An element the enum cannot parse throws, and the configuration manager
catches that and returns defaults, so one bad preset discarded every other
encoding setting.
- serialize EncoderPreset through a string surrogate in XML
- fall back to auto for an empty, unknown or out-of-range value
Dolby vision files having unexpected transfers now also marked as invalid, and it will now get its base video range from its base layer color transfer, as not all invalid dolby vision files are HDR now.
ImageHelper.GetNewImageSize passed the caller-supplied width/height straight
through to SkiaEncoder.EncodeImage, which allocates an SKImageInfo of exactly
that size. Nothing bounded those values against the source image, so a request
like Items/<id>/Images/Primary?width=23100&height=23100 made the server allocate
and resample a 23100x23100 surface from, say, a 600x336 poster: the reporter
measured 100% of a core for 10-15 minutes and 6-12 GB resident per request.
The item images endpoints do not require authentication, so any caller who knows
an item id can trigger this, and varying the size by one pixel misses the cache
every time.
Add DrawingUtils.ScaleDownToFit, which scales a size down uniformly until it
fits inside a bounding box and returns it unchanged if it already does, and
apply it in GetNewImageSize against the original image dimensions. Requests
that ask for more pixels than the source now get the source resolution back,
scaled to the requested aspect ratio. Downscaling paths are untouched, and
DrawingUtils.Resize keeps its existing behaviour for the transcoding callers in
EncodingJobInfo and StreamInfo, which legitimately size video output.
ResizeFill already refused to upscale; this makes width/height consistent
with fillWidth/fillHeight.
Fixes#17056.
Allow external subtitle files (SRT, ASS, PGS, etc.) to be muxed into
MKV output containers when the device profile requests Embed delivery.
Previously, the IsExternal guard in GetSubtitleProfile excluded external
subtitles from Embed consideration entirely, forcing them to be served
as separate sidecar files even when the output container supports
embedding.
Changes:
- Extract CanConsiderEmbedSubtitle in StreamBuilder to allow external
subs through when transcoding to MKV
- Add external subtitle file as FFmpeg input (-i) for Embed delivery
- Map external embedded subs from the correct FFmpeg input index
- Fix external audio map index to account for the new subtitle input
- Extract NeedsExternalSubtitleMuxing in EncodingHelper to deduplicate
the external subtitle input check
Fixes#16403
Wire up EnableSubtitleExtraction config to MediaEncoder.CanExtractSubtitles
so the setting is actually respected. Gate subtitle extraction check behind
PlayMethod.Transcode since DirectPlay has no competing ffmpeg process.
Add parameterized tests for StreamBuilder.GetSubtitleProfile covering
text and graphical codecs, profile format matching, and extraction
setting behavior. Remove misplaced SubtitleEncoder extraction test.
Previously, images with no language were ranked higher (score 3) than
English images (score 2), causing poorly rated languageless images to
be selected over well-rated English alternatives for posters and logos.
Swap the priority so English is preferred over no-language images.
Backdrop images are unaffected as they have their own dedicated sorting.
Add unit tests for OrderByLanguageDescending.
Fixes#13310
StreamInfo.ToUrl() generated URLs like `/master.m3u8?&DeviceId=...` (note `?&`)
because `?` was appended to the path and all parameters started with `&`. When
the first optional parameter (DeviceProfileId) was null, the result was a
malformed query string.
This is harmless when clients hit Jellyfin directly (ASP.NET Core tolerates `?&`),
but when accessed through a reverse proxy that parses and re-serializes the URL
(e.g. Home Assistant ingress via aiohttp/yarl), `?&` becomes `?=&` — introducing
an empty-key query parameter. ParseStreamOptions then crashes on `param.Key[0]`
with IndexOutOfRangeException.
Changes:
- StreamInfo.ToUrl(): Track query start position and replace the first `&` with
`?` after all parameters are appended, producing valid query strings
- ParseStreamOptions: Guard against empty query parameter keys
- Tests: Remove .Replace("?&", "?") workaround that masked the bug
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The DisplayTitle property was using .NET's CultureInfo.GetCultures(NeutralCultures)
to resolve language display names. Since zh-CN is a specific culture (not neutral),
it would fall back to the base 'zh' code, resulting in generic 'Chinese' instead
of 'Chinese (Simplified)'.
This change adds a LocalizedLanguage property to MediaStream that gets populated
via LocalizationManager.FindLanguageInfo() when streams are retrieved from the
database. This leverages Jellyfin's existing iso6392.txt mappings which correctly
map zh-CN to 'Chinese (Simplified)'.
The same pattern is already used for other localized strings like LocalizedDefault
and LocalizedExternal.
The DisplayTitle property was using .NET's CultureInfo.GetCultures(NeutralCultures)
to resolve language display names. Since zh-CN is a specific culture (not neutral),
it would fall back to the base 'zh' code, resulting in generic 'Chinese' instead
of 'Chinese (Simplified)'.
This change adds a LocalizedLanguage property to MediaStream that gets populated
via LocalizationManager.FindLanguageInfo() when streams are retrieved from the
database. This leverages Jellyfin's existing iso6392.txt mappings which correctly
map zh-CN to 'Chinese (Simplified)'.
The same pattern is already used for other localized strings like LocalizedDefault
and LocalizedExternal.
* Check MaxAudioChannels for directAudioStream candidates
The current stream builder logic does not check the channel limit when determining if the audio stream can be directly used, and this can cause some undesired effects:
- A high channel count surround sound stream might be picked even if a stereo one exists when the user requires stereo audio.
- The user's preferred audio codec might not be respected during the downmix because the requested codec is now forced to be the same as the original source.
Signed-off-by: gnattu <gnattuoc@me.com>
* Fix unit test
Signed-off-by: gnattu <gnattuoc@me.com>
* Set correct transcode reason and target channels for unit test
Signed-off-by: gnattu <gnattuoc@me.com>
* Match old stream selection behavior
Signed-off-by: gnattu <gnattuoc@me.com>
* Fix reason matching
Signed-off-by: gnattu <gnattuoc@me.com>
---------
Signed-off-by: gnattu <gnattuoc@me.com>
* Correct MIME types for comicbook file extensions
cb7, cba, cbr, cbt and cbz all refer to different types of digital
comicbooks. The last letter of the extension indicates the compression
algorithm that was used: 7zip, arc, rar, tar or zip.
All these filetypes used to have the `application/x-cbr` MIME type
assigned to them. However, that has since been deprecated and was
replaced with
- `application/vnd.comicbook-rar` for rar compressed files and
- `application/vnd.comicbook+zip` for rar compressed files.
Only these two are officially listed by IANA
https://www.iana.org/assignments/media-types/application/vnd.comicbook+zip
. cbr and cbz are by far the most common file extensions for comicbooks.
There's no official MIME type for cb7, cba or cbt files. However, with
rar being a proprietary compression algorithm, FOSS applications will
often refuse to handle files that identify themselves as
`application/x-cbr`, so I decided to assign extension specific MIME
types to them. I've seen these being used by other applications,
specifically comic book readers.
I've read through the docs on iana.org, but haven't figured out why they
chose `-rar`, but `+zip`.
* Add conversions from MIME type to file extensions for comicbook formats
cb7, cba, cbr, cbt and cbz all refer to different types of digital
comicbooks. The last letter of the extension indicates the compression
algorithm that was used: 7zip, arc, rar, tar or zip.
All these filetypes used to have the `application/x-cbr` MIME type
assigned to them. However, that has since been deprecated and was
replaced with
- `application/vnd.comicbook-rar` for rar compressed files and
- `application/vnd.comicbook+zip` for rar compressed files.
Only these two are officially listed by IANA
https://www.iana.org/assignments/media-types/application/vnd.comicbook+zip
. cbr and cbz are by far the most common file extensions for comicbooks.
There's no official MIME type for cb7, cba or cbt files. However, with
rar being a proprietary compression algorithm, FOSS applications will
often refuse to handle files that identify themselves as
`application/x-cbr`, so I decided to assign extension specific MIME
types to them. I've seen these being used by other applications,
specifically comic book readers.
* Update CONTRIBUTORS.md
unkin-agent
gnattu