I screen-recorded a course assignment, about fifteen minutes, and the export came out over 600 MB. The submission portal rejected it for exceeding the size limit; trying to send it to the group chat stalled on the upload too. The work itself was done — stuck at "file too big", which is a maddening place to be stuck.
The fix is simple: shrink the video. But before compressing I had two worries — will it come out too blurry to read, and will it take forever. Both turned out fine, and the key was understanding what "shrinking" actually cuts.
Video compression doesn't shave the picture evenly. What really saves size is allocating by content complexity: fast-moving, detailed segments keep more data, static segments waste less. ImgIng samples the video first — texture, motion and change at the start, middle and end — sets a pre-encode budget from that, and compresses with quality-first variable bitrate. So at a given size the parts that should stay sharp do, instead of the whole thing going soft together. For an assignment video where "you can read it" is the bar, that's plenty.
Concrete scale: that 600 MB fifteen-minute recording came out just over 100 MB as MP4, and the portal's limit was 200 MB — cleared it easily, with the code and UI clicks still perfectly legible. If the limit were tighter, say 50 MB, I'd drop resolution from 1080p to 720p and cut it in half again. How far to compress is set backwards from "where it's going and what that place caps at", not by guessing.
It also has a setting I found reassuring: the result won't be larger than the source. If compressing doesn't actually shrink it, it keeps the original or refuses to save — no "spent ages compressing and it got bigger". I didn't agonise over format either: it defaults to MP4, safest for submission — I don't know what the teacher opens it with, MP4 has the lowest chance of failing, just hand it in without studying format options.
And if it's still too big after that, don't hunt for a more aggressive codec — check whether resolution and frame rate can drop first. A screen recording from 1080p60 to 720p30 sheds a lot of size, and for watching a walkthrough it's plenty clear. Resolution and frame rate are more direct size levers than compression parameters, and assignment videos especially respond to it.
One more thing: the whole compression runs locally, the video never uploaded. That's not just privacy — a few-hundred-MB video takes ages just to upload to some online tool; compressing locally then uploading skips that slowest step, and the gap is bigger on a bad connection. On a deadline that time matters.
So "assignment video too big to submit" is nothing to dread: compress locally, content-awareness keeps the clarity where it belongs, drop resolution and frame rate if needed, keep the format at MP4. Work you finished shouldn't die at the upload step — it's a few minutes to fix.
Top comments (0)