Why your screen recording looks blurry, and how to fix it
Soft text in a screen recording comes from one of four places: the capture size, the export size, the bitrate, or the upload. Here is how to tell which, and what to change.
You recorded a demo, played it back, and the text is soft. Not broken, not blocky, just slightly out of focus in a way the app on your screen never is. It is the commonest complaint about screen recordings and it has four causes, each with its own tell and its own fix. Work through them in order, because the first one is the usual culprit and the last one is out of your hands.
1. It was captured at half the pixels
Open the file and look at its dimensions. On a Mac, select it in Finder and press Cmd+I, or open it in QuickTime Player and press Cmd+I there.
Now compare that with your display. System Settings shows a resolution like 1512 by 982 or 1440 by 900. That is the size in points, the unit macOS lays the interface out in. A Retina display has two pixels per point in each direction, so the panel actually has 3024 by 1964 or 2880 by 1800 of them, and the interface is drawn with all of that detail. The Retina and HiDPI entry has the longer version.
If the file is the point size, the recorder captured one pixel per point and threw three quarters of the detail away. Text drawn that way and played back on a Retina display is soft everywhere, evenly, all the time. That is the tell: the stillest part of the frame is as blurry as the busiest.
The fix is a recorder that captures at the display's full pixel size, or a setting on the one you have. Some call it "Retina", some "HiDPI", some "capture at 2x". The file should come out at 2880 by 1800 or whatever your panel is. Then, if you want a smaller file, scale it down on export: averaging four real pixels into one is sharper than capturing one.
2. It was exported smaller than it was recorded
The recording is fine and the export is soft. This happens when a 2880 by 1800 capture goes out as 720p, or when an editor's default export preset is smaller than the source and nobody looked.
The tell is the same even softness as above, in the exported file but not in the original. The check is the same too: compare the two files' dimensions.
The fix is to export at the source size, or at 1080p at the least. 1080p holds text well at normal reading sizes; 720p does not. The resolution entry goes through what to choose for where the file is going.
3. The bitrate is too low
This one looks different. The still parts of the frame are sharp. What goes wrong is around whatever moved: a scroll turns to mush for a moment and then snaps back, text goes soft while the pointer crosses it, a window drag leaves smears. That is the encoder running out of bits for the frames with the most change in them.
Bitrate is how much data each second gets. Screen content needs less than camera footage, because most of each frame is identical to the last, but it still needs enough for the moments when the whole frame changes at once. For 1080p at 60 fps, somewhere around 8 to 12 Mbps keeps text clean through a scroll.
The fix is a higher bitrate on export, or a quality setting a notch up if that is what the exporter exposes. The file gets larger in proportion. If the recorder does not expose it at all, exporting at a larger resolution usually raises it, because most encoders scale the bitrate with the frame size.
4. The platform re-encoded it
The file is perfect on your Mac and soft on YouTube, or in Slack, or on LinkedIn. Every platform re-encodes what you upload at its own bitrate, and that bitrate is chosen for its bandwidth bill, not your text. There is no setting that stops it.
What you can do is give it more to work with. Platforms put larger uploads in higher bitrate tiers: a 4K upload to YouTube is served at a considerably higher bitrate at every size than a 1080p upload is, including the 1080p version most viewers watch. Scaling a 1080p recording up to 4K before uploading is a real technique and it works, for that reason. A recording captured at full Retina size is already most of the way there.
For a chat or an email, where the file is played as sent, none of this applies: a good 1080p H.264 file arrives as it left.
The short version
Check the file's dimensions against the display's pixel size. If it is half, the capture is the problem and nothing downstream will fix it. If the export is smaller than the recording, export larger. If only the moving parts smear, raise the bitrate. If it was fine until it was uploaded, upload bigger.
Prequel records at the display's full pixel size, exports at that size or scaled to 1080p, 720p or 480p, and writes H.264 or HEVC at a bitrate that keeps text clean. The export page has the dialog.
Frequently asked questions
- Why is text blurry in my screen recording?
- Usually because the recording was captured or exported at fewer pixels than the display has. A Retina Mac draws the interface at twice the pixel size it reports, and a recorder that captures at the reported size records half the detail. Check the file's dimensions: a 1440 by 900 display should produce a 2880 by 1800 recording.
- Does a higher bitrate fix blurry video?
- Only if the bitrate was the problem. A recording that is already sharp but breaks into blocks around moving parts needs more bitrate. A recording where the text is soft everywhere was captured or scaled too small, and no bitrate recovers detail that was never recorded.
- Why does my video look fine on my Mac but blurry on YouTube?
- YouTube re-encodes every upload at its own bitrate, and for 1080p that bitrate is low. Uploading at 4K, even from a 1080p source scaled up, puts the video in a higher-bitrate tier, and the 1080p version YouTube serves from it comes out cleaner.