Broadcast outputs: DropFrame, NonDropFrame, and Offset Time Codes

Closed Captioning with CaptionSync allows you to request multiple broadcast formats. This article shows how to select the appropriate timecode flavor for broadcast formats.

One of the most confusing topics in video is time codes. Basically, there are three ways to reference a particular time in a video file:

• Real time timecode
• Drop Frame timecode
• Non Drop Frame timecode

Real time is exactly what is sounds like - an absolute measure of time in a video file. It can be expressed as HH:MM:SS:FF (hours, minutes, seconds, frames), or as HH:MM:SS.SSS where seconds is shown as a decimal number), or in similar forms.

Web video content is typically referenced in Real time timecode. So if you look in the caption files for web video players (e.g.: Windows Media, Flash, QuickTime, etc), you will see timecodes that are expressed in real time format.

For Broadcast video, the situation is more complex. In North America, broadcast video is typically represented as having 30 frames per second. But (due to changes made when color television was introduced), it is actually broadcast at 29.97 frames per second. The timecode formats are numbered as if there were 30 frames per second (ie: ff is a number from 0 to 29), so some adjustments are necessary to accurately map to real time. Drop and Non-Drop timecodes are both methods of counting video frames.

DropFrame timecode counts each video frame but when the slight difference between the 30 fps representation that it uses and the 29.97 fps that is actually broadcast adds up to a video frame, it skips ahead (or drops) a frame number. It does not drop a video frame, it just skips a number and continues counting.

NonDropFrame timecode counts every single video frame and does not make any allowances for the fact that it is really broadcast at 29.97 fps. This means that 108,000 frames in NonDrop (which would be 1 hour at 30 fps) is really 1:00:03.6 in real time. Yes, 01:00:00:00 NDF == 01:00:03.6 in real time.

Keep in mind that for a given hour of video there are the same number of frames in both DropFrame and NonDropFrame, there are just two different methods to label each of the 107,892 frames. 

For the most part, you do not have to know much of the detail behind this to use the CaptionSync system, but there are a few important points to keep in mind if you use the Broadcast output types:

• For the DVD outputs (.SCC, .STL, .ADB.TXT), CaptionSync will generate both a DropFrame and NonDropFrame version of the result for you. The more common DF output will be the normal extension (e.g., .scc), with the NDF output will contain NDF in the extension (e.g., .ndf.scc).
• If you need to use these outputs, ensure you use the variant that matches your media file. That is, if you are using DF media, make sure you use the DF caption files. DropFrame is more widely used than NDF, so if you don't know what you have, DF is the better bet.
• If you are using an Offset timecode when you submit your request, understand that the Offset will be interpreted as a DF timecode when producing DF results, and as a NDF timecode when producing NDF results. This means that in absolute terms, the offset timecode will have a different real time interpretation for each of the two (DF and NDF) output types.