All articles

TikTok Video Quality After Upload: Why Clips Look Blurry

Split illustration: a soft, heavily blurred scene on the left, separated by a vertical line from the same scene rendered crisp and sharp on the right.

Your clips look sharp in the editor and soft on TikTok because the platform re-encodes every file you upload, and it compresses harder when your file does not match its preferred specs. The fix is almost never in the edit, it is in the export. Below are the exact settings to export at, the habit that quietly destroys quality, and a five-point check to run before you publish.

Why a clip loses quality after upload

A finished clip passes through four encodes before a viewer sees it: your source footage, which has to reach you in one piece first, your export from the editing software, the platform's own encode at upload, and the version finally served. You only control two of those steps.

When your export matches the platform's format, its encoder has very little work to do and the loss is invisible. When it does not, a wrong resolution, a frame rate the encoder has to convert, a file that has already been compressed once, the encoder is forced to re-express the image, and it does that by discarding detail. Shadows start to smudge first, then fine texture, then faces.

TikTok re-encodes every upload no matter how clean the source. You cannot switch that off. You can only make sure it starts from a file it barely has to touch. One more thing explains why the same file can look fine on your desktop and bad on a phone: the app on the viewer's device renders the version the platform delivered, not the one you exported.

The export settings that actually matter

Most of the confusion comes from settings that change very little and one that changes everything. Here is the short list, in order of impact.

SettingUse thisWhy it matters
Resolution1080 x 1920Native vertical HD; 720p is accepted but visibly softer
Aspect ratio9:16, set on the canvasCropping later means a second encode
Frame rateConstant, same as your sourceMixed or variable rates cause judder and blur
CodecH.264, high profileThe format every platform reads without conversion
Bitrate8 to 12 Mbps for 1080pYouTube's own upload guidance for 1080p SDR is 8 Mbps; below that there is little left to preserve
AudioAAC, 128 to 192 kbpsClean and universally readable
ColourRec. 709 (SDR)What the platforms expect

Export once at these values and the platform's second encode starts from a clean file. YouTube's recommended upload encoding settings are the reference for the resolution and bitrate rows, and CapCut's own help page documents where resolution and bitrate sit in the export panel. It also pairs with exporting several shorts in one batch, so you render once instead of clip by clip.

If the repetitive part of this is what eats your evening, the clip production pipeline at ClipFinish renders each clip at 1080 x 1920 with burned-in captions, so the file you download is already at the specs above instead of being re-exported by hand.

The habit that costs the most quality

Downloading a clip you already published in order to cut it again, or exporting, cropping, then exporting a second time, compresses the image again and keeps the artefacts of the previous pass. Two generations out is where a clip starts to look muddy on a phone screen.

The rule is simple: edit the original and export once. Never make a published file the source of your next edit.

The same trap hides in the canvas. Exporting a 16:9 timeline and cropping it to vertical afterwards adds an encode and throws away resolution. Set the 9:16 canvas before the first cut; the method is the one in this guide to reframing 16:9 footage into 9:16.

Variable frame rate: the silent quality killer

Phone and screen recordings are usually variable frame rate (VFR): the file does not hold one steady rate, it speeds up and slows down from moment to moment. It looks fine on the device that recorded it. Feed it to a platform encoder and frames get duplicated or dropped to force a constant rate, which shows up as judder and a soft, unstable image.

HandBrake's documentation on frame rate explains the difference between variable and constant frame rate. The practical move is to match your source's rate and force a constant rate on export, so the platform never has to guess. Desktop editors often pass a variable-rate clip through untouched by default, which is why the problem shows up after upload rather than during editing.

What no setting can fix

Three things are out of your hands, and promising otherwise only sets you up to be blamed for them:

  • The platform's own compression. A perfect file still gets re-encoded.
  • A soft source. No export setting adds detail that was never captured.
  • Screen-recorded footage. The compression is already baked into every frame before you cut it, as this piece on cutting one master clip for every platform covers.

That is why the honest advice is to nail the export and stop there, not to keep re-exporting in the hope that a third pass will help. And judge the result on a phone screen, never on the desktop monitor you edited on.

A pre-upload check that takes one minute

Run this before every upload and the quality question stops being a mystery:

  1. Set the canvas to 9:16 (1080 x 1920) before the first cut.
  2. Match the frame rate to the source and keep it constant.
  3. Export once, at 8 to 12 Mbps for 1080p, H.264.
  4. Play the exported file on the phone you will publish from.
  5. Upload the exported file itself, never a download of an earlier post.
Why is my video blurry after uploading it to TikTok?
Because the platform re-encodes every upload. If your file is not already at its preferred resolution, frame rate and bitrate, the encoder has to compress harder and the loss becomes visible.
What is the best export setting for TikTok and Shorts?
1080 x 1920 (9:16), a constant frame rate matching the source, H.264, and a bitrate of 8 to 12 Mbps for 1080p. Export once and upload that file directly.
Does a higher bitrate always mean better quality?
No. Above roughly 20 Mbps for a 1080p vertical clip there is little left to gain, because the platform re-encodes anyway. The bitrate that matters is the one that keeps the encoder from discarding detail: 8 to 12 Mbps.
Can I fix a clip that already looks bad after posting?
Not by re-uploading it, which adds another compression pass. Go back to the original project, export once at the settings above, and publish the new file.

Quality loss after upload is a settings problem, not a talent problem. Export once at 1080 x 1920 in a constant frame rate, start from the original file, and give the platform's encoder as little to do as possible. Do that, and what you saw in the editor is what your viewers see. If the part that eats your evenings is the repetitive one, finding the moments, burning in captions, rendering every vertical clip, the ClipFinish clip pipeline handles that and hands back files that already match these settings.

You pick the moments. ClipFinish does the rest.

Drop your long video, tick the moments in the transcript, and get the whole batch back: framed, captioned, ready to post.

Try for free

5 free minutes of finished clips every month. No card. No watermark.