If FFmpeg cannot open audio named in a podcast playlist on your Linux VPS, first identify whether the file is an FFconcat script or an HLS M3U8 manifest. They use different path rules, so a concat-specific setting such as -safe 0 will not fix a missing HLS segment.
The title alone does not reveal your playlist, command, working directory, process account, or error. Treat the steps below as inspection patterns: run checks in your own environment and compare them with the command that actually launches FFmpeg.
Identify which kind of playlist you have
An FFconcat script is a text file that tells FFmpeg to read a sequence of media files as a virtual concatenation. It commonly contains a file directive for each audio file, and may begin with ffconcat version 1.0. An HLS manifest is an M3U8 text file whose entries refer to media segments, and may also include variant playlists and other HLS tags.
The file extension is a clue, not a diagnosis. A .ffconcat or .concat suffix suggests a concat script, while .m3u8 suggests HLS, but inspect the content and the command before changing anything. For an FFconcat script, look for file lines. For HLS, look for tags such as #EXTM3U and segment references that follow duration tags.
The distinction matters because the mechanisms differ. The concat demuxer reads its script and applies its own safe-path checks to the listed files. HLS uses references in a manifest and segment naming and location rules. FFmpeg documents these separately in its concat demuxer documentation and HLS muxer documentation.
Also establish what FFmpeg is doing with the playlist. It could be reading an existing manifest or script, or writing an HLS playlist and segments as output. A read-side “cannot open” error calls for checking the input reference and access. A write-side error calls for checking the output paths and whether the target directories exist and are writable.
Inspect the playlist syntax and file paths
For an FFconcat script, a basic example looks like this:
ffconcat version 1.0
file '/srv/podcast/audio/episode-01.mp3'
file '/srv/podcast/audio/episode-02.mp3'
The version marker is optional when the command explicitly selects the concat format with -f concat. If you rely on automatic recognition, FFmpeg requires the marker to be exactly at the start of the first line, without leading spaces or a byte-order mark. A text editor can introduce a byte-order mark or unexpected characters, so if FFmpeg fails to recognise the format, inspect the file’s first line and encoding rather than assuming the paths are the only issue.
Each file directive must refer to the intended file as the FFmpeg process can see it. Check spelling, capitalisation, extension, and directory layout. Linux filesystems commonly distinguish uppercase and lowercase names: Episode-01.mp3 and episode-01.mp3 can be different paths. Also check for stale paths left behind after moving audio or changing a mount location.
FFmpeg’s concat documentation describes quoting and escaping for spaces and special characters. For example, a path containing spaces can be enclosed in single quotes. A filename containing a single quote needs escaping within that syntax; do not assume shell quoting rules are interchangeable with the concat script’s own parsing rules. The playlist is input data read by FFmpeg, not a shell command line.
A useful inspection is to read the playlist itself, then check every referenced name against the directory layout. If the list is generated by a script or application, inspect the generated file on the VPS, not just the template on your laptop. A generated playlist can preserve a local path such as a Windows drive letter, or refer to a directory that does not exist in the VPS environment.
For HLS, inspect the manifest references instead of searching for file directives. Determine whether each URI is relative or absolute, and where it resolves in the context in which the manifest is consumed. Check whether the referenced segment files are present at those locations. If another programme or web server serves the playlist, a file existing on disk does not by itself prove that clients can retrieve it.
Check the launch directory and process user
Relative paths are a common source of confusion because your interactive shell and a service can start FFmpeg in different contexts. Do not infer the base directory for a path from where you last edited the playlist. Inspect the exact command, the process launch configuration, and FFmpeg’s observed behaviour on the installed build; the documentation does not establish one universal working-directory answer for every invocation and service arrangement.
First identify the account under which FFmpeg runs. A file readable in your login session may not be readable by a separate service account. Check the permissions on the playlist, each audio file, and every parent directory in the path. The process needs permission to traverse directories as well as read the file. For an output playlist or segment directory, it also needs the relevant write permissions.
Then establish the launch directory from the mechanism you use. If a service manager launches FFmpeg, inspect that service’s configuration and logs for its command and working-directory settings. If a shell script launches it, inspect the script and the directory from which it changes or starts the process. Do not presume a particular distribution, service manager, or default directory: those details have not been established here.
A careful check uses the same account and launch context as the real job. As a diagnostic pattern, an operator might inspect the effective user and current directory in the launch script, or temporarily log them immediately before FFmpeg starts. For access checks, run a narrowly scoped test as the service account against the playlist and media files, where your environment permits it. An ls or file check performed as an administrator can prove that a path exists, but not that the FFmpeg account can open it.
If relative entries are involved, record the command’s input playlist path and the working directory used at launch. Test what those entries resolve to under that exact context. Avoid changing several paths at once: otherwise, a successful retry will not tell you which difference mattered.
Use absolute paths where they clarify the layout
Absolute paths can make a playlist easier to inspect because each entry names its location from the filesystem root, rather than relying on context. For a fixed collection stored under /srv/podcast/audio, an entry such as file '/srv/podcast/audio/episode-01.mp3' makes the intended location explicit. Use the real path in your environment, not this illustrative one.
Absolute paths do not fix incorrect permissions, spelling, missing files, or an unavailable mount. Nor do they automatically pass the concat demuxer’s default safe-path check. If changing relative entries to absolute ones leads to an error mentioning safe filenames, that is a different issue from whether the file exists or can be read.
Relative paths may be more convenient when the playlist and media directory move together, but their resolution must be confirmed in the actual command context. A reliable choice is the one that remains correct when the service launches, not the one that works only after you cd into a directory by hand. Keep the layout consistent and document where the playlist and audio live so that later updates do not silently invalidate the entries.
For HLS, “absolute path” needs care: a filesystem path and a URI delivered to a player are not necessarily the same thing. A manifest consumed by a web server may need segment references that the server can serve, rather than private filesystem locations. Confirm the intended consumer and deployment layout before editing those references. The HLS output options, including -hls_segment_filename, control segment naming and location when FFmpeg writes an HLS output; they do not rewrite an unrelated FFconcat script.
If your underlying goal is simply to keep episodes in sequence, playlist syntax is only one part of the design. The OBS playlist source settings for podcast episodes describe a different playback approach; compare it only if OBS, rather than an FFmpeg process reading a concat script, is the software you intend to run.
Understand FFconcat safe-path behaviour
The concat demuxer’s safe option is enabled by default. FFmpeg’s documented safe-path rules reject names that do not meet its criteria, including absolute paths and paths with components outside its permitted portable character set. The option is specific to the concat demuxer, not a general setting for every file called a playlist.
That means an absolute path may be valid on the filesystem and still be rejected by the demuxer’s safety check. Conversely, setting -safe 0 can allow a name that the check would reject, but it cannot create a missing file or grant the process permission to read it. Make sure the error points to a safe-path rejection before treating this as the cause.
If you control the playlist’s contents and have checked every entry, -safe 0 may be appropriate for a concat script that needs absolute paths or other names excluded by the default. Keep it scoped to the intended concat input. FFmpeg documents that disabling the check accepts any filename; operationally, that is a reason to use it only with a playlist you trust and maintain. It is not a universal fix, and it does not apply to HLS segment URIs.
Another option is to arrange safe relative names under a suitable directory, then confirm that the process context resolves them correctly. This avoids disabling the check, but only works when the relative layout and launch context are dependable. Choose between those approaches based on the actual paths and who can modify the playlist, rather than adding -safe 0 pre-emptively.
Path acceptance is not the same as successful concatenation. FFmpeg’s concat demuxer documentation says, “All files must have the same streams (same codecs, same time base, etc.).” If the files differ, resolving every path can expose a new format or stream-compatibility error. For a podcast, check that the episode files have compatible audio streams and that their durations are usable; inaccurate duration information can also affect transitions and timestamps.
Review the FFmpeg command and error output
Read the full command as executed, including option order and the input it names. For an explicitly selected FFconcat input, a documented invocation pattern might look like this:
ffmpeg -re -f concat -safe 0 -i /srv/podcast/playlist.ffconcat -c:a copy -f mp3 OUTPUT_URL
This is an illustration, not a ready-made command for your channel. The output muxer, codec handling, real-time option, and destination depend on the streaming arrangement. In particular, do not copy -safe 0 just because it appears here: use it only if the input is a trusted concat script and the safe-path restriction is the problem you have identified.
Options such as -f concat and -safe apply to the concat demuxer. They do not make an HLS manifest behave like FFconcat. FFmpeg also has a concat filter and concat protocol, each with a different purpose. The FFmpeg concat FAQ explains the alternatives: the filter is relevant when re-encoding is needed, while the demuxer can avoid re-encoding where the inputs meet its requirements. Switching concat methods is not a substitute for correcting a bad path.
Read the first meaningful error in the log, along with the lines immediately before and after it. “No such file or directory” points you towards the exact path and launch context; a permission error suggests access; a safe-filename complaint points towards concat validation; and a format or stream error after opening files is a separate compatibility issue. These are clues, not a diagnosis of your VPS. Preserve the command and error output when comparing attempts so you can see what changed.
Check that the command names the playlist you edited. Multiple copies with similar names, a relative input path, or a service using an old configuration can leave you correcting a file FFmpeg never reads. If you generate the playlist, verify its output timestamp and contents after generation, while taking care not to expose stream keys or other credentials in shared logs.
When an error mentions a particular entry, inspect that entry before changing the whole list. One malformed quote or misspelt episode name can stop progress even when all other paths are correct. If files open but the stream has gaps or timing problems, investigate the media compatibility and duration assumptions rather than continuing to alter paths.
Test changes in the actual service environment
A useful test recreates the conditions that matter: the same FFmpeg binary, command, playlist, process account, and launch directory. A manual run as your own user is a preliminary check, not proof that the scheduled or supervised job will work. If the service runs continuously, plan a controlled test that does not interrupt a live broadcast unexpectedly.
Change one thing at a time. For example, first correct a filename, then retry and capture the result. If the next error refers to safe-path validation, decide whether to use safe relative names or disable that check for a trusted list. If the test then reaches an input compatibility error, treat it as a new issue rather than concluding that the path change failed.
For HLS output, test both file creation and delivery. Confirm that FFmpeg writes the manifest and segments to the intended locations, and that the manifest entries correspond to those segment names. The HLS muxer documents -hls_segment_filename for controlling segment filenames and -strftime_mkdir 1 for creating directories represented by expanded segment filename patterns. Those settings are relevant to output layout; use them only when your command writes HLS and the desired directory structure calls for them. If a web server or player consumes the output, test that side as well.
Keep a known-good copy of the playlist and command before edits. Once a fix works in the service context, record the exact path convention and permissions that made it work. That makes the next episode addition less likely to reintroduce a laptop-only path or a file the service account cannot read.
If the repeated operational burden is keeping a computer running and recovering a broadcast after interruptions, StreamNeo removes that particular machine-on requirement: you upload the video once, provide the YouTube stream key, and the channel can continue with your computer switched off. It does not change the need to prepare the media and channel correctly, and it is YouTube-only.
For related playlist issues, the FFmpeg playlist skipping files troubleshooting guide is relevant when FFmpeg opens a list but skips entries. If your concern is maintaining a continuous loop rather than path resolution, see how to keep a YouTube playlist stream running during an internet outage. If you are considering another playback approach, the article on repeating a product demo on YouTube Live without a gap covers a different loop problem.
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
Are FFconcat and HLS M3U8 playlists interchangeable?
No. FFconcat scripts list files with file directives for the concat demuxer, while HLS manifests refer to segments and sometimes variant playlists. Use the errors and command to establish which format FFmpeg is reading or writing before choosing a fix.
Does -safe 0 fix absolute paths in an FFconcat list?
It disables the concat demuxer’s safe-path rejection, which can allow absolute paths, but it does not fix missing files or permissions. Use it only for a playlist whose contents you control, and do not apply it as a fix for HLS references.
Are relative playlist paths based on the playlist file or the process?
Do not assume a universal base directory from the filename alone. Inspect the exact invocation and launch context, then test the entries as the account that runs FFmpeg; confirm the behaviour of the installed build in that environment.
Why do files open but the podcast still has gaps?
Path resolution only confirms that FFmpeg can find inputs. The concat demuxer expects compatible streams, codecs, and time bases, and timing can be affected by duration information, so inspect those properties if paths now work but playback does not.