RecordingControl 25.06 ONVIF™ www.onvif.org June, 2025 2008-2025 ONVIF™ All rights reserved. Recipients of this document may copy, distribute, publish, or display this document so long as this copyright notice, license and disclaimer are retained with all copies of the document. No license is granted to modify this document. THIS DOCUMENT IS PROVIDED "AS IS," AND THE CORPORATION AND ITS MEMBERS AND THEIR AFFILIATES, MAKE NO REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT, OR TITLE; THAT THE CONTENTS OF THIS DOCUMENT ARE SUITABLE FOR ANY PURPOSE; OR THAT THE IMPLEMENTATION OF SUCH CONTENTS WILL NOT INFRINGE ANY PATENTS, COPYRIGHTS, TRADEMARKS OR OTHER RIGHTS. IN NO EVENT WILL THE CORPORATION OR ITS MEMBERS OR THEIR AFFILIATES BE LIABLE FOR ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL, PUNITIVE OR CONSEQUENTIAL DAMAGES, ARISING OUT OF OR RELATING TO ANY USE OR DISTRIBUTION OF THIS DOCUMENT, WHETHER OR NOT (1) THE CORPORATION, MEMBERS OR THEIR AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES, OR (2) SUCH DAMAGES WERE REASONABLY FORESEEABLE, AND ARISING OUT OF OR RELATING TO ANY USE OR DISTRIBUTION OF THIS DOCUMENT.  THE FOREGOING DISCLAIMER AND LIMITATION ON LIABILITY DO NOT APPLY TO, INVALIDATE, OR LIMIT REPRESENTATIONS AND WARRANTIES MADE BY THE MEMBERS AND THEIR RESPECTIVE AFFILIATES TO THE CORPORATION AND OTHER MEMBERS IN CERTAIN WRITTEN POLICIES OF THE CORPORATION. 2.1 Jul-2011 Hans Busch Split from Core 2.0 without change of content. 2.1.1 Jan-2012 Hans Busch Change Requests 293, 297, 535 2.2 Apr-2012 Hans Busch Change Requests 608, 625, 636, 673 2.2.1 Dec-2012 Hans Busch Michio Hirai Change Requests 708, 709, 719, 759, 827, 845, 852, 866, 867, 870, 862, 872, 861 2.3 May-2013 Michio Hirai Change Request 934 2.4 Aug-2013 Michio Hirai Change Request 1073, 1086 2.4.1 Dec-2013 Michio Hirai Change Request 1148, 1189 2.4.2 Jun-2014 Michio Hirai Change Request 1292, 1298, 1304, 1412 2.5 Dec-2014 Hasan Timucin Ozdemir Added 5.21 ExportRecordedData command and corresponding capability flag in 5.24 Capabilitites Added 5.22 StopExportRecordedData Added 5.23 GetExportRecordedDataStatus 16.12 Dec-2016 Hans Busch Change Request 1991 17.06 Jun-2017 Stefan Andersson Hiroyuki Sano Update method layouts Change Request 1843, 2060, 2062 Change Request 2061, 2063, 2065, 2109 17.12 Dec-2017 Hiroyuki Sano Change Request 2178, 2179, 2185 18.06 Jun-2018 Hiroyuki Sano Change Request 2230, 2258 19.06 Jun-2019 Hans Busch Added Scheduled Recording 21.12 Dec-2021 Sergey Bogdanov Change the "role" attribute for request access class. 22.12 Dec-2022 Hans Busch Add support for event recording. 23.06 Jun-2023 Felix Schuetz Add annex on object storage recording. 24.12 Dec-2024 Sriram Bhetanabottla Clarify delete interfaces for object storage. 25.06 Jun-2025 Jean-Francois Levesque, Jose Melancon Add the possibility to temporary override of segment duration for Cloud Recording. This document defines the web service interface for the configuration of recording of Video, Audio and Metadata. Additionally associated events are defined. The overview section provides a definition of the ONVIF storage model. This is common for all ONVIF storage related services. Web service usage is outside of the scope of this document. Please refer to the ONVIF core specification. ISO/IEC 14496-12:2022 — Information technology — Coding of audio-visual objects — Part 12: ISO base media file format <> ISO/IEC 14496-14:2020 — Information technology — Coding of audio-visual objects — Part 14: MP4 file format <> ISO/IEC 23000-19:2020 — Information technology — Multimedia application format (MPEG-A) — Part 19: Common media application format (CMAF) for segmented media <> ISO/IEC 23001-7:2016 — Information technology — MPEG systems technologies — Part 7: Common encryption in ISO base media file format files <> ISO 8601-1:2019 — Date and time — Representations for information interchange — Part 1: Basic rules <> RFC 5234 — Augmented BNF for Syntax Specifications: ABNF <> RFC 6381 — The 'Codecs' and 'Profiles' Parameters for "Bucket" Media Types <> ONVIF Core Specification <> ONVIF Schedule Service Specification <>

Metadata All streaming data except video and audio, including video analytics results, PTZ position data and other metadata (such as textual data from POS applications). Recording A container for a set of audio, video and metadata tracks. A recording can hold one or more tracks. A track is viewed as an infinite timeline that holds data at certain times. Recording Event An event associated with a Recording, represented by a notification message in the APIs Recording Job A job performs the transfer of data from a data source to a particular recording using a particular configuration Track An individual data channel consisting of video, audio, or metadata. This definition is consistent with the definition of track in [RFC 2326] Video Analytics Algorithms or programs used to analyze video data and to generate data describing object location and behaviour.

This standard provides a set of interfaces that enable the support of interoperable network storage devices, such as network video recorders (NVR), digital video recorders (DVR) and cameras with embedded storage. The following functions are supported: Recording Control Search Replay These functions are provided by three interrelated services: Recording service enables a client to manage recordings, and to configure the transfer of data from data sources to recordings. Managing recordings includes creation and deletion of recordings and tracks. Search service enables a client to find information about the recordings on the storage device, for example to construct a “timeline” view, and to find data of interest within a set of recordings. The latter is achieved by searching for events that are included in the metadata track recording, Replay service enables a client to play back recorded data, including video, audio and metadata. Functions are provided to start and stop playback and to change speed and direction of the replayed stream. It also enables a client to download data from the storage device so that export functionality can be provided. WSDL for this service is specified in http://www.onvif.org/onvif/ver10/recording.wsdl. Prefix Namespace URI env http://www.w3.org/2003/05/soap-envelope ter http://www.onvif.org/ver10/error xs http://www.w3.org/2001/XMLSchema tt http://www.onvif.org/ver10/schema trc http://www.onvif.org/ver10/recording/wsdl

The storage interfaces in this standard present a logical view of the data on the storage device. This view is completely independent of the way data might be physically stored on disk. The key concept in the storage model is that of a recording. The term recording is used in this specification to denote a container for a set of related audio, video and metadata tracks, typically from the same data source e.g. a camera. A recording could hold any number of tracks. A track is viewed as an infinite timeline that holds data at certain times. At a minimum, a recording is capable of holding three tracks, one for audio, one for video and one for metadata. Some implementations of the recording service may support multiple tracks of each type. For example the same recording could hold two video tracks, one containing a low resolution or low frame rate stream and one containing a high resolution or high frame rate stream.

It is important to note that the storage interfaces do not expose the internal storage structures on the device. In particular, a recording is not intended to represent a single file on disk although in many storage device implementations a recording is physically stored in a series of files. For instance, some camera implementations realise alarm recording by creating a distinct file for each alarm that occurs. Although each file could be represented as a different recording, the intent of the model in this standard is that all these files are aggregated into a single recording. Within a recording the regions where data is actually recorded are represented by pairs of events, where each pair comprises an event when recording started and an event when recording stopped. A client can construct the logical view of the recordings by using the FindRecordings and FindEvents methods of the search service. If metadata is recorded, the metadata track can hold all the events generated by the data source (see the chapter on event handling and the MetadataConfiguration object). In addition, a device also conceptually records ONVIF defined historical events (see Recording Event Descriptions in the search service), this includes information like start and end of a recorded data range. A device may also conceptually record vendor specific historical events. Events generated by the device are not inserted in existing metadata tracks of recordings. The FindEvents method in the search service can find all the recorded events.

The recording service enables a client to manage recordings, and to configure the transfer of data from data sources to recordings. Managing recordings includes creation and deletion of recordings and tracks. Recording jobs transfer data from a recording source to a recording. A recording source can be a receiver object created with the receiver service, or it can be a media profile that encodes data on a local device. The media profile could be used as a source on a camera with embedded storage. To save data to a recording, a client first creates a recording and ensures that the recording has the necessary tracks. Then the client creates a recording job that pulls data from one or more sources and stores the data to the tracks in the recording. Clients may set up multiple recording jobs that all record into the same recording. If multiple recording jobs are active, the device uses a priority scheme to select between the tracks defined in the recording jobs. Clients may change the mode of recording jobs at any time, thereby providing means to implement features like alarm recording or manual recording. If a device supports scheduled recording, clients may configure scheduled recording by adding a scheduler token to the recording job. A recording job with a scheduler token will only record when the associated schedule is active. If the associated schedule of a recording job is inactive a job with lower recording priority may record. By default a recording job is continuously recording. Devices supporting the EventFilter can be configured such that they only record events. To provide some context time intervals before and after the event may be captured. The recording job relies on the receiver service for receiving the data from other devices through receiver objects identified by ReceiverTokens

The target interface allows configuration for devices that support recording to external storage targets. For authentication configuration see the related storage configuration APIs of the core specification. The target API defines for each recording a path on the storage device where the related recordings should be stored. In order to keep reasonably sized files the recordings are split into segments. This specification defines a date and time based index. Indexing according to events and video content is outside of the scope of this specification. An encryption configuration interface allows to encrypt the content according to well defined standards.

The recording service enables a client to manage recordings, and to configure the transfer of data from data sources to recordings. Managing recordings includes creation and deletion of recordings and tracks, as well as locking and unlocking ranges of recordings and deletion of recorded data. Recording jobs transfer data from a recording source to a recording. A recording source can be a receiver object created with the receiver service, or it can be a media profile that encodes data on a local device. The media profile could be used as a source on a camera with embedded storage. The term recording is used in this specification to denote a container for a set of audio, video and metadata tracks. A recording could hold any number of tracks. A track is viewed as an infinite timeline that holds data at certain times.

The figure shows three recordings, each with a video, a metadata and two audio tracks. Here second audio track is used for storing the audio backchannel. At a minimum, a recording shall be capable of holding three tracks, one for audio, one for video and one for metadata. Some implementations of the recording service may support multiple tracks of each type. All recorded data of a track shall have the same encoding. To save data to a recording, a client first creates a recording and ensures that the recording has the necessary tracks. Then the client creates a recording job that pulls data from one or more sources and stores the data to the tracks in the recording. Clients may set up multiple recording jobs that all record into the same recording. If multiple recording jobs are active, the device uses a priority scheme to select between the tracks defined in the recording jobs. Clients may change the mode of recording jobs at any time, thereby providing means to implement features like alarm recording or manual recording. A recording job with schedule token shall only be recording when the associated schedule is active. For the cases where media attributes of a source are changed for an active recording job, the recording state is outside the scope of this specification. The recording job relies on the receiver service for receiving the data from other devices through receiver objects identified by ReceiverTokens For the cases where a client uses a receiver object with a single recording job, the recording service can auto create and auto delete receiver objects. Autocreation is signalled with the AutoCreateReceiver flag in the recording job configuration structure. Receiver objects created this way shall be automatically deleted when no recording job uses them anymore. A receiver object that is automatically created shall have all its fields set to empty values. The client should configure the receiver object after it has created the recording job. The ONVIF view of recordings is a logical one which is independent of the way recordings are physically stored on disk. For instance, some camera implementations realise alarm recording by creating a distinct file on a FAT file system for each alarm that occurs. Although each file could be represented as a different ONVIF recording, the intent of the model in this standard is that all these files are aggregated into a single recording. By searching for the “DataPresent” event with the FindEvents method of the search service, a client can locate the times at which video started to be recorded and where video stopped being recorded. If Metadata is recorded, the metadata can also hold all the events generated by the data source (see section event handling of the ONVIF Core Specification and section on Metadata configuration in the ONVIF Media Service Specification). In addition, a device also conceptually record ONVIF defined historical events (see Recording Event Descriptions in the search service), this includes information like start and end of a recorded data range. A device may also conceptually record vendor specific historical events. Events generated by the device are not inserted in existing metadata tracks of recordings. The FindEvents method in the search service can find all the recorded events. Many device implementations will automatically delete the oldest recorded data from storage in order to free up space for new recordings. Locks provide a mechanism to allow a user to select ranges of data. A range of data that is locked does not get deleted automatically. Support for locks is reserved for future versions of the specification.

All the objects created within the recording service shall be persistent – i.e. they shall survive a power cycle. Likewise, all the configuration data in the objects shall be persistent.

The RecordingConfiguration structure shall be used to configure recordings through CreateRecordings and Get/SetRecordingConfiguration. MaximumRetentionTime specifies the maximum time that data in any track within the recording shall be stored. The device shall delete any data older than the maximum retention time. Such data shall not be accessible anymore. If the MaximumRetentionPeriod is set to 0, the device shall not limit the retention time of stored data, except by resource constraints. Whatever the value of MaximumRetentionTime, the device may automatically delete recordings to free up storage space for new recordings. None of the other fields defined in this structure shall be used by the device. Instead, it simply stores this information, and it shall return it through the GetRecordingConfiguration and GetRecordingInformation (see ONVIF Recording Search Service Specification) methods. A device may truncate any descriptive string without causing a fault if it exceeds the supported length. Descriptive strings are Location, Description and Content. A device signaling support for recording to external targets via the SupportedTargetFormats capability shall support the target configuration with the following parameters: Storage Token of a storage configuration. Format Format of the recording. Examples are MP4 and CMAF. Prefix Path prefix to be inserted in the object key. Postfix Path postfix to be inserted in the object key. SpanDuration Maximum duration of a span. SegmentDuration Maximum duration of a segment. Encryption Optional encryption configuration. See for ONVIF defined recording formats.

The TrackConfiguration structure shall be used to configure tracks using CreateTrack and Get/SetTrackConfiguration The TrackConfiguration contains the following fields: The TrackType defines the data type of the track. It shall be equal to the strings “Video”, “Audio” or “Metadata”. The track shall only be able to hold data of that type. None of the other fields defined in this structure shall be used by the device. Instead, it simply stores this information, and it shall return it through the GetTrackConfiguration and GetRecordingInformation (see ONVIF Recording Search Service Specification) methods.

The RecordingJobConfiguration structure shall hold the configuration for a recording job. Its UML diagram is hown in .

The RecordingJobConfiguration holds the following fields: RecordingToken: Identifies the recording to which this job shall store the received data. Mode: If it is idle, nothing shall happen. If it is active and the recording job has the highest priority, the device shall try to obtain data from the receivers. A client shall use GetRecordingJobState to determine if data transfer is really taking place. The only valid values for Mode shall be “Idle” and “Active”. Priority: This shall be a non-negative number. If there are multiple recording jobs that store data to the same track, the device shall only store data for the recording job with the highest priority. The priority is specified per recording job, but the device shall determine the priority of each track individually. If there are multiple recording jobs with the same highest priority it is undefind which of them is activated. The value 0 indicates the lowest priority. Higher values shall indicate a higher priority. ScheduleToken: This attribute adds an additional requirement for activating the recording job. If this optional field is provided the job shall only record if the schedule exists and is active. EventFilter: This set of parameters allows to control recording depending on a given set of event condititions. SourceToken: This field shall be a reference to the source of the data. The type of the source is determined by the attribute Type in the SourceToken structure. If Type is http://www.onvif.org/ver10/schema/Receiver, the token is a ReceiverReference. In this case the device shall receive the data over the network. If Type is http://www.onvif.org/ver10/schema/Profile, the token identifies a media profile, instructing the device to obtain data from a profile that exists on the local device. A device that includes the ONVIF Media Service shall support a Media Profile token and a device that includes the ONVIF Receiver Service shall support a Receiver token. AutoCreateReceiver: If a request includes this field set to true and no source token is provided, the device shall create a receiver object (through the receiver service) and assign the ReceiverReference to the SourceToken field. A device shall never report this parameter in a RecordingJobConfiguration. A device may reject a request that neither contains a SourceToken nor AutoCreateRecevier set to true. SourceTag: If the received RTSP stream contains multiple tracks of the same type, the SourceTag differentiates between those Tracks. Destination: The destination is the track token of the track to which the device shall store the received data. All tracks must belong to the recording identified by the RecordingToken. The TrackInformation field for a Track holds a single Source. In case multiple RecordingJobs with differing Source are recording to the same Track it is undefined which of them is reported in the corresponding TrackInformation of the the RecordingSearch API.

A device signalling support for EventRecording via its capabilities shall support controling recording job activity via the EventFilter with the following set of parameters: Filter One or more filter pairs containing a mandatory topic filter as defined in section 9.6.3 of the ONVIF Core Specification. It may be associated with an optional message content filter as defined in section 9.4.4 of the ONVIF Core Specification. Before Optional timespan to record before the actual event condition became active. After Optional timespan to record after the actual event condition becomes inactive. A device shall support filtering on topics and message source parameters. Filtering on message data values doesn't need to be supported since it may cause malfunctions. A device shall support Before and After durations when their limit is signalled via the respective capability. A device may adapt the Before and After duration values to internal quantization. A device shall at least record the event duration and the specified before and after timesspans. Due to the nature of GOP structures it may record more. Note that non-property events result in an infinite short timespan. In such cases at least one I-Frame shall be recorded, optionally extended by before and after timespans. A recording job of a device supporting both EventFilter and ScheduledRecording shall become active if both conditions are met.

CreateRecording shall create a new recording. This method is optional. It shall be available if the Recording/DynamicRecordings capability is TRUE. request RecordingConfiguration [tt:RecordingConfiguration] Contains the initial configuration for the recording. response RecordingToken [tt:RecordingReference] The reference to the created recording. faults env:Receiver - ter:Action - ter:MaxRecordings The device cannot create a new recording because it already has the maximum number of recordings that it supports. env:Sender - ter:InvalidArgVal - ter:BadConfiguration The RecordConfiguration is invalid. env:Receiver - ter:ActionNotSupported - ter:NotImplemented This optional method is not implemented. access class ACTUATE When successfully completed, the device shall have created one or more tracks with the following configurations: TrackToken TrackType VIDEO001 Video AUDIO001 Audio META001 Metadata The RecordingConfiguration shall have the MaximumRetentionTime set to 0 (unlimited) and all TrackConfigurations shall have the Description set to the empty string.

DeleteRecording shall delete a recording object. Whenever a recording is deleted, the device shall delete all the tracks that are part of the recording, and it shall delete all the Recording Jobs that record into the recording. For each deleted recording job, the device shall also delete all the receiver objects associated with the recording job that are automatically created using the AutoCreateReceiver field of the recording job configuration structure and are not used in any other recording job. This method is optional. It shall be available if the Recording/DynamicRecordings capability is TRUE. This method has no effect on the data stored in external targets for e.g. Object Storage S3. request RecordingToken [tt:RecordingReference] Identifies the recording that shall be deleted. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording env:Receiver - ter:ActionNotSupported - ter:NotImplemented The device cannot delete recordings. env:Receiver - ter:Action - ter:CannotDelete This specific recording cannot be deleted. access class ACTUATE

GetRecordings shall return a description of all the recordings in the device. This description shall include a list of all the tracks for each recording. request This is an empty message. response RecordingItem – optional, unbounded [tt:GetRecordingsResponseItem] Identifies a recording and its current configuration faults None access class READ_MEDIA

SetRecordingConfiguration shall change the configuration of a recording request RecordingToken [RecordingToken ] Identifies the recording that shall be changed. RecordingConfiguration [RecordingConfiguration] The new configuration for the recording. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter: BadConfiguration The configuration is invalid. env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. access class ACTUATE

GetRecordingConfiguration shall retrieve the recording configuration for a recording request RecordingToken [tt:RecordingReference] Identifies the recording for which the configuration shall be retrieved. response RecordingConfiguration [tt:RecordingConfiguration] The current configuration for the requested recording. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. access class READ_MEDIA

This method shall create a new track within a recording if the method GetRecordingOptions signals spare tracks for the recording. For a track to be created the SpareXXX (where XXX is the track type) needs to be set. This method is optional. It shall be available if the Recording/DynamicTracks capability is TRUE. request RecordingToken [tt:RecordingReference] Identifies the recording to which a track shall be added. TrackConfiguration [tt:TrackConfiguration] The configuration for the new track. response TrackToken [tt:TrackReference] Identifies the newly created track. A device shall ensure that the TrackToken is unique within the recoding to which the new track belongs. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. env:Receiver - ter:Action - ter:MaxTracks The new track cannot be created because the maximum number of tracks that the device supports for this recording has been reached. env:Sender - ter:InvalidArgVal - ter:BadConfiguration The TrackConfiguration is invalid. env:Receiver - ter:ActionNotSupported - ter:NotImplemented This optional method is not implemented. access class ACTUATE A TrackToken in itself does not uniquely identify a specific track. Tracks within different recordings may have the same TrackToken.

DeleteTrack shall remove a track from a recording. All the data in the track shall be deleted. This method is optional. It shall be available if the Recording/DynamicTracks capability is TRUE. This method has no effect on the data stored in external targets for e.g. Object Storage S3. request RecordingToken [tt:RecordingReference ] Identifies the recording from which to delete the track. TrackToken [tt:TrackReference] Identifies the track to delete. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. env:Sender - ter:InvalidArgVal - ter:NoTrack The TrackToken does not reference an existing track of the recording. env:Receiver - ter:Action - ter:CannotDelete This specific track cannot be deleted. env:Receiver - ter:ActionNotSupported - ter:NotImplemented This optional method is not implemented. access class ACTUATE

GetTrackConfiguration shall retrieve the configuration for a specific track. request RecordingToken [tt:RecordingReference ] Identifies the recording. TrackToken [tt:TrackReference] Identifies the track within the recording from which to get the track configuration response TrackConfiguration [tt:TrackConfiguration] The current configuration for the track. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. env:Sender - ter:InvalidArgVal - ter:NoTrack The TrackToken does not reference an existing track of the recording. access class READ_MEDIA

SetTrackConfiguration shall change the configuration of a track. TrackType shall be ignored by the device as it can’t be changed. The TrackConfiguration is the new configuration for the track. request RecordingToken [tt:RecordingReference ] Identifies the recording. TrackToken [tt:TrackReference] Identifies the recording within the recording from which to set the track configuration TrackConfiguration [tt:TrackConfiguration] The new configurartion for the track. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. env:Sender - ter:InvalidArgVal - ter:NoTrack The TrackToken does not reference an existing track of the recording. env:Sender - ter:InvalidArgVal - ter:BadConfiguration The contents of the configuration object are invalid. access class ACTUATE

CreateRecordingJob shall create a new recording job. A device shall support adding a RecordingJob to a recording for which it signals Spare jobs via GetRecordingOptions. A device should reject a configuration that neither includes a source with a source token nor AutoCreateReceiver set to true. If the configuration doesn not include any tracks a device should assign all tracks of the corresponding recording. request JobConfiguration [tt:RecordingJobConfiguration] The configuration of the new recording job. response JobToken [tt:RecordingJobReference ] Identifies the created recording job. JobConfiguration [tt:RecordingJobConfiguration] The configuration as it used by the device. This may be different from the JobConfiguration passed to CreateRecordingJob. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken given in the JobConfiguration does not reference an existing recording. env:Sender - ter:InvalidArgVal - ter:BadConfiguration The contents of the JobConfiguration are invalid. env:Receiver - ter:Action - ter:MaxRecordingJobs The maximum number of recording jobs that the device can handle has been reached. env:Receiver - ter:Action - ter:MaxReceivers If the AutoCreateReceivers flag is TRUE, this error can be returned if the receiver service cannot create a new receiver. access class ACTUATE If a request with AutoCreateReceiver is accepted the response shall provide the attached receiver token and the AutoCreateReceiver field shall be omitted. The device response shall include the complete JobConfiguration including the associated job token and the resulting recording track configuration.

DeleteRecordingJob removes a recording job. It shall also implicitly delete all the receiver objects associated with the recording job that are automatically created using the AutoCreateReceiver field of the recording job configuration structure and are not used in any other recording job. request JobToken [tt:RecordingJobReference] Identifies the recording job that shall be deleted. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter:NoRecordingJob The JobToken does not reference an existing job. access class ACTUATE

GetRecordingJobs shall return a list of all the recording jobs in the device. request This is an empty message. response JobItem– optional, unbounded [tt:GetRecordingJobsResponseItem] Identifies a job in the device and holds its current configuration. faults None access class READ_MEDIA

SetRecordingJobConfiguration shall change the configuration for a recording job. A device shall reject a request that tries to modify the RecordingToken. The JobConfiguration returned from SetRecordingJobConfiguration by a device shall be identical to the JobConfiguration passed into SetRecordingJobConfiguration, except for the ReceiverToken and the AutoCreateReceiver. In the returned structure, the ReceiverToken shall be present and valid and the AutoCreateReceiver field shall be omitted. request JobToken [tt:RecordingJobReference] Identifies the recording job to update. JobConfiguration [tt:RecordingJobConfiguration ] The configuration to apply to the recording job. response JobConfiguration [tt:RecordingJobConfiguration ] The JobConfiguration structure shall be the configuration as it is used by the device. This may be different from the JobConfiguration passed to SetRecordingJobConfiguration. faults env:Sender - ter:InvalidArgVal - ter:NoRecordingJob The JobToken does not reference an existing job. env:Sender - ter:InvalidArgVal - ter:BadConfiguration The contents of the JobConfiguration are invalid. env:Receiver - ter:Action - ter:MaxReceivers If the AutoCreateReceivers flag is TRUE, this error can be returned if the receiver service cannot create a new receiver. access class ACTUATE SetRecordingJobConfiguration shall implicitly delete any receiver objects that were created automatically if they are no longer used as a result of changing the recording job configuration.

GetRecordingJobConfiguration shall return the current configuration for a recording job. request JobToken [tt:RecordingJobReference] Identifies the recording job for which to retrieve the configuration. response JobConfiguration [tt:RecordingJobConfiguration} The current configuration of the recording job. faults env:Sender - ter:InvalidArgVal - ter:NoRecordingJob The JobToken does not reference an existing job. access class READ_MEDIA

SetRecordingJobMode shall change the mode of the recording job. Using this method shall be equivalent to retrieving the recording job configuration, and writing it back with a different mode. Note that the state of a recording job will only become active if the recording job has the highest priority of all active jobs of a recording. request JobToken [tt:RecordingJobReference] Identifies the recording job for which to change the recording mode. Mode [tt:RecordingJobMode] The new mode for the recording job. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter:NoRecordingJob The JobToken does not reference an existing job. env:Sender - ter:InvalidArgVal - ter:BadMode The Mode is invalid. access class ACTUATE

GetRecordingJobState returns the state of a recording job. It includes an aggregated state, and state for each track of the recording job. The RecordingJogState may change due to calls that effect the RecordingJobMode, e.g. SetRecordingJobMode, internal recording engine state changes, changes in the recorded local media profile or changes to the RTSP connection defined by the associated Receiver. request JobToken [tt:RecordingJobReference] Identifies the recording job for which to get the state. response State [tt:RecordingJobReference] The state of the recording job. faults env:Sender - ter:InvalidArgVal - ter:NoRecordingJob The JobToken does not reference an existing job. access class READ_MEDIA The UML representation of the RecordingJobStateInformation structure is:

State (as part of RecordingJobStateInformation) shall hold the aggregated state over the whole RecordingJobInformation structure. SourceToken shall identify the data source of the recording job. State (as part of RecordingJobStateSource) shall hold the aggregated state over all substructures of RecordingJobStateSource. SourceTag shall identify the track of the data source that provides the data. Destination shall indicate the destination track State (as part of RecordingJobTrackState) shall provide the job state of the track. The valid values of state shall be “Idle”, “Active” and “Error”. If state equals “Error”, the Error field may be filled in with an implementation defined value. Error, optional string describing the error state. The string should be in English. The following values are predefined: “Incompatible Stream” – The stream cannot be recorded because the encoding does not match to previously recorded data. A device shall apply the following rules to compute aggregate state Idle All state values in sub-nodes are “idle” PartiallyActive The state of some sub-nodes are “active” and some sub-nodes are “idle” Active The state of all sub-nodes is “Active” Error At least one of the sub-nodes has state “Error”

GetRecordingOptions returns information for a recording identified by the RecordingToken. The information includes the number of additional tracks as well as recording jobs that can be configured. This method shall be supported if the Options support is signaled via the capabilities. Note that this information is not static and is only guaranteed to be valid until the next modification of any recording jobs or tracks. The track options shall be supported if the device signals support for dynamic tracks. request RecordingToken [tt:RecordingReference] Identifies the recording. response JobOptions [trc:JobOptions ] Contains two attributes: Spare: Number of spare jobs that can be created for the recording. By setting none of the Spare attribute the device signals that no job can be created. CompatibleSources: A device that supports recording of a restricted set of Media/Media2 Service Profiles shall return the list of profiles that can be recorded on the given Recording. TrackOptions [trc:TrackOptions] Contains four attributes: SpareTotal: Total spare number of tracks that can be added to this recording. SpareVideo: Number of spare video tracks for this recording SpareAudio: Number of spare audio tracks for this recording SpareMetadata: Number of spare metadata tracks for this recording By setting none of the spare attributes the device signals that no track can be added. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingToken does not reference an existing recording. access class READ_MEDIA

ExportRecordedData exports the selected recordings to the given storage target. A device that indicates a capability of SupportedExportFileFormats shall support this command. For the parameter FileFormat it shall accept any format that it advertises via the SupportedExportFileFormats capability. request StartPoint – optional [xs:dateTime] Specifies start time for the exporting. EndPoint – optional [xs:dateTime] Specifies end time for the exporting. SearchScope [tt:SearchScope] Defines the selection criterion for the existing recordings. FileFormat [xs:string] Indicates which export file format to be used. StorageDestination [tt:StorageReferencePath] Indicates the target storage and relative directory path. response OperationToken [tt:ReferenceToken} The asynchronous operation token for associating the received event with this invocation. FileNames - optional, unbounded [xs:string] List of exported file names. A client can also use AsyncronousOperationStatus event to monitor the progress of ExportRecordedData operation. faults env:Sender - ter:InvalidArgVal - ter:BadConfiguration: The device cannot support the selected FileFormat for exported files. access class READ_MEDIA A device should return the resulting list of file names in the response. In cases where the retrieval of the file names is a longer lasting operation these file names may be reported only via the AsyncronousOperationStatus. The ExportRecordedDataResponse returns the unique operation token that is used by client to monitor the progress of this invocation. The progress of export recordings operation is obtained via an event (Core Spec, Monitoring/AsynchronousOperationStatus). A device can notify progress of exported files via optional FileProgressStatus (tt:ArrayOfFileProgress) element, containining an array of file name and completion percentage of file upload from device’s point of view. The value of completion percentage is normalized between 0.0 and 1.0 where 1.0 indicates 100% completion of file upload. The optional FileProgressStatus elemet is sent in Data section of a Message. If a delete recording request is issued during an export of recordings and there are common recordings, the device shall delete the relevant recording after completing the export of these relavant recordings.

Stops the ExportRecordedData operation that is started before. The response message lists the status of the exported files. A device that indicates a capability of SupportedExportFileFormats shall support this command. request OperationToken [tt:ReferenceToken] Identifies the ExportRecordedData operation to stop. response Progress [xs:float] Completion percentage of total file upload from device’s point of view. The value of completion percentage is normalized between 0.0 and 1.0 where 1.0 indicates 100% completion of total file upload. FileProgressStatus [tt:ArrayOfFileProgress] An array of file names and an individual progress for each file. The value of completion percentage is normalized between 0.0 and 1.0 where 1.0 indicates 100% completion of the file upload. faults env:Sender - ter:InvalidArgVal - ter:BadConfiguration access class READ_MEDIA The requested operation does not exist.

GetExportRecordedDataState returns the status of export operations. This interface allows client to poll the status information from the device. A device that indicates a capability of SupportedExportFileFormats shall support this command. request OperationToken [tt:ReferenceToken] Identifies the ExportRecordedData operation from which to get status. response Progress [xs:float] Completion percentage of total file upload from device’s point of view. The value of completion percentage is normalized between 0.0 and 1.0 where 1.0 indicates 100% completion of total file upload. FileProgressStatus [tt:ArrayOfFileProgress] An array of file names and an individual progress for each file. The value of completion percentage is normalized between 0.0 and 1.0 where 1.0 indicates 100% completion of the file upload. faults env:Sender - ter:InvalidArgVal - ter:BadConfiguration access class READ_MEDIA The requested operation does not exist.

OverrideSegmentDuration dynamically updates a RecordingConfiguration to produce smaller recording segments for a limited time. This allows for a client to request faster access to recorded data while reducing the overall storage costs during normal operation. When the temporary override is active, the device should attempt to close its currently recording segment and then shall record further segments with the new duration provided in this command. The override shall be active for the duration specified in the command, once the command expires the device shall resume recording with the original segment duration on its next segment. The override duration shall not exceed 1 hour. When a new override request is sent before the previous override has expired, the new override shall replace the previous override and the new expiration shall apply. The active override value, if any exists, shall be shown in the corresponding RecordingConfiguration and may not persist after a device restart. If the recording configuration is modified while an override is active, the override shall be removed. A device that indicates support for OverrideSegmentDuration shall support this command. request TargetSegmentDuration [xs:duration] The new segment duration to be used for the duration of the override request. Expiration [xs:duration] The duration for which the override request shall be active. RecordingConfiguration [tt:RecordingReference] The recording configuration that needs to use the new segment duration. response This is an empty message. faults env:Sender - ter:InvalidArgVal - ter:NoRecording The RecordingConfiguration does not reference an existing recording. env:Sender - ter:InvalidArgVal - ter:BadDuration The value of the duration parameter is invalid. env:Receiver - ter:ActionNotSupported - ter:NotImplemented This optional method is not implemented. access class ACTUATE

The capabilities reflect optional functions and functionality of a service. The information is static and does not change during device operation. request This is an empty message. response Capabilities [trc:Capabilities] List of capabilities as defined below. faults None access class PRE_AUTH The following capabilities are available: DynamicRecordings Indication if the device supports dynamic creation and deletion of recordings. DynamicTracks Indication if the device supports dynamic creation and deletion of tracks. Encoding Indication which encodings are supported for recording. The list may contain one or more enumeration values of tt:VideoEncoding and tt:AudioEncoding. For encodings that are neither defined in tt:VideoEncoding nor tt:AudioEncoding the device shall use the IANA defintions http://www.iana.org/assignments/media-types/media-types.xhtml. Note, that a device without audio support shall not return audio encodings. MaxRate Maximum supported bit rate for all tracks of a recording in kBit/s. MaxTotalRate Maximum supported bit rate for all recordings in kBit/s. MaxRecordings Maximum number of recordings supported. MaxRecordingJobs Maximum total number of supported recording jobs by the device. Options Indication if the device supports the GetRecordingOptions command. MetadataRecording Indication if the device supports recording metadata. SupportedExportFileFormats Indication that the device supports ExportRecordedData command for the listed export file formats. A device shall ony return this capability if it contains at least one export file format value. The value of ‘ONVIF’ refers to ONVIF Export File Format Specification. ScheduledRecording Indication that the device supports scheduled recording. EventRecording Indication that the device supports event triggered recording. BeforeEventLimit Indicates that the device supports before event durations up to the given value. AfterEventLimit Indicates that the device supports after event durations up to the given value. SupportedTargetFormats List of formats supported by the device for recording to an external target. See tt:TargetFormat for a list of definitions. EncryptionEntryLimit Number of encryption entries supported per recording. By specifying multiple encryption entries per recording, different tracks can be encrypted with different configurations. SupportedEncryptionModes Supported encryption modes. See tt:EncryptionMode for a list of definitions. OverrideSegmentDurationSupported Indicates that the device supports the OverrideSegmentDuration command.

Some of these events are similar to the automatically generated events that can be searched for by the FindEvents method in the search service. See ONVIF Recording Search Service Specification.

If the state field of the RecordingJobStateInformation structure changes, a device shall provide the following event: Topic: tns1:RecordingConfig/JobState ]]> The ElementItem Information shall be provided whenever the state of the different tracks is not unique. It can be omitted when the state of all tracks of a recording is consistent.

If the configuration of a recording is changed, a device shall provide the following event: Topic: tns1:RecordingConfig/RecordingConfiguration ]]> If the configuration of a track is changed, a device shall provide the following event: Topic: tns1:RecordingConfig/TrackConfiguration ]]> If the configuration of a recording job is changed, a device shall provide the following event: Topic: tns1:RecordingConfig/RecordingJobConfiguration ]]>

Whenever data is deleted, a device shall provide the following event: Topic: tns1:RecordingConfig/DeleteTrackData ]]>

Whenever a recording is created, a device shall provide the following event: Topic: tns1:RecordingConfig/CreateRecording ]]> Whenever a recording is deleted, a device shall provide the following event: Topic: tns1:RecordingConfig/DeleteRecording ]]> Whenever a track is created, a device shall provide the following event: Topic: tns1:RecordingConfig/CreateTrack ]]> Whenever a track is deleted, a device shall provide the following event: Topic: tns1:RecordingConfig/DeleteTrack ]]>

There are two steps involved. The first step is to configure the NVS This completes the configuration step. Finally, to really start recording, some entity calls to make the job active. This will cause the NVS to set up an RTSP connection with the device. Therefore, to start and stop recording, all that is needed is to call SetRecordingJobMode on pre-configured recording jobs. And since the embedded configuration objects are persistent, the configuration cycle only needs to be done once.

This example is very similar to example 1. The jobconfiguration will hold references to two receiver objects. Each receiver object is configured to receive from the same device, but from a different stream. To really start recording, some entity calls

This annex describes a scenario for Multiple Recording Jobs configured to record data into a single recording. Accordingly, a device supporting Multiple Recording Jobs is required to change the Job Modes of Recording Jobs with respect to Priority, as described below : Step 1 A Recording Job ‘J1’ with Priority ‘2’ is created in 'Active' mode Job State of Recording Jobs after Step 1: Recording Job ‘J1’ = ACTIVE Step 2 A new Recording Job ‘J2’ with Priority ‘4’ is now created in 'IDLE' mode Job States of Recording Jobs after Step 2: Recording Job ‘J1’ = ACTIVE Recording Job ‘J2’ = IDLE Step 3 Another Recording Job ‘J3’ with higher Priority ‘3’ is now created in 'Active' mode. Because it has a higher priority than J1, it takes precedence. Job States of Recording Jobs after Step 3: Recording Job ‘J1’ = IDLE Recording Job ‘J2’ = IDLE Recording Job ‘J3’ = ACTIVE Step 4 Job States of Recording Jobs after Step 4: Recording Job ‘J3’ is now deleted. 'J1' becomes active again since it is the active job with the highest priority. Recording Job ‘J1’ = ACTIVE Recording Job ‘J2’ = IDLE

This annex describes the requirements for recording from a device to an object storage for later use by a receiver. The device could be a camera, the receiver could be a video player. shows the data flow for object storage recording.

The device uploads recorded media as objects containing fragmented MP4 data according to ISO/IEC 14496-12, the ISO base media file format. In this annex, a media object containing fragmented MP4 data is called a segment. Each segment starts with a header containing initialization data, usually consisting of a FileTypeBox and a MovieBox. Each segment contains one or more fragments with recorded media, usually consisting of a MovieFragmentBox and a MediaDataBox. Each segment contains recorded media of a continuous period of time. Multiple successive segments without a time gap in between form a span. In this way, the size of segments can be limited when recording continuously for a longer period of time. For a single span, only a single segment contains recorded media of one point in time. Depending on the recording format, video, audio, and metadata are written into separate segments. In this case, the segments of each media type form separate spans. shows the logical model for object storage recording.

When a recording is activated, the device shall open new spans for this recording according to the rules of the configured Format. Inversely, when a recording is deactivated, the device shall close all open spans of this recording. When a SpanDuration is configured for a recording, the recording is active, and the duration of a span of this recording reaches the configured SpanDuration, the device shall close all open spans of the recording and open new spans according to the rules of the configured Format. All fragments of a span shall be independently processable and presentable with the information of a MovieBox in any of the segments of the span. If the information in the MovieBox required to process the fragments needs to change while the recording is active, the device shall close all open spans of the recording and start new spans according to the rules of the configured Format.

When a span is opened, the device shall also open a new segment for this span. Inversely, when a span is closed, the device shall close the open segment of this span. When a span is open and the duration of the open segment of this span reaches the SegmentDuration configured for the recording, the device shall close the open segment of this span and open a new segment for this span. When closing the open segment of a span and opening a new segment for this span, there shall be no time gap between the two segments. Note that the actual segment duration may vary and can be plus or minus one GOP size from the configured duration because a segment must start with an I-Frame. The device shall generate each segment as a fragmented MP4 file according to ISO/IEC 14496-12, the ISO base media file format. A segment shall contain recorded media for the time it was open. The recording formats provide specification about segment types, their contents, and restrictions on the ISO base media file format in the following sections.

The device shall upload each segment as one object, identified by its object key, to the object storage. The device shall generate the object key for a segment according to the following ABNF key rule (according to RFC 5234): key = lead "/" time "." counter "." extension lead = [prefix "/"] date ["/" postfix] counter = 1*DIGIT date = year "-" month "-" day time = hour "/" minute "-" second [secfrac] "Z" year = 4DIGIT month = 2DIGIT ; 01-12 day = 2DIGIT ; 01-31 hour = 2DIGIT ; 00-23 minute = 2DIGIT ; 00-59 second = 2DIGIT ; 00-60 (due to leap second rules) secfrac = "." 1*6DIGIT If a Prefix is configured for the recording, the device shall include prefix in lead and replace it with the configured Prefix. If a Postfix is configured for the recording, the device shall include postfix in lead and replace it with the configured Postfix. The device shall replace date and time respectively with the date and time portion of the first sample contained in the segment in Universal Coordinated Time (UTC), following the ABNF rules: year represents the 4-digit year month represents the 2-digit month (01 to 12) day represents the 2-digit day of the month (01 to 31) hour represents the 2-digit hour (00 to 23) minute represents the 2-digit minute (00 to 59) second represents the 2-digit second (00 to 60, due to leap second rules) secfrac represents the optional fractional seconds with a precision of up to 6 digits The device shall replace counter with a span-dependant counter set to 0 for a new span and incremented by 1 when closing a segment of this span. The device shall replace extension with the extension defined by the type of span the segment belongs to. Note: date is conformant to the extended date format defined in ISO 8601-1. time is not conformant to the extended time format defined in ISO 8601-1.

To store the end time of a span, the device should store an additional empty object when closing a span. The device shall generate the object key for this object according to the ABNF key rule and replacement rules defined in . The device shall replace date and time respectively with the date and time portion of the end of the last sample contained in the span in Universal Coordinated Time (UTC), following the ABNF rules. The device shall replace counter by a value equal to the counter of the last segment of the span plus 1. The device shall replace extension according to the following ABNF extension rule (according to RFC 5234): extension = span-extension "_end" The device shall replace span-extension with the extension defined by the span type.

If the Format configured for a recording is MP4, the device shall follow the rules given in this section. The device shall open a single span for an active recording containing all tracks of the recording. The device shall use mp4 as the extension for this span type. The device shall upload each segment object belonging to a span of this type with the content type video/mp4. The device should include the codecs parameter in the content type according to RFC 6381. The MovieBox contained in a segment shall not reference any samples. A segment shall contain a MovieFragmentRandomAccessBox as the last box. The MovieFragmentRandomAccessBox shall provide references for all tracks and their applicable MovieFragmentBox instances.

If the Format configured for a recording is CMAF, the device shall follow the rules given in this section. The device shall open up to three spans for an active recording, each containing a single track of the recording. If the recording contains a video track, the device shall open a video span containing this track. The device shall use m4v as the extension for this span type. The device shall upload each segment object belonging to a span of this type with the content type video/mp4. If the recording contains an audio track, the device shall open an audio span containing this track. The device shall use m4a as the extension for this span type. The device shall upload each segment object belonging to a span of this type with the content type audio/mp4. If the recording contains a metadata track, the device shall open a metadata span containing this track. The device shall use m4m as the extension for this span type. The device shall upload each segment object belonging to a span of this type with the content type application/mp4. The device should include the codecs parameter in the content type according to RFC 6381 for all span types defined in this section. The device shall generate each segment according to ISO/IEC 23000-19, the common media application format (CMAF) for segmented media, with one CMAF header and one or more CMAF fragments. A segment may contain a MovieFragmentRandomAccessBox as the last box. If a segment contains a MovieFragmentRandomAccessBox, every CMAF fragment shall be referenced in the MovieFragmentRandomAccessBox. Note: As the different tracks are contained in different spans, the segments do not necessarily align. This is illustrated in .

A device signaling support for recording to an external target with encryption shall support writing encrypted files according to ISO/IEC 23001-7 (common encryption in ISO base media file format files). Each Encryption entry configured for a recording covers a distinct set of tracks for which to apply the encryption, identified by the Track element. If an encryption entry contains no Track elements, it covers all tracks of the recording. If an encryption entry contains one or more Track elements, it covers the tracks indicated by the track tokens contained in these elements. The device shall encrypt the covered tracks with the scheme given by the Mode configured for the encryption entry: CENC AES-CTR mode full sample and video NAL Subsample encryption, defined in ISO/IEC 23001-7. CBCS AES-CBC mode partial video NAL pattern encryption, defined in ISO/IEC 23001-7. The device shall create an individual initialization vector for each segment. The device shall encrypt with the Key configured for the encryption entry as the encryption key. The device shall interpret the KID configured for the encryption entry as a 16-byte hexadecimal value and use this value as the key identifier. If an encryption entry is reconfigured for an active recording, the device shall continue to use the old configuration for any open segments and shall start to use the new configuration for new segments.

The device may store additional attributes defined in as object metadata when uploading segment objects. Note: When sending attributes as metadata HTTP headers, the attribute keys need to be prefixed with a value determined by the type of storage configured for the recording. If the storage is of type ObjectStorageS3, the attribute keys need to be prefixed with x-amz-meta- (for example x-amz-meta-Version: 1.0). If the storage is of type ObjectStorageAzure, the attribute keys need to be prefixed with x-ms-meta- (for example x-ms-meta-Version: 1.0). Audio-related attributes may be stored for segment objects containing only a video track and video-related attributes may be stored for segment objects containing only an audio track. Key Value Format Version 1.0 major "." minor Source Identifier of the physical device String EventTopic Topic of the event that triggered the recording job String Duration Duration of the segment ISO 8601-1 FullContentType Combined content type of all media types contained in the recording video/mp4 content type with codecs parameter according to RFC 6381 VideoChannel Video channel identifier String VideoBitrate Video bitrate in kilobits per second Integer VideoWidth Video width in pixels from VisualSampleEntry Integer VideoHeight Video height in pixels from VisualSampleEntry Integer VideoFramerate Average video framerate in frames per second Float AudioChannel Audio channel identifier String AudioBitrate Audio bitrate in kilobits per second Integer AudioSampleRate Audio sample rate in kilohertz Integer For storage target systems that do not allow attribute changes during write the attribute Duration should be omitted when the total duration of the file may change while it is written e.g. due to an override segment duration request.

Given the following storage configuration in a recording configuration: 1 MP4 site-2 camera-5 PT60S ]]> The key of a segment object starting on 2022-11-08 at 15:08:23.547 (UTC) is: site-2/2022-11-08/camera-5/15/08-23.547Z.0.mp4 The key of a span end object for a span consisting of 85 segments and ending on 2022-11-08 at 16:34:11.421 (UTC) is: site-2/2022-11-08/camera-5/16/34-11.421Z.85.mp4_end Given the following storage configuration in a recording configuration: 1 CMAF site-2 camera-5 PT5S ]]> The key of a video segment object containing starting on 2022-11-08 at 15:08:23.547 (UTC) is: site-2/2022-11-08/camera-5/15/08-23.547Z.0.m4v The key of a span end object for a video span consisting of 1020 segments and ending on 2022-11-08 at 16:34:11.421 (UTC) is: site-2/2022-11-08/camera-5/16/34-11.421Z.1020.m4v_end The key of an audio segment object starting on 2022-11-08 at 15:08:23.989 (UTC) is: site-2/2022-11-08/camera-5/15/08-23.989Z.0.m4a The key of a span end object for an audio span consisting of 1020 segments and ending on 2022-11-08 at 16:34:11.779 (UTC) is: site-2/2022-11-08/camera-5/16/34-11.779Z.1020.m4a_end The key of a metadata segment object starting on 2022-11-08 at 15:08:23.692 (UTC) is: site-2/2022-11-08/camera-5/15/08-23.692Z.0.m4m The key of a span end object for a metadata span consisting of 1020 segments and ending on 2022-11-08 at 16:34:11.871 (UTC) is: site-2/2022-11-08/camera-5/16/34-11.871Z.1020.m4m_end