Black bars around a loop video usually come from a mismatch between the source image, OBS’s frame, or the way the source is positioned. Before changing settings, find out whether the bars are already in the file, appear in the OBS preview, or show up only in YouTube playback.
If the source and target frame have different shapes, you must choose between retaining the whole image with unused space, cropping the edges to fill the frame, or distorting the image by stretching it. For a standard 16:9 stream, check the source, OBS canvas and scaled output against a matching 16:9 frame, then verify the result in a recording or YouTube playback.
Find where the black bars appear
Start by comparing three views of the same moment: the original video opened outside OBS, the OBS preview, and the stream as played on YouTube. This simple comparison narrows the cause before you adjust the scene. Pause on a frame with recognisable details near the edges, so you can see whether the picture is being cropped or scaled differently between views.
If the bars are visible in the original file, they are part of its image. OBS may display the file inside a larger frame, but changing canvas dimensions will not erase black pixels already encoded in the video. You would need to crop those pixels, re-edit the file, or accept the borders.
If the original file has no bars but the OBS preview does, look at the source transform and the relationship between its dimensions and the canvas. If OBS looks clean but YouTube shows bars, compare the encoded output with playback: a recording can reveal whether the bars entered the outgoing picture, while the player itself can introduce empty space around a correctly shaped video.
For a long-running loop, make the comparison while the stream is actually live or use a short local recording from the same scene. A still preview does not prove that the scaled stream matches the player’s presentation. Keep the original untouched while you test; you can return to it if a crop removes content you need.
Check the source file for embedded borders
Check the file’s pixel dimensions in its properties or a video editor, and open it in a player that can show the image without fitting it into a differently shaped window. Note both the dimensions and whether black areas touch the encoded image edges. The familiar dimensions 1920×1080 and 1280×720 are both 16:9 examples; the width alone does not tell you the shape.
A file can have a 16:9 frame and still contain borders. For example, an older video may have been placed inside a 16:9 export with black padding above and below. In that case the file dimensions match a 16:9 output, but the visible picture inside them does not. You can inspect a paused frame near the edges, or use an editor’s crop controls to determine whether the black area is image content or just the player background.
Also distinguish a border from dark content. A night scene, a fade, or a black title card may reach the frame edge legitimately. Check more than one point in the loop before cropping; a crop that fixes one shot can remove subtitles, a logo, or important material in another.
If the borders are baked into the file, a source transform can hide them only by cutting into the picture. To retain the complete image, re-export the video with a suitable frame or place it in a scene with a designed background. YouTube’s video resolution and aspect-ratio guidance advises against adding padding or black bars directly to a video, and explains how the player adapts to different shapes. That is a reason to prepare a clean source where possible, not a promise that every viewer’s player will look identical.
Compare source, canvas, and output shapes
In OBS, open Settings → Video and note Base (Canvas) Resolution and Output (Scaled) Resolution. The canvas is the frame in which you arrange sources; the scaled output is the frame OBS sends out. Compare their aspect ratios, not just their pixel counts. For a standard 16:9 presentation, both 1920×1080 and 1280×720 have the same shape, even though one has more pixels in each direction.
If canvas and scaled output use different aspect ratios, the stream can be reframed or scaled in a way that leaves unused space or changes what fits. Pick the intended output shape first, then make the canvas and output agree with it unless you have a deliberate reason to use another arrangement. A familiar resolution is not enough: dimensions need to be considered together. OBS’s Aspect Ratio Guide covers proportional scaling, cropping, stretching and ways to compose a scene when a source does not fill the frame.
There is a useful historical example in an OBS forum thread about YouTube black bars: the original discussion described a 1920×1080 canvas paired with an output reported as 1080×720. A reply suggested 1280×720 as the 16:9 alternative. Treat that as an illustration of why checking both dimensions matters, not as proof that the same setting explains every case.
Write down the source, canvas and output dimensions before editing anything. Then calculate or identify each shape: a 4:3 source, for instance, is narrower than a 16:9 frame, so fitting it proportionally leaves space at the sides; filling the wider frame requires removing some of the top and bottom or stretching. If you change the output, check your encoder and streaming settings too, because they need to be appropriate for the chosen resolution. YouTube’s live encoder settings guidance explains that YouTube detects incoming resolution and frame rate automatically by default and transcodes live streams for playback.
Fit the source to retain the full image
When every part of the video matters, select the media source in the OBS preview and use its proportional fit transform, commonly labelled “Fit to Screen” in the source context menu. Menu names can vary between OBS versions. A proportional fit keeps the source geometry correct and shows the entire image; if source and canvas shapes differ, it leaves unfilled space on two sides or above and below.
That empty area is not necessarily a defect. You can keep it simple, choose a background colour, or put a designed background behind the source. For a devotional loop, for example, you might preserve the full image of a temple and place a restrained colour or still background in the side areas. For a local news loop, retaining the full lower-third text may be more important than making the video touch every edge.
The choice is between frame coverage and image retention, not between a correct and incorrect setup. A fit that shows all of the source may look less immersive on a particular screen, but it avoids losing information. If you add a background or overlay, keep it separate from the source where possible so you can adjust the composition without editing the video itself.
After applying the fit, inspect the corners and any text near the edge. Confirm the result in a recording or stream, since your eyes may mistake unused canvas for bars in the source. If you are also tuning the encode, keep the picture-shape test separate from compression changes; this OBS loop-stream bitrate guide addresses bitrate choices, which do not fix an aspect-ratio mismatch.
Crop or enlarge only when edge loss is acceptable
If the frame must be filled and losing some content is acceptable, crop the source or enlarge it proportionally until the unwanted edges extend beyond the canvas. With a 4:3 source in a 16:9 frame, filling the width by enlargement cuts off part of the top and bottom. With a wider source in a narrower target frame, it cuts off the left and right instead. The exact amount depends on the source and target shapes, so inspect the crop rather than assuming the subject will remain in view.
In OBS, select the source and adjust the crop with the appropriate controls. The OBS guide describes holding Alt while dragging a source edge to crop, though controls can differ with platform or version. Use the preview to check the full frame, and keep captions, faces, logos and anything close to an edge visible. A loop can contain multiple scenes, so sample several points rather than judging a single frame.
Cropping and enlarging preserve geometry, but they do not preserve every pixel of the original picture. They are a reasonable choice for footage with expendable margins; they are a poor fit for a video where edge content carries information. If you need more control over an ongoing show, create a scene layout with the video and a background, rather than forcing a single transform to solve every shot.
Do not use an unconstrained stretch to make a non-16:9 image fill a 16:9 frame. It can make circles oval, people look wider or narrower, and text proportions change. In OBS, Shift-dragging can stretch a source; use proportional scaling instead unless you have a specific reason to alter geometry and accept the distortion. When the whole image, full coverage and undistorted geometry all matter, a different-shaped source cannot satisfy all three at once.
Check the encoded picture and YouTube player
Once the scene looks right, test what OBS is sending rather than relying only on its preview. Make a short local recording if that is available in your workflow, or inspect the live playback after it has settled. Compare a frame in the recording with the same frame in the OBS preview. If the recording contains bars that the preview does not, revisit output scaling, source transforms and scene layout.
If the recording is clean but YouTube playback appears to have bars, look at the player window and device. YouTube adapts its player to the video shape and the space available on a viewer’s screen; unused space around the player is not necessarily part of your encoded picture. Compare on another browser or device if practical, but do not change a clean source simply because a particular player window has a different shape.
YouTube’s live workflow detects resolution and frame rate automatically by default and transcodes the stream for playback. That means playback may not be a pixel-for-pixel view of the encoder input, but it does not establish that YouTube added bars to the image. Confirm that the encoded picture itself has the expected frame dimensions and that the source occupies the intended area before changing scene settings again. For account-side settings and stream configuration, refer to YouTube’s Manage live stream settings page.
A clean output that still looks wrong on a phone may reflect how the video is framed on that screen, rather than a fault in OBS. For a channel that needs to run unattended, save a scene and test it through a full loop before treating the change as finished. Other operational checks, such as recovering OBS after a power interruption, are separate from aspect ratio; see how to keep OBS streaming after a power cut in India.
Choose the fix that suits the channel
Use the source-to-output comparison to choose deliberately. The table summarises what each approach preserves and what it gives up; it is more useful than treating all black bars as a fault to eliminate.
| Approach | Full image retained | Frame filled | Geometry preserved | Main trade-off |
|---|---|---|---|---|
| Proportional fit | Yes | Not always | Yes | Unused canvas space when shapes differ |
| Crop or enlarge to fill | No | Yes | Yes | Edge content is lost |
| Stretch to fill | Yes | Yes | No | Shapes and proportions are distorted |
| Designed background behind fit | Yes | The scene can be | Yes | Requires a composed background or layout |
For a music or ambience station, a designed background may be less distracting than a crop that removes details from the scene. For a news loop, fit may keep all of the lower-third readable even if the sides are unused. A close-up that has generous margins may tolerate cropping well. There is no universal rule that the frame must be filled; decide what viewers need to see.
Keep a record of the original dimensions and the OBS canvas/output values before you make a change. Adjust one thing at a time, then check a recording and the actual YouTube player. This makes it easier to undo a crop that removed a logo, or to identify a ratio mismatch that was introduced while changing output settings.
For a 24/7 channel, the fix should be repeatable, not merely attractive in a preview. Reopen OBS, confirm the source and transform are still in place, and view more than one point in the loop. If the channel’s main difficulty is keeping a file playing and the broadcast running while your computer is off, StreamNeo removes that specific need to leave an OBS computer running; it does not repair bars already baked into a video or choose the right crop for you.
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
Why does my OBS preview look fine but YouTube has black bars?
First check whether a recording of the OBS output contains the bars. If it does, revisit the canvas, scaled output and source transform; if it does not, compare YouTube’s player space and device framing before changing a clean scene. The preview alone cannot tell you where the bars entered the path.
Will changing the OBS canvas remove black bars in the video file?
No. If the bars are encoded as pixels in the source, changing the canvas does not remove them. You can crop them out, edit and re-export the file, or preserve the full picture with the bars or a background around it.
Should I fit, crop or stretch a 4:3 video for a 16:9 stream?
Fit shows the complete picture without distorting it, but leaves unused space at the sides. Cropping or enlarging fills the frame while cutting off some top and bottom content; stretching fills it but distorts the image. Choose based on what the video needs viewers to see.
Can YouTube add black bars to a live stream?
Do not assume playback bars are part of the encoded stream: the YouTube player adapts to the video shape and viewer device. Compare playback with a local recording or another view of the output, and check current YouTube guidance if you are unsure.