> Is there an example of an encoder where per-frame control of keyframes isn't possible? Is there a risk with the per-frame API for keyframes, that the encoded stream has more visual artifacts as a result of the encoder having a harder time predicting how to allocate the target bitrate, than if it were to configure the encoder with a per-frame interval? I have seen behavior in libvpx suggesting this, but I'm not sure how much of it was due to using VPX_EFLAG_FORCE_KF and how much was due to a [libvpx bug](https://bugs.chromium.org/p/webm/issues/detail?id=1714). IIRC Firefox sets a keyframe interval of 10 seconds, for historical reasons [1], and occasionally recalculates the frame-interval based on recent incoming framerate, so it can configure the encoder with this. [1] our ebml writer only creates a new cluster on a keyframe, and the max timestamp within a cluster is 32 seconds. -- GitHub Notification of comment by Pehrsons Please view or discuss this issue at https://github.com/w3c/mediacapture-record/pull/216#issuecomment-1415802940 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-configReceived on Friday, 3 February 2023 12:24:45 UTC
This archive was generated by hypermail 2.4.0 : Saturday, 6 May 2023 21:20:00 UTC