You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a stream switching filter which returns S_FALSE from its Receive method on its input pin while it is initiating seek processing, before sending the seek upstream. Standard DirectShow protocol is that the upstream sending filter, in this case the LAV Audio Decoder should stop sending samples until the graph stops or completes a flush operation (which would happen during the subsequent seek). The LAV Splitter Source and LAV Audio Decoder just keep sending samples, including calling EndOfStream if it gets to the end of the stream.
LAV Filters 0.70.2 x64.
JW
The text was updated successfully, but these errors were encountered:
Unfortunately, this is just the way it is. There have been way too many filters that didn't handle this right that playback would often break if S_FALSE would be taken as a literal "stop sending" command.
If you are queueing a seek anyway, wouldn't you just keep rejecting samples and then perform the seek, and all would be the same?
Once the seek has reached the source filter and a BeginFlush was issued, then no further samples should be delivered until EndFlush and NewSegment are called.
I have a stream switching filter which returns S_FALSE from its Receive method on its input pin while it is initiating seek processing, before sending the seek upstream. Standard DirectShow protocol is that the upstream sending filter, in this case the LAV Audio Decoder should stop sending samples until the graph stops or completes a flush operation (which would happen during the subsequent seek). The LAV Splitter Source and LAV Audio Decoder just keep sending samples, including calling EndOfStream if it gets to the end of the stream.
LAV Filters 0.70.2 x64.
JW
The text was updated successfully, but these errors were encountered: