White balance controls how your camera renders the colour of the light in a scene. Set it with the lights you will actually use on stream, using a neutral reference or manual Kelvin adjustment while watching the image, then keep it fixed while those lights stay the same.
There is no single Kelvin value that is correct for every room or camera. Matching the light aims for neutral-looking whites and natural skin; choosing a different balance can create a deliberate look, but it does not improve poor lighting or guarantee accurate colour on every camera.
What white balance changes
Light sources have different colour characteristics. A camera uses white balance to compensate, so a neutral object under the chosen light is rendered closer to neutral rather than obviously orange, blue or green. Canon describes its camera processing as based on the colour temperature of the illumination. The practical result is visible in both skin and everyday reference objects: a white wall, paper or shirt may look tinted if the setting does not suit the light.
White balance is separate from exposure, focus and brightness. If your face is dark, increasing exposure or improving the light is the relevant fix; changing white balance will not make the scene brighter. Likewise, a balanced setting cannot remove a coloured reflection from a wall or make a weak lamp illuminate a room evenly. Treat it as one part of a stable image, not a correction for every lighting problem.
Automatic white balance estimates the light and can adjust as the scene changes. That may be useful if conditions vary, but the image can shift as a person moves, a monitor fills more of the frame, or daylight changes. A fixed custom or manual setting is often easier to repeat when your stream has a stable set and you want the colour to stay steady. Which mode makes sense depends on the camera, its software, and how much the scene changes.
Set it under your actual lighting
Start with the lighting arrangement you intend to use, not a setting copied from someone else's room. Turn on the key light, fill light and background lights that will be on during the broadcast. Put the camera in its final position and frame the same view you plan to stream. If you set balance near a window in daylight and later close the curtains and switch on a warm lamp, the original reading no longer describes the scene.
Let the camera show the same exposure and image path you expect on air. A reference captured in very different brightness conditions can be misleading; Canon notes that an exposure far from standard can prevent a correct custom balance. You do not need to buy a special card to begin, but avoid using a visibly tinted object as your supposed neutral target. Plain white card or paper can work if it is not coloured by printing, gloss or surrounding reflections; a grey card is a useful optional reference.
Make one adjustment at a time and give yourself a consistent view. If you are setting up a devotional channel, for example, sit at the place where you will speak or sing, with the actual lamp and any decorative lighting on. A camera pointed at an empty wall may see a different balance of colours than the one that appears when your face and clothing enter the frame. After you set the value, check the actual streaming preview, not only a camera menu or a separate monitor.
If you are building a long-running stream around recorded classes or a fixed scene, the same principle applies to the camera used for the live presentation or opening segment. The rest of the programme may be prerecorded, but any on-camera presenter should be balanced under their real lights. For a broader example of planning a continuous recorded programme, see how to run a 24/7 yoga and workout channel from existing classes.
Use a white or grey reference
A custom white balance procedure gives the camera a known neutral target under the current light. Hold a plain white or grey card where the subject will be, so it receives the same illumination as the face or object in the shot. Follow the camera's own instructions: some models ask you to fill a target area in the frame, while others register a captured reference image. The menu wording and steps vary by model.
Canon's EOS R50 manual permits a grey card or standard 18% grey reflector as an alternative to a white object. Panasonic also documents calibration using a white or grey card. Those are examples for the named makers and models, not a guarantee that every webcam has a custom-reference feature. Consult the manual for your camera before relying on a particular menu sequence. The Canon EOS R50 white-balance instructions explain how its custom balance is registered.
Keep the target evenly lit and avoid letting a lamp, window or bright reflection clip it to featureless white. If the card is coloured by a nearby wall or a coloured bulb, the camera will interpret that cast as part of the reference. Once registered, remove the card and inspect your face and a neutral object in the composition. The card is a calibration aid, not proof that every colour in the scene will be reproduced perfectly.
A useful routine is to note the light arrangement and the selected preset or custom slot after a successful setup. You can then repeat the arrangement later rather than trying to remember whether a different lamp was on. If you run a Hindi guided meditation or another spoken programme, check a natural skin tone and a white cloth or page before you settle into the frame; the practical visual checks matter more than a menu label that merely says “custom”.
Adjust manual Kelvin while watching
If your camera offers a manual colour-temperature control, use it as a viewing adjustment, not as a magic recipe. Kelvin expresses a colour-temperature scale, but the number is interpreted by the camera as a correction control. Different cameras may offer different ranges, increments and behaviour. Canon lists an approximately 2500–10000 K range in 100 K increments for the EOS R50; Panasonic gives a 2500–10000 K input range for the named LUMIX S5M2/S5M2X models. These are selectable ranges, not recommended values for every room.
Choose a value that looks close under the actual lights, then make small changes while watching the feed. Look at a neutral object and the subject's skin together. If white paper looks blue, the balance may be too cool for the light or the camera's processing; if it looks strongly orange, adjust the other way. Camera controls may label the direction in terms of warmer or cooler rather than making the visual logic obvious, so trust the preview and learn which direction moves the image on your device.
Do not chase a numerical target simply because a guide gives one. A room with daylight, an LED panel and a warm table lamp may not behave like a room lit by one source. A camera's Kelvin reading also does not account for every green or magenta tint produced by a particular fluorescent or LED fixture. Some cameras offer additional tint controls; where they exist, make cautious adjustments only after you have checked the light itself.
Once the image looks consistent, record the setting or save it to a camera preset if available. Lock it while the lighting remains unchanged. The image should not wander in colour as you move slightly or a pale object enters the background. If it does, verify whether automatic balance remains active somewhere in the camera or webcam application.
Match mixed or changing light
Mixed light is a common reason one setting seems right for the face but wrong for the background. A window may provide cool daylight on one side, while a warm bulb lights the other; a coloured LED can add another cast. The camera can only apply one balance to the whole frame in an ordinary setup, so balancing one source may make another look more strongly tinted. Panasonic's guidance notes that suitable balance under fluorescent and LED fixtures varies with the lighting type.
The practical remedy is usually to simplify before calibrating. Turn off unnecessary room lights, block or control daylight if it changes during the session, and use the same kind of light on the subject where possible. If the background has a warm decorative lamp by design, decide whether that look is intentional and keep it out of the face's key light. White balance will not make two differently coloured sources identical across the whole room.
For a channel that runs at different times of day, create a repeatable lighting state rather than recalibrating by guesswork each time. Close the blind, switch on the same lamps, set their brightness consistently if possible, then use the saved balance. If you move the camera or change the lighting mix, recheck the result. The YouTube live-stream setup guide is useful for the wider broadcast workflow, but colour matching still depends on what reaches your camera.
Auto balance has a place when conditions truly vary and you accept that the image may move with them. A camera near a window may respond to changing daylight; Elgato's Facecam Camera Hub, for instance, documents both automatic and manual white-balance controls for its product. That is a product-specific example, not a general recommendation. Check the Elgato Camera Hub settings overview if you use that camera, and check your own software's controls otherwise.
Check the image that reaches the stream
A camera setting only helps if it is the setting actually feeding the broadcast. Some cameras apply video settings to USB streaming; Nikon says that video-mode settings, including white balance, apply to the USB-streamed image on the Z5II. In other workflows, a webcam utility may expose its own white-balance control or override the camera's setting. FUJIFILM X Webcam, for example, documents white-balance mode and colour-temperature selection in its own guide.
Check the final preview in the software that sends your stream. Compare the same neutral object and skin tone there with the camera's view. If the colour differs, inspect the camera utility, capture software and any image filters rather than repeatedly changing the light. Nikon also cautions that some menu operations while streaming can cause a grey image to be sent, so verify the output after changing settings and before relying on it live. See the Nikon Z5II USB streaming guidance for its model-specific behaviour.
For a continuous channel, a stable file and stream path do not fix a camera setting that changes mid-session. StreamNeo removes the separate burden of keeping your own computer running for a file-based 24/7 YouTube broadcast, but any camera image you add still needs its own lighting and colour check. If you also use original music in a long-running channel, understand the separate rights issue described in how Content ID can claim your own music on YouTube Live.
Choose neutral matching or a deliberate look
The word “correct” can mean two different things. For accurate matching, you are trying to render a neutral reference as neutral under the scene's light and keep skin plausible. For a deliberate look, you may intentionally make the image warmer for a cosy room or cooler for a restrained studio feel. Neither goal has one universal number, and a deliberate cast should be a choice rather than an accidental result of leaving auto balance to drift.
A useful distinction is to establish a neutral baseline first. Set the custom reference or Kelvin value until the neutral object and skin look reasonable; save or note that result. If you then want a warmer or cooler mood, make a controlled change and check it across the full frame. This makes it easier to return to the balanced setting if the look distracts from a face, garment or product.
The table below is a decision aid, not a performance test. It summarises the trade-offs documented by manufacturers for available controls and the operational implications of using them.
| Situation | Sensible starting method | Trade-off to watch |
|---|---|---|
| Stable room and camera offers custom balance | Set a white or grey reference under the final lights | Repeatable when the light stays the same; redo after material changes |
| Stable room and camera offers Kelvin control | Adjust manually while viewing the feed | Direct control, but the useful value depends on the camera and light |
| Light changes during the broadcast | Consider auto balance or stabilise the light first | Auto may respond, but rendered colour can shift with the scene |
| Mixed bulbs or daylight on the subject | Simplify the sources, then balance again | One global setting cannot neutralise every differently coloured source |
| Camera image differs from stream preview | Check camera and software controls | The camera menu may not be the final control in the streaming path |
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
What Kelvin should I use for live streaming?
There is no single Kelvin value that suits every room, light source and camera. Set the value under your actual lighting while watching the image, or use a custom white-balance reference if your camera supports it. Manufacturer ranges describe what a camera allows you to select, not a universal recommendation.
Should I use auto white balance or manual white balance?
Use manual or custom balance when your lighting is stable and you want colour to remain consistent as the scene changes. Auto can be useful if the light genuinely changes, but it may shift the image as the camera sees different objects or daylight varies. Check the final stream preview whichever mode you choose.
Do I need a grey card?
No. A grey card is an optional neutral reference that can make custom calibration repeatable; some camera manuals also permit a suitable white object. Use a card or surface that is evenly lit and not clipped, tinted or reflecting a nearby coloured light.
Why does my camera look right but the stream look tinted?
The camera's setting may not be the last control in the path. A webcam utility or streaming application can have its own white-balance controls, and the output can differ from a camera monitor. Compare the same face and neutral object in the broadcast preview, then check for automatic settings or software overrides.