# Why is the recording file split? (/en/api-reference/faq/quality/record_split)

> For AI agents: see the complete documentation index at [llms.txt](/llms.txt).

During the recording process, if the recorded audio and video format is not the recording original audio and video data format, the following situations result in file splitting:

* In the individual recording mode, a user rejoins a channel after leaving the channel, or resumes sending audio/video after not sending for a period of time. The specific rules are as follows:

* For audio recording files:
  * The user sends the audio data again after not sending for more than 15 seconds.
  * The user rejoins the channel after leaving for more than 15 seconds.

* For video recording files:

  * The user sends the video data again after not sending for more than 15 seconds.
  * The user changes the video resolution.
  * The user leaves and rejoins the channel.

* In the composite recording, if the manual recording mode is set, pause and then resume recording.

If the file splits and the above situations do not apply, check the following:

* Ensure that the same channel profile is used between the Native/Web SDK and Recording SDK.
* Call `enableWebSdkInteroperability` in the Native SDK when the recording channel uses both the Native and Web SDK.