![]() Improved: Better firewall handling via the 'Configure Windows Firewall' button in Media Server Settings dialog. Improved: Faster processing speed for active playlists. New: Support for streaming offline videos on Twitch channel. New: Support for streaming raw h264 files (.h264 and. New: Support for streaming native M3U8 HLS video. New: Network adaptor selection in Media Server Settings dialog lets you choose which network you broadcast and communicate with Mezzmo server. See Transcoding Settings dialog to turn on hardware acceleration. Take advantage of your Nvidia NVENC-enabled graphic card, Intel Quick Sync-enabled CPU and Microsoft Direct3D-enabled graphics card. When files aren't being transcoded, the situation is completely different, because the file is available on disk and the size is known, so jumping to any specific position in the source file is easy.New: Fast hardware acceleration when transcoding videos. ![]() The data that has already been transcoded isn't stored either, so UMS doesn't even know where it is byte-wise at any given time. There's simply no way to translate a byte-offset into a specific position when the data is being generated on-the-fly. But, those that only support byte-based positioning regardless of if UMS had done all it should correctly, ultimately cannot seek when transcoding. ![]() UMS doesn't handle all of these details properly, so there are probably some renderers where it could work had UMS done it correctly where it's not available. That said, it's not quite that simple, as there are multiple details I left out, that can impact what the renderer supports and not. This means that renderers that use time-based positioning does support seeking while transcoding, but those that only accepts byte-based positioning doesn't. The duration in time is known since it is the same as the duration of the source file. ![]() When transcoding, the duration in bytes in undefined, there's no way to know that until you have transcoded the whole file (aka played through it). The fundamental problem is that some use byte-based positioning (by number x from the start of the file) and some use time-based positioning. In short, seeking while transcoding works on some renderers but not on others. That said, the duration (in bytes, not seconds) would still be unknown so I don't quite see how it would solve anything. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |