A concat playlist can send a prepared sequence of lofi tracks to a YouTube Live stream, but it does not make unlike files compatible or make the stream repeat itself indefinitely. Check the audio files, prepare the visual separately, test the loop and ingest settings, and clear live-use rights and Content ID allowlisting before you go public.
This guide focuses on a documentation-led workflow rather than a universal command: FFmpeg builds differ, and the playlist’s end-to-start behaviour must be verified in your own setup. A private test can reveal media and connection problems, but it cannot establish that a continuous broadcast will never stop.
Prepare and clear tracks and visual assets
Start with a library you are entitled to use for the exact purpose. The permission should cover a continuous YouTube livestream and, if you intend to keep the broadcast available, the resulting archive or replay. A licence for personal listening, background use in a one-off video, or a limited online placement may not grant those rights. Read the terms and ask the rights holder when the scope is unclear.
Keep evidence together: the licence, receipts or permission emails, track names, and any instructions about attribution or reporting. If the music comes from another creator or a licensing service, ask specifically whether uninterrupted live use is included. Do not assume that a label such as “royalty-free” means that YouTube’s Content ID system will never match it. The practical distinction between a licence and platform matching is explained in whether royalty-free tracks can still trigger Content ID.
Make a working folder with an audio subfolder, a video or artwork subfolder, and a playlist file. Use plain, predictable filenames such as track-001.wav; avoid changing file names after you write the playlist. Keep a copy of the original files in a separate location so you can return to them if a conversion or edit introduces a defect.
Decide what viewers will see. A still image, an animated visual loop, and a sequence of changing scenes are different video inputs; the audio concat list does not schedule video alongside its tracks. Build or select the visual as a separate input, then make sure its duration and repetition behaviour are understood. If you want a moving background, inspect the start and end frames for an obvious jump. A useful comparison of separate audio and video loop behaviour is why a repeating video can outlast its audio loop.
If the source music is mixed-format, do not start by guessing that FFmpeg can join it unchanged. First inspect streams and metadata, then choose either normalization to common properties or a workflow that decodes and re-encodes the media. Preserve a clean source copy and listen to a sample from each converted file, particularly around its start and end, where clicks, silence, or truncated fades are easiest to notice.
Check concat-demuxer file compatibility
The FFmpeg concat demuxer reads a script and opens the listed files in sequence. Its advantage is that compatible inputs can be joined without re-encoding. The condition matters: the files need matching streams, codecs, and time bases. The FFmpeg formats documentation describes the demuxer and its requirements; a playlist is not a conversion step and will not reconcile mismatched audio formats for you.
For an audio-only list, inspect whether each track exposes the same kind of audio stream with compatible codec and timing properties. For example, one file encoded as MP3 and another as AAC do not become a common stream simply because their names appear in the same text file. Different sample rates, channel layouts, or stream arrangements are also reasons to check the files rather than assume a clean join. The concat demuxer’s exact requirements are about stream compatibility; normalizing sample rate and channel layout is a practical way to create a consistent set when your sources vary.
A file inspection tool such as ffprobe can report each file’s streams, codecs, time bases, sample rate, channel count, and duration. Compare the reports across the whole library, not only the first two tracks. A single unusual file near the end can cause a failure after a long clean-looking start. If you find a mismatch, either convert the outlier or normalize all tracks to a defined common format and verify the resulting files again.
The concat demuxer and concat filter solve related but different problems. The demuxer is suitable when the inputs already have compatible streams and you want sequential concatenation without an intermediate re-encode. The filter operates on decoded media and is a better fit when you need to re-encode or process inputs to make them join. FFmpeg’s concat FAQ distinguishes these approaches. Re-encoding gives you a chance to standardize media, but it also adds an encoding step that should be checked for audible changes and adequate processing capacity.
Duration deserves its own check. The demuxer uses each file’s duration to position the next file’s timestamps. If duration metadata is wrong, the transition can have timestamp problems or artifacts; when stream lengths differ, gaps can appear. Do not add guessed duration values to the script to silence a warning. Use a duration directive only when you know the accurate duration, and confirm that the underlying audio actually ends where that value says it does.
Build and validate the concat playlist
A minimal script uses one file directive per track. Save it as a text file alongside the media, or write paths relative to a location you can keep stable:
ffconcat version 1.0
file 'track-001.wav'
file 'track-002.wav'
file 'track-003.wav'
When you rely on automatic format recognition, the signature must be the exact first line, with no leading spaces or byte-order mark. The file directive identifies each input. The FFmpeg concat syntax documentation covers quoting and path handling. Filenames with spaces or special characters need correct quoting or escaping; the simple example avoids those complications.
The demuxer’s safe-path option defaults to enabled. That setting rejects paths that are not considered safe and portable, so relative paths in a stable working folder are usually easier to manage than absolute paths. If your playlist is rejected because of path handling, first consider moving the script and assets into a suitable folder and using relative names. Do not disable path checks for a playlist you did not create or inspect; an input script can point to files beyond the media you meant to use.
Before running a long stream, validate the list as an ordinary media operation. Confirm every named file exists, the ordering is right, and the playlist opens from the folder where it will be used. Ask FFmpeg to read or produce a short output from the sequence and inspect the beginning and the transitions between tracks. Listen across a transition with headphones: a clean change in file order can still sound abrupt if the source tracks have different loudness or if a fade was cut off.
For a library with many tracks, generate the playlist from a reviewed track order rather than relying on a directory listing whose order may vary. Keep a human-readable order file or a dated copy of the playlist so that you can reproduce the intended sequence. If you revise the library, repeat the compatibility and playback checks for the revised list.
The concat demuxer reads the entries in order; do not confuse this with repeating the whole playlist. The documentation does not establish a universal “repeat forever” directive for every FFmpeg build and command arrangement. Confirm the looping option’s order and semantics against the installed FFmpeg version, then run a test long enough to reach the list’s end and verify it starts again cleanly. The separate problem of looping multiple pictures or clips is covered in how to loop multiple videos on YouTube Live.
Configure FFmpeg audio and video output
Treat the audio playlist and visual as separate inputs to the output. That makes it easier to reason about a still or looped video continuing while the audio advances. Specify the desired mapping deliberately: the outgoing stream needs the intended video and audio, not whichever streams FFmpeg happens to select by default. Check that the video has a sensible frame size and frame rate for the chosen source, and that the audio remains stereo if that is how the library is prepared.
The output stage is also where you choose encoding and rate control. YouTube accepts AAC or MP3 audio and recommends stereo audio at 128 kbps. For video, YouTube recommends constant bitrate (CBR) and a keyframe interval of two seconds, with a maximum interval of four seconds. These are platform recommendations, not a guarantee that a particular command will work with every FFmpeg build or encoder. Consult YouTube’s current live encoder settings and bitrate guidance before settling on parameters.
A practical configuration has to fit the input and the computer doing the encoding. A still image requires little visual processing compared with a moving loop, but the output still needs to be encoded at a stable mode that your encoder can sustain. If you choose a high-resolution or high-frame-rate output without checking available encoding capacity, a machine may fall behind even if the media files themselves are valid. Select a supported output mode appropriate to the visual and test it under the conditions you expect to use.
There is no single command here that can be labelled tested for every local installation. Encoder names, available hardware acceleration and option behaviour depend on the FFmpeg build and the machine. Build the command from your installed version’s help and the current YouTube recommendations, then test the exact command, playlist, visual, and destination you plan to use. Keep the command and FFmpeg version with your notes so a later update does not quietly change the workflow.
The local output should also be inspected before it is sent live. Check that both audio and video are present, the audio is not clipped or unexpectedly quiet, and the visual does not freeze or disappear. Confirm that the output duration and timestamps continue as intended during the test. For a deeper walkthrough of the same demuxer approach with another music library, see the Indian classical music concat setup.
Match output to YouTube ingest guidance
YouTube recommends RTMPS for ingest. Use the server address and stream key shown for the event in YouTube Studio, and keep the key private: anyone who obtains it may be able to send a broadcast to that destination. Confirm that the selected event is the one you intend to test, and do not publish the stream key in a command screenshot, shared document, or public support request.
The video recommendations include CBR and keyframes every two seconds, not exceeding four seconds. YouTube accepts AAC or MP3 audio and recommends 128 kbps for stereo. Match these to the encoder settings and inspect the ingest preview or stream health indicators for warnings. If a setting is unclear, use YouTube’s current documentation rather than copying an old command whose assumptions may no longer fit your stream.
Upload capacity is another input, not an afterthought. Test the connection from the location and connection you intend to use, and leave room for normal variation rather than aiming at the absolute measured maximum. A speed test is only a snapshot; it does not prove the connection will remain steady overnight. Monitor the event dashboard during the test and check for dropped frames, connection warnings, or audio and video issues.
YouTube transcodes live video into other viewing formats, but that does not remove the need to send a stable, supported ingest. A viewer’s playback quality and the broadcaster’s upload path are different parts of the chain. If an OBS-based test is stuck at startup or losing its connection, the troubleshooting steps in the OBS encoder and connection guide may help isolate that separate issue.
Test the stream before a continuous run
Create a private or unlisted test event before planning a public launch. Use representative tracks, including files with different transitions, and the actual visual input. Let the test reach the playlist boundary if you intend to repeat the whole sequence. Check whether the next cycle begins without silence, a gap, a timestamp jump, or an abrupt visual reset. A short opening sample cannot validate behaviour at the end of a long playlist.
Listen and watch locally as well as checking YouTube’s preview. Locally, you can identify a source file problem or an FFmpeg mapping mistake; in the event dashboard, you can see ingest and stream-health warnings that the local output cannot reveal. Check audio synchronisation, channel balance, and whether the artwork or animation stays present throughout the test. If you change output settings after a test, repeat the relevant checks.
A continuous run needs more than a correct playlist. The FFmpeg process can exit, the network can disconnect, or the host can restart. Plan how you will notice a stopped process and how it will be restarted, and test that recovery procedure separately from the media test. The consulted platform guidance does not promise uninterrupted operation or prescribe one required supervisor, so do not treat a clean test as evidence of guaranteed 24/7 service.
Keep a launch checklist that records the playlist version, FFmpeg build, output settings, event destination, and the time of the last successful test. When the broadcast has begun, periodically check the stream-health view and the actual playback. If more than one stream shares a connection, estimate their combined upload needs and read the guide to running two 24/7 streams on one connection.
Address Content ID and allowlisting before launch
Technical preparation and rights clearance are separate tasks. YouTube says live streams are scanned for third-party content. If a match is detected, the platform may replace the stream with a placeholder and warn you; continued use can lead to temporary interruption or termination. A music licence does not itself prevent automated matching.
YouTube Help specifically warns that a licensed track can still interrupt a livestream if the rights owner has not added the channel to its Content ID allowlist. Contact the owner before launch, provide the channel details they need, and request confirmation that the channel is allowlisted for the music in your stream. Do not assume that a permission email or a successful private test means allowlisting has happened.
Keep written records of the rights granted, the tracks covered, the permitted use, and the allowlisting request and confirmation. If the library changes, check whether the new tracks have separate terms or rights owners. If any track is outside the permission scope, remove it or obtain the appropriate permission before including it. No workflow can guarantee that Content ID will not flag material, and the channel owner remains responsible for checking current YouTube requirements and the relevant licences.
When the burden is keeping a prepared file and channel connected without leaving a personal computer running, StreamNeo removes that specific operating chore by turning an uploaded video into a YouTube broadcast that continues with your computer switched off. That does not clear music rights or replace the checks above. Whichever operating approach you use, compare the practical requirements against your own ability to monitor and recover a process.
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
Can FFmpeg concat automatically repeat the playlist forever?
Do not assume that a concat script itself repeats. It lists files in sequence, and repetition depends on the exact FFmpeg options and how they are applied in your installed version. Test through the end of the list and into the next cycle before relying on it.
Do all lofi tracks need the same codec and sample rate?
The concat demuxer requires matching streams, codecs and time bases, so mixed files need inspection and may need normalization. Consistent sample rate and channel layout are useful preparation choices, but they do not replace checking the resulting stream properties. Use the concat filter when your workflow needs decoded inputs and re-encoding.
Does having a licence stop Content ID interruptions?
Not necessarily. YouTube says licensed third-party content can still interrupt a live stream when the channel has not been added to the rights owner’s Content ID allowlist. Confirm both the licence scope and allowlisting with the rights holder before launch.
Is a successful private test proof that the stream will stay up all night?
No. It can help expose media, ingest, and loop problems under the conditions of the test, but it does not prove long-term network or process reliability. Plan monitoring and recovery, and check YouTube’s current stream-health guidance during operation.