Release notes from viseron

3.2.3 - Hotfix for camera disconnections

2025-08-25T06:33:33Z

This release adresses the issues #1075, #1076 and #1096

When a camera disconnected, the retry loop would sometimes get stuck, leaving cameras disconnected

Note: This is a re-release to trigger new builds due to failures of the previous one

What's Changed

fix race condition when stopping and starting a camera rapidly by @roflcoopter in #1097
attempt to free up space on azure agent by @roflcoopter in #1098

Full Changelog: v3.2.1...v3.2.3

3.2.2 - Hotfix for camera disconnections

2025-08-24T20:41:10Z

This release adresses the issues #1075, #1076 and #1096

When a camera disconnected, the retry loop would sometimes get stuck, leaving cameras disconnected

What's Changed

fix race condition when stopping and starting a camera rapidly by @roflcoopter in #1097

Full Changelog: v3.2.1...v3.2.2

3.2.1 - Hotfix for tier check exception

2025-08-03T22:07:19Z

This release addresses the issue with the uncaught exception in #1068, which also caused object detection to stop working in some circumstances

Full Changelog: v3.2.0...v3.2.1

3.2.0 - Proper live streaming

2025-07-17T18:26:30Z

Proper live streaming is now available in Viseron, utilizing the great go2rtc application.
Previously Viseron resorted to using the MJPEG stream for live streaming, which results in low quality and high latency.
With the new go2rtc component, Viseron can now stream using WebRTC or MSE, which gives a much better experience.
The new Live page is available at /live and can be accessed from the sidebar.
The Live page is very similar to the Events page, using the same responsiveness so that you can view multiple cameras at once.

In order to use it, you will need to update your configuration to include the new go2rtc component.

Breaking changes

port in webserver is now deprecated. In order to support the new go2rtc streaming server, an NGINX reverse proxy is now included in the container, which listens to the old default port 8888

New features

A new Live Page is now available, which allows for better live streaming and support for more protocols.
- The Live Page is available at /live and can be accessed from the main menu. It utilizes the new go2rtc component to support WebRTC/MSE streaming.
- The MJPEG stream can now be accessed at the Live page by using the toggle in the extra options menu of the video player.
A new go2rtc streaming server is now included in the container. This allows for better live streaming and support for more protocols.
NGINX reverse proxy - A reverse proxy is now included in the container to enhance performance and provide live streaming through go2rtc.
A new YOLOcomponent which uses Ultralytics YOLOv8 for object detection is now available.
You can now directly go to a specific timestamp on Events Page by using query parameters. More information on this can be found in the documentation.
The EdgeTPU component can now utilize multiple EdgeTPU devices for better performance. See the updated documentation for more information.
Fullscreen button is now available on the Events Page (and the new Live Page) to allow for a better viewing experience.
Does not work in iOS, because Apple.
New way to calculate which files to move or delete in the tiered storage system.
The new implementation uses numpy instead of pure SQL queries, which should improve performance at the cost of slightly (temporary) increased memory usage.
You can limit the CPU usage of this calculation by setting the new tier_check_cpu_limit config option.
You can further tune the workload using the config options tier_check_batch_size, tier_check_sleep_between_batches and tier_check_workers.

Changes

HLS.js upgraded to 1.6.2
If you have a substream with different resolution than the main stream, the video player now gets the correct resolution from the main stream
EdgeTPU devices will now be restarted in case of multiple errors, helping to recover from temporary issues.
Segment processing is now done in a separate process, for better performance and stability of the main process.
Some unusual circumstances may cause the size based cleanup to fail. If thats the case and the first tier is full, a tier check as well as a cleanup job will be scheduled to remedy that.

Fixes

Camera name overlapping the video stream controls on the Events Page
PTZ component failing to initialize due to missing devicemgmt.wsdl file
Fix for DeepStack face recognition not working
Includes coco.names file in the RPI3 build, which is required for the Darknet object detection model.
Properly resubscribe to MQTT topics when the connection is lost.

Developer experience

Fixes a long standing issue with the Devcontaner failing to start due to a missing config folder.
When cloning using the instructions in the Developer docs, the config folder was not present on the host since cloneInVolume was used.
The config folder is now dynamically set to the correct path in the Devcontainer using the env variable VISERON_CONFIG_PATH.

Docker images will be on Docker Hub shortly

All Changes

pin gpac version to 2.4.0 by @roflcoopter in #983
set requested timestamp from query parameter by @roflcoopter in #984
expose main stream resolution to get correct aspect ratio for streams by @roflcoopter in #988
ptz: find location of .wsdl file by @roflcoopter in #989
Go2rtc and NGINX by @roflcoopter in #991
implement EdgeTPU support for multiple devices by @roflcoopter in #992
add title to config.yaml code blocks by @roflcoopter in #993
edgetpu restart on failure by @roflcoopter in #994
read frames in subprocess by @roflcoopter in #995
merge master back into dev by @roflcoopter in #1002
Live page by @roflcoopter in #1007
run fragmenter in subprocess by @roflcoopter in #1008
trigger tier check when fragmenter cant move file to first tier by @roflcoopter in #1014
remove config bind mount and change its location in devcontainer by @roflcoopter in #1015
upgrade to hls.js 1.6.5 by @roflcoopter in #1017
camera snapshot auth using cookies by @roflcoopter in #1016
add ssh to devcontainer by @roflcoopter in #1022
properly close queues in ChildProcessWorker by @roflcoopter in #1023
wrap start of queue manager server in a lock by @roflcoopter in #1024
include logo as png by @roflcoopter in #1025
catch TypeError in duplicate log filter by @roflcoopter in #1026
update throttling mechanism with unique throttle key by @roflcoopter in #1028
add index to Files table to improve tier check performance by @roflcoopter in #1029
make sure only one cleanup job can be started at a time by @roflcoopter in #1032
delete unneeded binaries and docker images in ci by @roflcoopter in #1037
fix timing issue with _concatenate_fragements by @john- in #1039
Tier check file processing by @roflcoopter in #1040
copy coco.names to rpi image by @roflcoopter in #1042
fix: update deepstack face detection coordinates to use correct keys by @roflcoopter in #1041
yolo component by @john- in #1033
extract tag from docker compose file before tagging by @roflcoopter in #1043
resub on mqtt reconnection by @roflcoopter in #1046
make mjpeg stream viewable from Live page by @roflcoopter in #1047
force pytorch CPU install in CI by @roflcoopter in #1048
cap HLS error to max 200 chars by @roflcoopter in #1051
enable react query retries by @roflcoopter in #1052
fix cleanup jobs to run on all tier paths by @roflcoopter in #1053
add live view documentation by @roflcoopter in #1054

New Contributors

@john- made their first contribution in #1039

Full Changelog: v3.1.2...v3.2.0

3.1.2 - EdgeTPU Image classification fix

2025-05-15T06:45:03Z

Resolves an issue with EdgeTPU image classification which caused it to not function at all. Closes #1000

Full Changelog: v3.1.1...v3.1.2

3.1.1 - Camera storage config fix

2025-04-17T20:49:01Z

Hotfix release to resolve an issue where you couldnt specify recorder or snapshots without also specifying the other under a cameras storage config

What's Changed

fix undefined check for recorder and snapshot tiers by @roflcoopter in #981
3.1.1 by @roflcoopter in #982

Full Changelog: v3.1.0...v3.1.1

3.1.0 - User and Role Management

2025-04-16T10:01:51Z

Multiple users, finally! Viseron has had authentication for quite some time, but you could only create one initial admin user during the onboarding.

This release brings a new User mangement page to the frontend, and with it a new Settings page.

Roles

Users can be assigned different roles: admin, write or read.
See the updated documentation: https://viseron.netlify.app/docs/documentation/configuration/authentication#roles

Assign cameras

If you want to restrict what cameras each user can see, you can do so by assigning specific ones to each user.
See the updated documentation: https://viseron.netlify.app/docs/documentation/configuration/authentication#assigning-cameras

New features

Masks can now be configured for Post processor. They are configured the same way as masks for object/motion detection. A new config option draw_post_processor_mask is also available for the MJPEG streams in order to debug the masks
Discord notifications component, provided by @InterN0te
Gotify notifications component, provided by @InterN0te
New config option use_last_snapshot_on_error for FFmpeg/GStreamer. It allows for returning the previously retrieved snapshot for a camera incase of a flaky connection
New filter option on Events page to group cameras.
New filter option on Events page to adjust the requested timestamp with the configured lookback.

Fixes

Wait for file updates when shutting down and using move_on_shutdown: true
Fix Docker compose documentation examples
Fix Codeproject AI Face recognition training. Previously faces were never recognized when training
Fix Compreface failing to start when use_subjects was set to false
Avoid reloading page when submitting login form. This caused some issues where you sometimes had to login twice
Events are now properly sorted by timestamp in the details popup on the Events page
When clicking on an event in the Events page, a snapshots is shown instead of video if its not available.
This calculation is now more forgiving and now has a 10 seconds grace period

Docker images will be on Docker Hub shortly

All Changes

fix typos in docker compose examples by @roflcoopter in #930
catch OSError when closing ffmpeg pipe by @roflcoopter in #935
make sure we wait for file size update timers on shutdown by @roflcoopter in #934
fix cpai face recognition training by @roflcoopter in #939
new config option use_last_snapshot_on_error for still_image by @roflcoopter in #940
fix mask image coordinates being reversed by @roflcoopter in #943
Add Notification Category to Components Explorer by @InterN0te in #947
Discord component for notifications by @InterN0te in #944
Add Gotify notification component by @InterN0te in #941
Post processor masks by @roflcoopter in #953
update docs on how to add timestamp to recorded video by @roflcoopter in #954
allow for grouping events for different cameras by @roflcoopter in #955
better handling of compreface initialization by @roflcoopter in #962
User management by @roflcoopter in #965
Feature/auth improvements by @roflcoopter in #968
assign cameras to users by @roflcoopter in #969
properly cleanup axios interceptor and add tests for AuthContext by @roflcoopter in #970
updated required roles for some endpoints by @roflcoopter in #972
docs: clarification on events and timeline by @roflcoopter in #973
add lookback adjustment when clicking on an event by @roflcoopter in #974
Fix event sorting on the Events page by @roflcoopter in #975

New Contributors

@InterN0te made their first contribution in #947

Full Changelog: v3.0.0...v3.1.0

3.0.0 - Continuous Recordings

2025-03-15T08:04:29Z

This release is a big one. It mainly focuses on continuous recordings, but also includes a lot of other features and fixes.

I have been working on this release for a long time and I am very happy with the result. I hope you will enjoy it as much as I do.
The release cadence has been very slow lately and for that I am sorry. I will try to make smaller releases more often in the future.

Breaking changes

New docker volumes are required. Please see the updated docker documentation
Old recordings will not be compatible with the new version and wont be viewable in the UI
Files will no longer be stored in the /recordings folder, and should be removed from your docker volumes
Recordings are now in HLS format. If you need to have the recordings saved as full MP4 files you can use the new config option create_event_clip, or use the new Export feature in the UI.
mp4 file extension is now forced and extension is removed
segments_folder is removed in favor of tier paths from the new storage component
Thumbnails are now stored in /thumbnails folder (depending on the tier config)
filename_pattern in thumbnail is removed, stored as ID of the recording
audio_codec behavior changed. It is now only used as stream information. To transcode Audio, set audio_codec under recorder
trigger_recorder is removed in favor of trigger_event_recording
The FFmpeg command has been changed, so if you use raw_command you need to update it

New features

24/7 recordings

The main feature included in this release is the option of having 24/7 recordings.
With this comes a new component, storage which will allow you to retain data based on time (like it is today) and based one consumed space.
This allows you to use your storage to its maximum.

Documentation for how the recordings work can be found here

Updated screenshots for the frontend can be found here

Storage Tiers

Another feature of the storage component are tiers.
They allow you to store recordings on multiple media.
Say you have one 256GB SSD and a 1TB HDD, you can keep the most recent recordings on the SSD for fast access while offloading the older recordings to the HDD for archiving.

Timeline

To easily view 24/7 recordings, a new Timeline view is available.
It currently shows continuous recordings and different events such as motion, detected objects, face recognition etc.

Events

The Events view is a more concise view than the Timeline.
Here you can scroll through and view all the Events.

Other new features

A new safe mode has been added. If Viseron cant startup due to config errors, it will restore the last known good config and start up with a minimum amount of components. This will allow you to fix the config in the UI and restart Viseron without having to manually restore the config.
Allow empty password for cameras
A logfile is now created in the /config folder. This will allow you to see the logs without having to enter the container.
New telegram component provided by @smelis, which allows sending of clips and thumbnails using Telegram.
New ptz component provided by @smelis, which can be used by the Telegram component to control a camera.
The PTZ functionality will be ported to the web UI as well at some point.
max_recording_time has been added to the recorder component. This will allow you to set a maximum recording time for each recording. This is useful if you have a lot of motion events and want to keep the recordings short.
record_only has been added to the recorder component. This will allow you to record without running any detectors.
still_image_width and still_image_height has been added to the camera component. This will allow you to set the resolution of the still images.

Changes

A PostgreSQL database is now running inside the container. It is not possible to use an external database at the moment.
Upgraded to Ubuntu 22.04 which gives updated drivers to support newer hardware
Updated to React 18
Updated to Python 3.10
Updated to OpenCV 4.10.0
Updated to FFmpeg 5.1.2
Updated CUDA
Updated DLIB
Updated OpenCL
Updated VA-API
Updated all Python dependencies
The documentation has been updated with a lot of new chapters

Fixes

Camera passwords are now escaped, meaning that passwords with special characters are now supported.
This has been a long standing issue which caused FFProbe to fail when trying to connect to the camera.
Properly redact sensitive information in logs
Reduce the occurence of KeyError: UUID in the logs
Fix a long standing bug where frames with drawn items (like bounding boxes, masks etc) would bleed through to other frames.

Docker images will be on Docker Hub shortly

v3.0.0b13 - Final beta

2025-03-10T20:49:20Z

This beta will be the last one before the release of 3.0.0, unless some big error is found.
The release focuses on stability improvements and polishing of existing features.
Some new features are also included.

New documentation chapters have also been added which aims to explain how recordings and Viseron in general is working.
New sections will be added in the future to highlight common usecases.
https://dev--viseron.netlify.app/docs/documentation/configuration

Breaking changes

The /recordings folder is no longer used, and should be replaced with the /event_clips folder.
This namechange was made in order to clear up some confusion on what this folder is storing in v3.
Docker instructions have been updated: https://dev--viseron.netlify.app/docs/documentation/installation#running-viseron
trigger_recorder has been deprecated in favor of trigger_event_recording

New features

Faces to recognize can now be pulled from Compreface directly
A logfile is now created in the config directory. Docs have been updated accordingly: https://dev--viseron.netlify.app/components-explorer/components/logger#view-logs
You can now download events and snapshots from the Events page.
You can also specify a timespan for 24/7 recordings which will be downloaded as an MP4 file.
Documentation can be found here: https://dev--viseron.netlify.app/docs/documentation/configuration/recordings#downloading-recordings
New config option record_only which makes it possible to create continuous recordings without running any type of detectors.
New config option width and height for still_image which allows setting the resolution of the still image
You can now click individual events in the popup windows on the Events/timeline view to seek to the event
New telegram component provided by @smelis, which allows sending of clips and thumbnails using Telegram.
New ptz component provided by @smelis, which can be used by the Telegram component to control a camera.
The PTZ functionality will be ported to the web UI as well at some point.

What's Changed

add empty config folder to allow devcontainer to start by @roflcoopter in #843
remove devcontainer config mount by @roflcoopter in #846
pin types-requests to 2.31.0.6 by @roflcoopter in #847
added new functionality to use compreface subjects for facial recognition by @tantonj in #845
Documentation updates by @roflcoopter in #850
add logging to file by @roflcoopter in #852
wait for move_on_shutdown to complete by @roflcoopter in #853
allow dates with only recorded video to be selected by @roflcoopter in #854
use yield_per to avoid cursor closing in _check_tier by @roflcoopter in #856
dont materialize when using yield_per by @roflcoopter in #857
fetch all rows to avoid cursor closing by @roflcoopter in #858
Avoid importing dlib when it is not used by @roflcoopter in #865
Allow download of events and snapshots by @roflcoopter in #869
check for nan value when clicking the LIVE button by @roflcoopter in #870
reload hls every 5 seconds on fatal errors by @roflcoopter in #871
reload page when getting a vite preloadError by @roflcoopter in #872
use getCameraNameFromQueryCache in Breadcrumbs by @roflcoopter in #873
apply object detection mask the same way as motion detection by @roflcoopter in #874
properly cleanup after test_setup_components by @roflcoopter in #875
add category and subcategory to OrphanedRecordingsCleanup query by @roflcoopter in #876
allow pressing enter to login by @roflcoopter in #877
remove future=True from sessionmaker by @roflcoopter in #878
add screenshots and videos to event download instructions by @roflcoopter in #879
change bash to shell in docs codeblocks by @roflcoopter in #880
Continuous recording config parameter by @roflcoopter in #881
Async Websocket API by @roflcoopter in #883
improve shutdown sequence and make sure all threads are stopped by @roflcoopter in #884
Doc typos and FFmpeg timestamp example by @roflcoopter in #885
Update docker.mdx - --env-file vs --env by @nneul in #887
fix ffmpeg log pipe closing by @roflcoopter in #888
remove files_meta table to avoid costly joins by @roflcoopter in #890
add gitsponsors to README by @roflcoopter in #891
record_only config option by @roflcoopter in #892
fix mjpeg stream exception by @roflcoopter in #893
further asyncio event loop improvements by @roflcoopter in #894
upgrade mypy to 1.14.1 by @roflcoopter in #895
Mount config folder in repo to /config in devcontainer by @roflcoopter in #896
fix so that opening drawer doesnt rerender page by @roflcoopter in #897
add config option to set still image resolution by @roflcoopter in #898
remove gitsponsors by @roflcoopter in #899
add error boundaries to frontend by @roflcoopter in #900
Events and timeline navigation improvements by @roflcoopter in #904
deprecate trigger_recorder in favor of trigger_event_recording by @roflcoopter in #905
Component explorer link fix by @roflcoopter in #911
Rename /recordings to /event_clips by @roflcoopter in #912
add direct navigation to timeline by @roflcoopter in #913
allow overrides for each snapshot domain by @roflcoopter in #916
PTZ and Telegram Components by @smelis in #788
Fixes to the new PTZ and Telegram components by @roflcoopter in #918
update images and descriptions for ptz and telegram by @roflcoopter in #919

New Contributors

@tantonj made their first contribution in #845
@smelis made their first contribution in #788

Full Changelog: v3.0.0b12...v3.0.0b13

v3.0.0b12 - Bugfixes and cleanup jobs

2025-02-09T21:32:39Z

This release includes a multitude of fixes and changes to improve the stability of Viseron.
It also adds multiple clean up jobs to keep the database and filesystem in sync in case of any crashes.

If you have tried the v3 beta before and experienced issues, i highly suggest trying this release out since it contains loads of fixes.

Fixes

Remove volumes from the Dockerfile to avoid issues with bind mounts
Use the process watchdog for Darknet so that it is restarted in case of crashes
Add extra logging to shutdown procedure to pinpoint threads and processes that is slowing it down
Allow empty passwords for cameras
Run every thread using RestartableThread so that it gets awaited during shutdown
Fixes the move_on_shutdown: true config option. Previously Viseron did not wait for the callback to finish before exiting.
A condition was missing when searchng for files for move_on_shutdown causing files to be moved multiple times
Delete from FilesMeta table manually instead of using a trigger
Properly throttle the function that checks if any files should be moved. This cause issues with attempting to move files multiple times
Event clips are now generated in the servers timezone instead of UTC

Cleanup jobs

Multiple cleanup jobs now run regularly to make sure to keep the database and filesystem clutter free.
For instance if you delete files manually while Viseron is not started, the cleanup jobs will make sure that they are removed from the database.

They are intentionally throttled to run quite slow to not degrade performance of your host.

OrphanedFilesCleanup (Mondays, 3 AM)
- Removes files that are no longer associated with any recordings or events
OrphanedDatabaseFilesCleanup (Wednesdays, 3 AM)
- Cleans up database entries that point to non-existent files
OrphanedFilesMetaCleanup (Fridays, 3 AM)
- Removes metadata for files that no longer exist
EmptyFoldersCleanup (Daily at midnight)
- Removes empty folders to keep storage structure tidy
OrphanedThumbnailsCleanup (Daily at midnight)
- Removes thumbnail images that are no longer linked to any recordings
OrphanedClipsCleanup (Daily at midnight)
- Removes video clips that are no longer associated with any events
OrphanedRecordingsCleanup (Daily at midnight)
- Removes recordings that are no longer linked to any events
OrphanedPostProcessorResultsCleanup (Daily at midnight)
- Removes post-processing results that are no longer needed
OrphanedObjectsCleanup (Daily at midnight)
- Removes detected object data that is no longer associated with any events
OrphanedMotionCleanup (Daily at midnight)
- Removes motion detection data that is no longer linked to any events
OldEventsCleanup (Daily at midnight)
- Removes expired system events based on retention settings

For discussions, please use this discussion: #721

Full Changelog: v3.0.0b11...v3.0.0b12