VidiCore Release Notes

The following items on the list encompass breaking changes, features, and fixes that are relevant for the major release.

Visit this page to get an overview of all releases and patches.

Patches for this major release

Patches are shown based on recency and will appear below as soon as they are issued.

Release 25.4.0-0-gfd6ad501af on 26 Jan 2026

Version information

Component	Version
VidiCore	25.4.0-0-gfd6ad501af
VSA Docker Image	25.4.0-0-g964ca6c5-25.4.23-20260129.2
Transcoder	25.4.19

VidiCore Prerequisites

For more information on the prerequisites, please check the Prerequisites section of the API doc:

Link to VidiCore Prerequisites.

Cancellations & Deprecations

The MXFServer integration is no longer supported and will be removed in a future VidiCore release.

Features

New implementation of search and indexing (“Search2” - Beta)

The new implementation of searching and indexing (“Search version 2”) introducted in VidiCore 25.2 was further improved and documentation clarified - see table below for details.

A major new feature in Search2 is the support of separate indices for multiple languages. It allows to specify a language for search calling, causing VidiCore to use the search index for the corresponding language. This feature is turned off by default to maintain compatibility with existing installations. Refer to VidiCore’s API documentation for details and how to enable the feature.

Item#	Item
239717	Search2: Implement multi-language search
242158	Search2: Document limitation on matching different groups
242032	Search2: Adapt wild card search to work on the original string
241353	Search2: Highlighting generic and timed fields
241271	Search2: Document that ranges are possible with keywords but unreliable
241231	Search2: Use global aggregation to implement facet excludes
240806	Search2: Improve exception handling
240634	Search2: Support external IDs in Search2 when filtering for collection-id related transient fields
240633	Search2: Improvements when doing large reindexing tasks
239877	Search2: Ensure that re-indexing also works in Search2 when using external IDs in API calls
242297	Search2: Implement autocomplete call
241781	Search2: Text sorting throws exception
240668	Search2: Examples in documentation of new search endpoint are not working
239931	Search2: Range queries do not work
239928	Search2 only returns relative thumnail URIs
242548	Search2: Add search call comparison for migration documentation

Features and improvements for VSA, media operation, and file state handling

Growing-file support was significantly improved for MXF files. Please ensure to use the matching VidiCoder version for using the full potential and refer to the VidiCoder changes below for configuration.

Item#	Item
242233	Improved growing-file handling for VidiCore with Nablet demuxer (avoid `THUMBNAIL` jobs to fail on growing files)
242184	Add `resourceTag` parameter for API call for creating thumbnails on a shape
242102	Prefer own storages for VSA-to-VSA inter-storage transfer jobs
241561	Prefer own storages for VSA-to-VSA direct transfer jobs
241402	Direct transfers between VSAs should be distributed across all matching source VSAs
241902	Speed-up processing of growing input files read via http proxy
241610	Implement FTPS transfer
241197	Improve VSA direct upload speed by caching authentication results
241390	Improve VSA direct upload and WebDAV writing speed by increasing write buffer size to 1 MByte and making it configurable
241889	VSA setting `useWatchService` should default to `false` if `operationMode=VSA-VS`
240796	Support cix&tix creation on growing chunked proxy on S3 bucket
241352	Support `TRANSCODE`/`CONFORM` on growing chunked source files on S3 buckets
238347	Pass stream pids to VidiCoder `ComplexJob` and `NLEJob` for Nablet demuxer support
236883	Implement API to normalize media files
235289	VSA transcode/conform jobs should work even with `vxa://` URIs on multiple VSAs
241564	Monitoring of VSA storages stops when files with non-ASCII characters are on the storage
242072	Support scanning of files and folders with non-ASCII characters on VSA storages
242261	VSA http proxy should respect VidiCore file state for growing files (`CONFORM` on small source files times out via VSA http proxy)
242501	API call for raw file data upload directly via VSA
241972	VSA is creating additional storage for existing share in VidiCore during startup (only when `useWatchService=false`)
241970	VSA shares created for newly added storages during VSA startup are deleted directly after creation
242054	http proxy is waiting too much if growing input file is read at the end (instead of sequentially)
239691	VSA settings `operationMode=VSA-VS` and `useWatchService=false` contradict each other
240324	Raw-passkey import does not work with multiple VSAs
240248	(S)FTP and file size detection does not work on VSA storages with new WebDAV implementation
235206	`.srt` files are not listed on VSA storages (make file extensions excluded from WebDAV file systems listing configurable)
242469	Thumbnail resources on a VSA do not work with new WebDAV implementation
242462	Storage wrongly is added for pre-configured thumbnail resource during VSA startup
242461	VSA shares created for newly added thumbnail resource/export locations during VSA startup are wrongly deleted directly after creation (when `useWatchService=false`)
242526	Actual error message gets lost in passkey upload via VSA
241842	VSA is leaking threads when reconnecting to VidiCore
240594	Instable/overloaded VSA-VidiCoder connection (results in transcoder resource intermediately going offline)
243199	Allow log level adjustment of VidiCoder (when running inside VSA)
242838	Set originalFilename system metadata field when doing direct VSA uploads into a file
242361	Support VSA to VSA transfer for growing files
241982	Allow VSA to be in multiple agentGroups

General features

DIVArchive can now be controlled by VidiCore using the current REST API instead of the previous JAVA API. Improvements for handling weak references are implemented.

In the VidiCore datamodel of previous version, the table t_bigtext was used to stored all larger text information and it was referenced from many other VidiCore database tables. With VidiCore 25.4, all these tables are extended with explicit columns to store the text information. The migration of the data can be done in two ways, that complement each other:

VidiCore will automatically migrate data from t_bigtext to the new explicit columns when it writes or updates text information.
Additionally, a new VidiCore cruncher service is available to migrate all data in a resource-efficient approach.

Item#	Item
242005	DIVArchive integration based on REST API
242315	Remove the need for `browse=true` on DIVA archive storages
241890	Improve logging of `StorageSupervisor` for deletion failures
241843	Add support for OpenSearch 2.19
241400	Performance improvements in `AccessControlCruncher`
241240	Change id column of `t_taskdefinition` to `bigint`
241060	Remove notification queue entry when it cannot be deserialized (avoid broken notifications blocking the system)
240936	Allow renaming of weakly referenced metadata groups
240893	Weak references to metadata groups in metadata documents can be specified via a group path instead of a uuid
240911	Re-enable logging to S3 bucket (got lost in 25.2)
240898	Allow creating OWNER ACL when entities have already other ACLs
240669	API to set owner on a entity without any access rights
240851	Select resource if `resourceTag` query parameter does not match any of the resource tag keys on the resource
239695	Replace `t_bigtext` for new systems
239974	Implement on-the-fly migration of `t_bigtext` contents in existing systems
237685	Allow start timecode override when importing STL subtitles (for STL files with wrong TCP value)
197837	Improve performance in FileSizeCruncher
242385	Add support for mime type image/webp
239036	Adding a child collection to a parent collection takes very long when parent collection has many items
242695	Add option to `POST /API/storage/(storageId)/file` for letting VidiCore create the filename
242559	Issue warning during VSA startup if entropy is too low

Documentation and configuration

These tickets were realized to simplify the configuration of VidiCore and improve the documentation:

Item#	Item
241220	Update prerequisites for VidiCore & VSA to RHEL9
241057	APIdoc: Clarify need for trailing slashes in storage URIs (also on `vxaLocalPath`)
241261	Clarify some aspects of subclip API usage

Fixes

Item#	Item
242176	`TranscoderResourceTag` problem caused by offline transcoder does not get resolved when transcoder goes online again
241980	MetadataFieldGroup cannot be updated after child group was removed from it (namespace General not found)
241939	Rendering sequences with external media fails
241779	Destination shape of `TRANSCODE`/ `CONFORM` job may have wrong technical metadata if destination file is on block storage and chunking is used
241718	Default database transaction timeout is not 5 minutes anymore
241681	Metadata of child collections are not included when fetching child collections of a collection
241655	Destination file stays in state `TO_APPEAR` when copying to a storage with `browse=false` (or chunkedWrite even if `browse=true`)
241651	Metadata inheritance rules do not work when condition is applied to field with dataset)
241628	JavaScript job steps running more than 5 minutes cause the job step to go into state `DISAPPEARED`)
241623	File uri missing from `FileDocument` when storage is offline
241453	VidiNet hash jobs take very long to update `FILE` entities after completion
241435	Large files on GCP storages lead to failing `PLACEHOLDER_IMPORT` jobs
241428	Resource tags are not respected in direct transfers between VSAs
241364	No longer possible to recreate search index
241297	Subtitle language specified via jobmetadata is not respected for VTT/SCC file import
241222	Thumbnails are not created in `/API/import/placeholder/(itemId)/container/raw/` even if `createThumbnails=true`
241056	New files in watch folders are not detected for VSA remote storages (FTP, SFTP, S3, DS3))
241012	Chunked file copy fails when source storage is of same type as destionation storage)
240957	`COPY_FILE` job in `WAITING` state not retried until source `FILE` state was set from `OPEN` to `CLOSED`
240905	Selftest API reports errors as version information can not be parsed)
240830	VidiCore fails with wrong media type on DV files imported with 25.1 or older (and other types with no pid in MediaComponents)
240514	Storage monitoring (`StorageSupervisorWorker`) dies silently, files are not deleted anymore
240353	Proxy index creation: physical files are not created
240054	Proxy index creation places index file into wrong path
239505	`GET /API/item/VX-1/uri` does not return an URI when container component has multiple formats
238688	Deletion locks are not properly inherited when locks are updated
238571	`CONFORM` should use fps from shape tag instead of 1st clip of timeline
236587	`__deletion_lock_expiry` is not triggering a notification upon modification
230857	OpenSearch client disconnects under load
196969	Subtitle language specified via `jobmetadata` is not respected for TTML file import
242649	Missing database retry results in occasional 500 errors on API calls
242616	Increased number of optimistic locking exceptions and `ConcurrentModificationException`s during job execution
242541	Jobs have `TranscoderBusy` problem although less than `transcoder.maxJobs` jobs are running on this transcoder (transcoder is not freed when last job step is asynchronous)
242535	`JobCruncher` dies due to `CancellationException` in `tickAll()`
242526	Actual error message gets lost in passkey upload via VSA
242120	ACL-Reindex stalls system

VidiCoder Features

Nablet Demuxer

VidiCoder now supports demuxing using Nablet. This highly improves processing of growing MXF files for transcoding, thumbnail creation, and further operations done by VidiCoder.

Please refer to VidiCore’s API documentation to enable this feature.

WebP Image Encoding

VidiCoder now can encode WebP image files. This can be especially useful for proxy images, as WebP files can maintain quality at higher compression, leading to significantly reduced file sizes compared to PNG and JPEG (depending on quality settings).

Please refer to VidiCore’s API documentation to use this feature.

VidiCoder Changes

Item#	Item
242190, 242311, 240006	Improve interoperation with http proxy in VidiCore/VSA for processing growing files
238369	Fixed MXF source file origin is not taken into account
242411	Fixed 16 bit AES audio in nablet demuxer
242163	Added support for WebP encoding
242080	Fixed double flush in Nablet h264 encoder
241102	Optimized nablet color conversion filter
241839	B-Frame support in CIX/TIX
241102	Very large performance optimization
241393	Prores framecopy iterations now run in parallel
241102	Fixed interval end evaluation for intra-frame coding
240006	Cix/Tix creation should make use of a smaller blocksize for reading
241008	Shape deduction for P2 card spanning clips
240055	Adding smart normalize to HDF01,02,03 and XDF01a
235861, 240177	Fixed HEIC image files not decoding when taken on ios18+ iPhone
240004	Implemented color space metadata handling in NabletMXFDemuxer
240005	Fixed handling of SRT CC in MXF
238302	Fixed extradata bug in Nablet MPEG-2 video encoder
242960	Fixed RDD25 proxy transcoding of 50p videos: use each frame instead of duplicating every second frame
242960	Videos with 50P are transcoded wrongly to Proxy
242190	THUMBNAIL job on growing/chunked proxy (s3) stops too early - NabletMXFDemuxer read timeout