TIMECODE AND METADATA

Only one ex-girlfriend and a few friends ‐ artists, mostly musicians ‐ have ever sent me digital information as a screenshot. A picture.

Very recently, during what was supposed to be professional communication about a television programme, a young woman sent me a phone photo of her monitor. That is what prompted this post.

Or: copy and paste text instead of applying convolution and correlation algorithms to a raster image ‐ if we can manage that as we enter the second quarter of the 21st century.

What is metadata?

Metadata is information about content ‐ text, numbers or other structured data. It may be stored inside the media file, in the project or in a separate sidecar file. Modern software, both DAWs and NLEs, can read and use it to “communicate” with other applications, provided both systems support the relevant fields and interchange format.

Back around 1992, at our youth science and technology club, we called data arranged in rows and columns an “electronic spreadsheet”. Agroplan was popular then; today it is Excel. My subject here is video ‐ specifically, the Time Code column in its metadata.

Timecode: the address of a frame

Timecode was created and used for decades before the digital media boom, as a way to address individual frames and align audio and video in time.

Standardised by SMPTE (the Society of Motion Picture and Television Engineers), timecode ‐ TC for short ‐ is an industry system for labelling frames in hours, minutes, seconds and frames: HH:MM:SS:FF. A timecode address is meaningful together with the frame rate, counting mode and source identifier.

Without going into the details of carrying timecode in an audio/video signal or digital stream, I want to look at some practical uses of this BASIC tool.

This is probably the place to discuss the timebase and frame rate: how many frames are played each second.

For the romantically inclined, here is the short version: it is the magical link between frames, sound, pictures and text that makes studio life and communication easier. Set up correctly, it saves dozens of hours of manual work ‐ often unpaid ‐ and reduces the risk of errors in editing, project transfers and picture/sound synchronisation.

Frame rates, PAL, NTSC and SECAM

Analogue television developed families of systems with different frame and field rates. PAL, NTSC and SECAM are colour encoding systems, not universal definitions of frame rate in themselves. In their most widespread historical implementations, the relationship is as follows:

PAL (Phase Alternating Line): most commonly 625 lines, 50 fields and 25 frames per second. Used across much of Europe and in numerous other regions.

NTSC (National Television System Committee): for the colour system, most commonly 525 lines, approximately 59.94 fields and 29.97 frames per second ‐ exactly 30000/1001 fps, rather than exactly 30 fps. Used in the United States, Japan and other regions.

Bulgaria historically used SECAM (Séquentiel couleur à mémoire), a French colour encoding system, also in a 625/50 version with 25 frames per second, adopted in the USSR as well. Which gave the old regime’s blockheads a convenient way to flood us with Soviet nonsense every Friday. That analogue system is now history.

HD and UHD continue to use rates from the 25/50 and 29.97/59.94 families, among others. PAL and NTSC are often used as informal labels for them, although those terms no longer describe their digital colour encoding.

I will not go into cinema’s 24 fps in detail here. One important clarification: 24 fps and 23.976 fps (24000/1001) are different rates. 23.976 is not “drop-frame”, while 29.97 fps may use either drop-frame or non-drop-frame timecode. Drop-frame skips timecode numbers, not picture frames. No cinema romance here ‐ just markets and distribution. For keycode and feet of film, send me a private message.

Historically, the choice of 50 and 60 fields per second was linked to mains frequency and the limitations of analogue equipment. Europe uses 50 Hz and the United States 60 Hz; Japan has both 50 Hz and 60 Hz regions. Colour NTSC’s fractional rates resulted from compatibility requirements between colour and monochrome signals.

Since many colleagues have never dealt with an analogue monitor or signal, I will leave the finer details for a later post about types of timecode. Here I will continue with metadata’s practical role in collaboration between different NLEs and DAWs.

For television work in Bulgaria, the 25/50 fps family is common. The actual format, however, is determined by the production specification and the recipient’s delivery requirements, not geography alone. The historical EBU Tech 3097 specification covers timecode for 625-line television systems.

How metadata saves work

The concept is simple: different versions of the same media, including different resolutions, can be linked to the corresponding frames through preserved identifiers and timecode. This makes proxy editing and a later conform to the originals possible. Matching timecode alone does not guarantee that two files contain the same picture or sound.

Picture, sound and subtitles ‐ and, with a suitable interchange, colour corrections, CDL and other data ‐ can move between applications. This requires preserving and checking the timebase, start timecode, DF/NDF mode, source names and identifiers, audio parameters and interchange compatibility.

From acquisition to final export, however many departments a project passes through, the goal is to preserve synchronisation and traceability to the sources with as little unnecessary work as possible. Resolution may intentionally change for proxies or different deliverables; what matters is that the change is controlled and checked.

In a professional environment, these basics should go without saying. But this is Bulgaria: a place where you can sometimes leave university armed with nothing but talent ‐ which is not guaranteed either ‐ and without the technical preparation needed to present it to an audience, large or small.

Why is SMPTE timecode so important? Accuracy and repeatability. Within a correctly identified source and a known timebase, we can return to the same frame using its address. Timecode is not globally unique: different files and tapes can carry identical values, which is why reel/tape names and other identifiers matter too.

In my experience, far too many film, sound and motion graphics graduates are unaware of metadata’s significance or how to use it.

A workflow protocol is not a whim

In 2004, I worked on Bulgaria’s first reality show, Big Brother, and everything followed a protocol: timecode, tape name, reel name and so on. We knew the drill ‐ otherwise the tape machines and monitors would not play along. And we still went through training.

Yet in 2025, while helping with screenings at a small film festival, I had to wrestle with media reporting extraordinary frame rates outside any expected delivery standard ‐ something like 33.314662 fps. Such a number may be an average for VFR media or a consequence of how the file is analysed, so the file needs investigation rather than a diagnosis from one number. But for a festival delivery, I expect an agreed, checked format. And this is someone who is supposed to be a colleague?

This “author” has been assessed by a high-level panel and somehow received a mark, which rather calls the assessors’ competence into question. But even that is not the main point.

I genuinely do not understand how academies, universities, colleges and other institutions can flood our market with talent while leaving out basic, hobby-club-level knowledge. I know more engineers for whom this editing “enigma” is second nature, despite having no need for it, than practising editors, colourists and other artists proudly waving diplomas who understand these basic working principles.

It makes no sense for six shooting days abroad to sync in ten minutes, while one shooting day here costs three unbudgeted days of post-production. I do not want to sound sexist, but I notice a recurring pattern in projects led by biologically non-male individuals, regardless of self-identification. And since we are cultural professionals rather than Big Brother contestants, we have no defence against aggressively emancipated ignorance. In the old days it was simple: you were not up to the job, and that was that. Now you have to be careful not to call someone a thick-headed fool, because he is a man. And so is she… ;)

Anyway. My point is that, before starting a job, I would prefer us to know how it works. “I just click around,” someone says. You click around, my friend; I raise children and pay bills. What you call clicking around involves real professions, person-hours, studio time and other costs.

No business in the world can rely on wasted time and call it talent. Here, that has left only the show in show business ‐ and it is no longer funny.

The more skilled you are, the easier it is to make dreams real. That is it. Let viewers, listeners or readers be the judges of talent.

In conclusion

Whenever the software allows it, send timecode as copyable text: eight digits in HH:MM:SS:FF format, or HH:MM:SS;FF in some DF displays. Include the frame rate, DF/NDF mode and clip or version name. A screenshot can provide useful context, but it should not be the only way to access the numbers.

Please, colleagues: do not photograph text with your phone when you can copy and paste it.

And do not choose a working format by its extension alone. MP4, AVI and WMV do not guarantee the metadata or reliable interchange required by your particular applications. MP4 may be perfectly suitable for review or proxy use in a tested workflow. For editing, conforming and mastering, agree on the container, codec and metadata for the job ‐ for example, MOV or MXF with a suitable codec ‐ and test the handover. The extension alone guarantees nothing.

Ruslan Grudev ‐ Rocko

Technical sources: EBU Tech 3097 · SMPTE ‐ Understanding Standards: Time Code