
An EDL, or Edit Decision List, is a plain-text file that records every cut in a sequence as a pair of timecodes: where each shot starts and ends in the source footage, and where it lands on the finished timeline. No video, no audio, no thumbnails. Just numbers, in a format simple enough to read in a bare text editor and old enough that the machine it was invented for has been in a museum for decades.
That combination, dumb and universal, is why it never fully died. When two systems that agree on nothing else need to hand off a cut, an EDL is usually the one thing both of them can open. It has outlasted the hardware that created it, the company that built that hardware, and several generations of the software that replaced both.
The format traces back to 1971 and a company called CMX Systems, a joint venture between CBS and Memorex, hence the name: CBS plus Memorex plus eXperimental. Their first machine, the CMX 600, ran on a DEC PDP-11 minicomputer wired to modified Memorex disk packs standing in for videotape, and CMX marketed it as a RAVE, a Random Access Video Editor. It was built for offline editing: cut a rough version fast on disk, then hand the tape department a list of exactly which frames to conform for air. That list was the first EDL, and it existed for a mundane reason, disk storage in 1971 was small and expensive, so the finished program still had to be assembled from the original tapes rather than played straight off the editing machine.
The 600 was followed by the CMX 3600, which became the dominant edit controller of the linear-tape era, and its list format became the de facto standard so completely that editors still call any EDL a "CMX3600" regardless of what software actually wrote it. Within roughly eighteen months of a later CMX model shipping, CMX systems reportedly handled the large majority of broadcast videotape editing in the US. A file format born to talk to a room full of 1970s tape machines is still the thing your NLE offers as a default export option in 2026.
An EDL is an ordered list of events, one line per cut, and each line reads like a short recipe. A typical CMX3600 event looks like this:
014 A004C012 V C 09:12:04:11 09:12:09:21 01:00:22:14 01:00:27:24
Event 14, source reel A004C012, video track, a straight cut, pulled from 09:12:04:11 to 09:12:09:21 in the source, and placed at 01:00:22:14 to 01:00:27:24 on the master timeline. That is five seconds and ten frames of one clip, told entirely through numbers. A dissolve gets written as two linked events sharing a matching in point instead of one plain cut; a wipe adds a SMPTE-defined effect number for the transition style; a speed change gets its own event flagged with a motion marker. Nothing about the shot itself, color, framing, performance, is in the file. An EDL describes where to find something, never what it looks like, which is exactly why it is so small and so durable.
Stack a handful of these lines and you have described a real cut without moving a single frame of picture. Three events in a row might read as: open on reel A002C004 for four seconds, cut to A004C012 for five and a half, dissolve over half a second into A006C001 to close. Every NLE that reads CMX3600 turns that same three-line stack into an identical sequence of cuts, because the format leaves no room for interpretation. That predictability, more than anything else, is the whole reason it survived four decades of hardware and software turnover that broke almost everything else from the tape era.
The limits below are not bugs. They are the ceiling of 1970s hardware, carried forward unchanged, and they still bite people who do not know to check for them before an export.
The place EDLs are least optional is the color suite, in a workflow colorists call pre-conforming. The editor exports a single reference movie of the locked cut plus an EDL. The colorist drops the reference file into DaVinci Resolve, imports the EDL, and Resolve reads the timecode pairs and slices the reference into individual clips on a fresh timeline, each one now addressable and gradeable on its own, without the colorist ever touching the original camera files.
Nikolai Waldman, a senior colorist and workflow trainer at the International Colorist Academy with more than three decades in the industry, put it plainly on his own site: "CMX and their EDL format has become a standard we still use today." The workflow he and other working colorists describe solves a real, unglamorous problem: a colorist grading a project remotely, with no practical way to receive the original camera files, or no time to wait for a data drive to arrive, can still conform an accurate cut from nothing but a reference movie and a small text file. That is the entire value proposition of EDL in one sentence: structure, without the weight of media.
Color grading is the most visible use today, but it is not the only one. Broadcast and cable networks have long used EDL-style logs to mark exact commercial-break boundaries for ad-insertion and skip-detection systems, because a timecoded list of in and out points is precisely what those systems need and nothing more. Archivists and rights teams use them the same way: a short text file that proves which frames of a master were used where, without shipping the master itself back and forth. None of these use cases need a single frame of video attached. They just need the numbers to be right, which is the one thing this format has never gotten wrong in fifty years.
Three interchange formats show up in almost every cross-platform handoff, and despite the overlap in what they get used for, they are not interchangeable with each other.
| Format | Carries | Best for | Weak point |
|---|---|---|---|
| EDL (CMX3600) | Cuts, basic dissolves/wipes, source and record timecode, reel names | Widest possible compatibility, color conform | No effects, 1-2 video tracks, 999-event cap, 8-character reel names |
| XML (FCPXML / others) | Multiple tracks, effects, titles, markers, structure, rich metadata | Editorial handoff between modern NLEs | Dialects vary by app and version; not every field survives every import |
| AAF | Timeline structure plus the option to embed or reference actual media and audio | Avid workflows, audio round-trips into Pro Tools | Heavier files, less universal outside Avid-centric pipelines |
Reach for EDL when the receiving system is unknown or ancient and only the cuts matter. Reach for XML or AAF when the edit's detail, effects, layers, or audio, has to survive the trip intact. The Advanced Media Workflow Association, which maintains the AAF specification, still positions it as a format built to carry both metadata and media together, which is exactly what an EDL was never designed to do.
ScriptCut is not an NLE and does not pretend to be one. It lives at the stage before the timeline, where an editor turns a transcript into a paper edit: read the words, highlight the lines worth keeping, drop the fillers, arrange the order, and get a client's sign-off before anyone opens a cutting tool. See text-based editing for how that transcript-first approach actually plays out day to day, and organizing interview footage before you cut for the step right before this one.
Because every selected line in ScriptCut carries a word-level timecode, the export at the end of that process is not a pile of loose clips, it is an already-ordered sequence. Send it as EDL and it opens as a real conform in Resolve, Avid, or an older tape-era system that has no idea what a transcript is. Send it as XML when the destination is Premiere or Final Cut and the extra structure is worth keeping. ScriptCut can also hand off subtitles and audio alongside the cut list, but the EDL itself stays what it has always been: a picture-cut instruction sheet, nothing more. ScriptCut's job ends where the NLE's job begins. It decides the story and the order, then hands off a format the next tool can actually read, the same problem CMX solved with a room of disk packs in 1971, just without the disk packs.
An EDL survives because it asks for almost nothing: a text file, a shared sense of timecode, and a reel name under nine characters. That minimalism is a limitation everywhere except the one place that matters most, when you genuinely do not know what system is going to open your cut next. Use XML or AAF when detail has to survive the handoff. Use EDL when only the cuts do. And do the actual storytelling earlier than either, on the transcript, where changing your mind still only costs a few dragged lines instead of a rebuilt timeline. If you want to see what that discipline looks like end to end, paper cut vs paper edit is the next place to look.
An EDL is a plain-text file listing every cut in a sequence by source and record timecode, letting any compatible system rebuild that exact edit from the original footage. It contains no video or audio, only timecode instructions.
CMX3600 is the specific EDL format that became the industry standard, named after the CMX 3600 edit controller from the 1970s. When editors say "EDL," they almost always mean a CMX3600-formatted file.
Each event line lists an event number, the source reel name, the track type, the edit type (cut, dissolve, or wipe), the source in/out timecode, and the record in/out timecode on the master timeline.
The limit comes from the original CMX hardware and was never expanded as the format carried forward. Long clip or camera-card names get truncated to fit, which is a common source of name collisions during conform.
No. A standard CMX3600 EDL is limited to cuts and basic transitions on one or two video tracks, plus up to four audio channels. Effects, layered timelines, and complex audio mixes do not survive the export.
Use an EDL when you need the widest possible compatibility and only the cuts matter, such as handing a locked cut to a colorist for conform. Use XML or AAF when effects, multiple tracks, or audio detail need to make the trip too.
ScriptCut is not an NLE, but it exports an EDL, along with XML, subtitles, and audio, once a transcript-based paper edit is arranged and approved. The export opens as an already-ordered sequence in Resolve, Premiere, Final Cut, or Avid.