One long recording goes in and a folder of short clips comes out, ready to post. Choose a length, a shape and optionally a logo, and every clip is reshaped and watermarked as it is encoded. The set arrives as a single ZIP. Nothing is uploaded — the work happens on your own device.
Clips are encoded one after another, and each finished clip is held in memory until the ZIP is written. That, rather than the length of the source, is what limits a run: a twenty-minute video at sixty seconds a clip is comfortable, while the same video cut into fifteen-second pieces is not. The number of clips and the estimated size are shown before anything starts, and a run that would be too much is refused with the settings that would fix it.
Getting clips worth posting
-
Blurred fill keeps the whole picture:
turning a widescreen recording into a vertical clip throws away most of the width if you fill and crop. Blurred fill scales the frame to fit and puts a blurred copy of it behind, so nothing is lost and no black bars appear.
-
The logo is drawn onto every frame:
it is part of the picture rather than an overlay, so it survives being reposted. Its size is a share of the frame width, which means the same setting looks right whether the clip is 720 or 1080 wide.
-
A short remainder is dropped:
splitting rarely divides evenly, and a two-second tail at the end is a mistake rather than a clip. Anything under two seconds is discarded rather than handed to you.
-
Stopping keeps what is finished:
the clips already encoded are packed into a ZIP as normal, so a setting you got wrong costs the waiting rather than the work.
What actually limits how many clips you can make
Every clip the [msm_tool slug="video-splitter"] produces is held in memory until the whole set is packed into a ZIP, and that is what caps a run — not the length of the source video. A twenty-minute recording split into sixty-second clips is a comfortable, ordinary job; the same recording cut into fifteen-second pieces produces roughly four times as many clips, and a long or large source hits the memory limit sooner than a short, small one. The clip count and an estimated size are shown before anything runs, and a setting that would not fit is refused with the specific number that would.
Turning widescreen into vertical without losing the subject
Fitting a widescreen recording into a vertical frame has to do something with the extra width, and the two options here handle it differently. Fill and crop scales the footage to fill the vertical frame and cuts off whatever does not fit, which can crop a subject out of the picture if the recording was not shot with a vertical crop in mind. Blurred fill instead scales the whole frame to fit within the vertical space and fills the empty area behind it with a softened, enlarged copy of the same footage — nothing is cropped and nothing is left black, which makes it the safer default for a recording that was not framed vertically to begin with.
The logo is part of the picture, not a layer on top
Rather than compositing as an overlay, the logo is drawn directly onto every frame during encoding, which is why it survives a clip being downloaded and reposted somewhere else. Its size is set as a share of the frame width rather than a fixed pixel count, so the same setting looks right whether the output is 720 or 1080 pixels wide.
Why the last few seconds sometimes disappear
A long recording rarely divides evenly by a fixed clip length, and whatever is left over at the end is usually only a second or two — a remainder that short reads as a mistake rather than a usable clip, so anything under two seconds is dropped rather than handed to you as a final, oddly short file.
Choosing a length
Sixty seconds fits comfortably within Shorts, Reels and TikTok. Thirty suits material aimed at a shorter attention span. Fifteen produces a great many clips from anything longer than a few minutes — usually more than would actually get posted, though it is there for whoever genuinely wants that many.
Where the sound can go missing
MP4 audio needs the AAC codec, which Firefox cannot encode on any platform and no browser can encode on desktop Linux. The page states this before a run starts rather than leaving it to be discovered afterwards. Chrome, Edge and Safari keep the sound in every clip.