MJPEG is a lossy codec but the visual quality will be the MJPEG would be a good codecįor that, YouTube definitely supports that in an. YouTube and not about TV broadcast or Cinema.īut if you really want to do it the "perfect" way, use a productionĬodec and not a consumer codec like h.264. Pracitically I'd say it really doesn't matter as we are talking about But please note this is all just a theoretical thing, The last step in production and during production you want to stay Output into a delivery/consumer format and the upload to YouTube is Theoretically best quality for your uploads choose a lossless codecįor uploading or atleast visually lossless. YouTube will ALWAYS encode your video once its uploaded, doesn't Nice after it got encoded by the YouTube servers. YouTube will accept this just fine and it will look Use h.264 Level 3.1/4.1 with Main Profile for SD or High Profile for That depends, if you are limited/concerned by your upload speed then What codec/container should you use for uploading? In regard of your linked question (maybe they should be merged?) Videos (going by the videos I saw encoded in WebM) so its not certain Some videos in WebM after they got uploaded and mostly only popular Was introduced with the HTML5 version of YouTube. Even on YouTube only very few videos are encoded in WebM, WebM Standard video codec for HTML5, though the support for it is not veryīig. WebM is a format developed by Google and was meant as a The other codec used is VP8 which is coming in the WebM containerįormat. Mobile devices (long before the smartphone era), it also comes in the ForĢ40p YouTube is using 3gp which is an rather old codec meant for Version for the highest available resolution, either 360p or 480p. Though every video below 720p will also have a mp4 Meaning only 360p and 480p will exist in an Generally,Įvery of your uploaded video will be encoded in h.264 and will be Video what codecs are used for your video (see below why). Some general info about the formats used: YouTube uses 4 containerįormats and 3 diffrent codecs. There is a great answer on for YouTube-codecs and more 2016, Optimizing Transcoder Quality Targets Using a Neural Network with an Embedded Bitrate Model So even if you had a well-compressed input file, YouTube would have to generate dozens of alternative, scaled versions anyway. HTTP Adaptive Streaming as it's used these days requires generating different sets of videos for different client types (e.g., 4K streaming to a modern browser on the desktop is different from streaming low-res video to a legacy Android device). They wouldn't just take your video as-is. They spend massive amounts of CPU time and human efforts into optimizing that aspect. Optimization of quality: YouTube and others know best how they need to compress videos such that the output is of good quality and low bitrate. So you'll convert the incoming video to that format and any further processing is going to be easier. You want to have a single kind of video format (an "intermediate format" or "mezzanine") where you know that it's not going to break your existing infrastructure. For example, decoding might not properly work, there may be wrong timestamps, and, perhaps a user was trying to upload a malicious file, trying to exploit bugs in the software. Lots of different encoders exist, and they might produce bitstreams with special attributes that may not be supported by all components of the streaming architecture. Video codecs/formats can be weird (for the lack of a better word). normalize files before transcoding by creating a high-bitrate, constant-frame-rate mezzanine. This has been confirmed by YouTube researchers. You cannot avoid YouTube or others recompressing your input video.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |