Video
When triggered. Video is only recorded when the camera is in a triggered state. This can greatly save on storage space over continuous recording. The pre-trigger video buffer applies only to this type of video recording. By default, multiple trigger events are stored into one file—see Combine or cut below.
When alerted. Much like When triggered, but video is only recorded when the camera is in a triggered state and the alert has been confirmed by AI or otherwise.
Continuous. Record video all of the time the camera is online and active.
Periodic. Record video in defined intervals in a discontinuous or time-lapse manner— meaning that playback gaps are removed.
Triggered + periodic. Combines when triggered with periodic. All frames are recorded when the camera is in the triggered state.
Continuous sub + alerts. When used with a dual-streaming camera along with direct to-disk, the result is a BVR file which will contain the sub-stream continuously recorded, but the main-stream only recorded when the camera is in an alerted state. During main-stream playback, the sub-stream will be upsampled whenever the mainstream is not available. This dual-stream hybrid may be exported only when reencoding is selected, although you may still export the main-stream in segments when not re-encoding. When not using dual-streams or direct-to-disk, recording of
the sub (or only) stream is continuous.
Continuous sub + triggered. Similar to Continuous sub + alerts, but main stream recording occurs upon trigger, not just alerts, ensuring that the main stream is recorded even for canceled alerts.
Now this text looks to contain your issue:Direct-to-Disk
This option may be used to save considerable CPU time. Video taken from a network IP camera stream is directly saved to disk without re-encoding it. There are some disadvantages to this however:
• Recording may only begin on a key frame boundary. Check that your camera is sending sufficient key frames on the Cameras page in Status (at least 0.50 per second).
• Video overlays are saved as meta-data, not directly burned into the video. This may be an advantage however as you may toggle from on/off during playback.
• Cannot be used with time-lapse options. All video frames must be saved, as each frame relies on the previous one in order to be decoded for display.
• You may want to alter the quality, size, or other video characteristics.
• May only be used by network IP cameras with H.264 or H.265 streams.
• May not be used by analog or USB cameras, unless it’s an uncommon source that supplies H.264 or H.265 video.