0.11.0 Release Candidate 1

[blakeblackshear]

Major Changes for 0.11.0

Link to updated docs: https://deploy-preview-2829--frigate-docs.netlify.app/

Changes since beta 7

• Fixed "Send to Frigate+" for addon users

BREAKING CHANGES

• Update to ffmpeg5 will break any custom input arguments using the stimeout parameter. This needs to be renamed to timeout in your config.
• Some amd64 users may need to set the environment variable LIBVA_DRIVER_NAME_JELLYFIN=i965 to get similar hardware acceleration performance to previous builds.
• Upgrading to the beta will run migrations on the database file that are not backwards compatible. In order to rollback to 0.10.0, you will either need to restore an old database file or delete the database files so they can be recreated.
• This version includes a significant rework of the docker build. This may have broken ffmpeg hardware acceleration for some hardware that was working previously. Please open an issue if this happens so we can try and track down what's missing.
• Some deployment approaches are overwriting the PATH environment value to an old value. If you see errors like No such file or directory: 'ffmpeg', then you need to stop overriding your PATH. This happens with both Proxmox LXC and Portainer ([Support]: Beta11 Can't Find FFMPEG #3223)
• The max_seconds config setting under events has been removed as it no longer has any effect

Frigate+ Integration

With this release, users can easily send snapshots to Frigate+ for labeling directly from the Events view. Note that Frigate+ is still in development, and custom models are not yet available. There is no cost to create an account and submit labeled images. Making it simpler to submit images should accelerate the build out of a robust training set.

To enable the Frigate+ functionality, you need to obtain an API Key from your Frigate+ account on the settings page and set it via the PLUS_API_KEY environment variable. I am working on getting this added into the Home Assistant Addon as well.

Here is the roadmap for Frigate+:

• Launch web version allowing manual upload of training examples and labeling
• Integrate submission in Frigate
• Launch Frigate optimized models
• Launch individually optimized models
• Integrate labeling and suggestions in Frigate

Updated Events View

• Speed. The events view has been completely revamped and should be much faster than it was before.
• Slightly larger thumbnails.
• You no longer have to refresh the page to see the latest events.
• You can "star" or "favorite" events to tell Frigate to keep associated recordings and the snapshot indefinitely.
• Sub labels have been added to allow setting arbitrary labels on events. For example, you can use deepstack to run facial recognition and update the event with the name of the person detected. (FEAT: Ability to set sub labels for specific events #2949)

Updated Recordings View

• Speed. The recordings view doesn't look different, but it has been completely overhauled. It should be much faster, and event data and thumbnails are lazy loaded to reduce the amount of data that is loaded when visiting the page.
• Smoother seeking and transitions. Re-renders of the player have been reduced and you should see smoother seek behavior when selecting events.
• 16x playback was added for browsers other than Firefox.

Updated Dockerbuild

• Frigate is now built on top of plain Debian containers and consolidated into a single multiarch image.

Experimental Timeline UI

Not recommended for use yet, but an alternative UI has been added. It has very limited browser/device support. #2830

Other changes

• MQTT sensors for any object type FEAT: MQTT support for all objects for each camera & zone #2908
• Update latest object endpoints to reference most recent event from database (you will no longer see black images for objects under the camera in live view) FEAT: Replace best jpg endpoint #2944
• Added a new ratio filter for object filters Add object filter ratio #2952
• Camera recordings always in sidebar Always show camera recordings in sidebar regardless of state #2846
• Enable and set mode per camera for birdseye Allow birdseye to be overridden at the camera level #3083
• Customize order and hide cameras from the UI FEAT: Ability to reorder & ability to hide Cameras in UI #2981
• Enable improve_contrast via MQTT Allow improve_contrast to be toggled via mqtt #3011
• Enable/disable motion detection independent of object detection via MQTT Ability to enable / disable motion detection via MQTT #3117
• MQTT topic for motion detection Send mqtt message when motion is detected #3152
• Batch deletes of recordings when more than 100,000 recordings need to be deleted
• Switch to jellyfin ffmpeg
• Improve cache handling for thumbnails and snapshots of completed events
• Fixes hwaccel issues for Raspberry Pi users
• Updates to debian 11 as a base image
• Consolidate Dockerfile to enable multiarch images
• Birdseye icno can now be customized FEAT: Ability to set custom birdseye icon and birdseye docs #2979
• Frigate now checks for a newer release on startup and exposes that in the stats endpoint Add latest frigate version to stats endpoints #3038
• Initial implementation of mypy and type checking
• Images from backend now have appropriate cache headers
• Motion contour_area and threshold values can now be adjusted dynamically via mqtt (Adjust camera motion threshold and contour area with mqtt #3154)

Full Changelog: v0.10.0...v0.11.0-rc1

Image: blakeblackshear/frigate:0.11.0-rc1

---

This discussion was created from the release 0.11.0 Release Candidate 1.

[ediguidedog]

Just a comment / observation
I run frigate (in its own docker) on a Beelink GK55 which has a J4125 CPU which is capable of both VAAPI and QSV hardware encoding/decoding. On previous versions QSV was slightly quicker than VAPPI with very slightly faster inference times and lower cpu usage

In the updated docs I see newer slightly different config file entry for hw acceleration

Intel-based CPUs (<10th Generation) via Quicksync (should be vaapi)
hwaccel_args: -hwaccel vaapi -hwaccel_device /dev/dri/renderD128 -hwaccel_output_format yuv420p

Intel-based CPUs (>=10th Generation) via Quicksync
hwaccel_args: -c:v h264_qsv

NOTICE: With some of the processors, like the J4125, the default driver iHD doesn't seem to work correctly for hardware acceleration. You may need to change the driver to i965 by adding the following environment variable LIBVA_DRIVER_NAME_JELLYFIN=i965 to your docker-compose file.

I've added LIBVA_DRIVER_NAME_JELLYFIN=i965 to my docker-compose environment section and recreated the container

What I have noticed is vaapi now appears to be much faster that before. Before they were much of a muchness and as QSV was the future I went with that but now using QSV I get inference speeds of 16 and Beelink CPU usage around 45% but with Vaapi I see inference speeds around 12 and CPU usage of 23%

I do notice running intel_gpu_top that
QSV I see around 10% utilisation on Video and zero elsewhere with 120 irqs/s
Vaapi I see 8% utilisation on video and the same on Render/3D with 260 irqs/s

As I say its as much an observation as anything else and I'm nor really sure when things changed but unless I'm missing something I'll be sticking with vaapi for now.

[NickM-27]

This changed with the latest release as we are now using the jellyfin build of ffmpeg. Very happy to hear that this env variable improves things this much, whichever works there shouldn't be an issue either way. QSV uses VAAPI under the hood anyway

[blakeblackshear]

Just wanted to clarify that hardware acceleration is only related to decoding the video stream. It has nothing at all to do with inference speeds which happens later in the process when running object detection. Changing hardware acceleration parameters doesn't have any impact on inference speeds.

[ediguidedog]

Just wanted to clarify that hardware acceleration is only related to decoding the video stream. It has nothing at all to do with inference speeds which happens later in the process when running object detection. Changing hardware acceleration parameters doesn't have any impact on inference speeds.

I thought that but my experience seems to fly in the face of that. Turning off hardware encoding pushed inference speed even higher.20 if I remember. I assumed the inference speed count must have included some element of cpu activity in the timings so while the actual time the task is with the coral would be independent of how busy the CPU is, the total end to end task time including sending and receiving would be affected by a busy CPU.

[blakeblackshear]

You might see slight variation like that due to CPU usage, but it's an indirect relationship. Running something with high CPU usage outside of frigate should have a similar impact.

[ediguidedog]

And I suspect that's it. The only things running on the box are frigate related. Mosquitto and an instance of homeassistant whose only job is to parse the mqtt events and send the odd telegram notification.
As you said it's an indirect relationship but lower CPU usage seems to give lower indicated inference times.

[markfrancisonly]

here's a few issues that I've found, perhaps these have been reported elsewhere.

in recordings, while clicking rapidly between event clips, I stumbled on an error message "Playback cannot continue. No available working or supported playlists." that appears with a spinner containing an X in the center. When this happens, the recordings page locks up, and it's necessary to navigate away from the camera page.

Playing recordings at 16x 'buffers' approximately every 3 seconds for about 1.5 seconds. this is 1080p video read from an ssd drive viewed on chrome with 40% cpu load.

Most important issue for me is a show stopper -

I proxy cam streams through rtsp-simple-server in frigate v0.10 successfully. in v0.11, all of my amcrest cameras green screen.

      output_args:
        record: -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -acodec aac -vcodec copy 
      input_args: -avoid_negative_ts make_zero -fflags +genpts+discardcorrupt -rtsp_transport tcp -timeout 5000000 -use_wallclock_as_timestamps 1

[2022-07-24 20:39:20] watchdog.baby                  ERROR   : Ffmpeg process crashed unexpectedly for baby.
[2022-07-24 20:39:20] watchdog.baby                  ERROR   : The following ffmpeg logs include the last 100 lines prior to exit.
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Guessed Channel Layout for Input Stream #0.1 : mono
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Input #0, rtsp, from 'rtsp:https://10.10.1.100:8554/baby':
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Metadata:
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     title           : Stream
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Duration: N/A, start: 1658709540.355375, bitrate: N/A
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:0: Video: h264 (High), yuvj420p(pc, bt709, progressive), 1920x1080 [SAR 1:1 DAR 16:9], 22 fps, 22 tbr, 90k tbn
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:1: Audio: pcm_mulaw, 8000 Hz, mono, s16, 64 kb/s
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Stream mapping:
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:0 -> #0:0 (copy)
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:1 -> #0:1 (pcm_mulaw (native) -> aac (native))
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:0 -> #1:0 (h264 (native) -> rawvideo (native))
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Press [q] to stop, [?] for help
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [aac @ 0x559f67980240] Too many bits 8832.000000 > 6144 per frame requested, clamping to max
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [segment @ 0x559f679ee600] Opening '/tmp/cache/baby-20220724203901.mp4' for writing
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Output #0, segment, to '/tmp/cache/baby-%Y%m%d%H%M%S.mp4':
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Metadata:
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     title           : Stream
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     encoder         : Lavf59.16.100
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:0: Video: h264 (High), yuvj420p(pc, bt709, progressive), 1920x1080 [SAR 1:1 DAR 16:9], q=2-31, 22 fps, 22 tbr, 11264 tbn
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #0:1: Audio: aac (LC), 8000 Hz, mono, fltp, 48 kb/s
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     Metadata:
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :       encoder         : Lavc59.18.100 aac
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [mp4 @ 0x559f68867000] Application provided invalid, non monotonically increasing dts to muxer in stream 1: 14400 >= 1222
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : av_interleaved_write_frame(): Invalid argument
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f695594c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f69583400] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f695c44c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f69607440] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f6964a3c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f6968b440] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f696ce3c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f69711340] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f697523c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f69795340] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f697d82c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f69819340] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f6985c2c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f6989f240] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f698e02c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f69923240] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [swscaler @ 0x559f6954c500] [swscaler @ 0x559f699661c0] deprecated pixel format used, make sure you did set range correctly
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Finishing stream 1:0 without any data written to it.
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Output #1, rawvideo, to 'pipe:':
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Metadata:
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     title           : Stream
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     encoder         : Lavf59.16.100
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :   Stream #1:0: Video: rawvideo (I420 / 0x30323449), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], q=2-31, 124416 kb/s, 5 fps, 5 tbn
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :     Metadata:
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   :       encoder         : Lavc59.18.100 rawvideo
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : frame=    9 fps=7.3 q=-1.0 Lq=0.0 size=N/A time=00:00:02.08 bitrate=N/A speed=1.69x    
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : video:167kB audio:7kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : [aac @ 0x559f67980240] Qavg: 51890.664
[2022-07-24 20:39:20] ffmpeg.baby.detect             ERROR   : Conversion failed!

I was able to verify that the rtsp-simple-server issue is only present in cameras with audio streams. if I disable the camera audio stream everything works normally. if I enable pcm ulaw or aac audio, frigate is unable to access the camera feeds through rtsp-simple-server.

for my baby cam, audio is a requirement. I primarily consume my baby monitor feed using webrtc from multiple tablets concurrently, but also want to use frigate for recording and sensing.

Unfortunately the cameras don't have enough reliable wireless bandwidth to stream more than one stream concurrently, and rtsp-simple-server is required.

this issue appears to be an incompatibility between the jellyfin ffmpeg builds and rtsp-simple-server code. rtsp streams containing audio work fine when proxied from rtsp-simple-server in frigate v0.10

[LordNex]

Ok now when I try and pull streams from frigate to double stack sitting inside Home Assistant I keep getting these errors

Logger: libav.rtmp
Source: components/stream/worker.py:496 
First occurred: 4:55:56 PM (41 occurrences) 
Last logged: 6:13:26 PM

Cannot read RTMP handshake response

And

Logger: homeassistant.components.stream.stream.camera.sidedoorbell
Source: components/stream/__init__.py:416 
Integration: Stream (documentation, issues) 
First occurred: 4:55:56 PM (31 occurrences) 
Last logged: 6:13:26 PM

Error from stream worker: Error opening stream (EOF, End of file) rtmp:https://Frigate_IP:5000/live/sidedoorbell

This one is a AD410 and I get streams fine from the proxy. I'm wondering if it's also because I'm using the HA Integration as well to pull the streams into HA. But if I still have it set to send via MQTT and that seems to get picked up as well. I'm also trying to get DoubleTake training down so that it's more accurate. My config is pretty detailed. But that's someone else's issue.

Off topic
I like where your going with the Frigate+ but I don't like the idea of using a cloud service as a facial recognition source. I mean as where the processing and storage of certain personal data in the cloud. I use my system as a backup security system that unlocks the door if that person is recognized and their HA tied phone is within the gps fence. So it probably wouldn't be something I would sign up and pay for. Now if it contained lists that I could subscribe too like known sexual offenders or a "think Facebook" algorithm could parse the data from online resources and show you the relevant data. So no matter who shows up at your door. You have more information about who that actually is before opening the door.

I mean that's why we put a lot of these cameras in is security for ourselves and our families. Anyway, I don't mind running local neural net AI with socialized hardware. But a source to pull from would be nice. Might be a project worth investing into. Anyway, thanks again!

[LordNex]

Sorry. I keep forgetting the formatting on here for code blocks. Thought it was 3 '

[NickM-27]

It's three backticks ```

[LordNex]

It's three backticks ```

Thanks

[NickM-27]

Do you have any errors in the frigate logs? I don't think pulling from frigate RTMP multiple streams should matter, but not 100% sure. If they're the same stream anyway, not sure why that is needed?

I like where your going with the Frigate+ but I don't like the idea of using a cloud service as a facial recognition source. I mean as where the processing and storage of certain personal data in the cloud. I use my system as a backup security system that unlocks the door if that person is recognized and their HA tied phone is within the gps fence. So it probably wouldn't be something I would sign up and pay for. Now if it contained lists that I could subscribe too like known sexual offenders or a "think Facebook" algorithm could parse the data from online resources and show you the relevant data. So no matter who shows up at your door. You have more information about who that actually is before opening the door.

I mean that's why we put a lot of these cameras in is security for ourselves and our families. Anyway, I don't mind running local neural net AI with socialized hardware. But a source to pull from would be nice. Might be a project worth investing into. Anyway, thanks again!

I think you must have misunderstood something? Frigate+ isn't for facial recognition directly, it is just better models for detecting objects like person, dog, cat, UPS, FEDEX, etc face is on that list but just as a general face (which will make DoubleTake work better since it won't run detection on all people, only when it sees a face).

[yayitazale]

Hi:

I only see one tag in this RC. In the new docs I understand that from now on using nvidia hard is going to be defined on the config file, so there is no more need for separated tags for each architecture?

But at the installation page I read :

Running in Docker directly is the recommended install method.

Make sure you choose the right image for your architecture:

Arch	Image Name
amd64	blakeblackshear/frigate:stable-amd64
amd64nvidia	blakeblackshear/frigate:stable-amd64nvidia
armv7	blakeblackshear/frigate:stable-armv7
aarch64	blakeblackshear/frigate:stable-aarch64

[LordNex]

alias: Notifications - Security Breach
trigger:
  - platform: state
    entity_id:
      - sensor.double_take_unknown
condition:
  - condition: template
    value_template: "{{ trigger.to_state.state != trigger.from_state.state }}"
    enabled: true
action:
  - service: notify.mobile_app_iphone_inator
    data:
      message: |-
        {% if trigger.to_state.attributes.match is defined %}
          {{trigger.to_state.attributes.friendly_name}} is near the {{trigger.to_state.state}} with {{trigger.to_state.attributes.match.confidence}}% probability by {{trigger.to_state.attributes.match.detector}}:{{trigger.to_state.attributes.match.type}} and it took {{trigger.to_state.attributes.attempts}} attempt(s) @ {{trigger.to_state.attributes.duration}} sec
        {% elif trigger.to_state.attributes.unknown is defined %}
          An unknown person is near the {{trigger.to_state.state}} with a confidence of {{trigger.to_state.attributes.unknown.confidence}}% by {{trigger.to_state.attributes.unknown.detector}}:{{trigger.to_state.attributes.unknown.type}} and taking {{trigger.to_state.attributes.attempts}} attempt(s) @ {{trigger.to_state.attributes.duration}} sec
        {% endif %}
        attachment:
          url: |-
            {% if trigger.to_state.attributes.match is defined %}
              http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}
            {% elif trigger.to_state.attributes.unknown is defined %}
               http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.unknown.filename}}?box=true&token={{trigger.to_state.attributes.token}}
            {% endif %}
        actions:
          - action: URI
            title: View Image
            uri: |-
              {% if trigger.to_state.attributes.match is defined %}
                http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}
              {% elif trigger.to_state.attributes.unknown is defined %}
                 http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.unknown.filename}}?box=true&token={{trigger.to_state.attributes.token}}
              {% endif %}
  - service: notify.mobile_app_fbi_investigative_agent_0451
    data:
      message: |-
        {% if trigger.to_state.attributes.match is defined %}
          {{trigger.to_state.attributes.friendly_name}} is near the {{trigger.to_state.state}} with {{trigger.to_state.attributes.match.confidence}}% probability by {{trigger.to_state.attributes.match.detector}}:{{trigger.to_state.attributes.match.type}} and it took {{trigger.to_state.attributes.attempts}} attempt(s) @ {{trigger.to_state.attributes.duration}} sec
        {% elif trigger.to_state.attributes.unknown is defined %}
          An unknown person is near the {{trigger.to_state.state}} with a confidence of {{trigger.to_state.attributes.unknown.confidence}}% by {{trigger.to_state.attributes.unknown.detector}}:{{trigger.to_state.attributes.unknown.type}} and taking {{trigger.to_state.attributes.attempts}} attempt(s) @ {{trigger.to_state.attributes.duration}} sec
        {% endif %}
        attachment:
          url: |-
            {% if trigger.to_state.attributes.match is defined %}
              http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}
            {% elif trigger.to_state.attributes.unknown is defined %}
               http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.unknown.filename}}?box=true&token={{trigger.to_state.attributes.token}}
            {% endif %}
        actions:
          - action: URI
            title: View Image
            uri: |-
              {% if trigger.to_state.attributes.match is defined %}
                http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}
              {% elif trigger.to_state.attributes.unknown is defined %}
                 http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.unknown.filename}}?box=true&token={{trigger.to_state.attributes.token}}
              {% endif %}
  - service: notify.mobile_app_princess_consuela_bananahammock
    data:
      message: |-
        {% if trigger.to_state.attributes.match is defined %}
          {{trigger.to_state.attributes.friendly_name}} is near the {{trigger.to_state.state}} with {{trigger.to_state.attributes.match.confidence}}% probability by {{trigger.to_state.attributes.match.detector}}:{{trigger.to_state.attributes.match.type}} and it took {{trigger.to_state.attributes.attempts}} attempt(s) @ {{trigger.to_state.attributes.duration}} sec
        {% elif trigger.to_state.attributes.unknown is defined %}
          An unknown person is near the {{trigger.to_state.state}} with a confidence of {{trigger.to_state.attributes.unknown.confidence}}% by {{trigger.to_state.attributes.unknown.detector}}:{{trigger.to_state.attributes.unknown.type}} and taking {{trigger.to_state.attributes.attempts}} attempt(s) @ {{trigger.to_state.attributes.duration}} sec
        {% endif %}
        attachment:
          url: |-
            {% if trigger.to_state.attributes.match is defined %}
              http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}
            {% elif trigger.to_state.attributes.unknown is defined %}
               http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.unknown.filename}}?box=true&token={{trigger.to_state.attributes.token}}
            {% endif %}
        actions:
          - action: URI
            title: View Image
            uri: |-
              {% if trigger.to_state.attributes.match is defined %}
                http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}
              {% elif trigger.to_state.attributes.unknown is defined %}
                 http:https://frigateip:5000/api/storage/matches/{{trigger.to_state.attributes.unknown.filename}}?box=true&token={{trigger.to_state.attributes.token}}
              {% endif %}
mode: parallel
max: 10

[LordNex]

I'm thinking of changing it to

http:https://frigateip:5000/api/{{trigger.to_state.state}}/{{trigger.to_state.attributes.match.filename}}?box=true&token={{trigger.to_state.attributes.token}}

[LordNex]

That looks like it is the doubletake API https://github.com/jakowenko/double-take#home-assistant

Yep it is

[LordNex]

That looks like it is the doubletake API https://github.com/jakowenko/double-take#home-assistant

Yep it is

So should I be defining the ip and port of the docker container DT is running in? Or would there be a simpler method.

[NickM-27]

So should I be defining the ip and port of the docker container DT is running in? Or would there be a simpler method.

Yes, DT ip and port

[LordNex]

Would this need to be a complete reinstall for a RPi4 8gig with USB TPU? Or would just updating the docker compose file to pull down the latest rc be enough. I almost want to pull the trigger with this but I mainly use the Home Assistant Proxy for viewing and I'm not sure if that parts ready or not. But this looks really cool. Integrates very well with Double-Take for facial recognition training and support.

[LordNex]

@LordNex You'll need to go to the advanced settings and enable the protocols that are used for viewing the stream as they are disabled by default

Where is that at? I'm not seeing it in the app. Web client or do I need another tool?

[NickM-27]

Accessing via the web it is Network Settings -> Advanced -> Port Settings

[LordNex]

Accessing via the web it is Network Settings -> Advanced -> Port Settings

I'll check that. Also, the new ReoLink specific page is 404

Reolink Cameras with Frigate

[NickM-27]

https://deploy-preview-2829--frigate-docs.netlify.app/configuration/camera_specific#reolink-410520-possibly-others

works fine for me

[LordNex]

https://deploy-preview-2829--frigate-docs.netlify.app/configuration/camera_specific#reolink-410520-possibly-others

works fine for me

Your the man. I didn't even think to check the web page to see if those settings got turned off. Once I reenabled those and reset up my stream settings, it's back on. Many thanks man!

[joelegg]

how long will it be until frigate 0.11 is released to main? (I ask as i'm after the coral usb compatibility in dsm7 synology) thank you

[joelegg]

looks like you can override the manifest.json - but i'm not entirely sure what to change there is no tag/image settings there. anyone here know how to override it to use the pre release? thank you

[NickM-27]

No no, there's a different addon in the repository for the beta, you don't need to edit any files just uninstall the current addon then install the beta addon

[LordNex]

No no, there's a different addon in the repository for the beta, you don't need to edit any files just uninstall the current addon then install the beta addon

I think they are referring to the Full Frigate install for Home Assistant. Not the proxy like we use. I don't believe you can goto HACS and flip the beta tag either. I tried that and the RCs didn't show up. You'd have to push to to HACS as a beta version for them to install it since Home Assistant Supervisor handles all the docker containers. Unless they feel comfortable plucking around with portainer or something.

[NickM-27]

@LordNex no you are incorrect, we have separate beta and stable addons available in the same repository. It is easy to switch between them

https://github.com/blakeblackshear/frigate-hass-addons

Also, I run HA in docker so I don't use any addons.

[NickM-27]

Also HACS is for integrations, the addon store is for addons, important not to mix those up.

[esmoyer]

Random; I upgraded to the RC2 and made the environment variable changes for LIBVA, but have noticed weird "artifact-ing" both live and recordings. Anyone else getting this? I am guessing it's transcoding related? I can submit a formal issue but wanted to check in case; I'm guessing I might have missed something. RC1 was working great.

[LordNex]

Recordings are pulled directly from the camera and no transcoding is done. So if it is occurring on the recordings it is most likely a camera issue.

Probably best if you create an issue and answer all the prompts.

I thought the recordings were pulled from wherever you specified in the docker compose file. For instance I have mine pointing to my NAS and I can go there and pull up all the recordings directly and not from the camera. This was the main reason I haven't put SD Cards I each camera since the feed is being recorded and sent to the NAS 24/7.

[esmoyer]

I think he's talking about the "stream" comes directly from the camera and is saved to disk and it doesn't go through any transcoding before being saved.

[NickM-27]

I thought the recordings were pulled from wherever you specified in the docker compose file. For instance I have mine pointing to my NAS and I can go there and pull up all the recordings directly and not from the camera. This was the main reason I haven't put SD Cards I each camera since the feed is being recorded and sent to the NAS 24/7.

You're reading that wrong. I'm saying the recording is pulled from the camera stream when creating the recording file and the recording stream is not transcoded. I was not talking about where recordings come from when viewing a recording in the frontend (which is of course stored in the directory set in docker compose)

[NickM-27]

I think he's talking about the "stream" comes directly from the camera and is saved to disk and it doesn't go through any transcoding before being saved.

Yes exactly

[LordNex]

I think he's talking about the "stream" comes directly from the camera and is saved to disk and it doesn't go through any transcoding before being saved.

Yes exactly

Sorry. Thanks for the clarification!

[J-Prince]

I'm running Frigate on Unraid 6.11 rc5 with Intel 12th gen CPU, Alder Lake.

I'm unable to get hardware acceleration to work with Frigate, only get green cameras (Hikvision).

hwaccel_args: -c:v h264_qsv

Using the same setup, Plex does hardware acceleration correctly so I feel like the host is Ok.

Any pointers?

[NickM-27]

Make an issue with the hwaccel support template and provide all the requested info