Closed Captioning Best Practices

Guidelines for Closed Captioning

The following guidelines will ensure that your closed captioned files will merge and display properly. Closed Captioning can be supplied either as embedded data / VBI (SD) or as a separate SCC caption file. Merging SCC files with media files are subject to an additional per spot merge fee and will incur a small delay in processing time.

If a spot is closed captioned, it must be noted in the platform metadata.


00:00:00:00 Time code must start at zero.
00:00:07:05* (or later) First worded “paint-on” caption
(should not be more than one line of text).
00:00:08:00* (or later) First worded “pop-on” caption, one line only
(needs a minimum of :01 second to display correctly).

*Specific start time will depend on the length of the caption. Longer caption lines must start later.


The Last Caption Must End Early

The final caption’s display time must not exceed the duration of content. At the very minimum, it should end 5 frames before the end of the content. There must be a final out-time code at the end of the file (rather than leaving the last caption with just an in-time).

General Requirements

  • Captions should always have at least one second of display time to avoid timing errors that cause improper display.
  • First worded caption should be one line only, if possible. If two lines are necessary, make sure to give enough time for it to build and display.
  • Allow a minimum of 1.5 seconds duration for 2-line captions.
  • Allow a minimum of 2 seconds duration to display 3-line captions.
  • Use sentence case text for all captions. Uppercase text is considered SHOUTING!
  • Do not use bold text in captions, some decoders will display this as blinking text.
  • Italics are often reserved for voiceover instead of identifying “Announcer:” in the captions.
  • Avoid covering existing supers and logos with captions.
  • So your text can be read quickly, break long sentences up into multiple captions according to grammatical breaks and natural phrasing.
  • If it’s not obvious who is speaking, a speaker identification should be used.
  • Description of sound effects [ex. “(Thunderclap), (Roaring), (Birds chirping)”] is not mandatory, but should be used when characters are reacting to it and it becomes part of the “plot.”

Common Reasons Why Captions Are Rejected

  • Supers, logos or legal text are covered by captions
  • Spelling and grammatical errors
  • Timing issues with first or last captions, typically causing a failure to decode
  • Lagging captions or captions not in sync by more than two seconds
  • Captions do not appear
  • Missing words in captions
  • Captions don’t match audio content
  • Transparent background supplied instead of an opaque background

Technical Requirements

  • CEA-608 is required (CEA-708 is optional). Assets without 708 will automatically derive 708 from 608.
  • The first caption should be “paint-on” and occur on or after the fifth frame of content or there must be a one-second delay if the first caption is “pop-on”.
  • Files must have a frame rate of 23.976 fps, 29.97 fps or 59.94 fps.
  • For MPEG files, CEA-608 data must be compliant and present (SMPTE 436M for MXF file types and QuickTime closed captioning track for MOV file types).
  • SD formats with embedded EIA-608 closed captioning data and QuickTime closed captioning track for MOV files are accepted. Line 21 captions are only accepted on 720x486 and 720x512 formats.
  • If a spot is closed captioned, it must be indicated in the metadata and should be included on the slate.
  • All spots intended for broadcast in Canada must contain closed captioning.

“Pop-on” captions may require up to a full second to pre-load into broadcast decoders before the caption can be displayed. For this reason, “pop-on” captions that start too early in the spot may fail to display when the pre-roll frames are removed during broadcast. Using “paint-on” for the first caption allows that caption to start earlier, as the caption data will stream to the decoder without the caching delay of “pop-on” captions.

Recommendations for Embedding Captions in Adobe Premiere

  • Always use “paint-on” for the first line. The blank trigger / “pop-on” method is problematic and does not provide consistent results. The “pop-on” display mode type should never be used for the first line; however, the “pop-on” type is fine to use for every subsequent caption block.
  • Caption text must be aligned center. Caption blocks must be positioned center. Positioning may be top, center or bottom, but always center-justified.
  • Make sure both 608 and 708 tracks share the same caption type (“pop-on” / “paint-on”) when choosing to submit both tracks on the finished asset.
  • Do not exceed 26 characters per line.
  • Set font to Arial.
  • Use an opaque background.
  • Save the Track Style once created for all projects.
  • Adobe Premiere only supports exporting 608 caption data for MOV files. 
  • MXF OP1a is the only format that supports both 608 and 708 closed captions.

Additional Resources


CAB Closed Captioning Standards and Protocol Documents

Caption vendors are responsible for quality control of the files they submit, so please ensure the above guidelines are adhered to. Failure to follow these guidelines may result in files being rejected or processed slowly by the Extreme Reach platform.

Some types of errors may be corrected by Extreme Reach for an additional cost. Please ask your Video Account Manager for details.


Any questions?

Contact us at or 800.324.5672 for a more personal discussion regarding your workflow and needs.


Items in this section

Was this article helpful?
0 out of 0 found this helpful

Quick links