Is Your Video's Bitrate Too Low? How to Tell Without Guessing

Someone tells you a video is “5 Mbps”. Is that good?

You cannot answer that. Not because you lack experience — because the question is incomplete. 5 Mbps is generous for 720p at 25fps and thin for 4K at 60. The same number describes a comfortable file and a starved one depending on facts nobody mentioned.

This is the single most common confusion about video quality, and it has a clean fix.

One number that works for everything

Divide the bitrate by the number of pixels it has to carry per second:

bits per pixel per frame = bitrate ÷ (width × height × fps)

That is it. Now the number means the same thing everywhere, because the thing it is measuring is how many bits each pixel actually gets.

Run the published recommendations through it and they collapse onto roughly the same value:

So around 0.10 is “as recommended”, whatever the resolution. Below 0.05 the file is carrying about half the data its dimensions imply. Below 0.025 it is carrying a quarter, and you can usually see it: blocking in the shadows, smeared detail wherever the picture moves.

Worked example. A 1080p30 export at 1.5 Mbps:

1,500,000 ÷ (1920 × 1080 × 30) = 0.024

A quarter of the recommendation. The file will look fine on a phone and fall apart on a monitor.

Why this matters more than it used to

Every platform re-encodes what you upload. That is not a conspiracy, it is bandwidth arithmetic — they cannot serve your master file to everyone.

What it means practically is that your upload is the input to someone else’s compressor. A file that is already thin gets compressed again from an already-degraded starting point, and the second pass has less to work with than the first. Detail you gave away at export cannot be recovered at any later stage.

Uploading well above the recommended rate is not wasteful. It is the only part of the chain you control.

The other thing to check: is the frame rate real?

Here is a failure that hides better than bitrate does.

A file can say 60fps in its metadata and carry 30 distinct pictures, each shown twice. This happens constantly:

• footage shot at 24 or 30 and conformed to a 60fps timeline
• screen recordings that captured fewer frames than they claimed
• “smoothing” or frame-rate conversion that duplicated rather than interpolated
• exports where the project frame rate did not match the source

Nothing in the file’s properties reveals it. The container says 60. The player says 60. The only way to see it is to compare consecutive frames and count how many are identical to the one before.

It matters because you are paying for those duplicates. At 60fps with half the frames repeated, a meaningful share of your bitrate is spent encoding “the same as last time” instead of new detail — and you would have been better off exporting honestly at 30 and giving every frame twice the data.

What a low bitrate actually looks like

If you want to confirm by eye rather than by arithmetic, three signatures, in the order they appear as the rate drops:

1. Shadows go blotchy first. Dark gradients are expensive to encode and the first thing an encoder gives up on. Look at a night scene or a dark corner and you will see patches rather than a smooth ramp.
2. Motion smears. Pause during a pan. Fine texture — foliage, fabric, hair — turns into shifting mush that reassembles when the camera stops.
3. Then blocking, everywhere. Once you can see 8-pixel squares at normal viewing size, the file is far past thin.

The first two show up well before anyone would call the video “bad”, which is exactly why measuring beats squinting.

Bitrate is not the whole story

A generous bitrate cannot rescue footage that was soft when it was shot, and it cannot un-blow a clipped highlight. Encoding quality and capture quality are independent, and a useful check reports them separately rather than collapsing them into one verdict.

That separation matters when you decide what to do. “Thin bitrate” means re-export. “Soft focus” means reshoot. A single overall score would tell you neither — and if it averaged a perfect bitrate with a hopeless focus, it would tell you something actively misleading.

Check one video free: DiffALL’s video quality checker samples your clip second by second and reports bitrate against the recommended rate, focus, noise, compression, clipping, and whether the frame rate is genuine. No score out of 100 — you get the weakest dimension named, plus the worst moment.

Got the master as well? Compare the two videos for per-second SSIM, a difference heatmap, and the exact moment they diverge — which tells you precisely what an export or a platform re-encode cost you.