What Happens Before 'Motion Detected'? The CCTV Timing Test Worth Running

A clip that begins with a hooded figure already at the door, or a van already halfway off the drive, is not really evidence of the approach. It is evidence that the camera woke up late. The frustrating part is that this usually isn't a fault. It is a setting — or a default — quietly deciding how much of the story you get to see.

The gap is not a glitch, it is a setting

Most home CCTV and doorbells do not record everything all the time. They wait for something to change: a person-shaped blob, a burst of heat, a bit of movement in a chosen zone. When that trigger fires, the camera starts writing a clip.

If the system has no pre-record buffer, the first frame of that clip is the moment after the trigger. By then, the person may already be at the door. The van may already be turning. The parcel may already be gone.

A pre-record buffer — sometimes called pre-event recording, pre-roll or lookback — keeps a few seconds of footage in memory before the trigger. When motion is detected, the recorder stitches that earlier footage onto the start of the clip. That is the difference between seeing someone walk up the path and seeing them standing on the step.

Pre-roll, post-roll and the bit in between

There are two ends to this. Pre-record decides how much footage you get before the trigger. Post-motion recording decides how long the camera keeps filming after it thinks the action has stopped.

A short pre-roll is common on budget cameras and battery doorbells because keeping a rolling buffer costs power and storage. Mains-powered cameras and recorders often let you set it properly. A useful starting point for a driveway or shop front is 10 to 15 seconds of pre-record and 20 to 30 seconds of post-motion. That is not a universal rule, but it is a sensible place to test from.

The trade-off is real. More buffer means more data, and more data means either less retention or a bigger drive. On a battery camera, it may mean noticeably shorter battery life, or it may not be available at all. That is worth knowing before you assume the setting is hiding somewhere.

Why a late clip can be worse than no clip

A clip that starts too late can create false confidence. You know something happened, but you do not know who did it, which way they came from, or what they were carrying. In a shop, the till clip may start after the handover. At a doorstep, it may start after the parcel has been picked up. On a driveway, it may catch the rear of a vehicle and nothing useful about the front.

That matters because the first few seconds often carry the most identifying detail: a face turned towards the camera, a registration plate, a distinctive coat, the direction of travel. A late clip can also be more awkward to explain than a missing one, because it looks like you have footage until someone asks what it actually shows.

That said, longer is not automatically better. A huge pre-roll can capture more of the pavement and more of next door, which raises privacy questions and makes reviewing clips slower. The aim is not maximum footage. The aim is enough footage to answer the obvious questions.

The weekend walk test

You do not need specialist kit to find out what your camera is doing. You need ten minutes, a phone and a bit of patience.

First, check the camera's timestamp is correct. A clip is hard to test if the clock in the corner is telling a different story. Then pick a route that mimics a real approach: from the pavement to the front door, from the gate to the shed, from the shop door to the till.

Start a stopwatch on your phone, walk slowly into the camera's view and note the time you enter the frame. Carry on to the door or target, then wait half a minute and stop the recording.

Now review the clip. Find the first frame where you appear and compare it with the timestamp on your phone. If the clip starts after you have already entered the frame, your pre-record is too short or switched off. If it starts a few seconds before you appear, you have some buffer. If it starts well before, you have plenty.

Repeat the test at night, and again in the rain if you can face it. Motion detection behaves differently in low light, and a camera that catches you in daylight may be slower after dark. Also test a side-to-side walk as well as a walk straight towards the camera. Some systems are much better at one than the other.

What to change if the clip starts late

Look through the camera or recorder settings for pre-record, pre-event, pre-alarm or lookback. Not every app uses the same word. Increase it in small steps, then run the walk test again.

Check your motion zones too. A zone that excludes the path to the door may be doing exactly what you asked, but it also means the camera only triggers once you are on the step. Include the approach if that is the bit you need.

Sensitivity is the other lever. Crank it up and the camera may trigger earlier, but it will also trigger on cats, bins, branches and passing headlights. The better fix is usually a slightly longer pre-record plus a zone that covers the route in, not just the destination.

Finally, check the post-motion setting. A clip that ends the moment someone turns away can be just as annoying as one that starts late. Give it enough tail to show the exit.

The limits of the fix

Some cameras genuinely cannot do true pre-roll. Battery-powered doorbells often wake on a heat sensor, and by the time they wake, the person is already in view. No setting will invent footage that was never captured. In that case, the practical answer is placement: angle the camera so the approach is visible, or accept that the camera is a doorbell first and a surveillance system second.

If the first fifteen seconds matter to you — and for parcel theft, shop incidents and driveway damage, they usually do — test what you have rather than assuming. A camera that looks up late is still a witness, but it is a much less helpful one.