Question / Help What's the difference between all the output formats?
- Thread starter Testnick
- Start date
Is there somewhere on here or elsewhere that I could educate myself on the differences between the different formats and what the properties, pros and cons of each are? I'm a clueless newb, so when it says "save to mkv and use muxer..." I don't know what mkv is or what a muxer even is (beyond the obvious) or how I would do any of that. This is the first time I've ever experimented with live streaming or video, I've only ever done a few slideshows to post mixes on YouTube, so this is totally uncharted territory. I need like remedial classes lol. I'll do some google research and search here meantime.
Some background: I am a professional EDM DJ, but am pretty much functionally retired from playing gigs, for the foreseeable future, so I am trying to get into live-streaming as a way to still reach my audience. I am currently specifically trying to understand what are the best options for separating audio from a video, (so I can have the mix in mp3 for my ipod and uploading to music streaming sites) and the best settings to facilitate that, and why, and also I need to edit the videos afterward for uploading to YouTube etc., and I see that iMovie does not support flv, as well as some other programs/websites do not seem to support it for uploads, although YouTube does. I got "FlvCrunch" but the converted videos are messed up, with the sound playing at the wrong speed and then skipping like it's trying to sync up but can't. Idk why, but I decoded it might be better to save it to another format that is more widely accepted by the sites I'm using, and also by editing software, since FlvCrunch is supposed to be the best free app lol. Or perhaps you know of a better (free, I'm dirt poor) video editing app.
So these are things I need to understand better; can you tell me some resources (and save yourself explaining it all to me lol)? And please excuse my obtuseness if it is already covered elsewhere here! I appreciate your input/patience.
koala
Active Member
Every "output format" describes the container format the video data is stored in. The compressed video frames are all the same in every container, but the format of the envelope that wraps the frames is different for every container format.
The container formats differ in usability. Some software supports this container format only, some software supports that format only. Not every compression style can be stored in every format. For example, you cannot store h.264-compressed frames in an .avi container. Some container can only store fixed-fps videos, some other container can also store variable-fps videos. There is also a difference of robustness between container formats: if you truncate a video file, the whole contents is lost for some container formats, even if the beginning seems intact.
The .mp4 format has the advantage that it is supported by the vast majority of devices and applications. It is indexed, so random access is fast. It is also able to store the current standard compression style (h.264+aac), supports multiple audio tracks and supports variable-fps video. It has the caveat that if you record a .mp4 and don't gracefully terminate the recording, for example due to a crash or disk full situation, the whole file is unusable and completely lost. This is a showstopper for important recordings that cannot be rerun.
From a technical point of view, the best format is .mkv, because it can store every combination of audio and video compression style, can do everything any other container format can, and it is robust against defects. Even if you truncate a .mkv file, the rest is still usable. It's the universal format. Unfortunately, it is not widely supported in applications and devices.
The bullet-proof workflow is to let OBS create *.mkv files and remux (convert) the file afterwards to the container format of your liking, most often .mp4. You do this with File->Remux Recordings. Remuxing only changes the container format, it does not recompress the video or audio. So quality is preserved. However, it fails, if the destination container format doesn't support the compression style that was used in the .mkv.
The container formats differ in usability. Some software supports this container format only, some software supports that format only. Not every compression style can be stored in every format. For example, you cannot store h.264-compressed frames in an .avi container. Some container can only store fixed-fps videos, some other container can also store variable-fps videos. There is also a difference of robustness between container formats: if you truncate a video file, the whole contents is lost for some container formats, even if the beginning seems intact.
The .mp4 format has the advantage that it is supported by the vast majority of devices and applications. It is indexed, so random access is fast. It is also able to store the current standard compression style (h.264+aac), supports multiple audio tracks and supports variable-fps video. It has the caveat that if you record a .mp4 and don't gracefully terminate the recording, for example due to a crash or disk full situation, the whole file is unusable and completely lost. This is a showstopper for important recordings that cannot be rerun.
From a technical point of view, the best format is .mkv, because it can store every combination of audio and video compression style, can do everything any other container format can, and it is robust against defects. Even if you truncate a .mkv file, the rest is still usable. It's the universal format. Unfortunately, it is not widely supported in applications and devices.
The bullet-proof workflow is to let OBS create *.mkv files and remux (convert) the file afterwards to the container format of your liking, most often .mp4. You do this with File->Remux Recordings. Remuxing only changes the container format, it does not recompress the video or audio. So quality is preserved. However, it fails, if the destination container format doesn't support the compression style that was used in the .mkv.
Last edited:
Thanks a lot!! Very helpful! So, I tried to do the File->Remux option for a video I made last night, following instructions on a YouTube tutorial video. But in the second box where it says "target" I notice that the the extension is still .flv. So it's going from .flv to .flv apparently lol. I'm assuming I need to make that say .mp4, but I can't find an option in the settings to change the remux output, am I missing it? Or can I just type .mp4 in that box? The video I saw didn't cover it, he made it seem like it would be already set to mp4 by default, but mine is evidently not.
Also do you know about the multiple audio tracks option? Why would you want multiple tracks, and does it separate into stereo channels? Or are they just duplicate mono channels? I can't imagine why you would need/want to have 5 redundant audio tracks, so there must be some kind of breakdown, no? Thanks again!!
Also do you know about the multiple audio tracks option? Why would you want multiple tracks, and does it separate into stereo channels? Or are they just duplicate mono channels? I can't imagine why you would need/want to have 5 redundant audio tracks, so there must be some kind of breakdown, no? Thanks again!!
koala
Active Member
Just type .mp4 in the output box.
Multiple audio tracks is for writing different audio sources to separate tracks. For example your game audio, your mic audio, your desktop audio, your voip audio. You use multiple audio tracks if you want to postprocess your video and postprocess every audio track separately. You might duck a background audio track, remove noises from vocal tracks, normalize and compress them, etc. As final postprocessing step, you mix the separate tracks and replace them in the video with the mixed track.
Multiple audio tracks is for writing different audio sources to separate tracks. For example your game audio, your mic audio, your desktop audio, your voip audio. You use multiple audio tracks if you want to postprocess your video and postprocess every audio track separately. You might duck a background audio track, remove noises from vocal tracks, normalize and compress them, etc. As final postprocessing step, you mix the separate tracks and replace them in the video with the mixed track.