Two rooms both say HD and look nothing alike

What the HD filter is actually filtering

Tick the HD box on any cam site and the list shortens. It feels like a quality control, and that is why it is the most misread button on the page. The filter is reading one number that the broadcasting software reports: the size of the frame being sent. It checks whether that number clears a threshold and hides everything that does not.

That is a fact about the stream. It is just a very thin one. It says nothing about how much data is being spent on those pixels every second, nothing about the camera that captured them, and nothing about whether there is a single usable light source in the room. All three of those matter more to how a stream looks than frame size does, and none of them have a badge.

Think of resolution as the size of the canvas and bitrate as how much paint is available. A large canvas with not enough paint is a thin, smeared picture stretched wide. A smaller canvas with plenty of paint is a small, clean one. Cam sites badge the canvas. Flick between two badged rooms right now and you will usually find one of each.

1. Resolution

   How many pixels wide and tall the frame is. It sets the ceiling on detail and nothing else. A large frame full of soft, noisy pixels is still large and still soft.

2. Bitrate

   How much data per second is spent describing those pixels. When there is not enough, the encoder discards detail to fit — and discarded detail is what you are looking at when a room turns to mush mid-movement.

3. Everything in front of the lens

   Light, distance, background, the sensor itself. No amount of bandwidth recovers detail the camera never captured in the first place. This is where most of the difference actually lives.

The four things that decide how a room looks

Almost every visible difference between two streams comes down to four variables, and you can identify which one is misbehaving without knowing a single technical term. You only have to know what each failure looks like.

The useful part is that these fail in different ways at different moments. A bitrate problem comes and goes with movement. A camera or lighting problem is there the whole time, steady and unchanging. That single distinction tells you most of what you need, and two minutes in an open room is enough to run it.

Why the picture predicts the rest of the room

A stream that looks good is rarely an accident of expensive equipment. It is the visible end of a series of small decisions somebody made on purpose: a lamp moved off the ceiling and onto a stand, a camera raised to eye level instead of left on a laptop lid, a background cleared, a resolution chosen deliberately because it holds steady rather than because it is the biggest number available.

Those are all the same kind of decision — someone thinking about what the person on the other end actually sees. That habit tends not to stop at the camera. It usually shows up again in chat that gets answered by name, in rates written down plainly instead of "ask me", and in a show that starts when it says it will. The correlation is not a law, but it is strong enough to be worth using as a first filter.

The obvious counterexample is real and worth stating: a beautifully lit room can be run by somebody who never speaks, and a merely adequate picture can belong to someone who is completely present. Picture quality is a filter, not a verdict. Narrow the list with your eyes, then decide with the chat window .

The half of the problem that belongs to you

Before blaming a room, it is worth knowing how much of what you see is produced on your end. The picture that arrives on your screen has been through your connection, your browser, your graphics hardware and your window size, and every one of those can degrade it after it left the performer perfectly intact.

The quick test is comparison. If one room looks poor and the next three look fine, the problem is at their end. If every room you open looks the same kind of soft, the problem is yours, and it is usually something dull and fixable. That is the whole subject of the setup guide, and it is the check almost nobody runs before deciding a performer has a bad camera.

Judge the picture before you pay for it

There is a practical reason to do all this in the free part of the site. A private show uses the same camera, the same lighting and the same encoder as the public room it came from, so whatever the picture looks like now is what it will look like then — except you will be paying 6–90 tokens a minute to watch it. Tokens land around $0.08–0.12 each depending on bundle size, so the picture you are about to buy is worth ten free seconds of attention.

The same goes for tips. A public tip of 15–100 tokens lands hardest in a room with roughly 20–200 viewers, where a response is actually possible, and a room like that is also small enough that you can see how much care has gone into the setup. Look at what you are tipping into first.

You can do every bit of this signed out. Registering is free and adds chat, follows, notifications and a block of starter tokens, which is the point at which a good-looking room becomes a room you can actually talk to.

A ten-second pass, in order

• Wait for movement. A still frame hides every bitrate problem there is. Watch what happens when the person moves.
• Look at the shadows. Speckled, crawling dark areas mean the camera is straining for light that is not there.
• Check the eye line. A camera at eye level was placed. A camera looking up from a desk was not.
• Check the distance. Pixels spent on an empty half of the room are pixels not spent on anything you came for.
• Then open the next room. One stream on its own is unreadable; three in a row are obvious.

The full version of that pass, including what each artefact actually tells you about the connection behind it, is on the stream quality page. The short version works fine on its own — try it across whoever is streaming now .

Frequently asked questions

Does the HD badge mean anything at all?

It means the broadcast is being sent at a resolution the site counts as high definition. That is a real fact about the stream, and it is roughly the only fact it tells you. It says nothing about how much data is behind those pixels, what camera captured them, or whether the room has any light in it.

Why does a smaller stream sometimes look sharper than a bigger one?

Because a lower resolution asks less of the connection and the encoder, so the detail it does capture survives intact. A stream sent at a size the upload cannot support has to throw away information to fit, and throwing away information is exactly what you see as mush during movement.

Is a blurry room always the performer’s fault?

No, and assuming so is the most common mistake viewers make. Your connection, your browser, how many tabs you have open and whether your device is throttling all change what arrives on your screen. Check your own side before you write a room off.

Do I need an account to look at any of this?

No. Public rooms open and play with no account and no card, which is what makes judging picture quality free. Registering is also free and adds chat, follows, notifications and a block of starter tokens.

Does a private show look better than the public room?

It is the same camera, the same light and the same encoder, so expect the same picture. That is the argument for judging the image in the free public room first, before you are paying 6–90 tokens a minute to look at it.

Does a good-looking room mean a good room?

It is a strong hint, not a guarantee. Picture quality is the visible end of a set of decisions somebody made on purpose. Plenty of well-lit rooms are dull, and the occasional soft picture belongs to someone who is genuinely present. Use the image as a first filter and the chat as the deciding one.