name from the competition registration list, the crest from the club record, the accent colour from the club theme, the event time from the match clock. It is a function of the data rather than a document somebody filled in. Five things change.
- Correctness becomes structural. If a player is spelled correctly in the competition database, they are spelled correctly in every graphic for the rest of the season, on every feed. One correction propagates.
- The graphic can trigger itself. A goal event already contains the scorer and the time, so the scorer lower third builds and fires from that event with nobody pressing anything.
- You have to design a no-data state. The part most teams miss. Real data has holes: a substitute with no photo, a player with no listed position, an event with no attributed scorer. A bound template must define what happens when a field is empty, and the answer is almost always to collapse that element rather than print a placeholder.
- Layout has to be elastic. A typed lower third is tuned to the name in front of you. A bound one must hold every name in the competition. Reserve space for the worst case or auto-fit with a minimum size floor, and test against the real squad list rather than sample text.
- Theming becomes a variable. Because colour, crest and typography are fields, one design serves every club. That is the difference between designing a graphic once and redrawing it forty times.
How LIGR builds a data-bound lower third
The designs are authored in Rive, which matters for one reason: in, hold and out are states in a state machine rather than three clips somebody has to sequence. A graphic can be told to enter, hold while its data is live, and exit on command.
Fuse is where those designs are bound to live data and published. Each text field, colour and image slot is connected to a match data field and themed per competition and per club, with no code. One package then covers every fixture and every team.
The data comes from an official feed where the competition has one, and from the LiveScore app where it does not, which below the top tier is nearly everywhere. That app is why the rest works: the scorer presses buttons for real match events rather than nudging a number, so a goal arrives as a scorer, a team and a timestamp. Game Plans decides what fires and when, and the Control Room covers the moments a human should be in the loop.
The timestamp is the quiet payoff: the event that fires the scorer lower third also marks that point in the recording, which is what lets a highlight package assemble itself from graphics you were making anyway. The wider platform picture is on the broadcast graphics software page, and the companion piece is the guide to sports scoreboard graphics.
Frequently asked questions
What is a lower third?
A lower third is an on-screen graphic in the lower portion of the video frame that identifies who or what is on camera. It usually carries a name and a role or title, sometimes with a brand mark and a colour accent. It animates on, holds for a few seconds while the viewer reads it, then animates off. In live sport it is used for player introductions, scorers, substitutions, cards and interviews.
Why is it called a lower third?
Because it sits in the bottom horizontal third of the frame. The name describes position, not size. Most modern lower thirds occupy far less than a third of the frame height, usually a strip between a tenth and a fifth of it, anchored low and to one side. A graphic that actually filled the lower third of a live sport picture would be covering the field of play.
How long should a lower third stay on screen?
Long enough for a viewer to read the longest line twice at a comfortable pace, because the first read is often interrupted by the action. For a name plus a role that is usually between four and seven seconds of hold, excluding the animation in and out. Shorter and anyone who glanced away has lost it. Much longer and it risks still being on screen after the shot has cut away from the person it names.
What is the difference between a lower third and a chyron?
Nothing, in practice. Chyron is a genericised trade name, from the Chyron Corporation whose character generators were standard in American newsrooms. British crews say aston, from Aston Broadcast Systems, for the same reason. You will also hear super, name super, CG and name bar. They all describe the same class of graphic, so if one person asks for an aston and another asks for a lower third, build the same thing.
What size should lower third text be?
Size it as a percentage of frame height rather than in pixels, because the pixel value stops meaning anything once the stream is rescaled. A useful working floor is about 3 per cent of frame height for the primary name and about 2 per cent for the secondary line, roughly 32 and 22 pixels on a 1080-line frame. Then check it on a phone at arm's length, since that is where most live sport is actually watched.
Where is the safe area for a lower third?
Keep the whole graphic within 5 per cent of every frame edge. The older SMPTE convention put action safe at about 5 per cent and title safe at about 10 per cent, while EBU R95 for 16:9 HD is tighter at 3.5 per cent action safe and 5 per cent graphics safe. The modern constraint is not overscan but platform interface: player controls, video titles and automatic captions cover the bottom 8 to 12 per cent of frame, so sit the lower third above that band.
What is a data-bound lower third?
A data-bound lower third has no text fields for anyone to type into. Each element is connected to a field in the live match data: the name from the competition registration list, the crest from the club record, the accent colour from the club theme, the event time from the match clock. Spelling is then correct everywhere at once, the graphic can trigger itself from a match event, and one design serves every club instead of being redrawn per team.
Do lower thirds work in vertical video?
Only if you build a vertical version. Cropping a 16:9 frame to 9:16 removes most of the width, which takes the ends of a horizontal lower third with it, and what survives is usually too large relative to the frame. Build the 9:16 variant as its own layout, normally stacked and narrower, and bind it to the same data so both versions stay correct without being maintained separately.
Lower thirds that build themselves
Broadcast-grade lower thirds, score bugs and sponsor graphics, bound to your match data and fired automatically across every fixture in the season.