Drop frame timecode (DF) skips specific frame numbers to keep the running clock synchronized with real time. Non-drop frame timecode (NDF) counts every frame sequentially but runs 3.6 seconds short per hour of 29.97fps video. For broadcast, use drop frame. For frame-accurate editing, use non-drop frame.
Основные выводы:
| Drop Frame (DF) | Non-Drop Frame (NDF) | |
|---|---|---|
| Notation | HH:MM:SS;FF | HH:MM:SS:FF |
| Real-time accurate? | Да | No (runs 3.6 sec short per hour at 29.97fps) |
| Frame-count accurate? | Нет | Да |
| Лучшее для | Broadcast, streaming delivery | Film editing, post-production |
| Frame rates | 29.97fps, 59.94fps | 23.976fps, 29.97fps, 59.94fps |
| PAL (25fps, 50fps)? | Not applicable | Standard (no drop frame needed) |
Timecode is a precise method for counting and labeling the frames in a recording so you know their exact location. Think of video timecodes like URLs on the web: you rarely notice them while browsing, but you need the exact address to share or reference a specific page. Timecode does the same thing for video frames, and is used to reference and synchronize all types of audio-video files.
SMPTE timecode is the global standard for labeling video and film. Developed in the 1960s by the Общество инженеров кино и телевидения (SMPTE), it enables accurate editing, identification, and synchronization of media. SMPTE timecode is expressed in HH:MM:SS:FF format (hours, minutes, seconds, frames).
Two foundational rules govern how timecode works:
When color television was introduced in 1953, the NTSC standard was modified to accommodate color in existing black-and-white receivers. The frame rate was reduced from 30fps to 29.97fps, creating a small but compounding disparity between real-world clock time and video time.
Because timecode is still numbered as if video runs at 30fps, a 29.97fps recording generates only 107,892 frames per hour instead of the 108,000 frames the counter expects. That gap equals 3.6 seconds per hour. Drop frame timecode was created specifically to close it.
Drop frame timecode matches real time. Non-drop frame timecode does not. The non-drop frame format counts frames accurately but drifts from the clock; drop frame corrects that drift by skipping specific frame numbers (not actual frames) in the count.
To put it simply, each format serves a different primary purpose:
Таймкод с пропуском кадров (DF) устраняет расхождение со скоростью 29,97 кадров в секунду за счёт пропуска определённых номеров кадров, благодаря чему счётчик таймкода остаётся синхронизированным с реальным временем. При этом из записи не удаляются никакие фактические кадры.
The math: at 29.97fps, there are 107,892 frames per hour instead of 108,000. The difference is 108 frame numbers, or 3.6 seconds. Left uncorrected, a real-time hour of 29.97fps video would display on the counter as 01:00:03;18.
The correction follows a specific pattern: the first two frame numbers (00 and 01) are dropped from the count at the start of every minute, except every tenth minute (minutes 00, 10, 20, 30, 40, and 50). Frame numbers are dropped from minutes 01-09, 11-19, 21-29, 31-39, 41-49, and 51-59. By the time one hour elapses, exactly 108 frame numbers have been removed, and the counter reads 01:00:00;00 in real time.
Drop frame timecode only removes numbers from the count. Your recording is completely unaffected.
Non-drop frame timecode (NDF) counts every single video frame without any relabeling. The ratio of timecode to frame count is 1:1, which makes it straightforward and frame-accurate.
The trade-off: at NTSC’s 29.97fps playback rate, an hour-long recording in non-drop frame is not an hour in real time. It is 3.6 seconds shorter, because the counter tallies 3,000 frames per 100 seconds while the actual playback rate delivers only 2,997. A one-hour program using non-drop frame timecode ends at 00:59:56:12 rather than 01:00:00:00.
Neither format alters the visual image. Most modern editing systems handle both, and you can switch between them, though consistency within a project is strongly recommended.
As a general rule:
Not sure which format you are looking at? The notation gives it away. Non-drop frame uses all colons (HH:MM:SS:FF). Drop frame uses a semicolon or period before the frame count (HH:MM:SS;FF or HH:MM:SS.FF).
The 29.97fps case gets most of the attention, but filmmakers and video producers work across several frame rates. Here is how drop frame applies to each.
No drop frame variant exists at 23.976fps. This frame rate divides cleanly enough into real time that the drift is negligible for most production contexts, and the industry has standardized on non-drop frame for cinema and streaming originals. As of 2026, 23.976fps is the most common delivery frame rate for streaming platforms.
This is the primary use case for drop frame, as described above. Both DF and NDF exist at this rate. Broadcast uses drop frame; post-production often uses non-drop frame.
Drop frame exists at 59.94fps and is used for sports and live broadcast. The math scales proportionally from the 29.97fps case: the same pattern of dropping frame numbers applies, just at double the rate.
No drop frame exists for PAL frame rates. Because 25fps divides evenly into the 50Hz electrical standard used in Europe and PAL regions, there is no disparity between frame count and real time. This is a frequent point of confusion: editors who learned timecode on NTSC systems sometimes look for a drop frame option in PAL projects and find none. That is by design.
The setting lives in your sequence or project settings, not in the export dialog. Here is where to find it in the three most common non-linear editors.
If you are not already captioning your video content, it is time to start. Closed captions are a recommended best practice for digital accessibility under the Руководство по доступности веб-контента (WCAG), and subtitled videos consistently earn more engagement in comments, likes, and shares.
When captioning, knowing whether your file is drop frame or non-drop frame is essential. If you caption a drop frame video with non-drop frame captions, the sync will be off from the start and the drift will compound as the video progresses. A one-hour program could end up with captions that are 3.6 seconds out of sync by the final scene.
Sonix handles this automatically. Its автоматическое создание субтитров syncs captions to your video’s timecode format, whether drop frame or non-drop frame, so you can make precise manual adjustments without starting from scratch. For транскрипция видео workflows, Sonix exports to SRT, VTT, and other caption formats that carry the correct timecode through to your NLE.
Sonix is the world’s most advanced automated transcription and subtitling platform. Supporting 53+ languages, it delivers fast, accurate transcripts at an affordable price, and handles timecodes in both drop frame and non-drop frame formats.
The in-browser editor works with multiple file types and integrates with Adobe Premiere Pro, Adobe Audition, and Final Cut Pro for seamless transcript export. Whether your project is a broadcast documentary or a streaming series, Sonix fits into your existing workflow without adding steps.
Попробуйте Sonix бесплатно and get 30 minutes of автоматические субтитры и титры included with your account.
No. Drop frame timecode only skips frame numbers in the count; it does not delete any actual frames from your recording. Your video content is completely unaffected regardless of whether you use drop frame or non-drop frame. The term is misleading: think of it as “drop frame number” rather than “drop frame.”
Точка с запятой (или точка) между секундами и кадрами в отображении таймкода, например 01:00:00;00, указывает на таймкод с пропущенными кадрами. Все двоеточия, например 01:00:00:00, указывают на формат без пропущенных кадров. Это различие в обозначениях — самый быстрый способ проверить, какой формат используется в файле, не открывая настройки проекта.
Да, для большинства рабочих процессов потоковой передачи использование метода «drop frame» является правильным выбором. Потоковые платформы ожидают, что продолжительность контента будет соответствовать реальному времени, а метод «drop frame» гарантирует точность длительности вашей программы. Если исходный материал имеет частоту 29,97 кадров в секунду, передавайте его в режиме «drop frame». Если частота вашего материала составляет 23,976 кадров в секунду, режим «drop frame» не применяется; используйте режим «non-drop frame».
Нет. Пропуск кадров применяется только при частоте 29,97 кадров в секунду и 59,94 кадров в секунду (частоты, производные от стандарта NTSC). При частоте 23,976 кадров в секунду (киноформат) отклонение настолько мало, что коррекция не требуется, и в отрасли по традиции используется метод без пропуска кадров. При частоте 25 кадров в секунду (PAL) частота кадров равномерно делится на электрический стандарт 50 Гц, поэтому нет никакого расхождения, которое нужно было бы корректировать, и опция «drop frame» отсутствует.
If your video is drop frame and your subtitle file uses non-drop frame timecodes (or vice versa), the captions will drift progressively out of sync. The drift starts small but reaches 3.6 seconds by the end of a one-hour program. Always confirm your subtitle export format matches your video’s timecode format. Sonix automatically syncs subtitles to your video’s timecode, eliminating this problem.
Mixing formats within a project creates sync errors that compound over time and can be difficult to diagnose. Most NLEs will flag the mismatch, but some will silently convert, which can introduce subtle timing errors. The safest practice is to set your project’s timecode format before you begin and keep all media consistent. If you inherit mixed-format footage, conform it to a single format during ingest.
Протокол Model Context Protocol меняет подход к подключению ИИ-помощников к внешним инструментам и подкастам…
Судебные стенографисты, ежемесячно проводящие десятки допросов, сталкиваются с новым вопросом: как помощники на базе искусственного интеллекта могут…
Ваш ИИ-помощник — умный. Записи ваших встреч содержат массу полезной информации. Но чтобы извлечь из них…
У вас есть 80 часов видеозаписей интервью, приближающийся крайний срок и ИИ-помощник, который…
Помните, как раньше анализ подкаста сводился к тому, чтобы копировать фрагменты стенограммы в ChatGPT и повторять этот процесс…
Раньше поиск подходящего решения для транскрипции в сфере управления персоналом и подбора кадров означал необходимость одновременно пользоваться несколькими разными инструментами…
На этом сайте используются файлы cookie.