OBS Studio: How to record Podcasts & Interviews w/ Multitrack WAV Audio! (OBS Studio Multichannel)

OBS Studio: How to record Podcasts & Interviews w/ Multitrack WAV Audio! (OBS Studio Multichannel)


If you want to record a podcast or interview,
often you will want to record in the highest possible audio quality format first, and then
compress later. Recording to a compressed file can result
in artifacting and noise – and while you might not hear it in the initial recording, it can
definitely come up if you’re applying any post-processing to the audio or compressing
it again for posting to a podcast host. BUT you probably want to record yourself and
the person you’re speaking with – or other sounds – to separate tracks to mix best later
on, too. Short of buying an expensive multi-track audio
recorder like my Sound Devices MixPre-3, you need to use a software solution. But Audacity by default doesn’t record system
sound and microphone audio to separate tracks, and it’s hard to find other software to
do it. BUT I’ve finally found (or been pointed
to by the developer in the OBS Discord server) a build of OBS called “OBS Studio Multichannel”
that might solve this dilemma. While I wish OBS Studio could have an alternate
“audio mode” where it only records audio and records multiple uncompressed WAV audio
tracks, this is the next best thing. OBS Studio Multichannel is a build of OBS
Studio that uses more updated FFMPEG libraries and enables multiple audio track recording
for the Custom Output (FFMPEG) recording mode. This means you can tell it to use your normal
X264 or Nvenc or whatever video format and PCM WAV audio and still have multiple tracks. (In case you were unaware, normal OBS Studio
only allows for single audio track recording in Custom FFMPEG mode.) So if you wish to record a podcast or interview
with uncompressed audio, download the OBS Studio MC installer from Github and install
it. Now close whatever OBS it opens up – it opens
your normal OBS installation, not the new one. You have to go to ‘C:Program Files (x86)obs-studio-MCbin64bit’
and run the OBS64.exe in that folder. Make a shortcut to it, if needed, to be safe. This launches the custom build. Now duplicate your normal profile you’d
use in this scenario, give the new one a name like “MultiChannel” or something. Open up the Setting. Go to Output, Recording, and now change “Standard”
to “Custom Output (FFMPEG).” Now things get a bit tricky here because you
have to navigate the custom modes to re-create your normal video settings. That is, if you care about how the video looks. I have attempted to re-make my normal lossless
Nvenc recording mode with these settings for you to compare. BUT if you don’t care about video at all,
set it to libx264 and just kind of ignore the video settings. However, you will need to choose “Matroska”
as your “Container Format.” This is MKV. MP4 can’t usually hold WAV audio streams. Under Audio Encoder, choose “pcm_s16le”
– that’s normal uncompressed WAV mode. Now you can check the different Audio Track
boxes for the tracks you wish to have recorded and it will put them all in one MKV file. I also went ahead and put the audio bitrate
to the maximum of 4096 just to be safe, but that shouldn’t affect anything for the PCM
encoder. Might as well be safe anyway. Hit Apply and theoretically you should be
good to go. Custom Output mode can be confusing and the
support team generally doesn’t support it – but it has been working so far for me. An issue that I have run into, however, is
that you can’t remux the MKV to a MP4 – it turns out to be 0 bytes. This, I believe, is because the MP4 container
can’t hold WAV audio tracks. In Premiere Pro CC 2018, as of the April update,
you can just edit the MKV file itself and use Premiere to separate out the audio tracks. Otherwise you have to get familiar w/ FFMPEG
scripting to extract the individual tracks to separate WAV files. This isn’t the most elegant solution, but
I do think it’s a good option for those who just want to record multi-track WAV for
podcasts or interviews. I would love to see a dedicated “Audio Mode”
for OBS Studio, listed in my “7 Features I want in OBS Studio” video – but for now,
this is the closest we can get. If you like this video, you know what to do. Also hit Subscribe for more tech education
and OBS tutorials. I’m EposVox, here to make tech easier and
more fun, and I’ll see you next time.

23 thoughts on “OBS Studio: How to record Podcasts & Interviews w/ Multitrack WAV Audio! (OBS Studio Multichannel)

  1. I have to disagree about audio compression adding noise and artifacts. In fact, an MP3 at 128kbps has less noise than a WAV. The audio compression acts as a low pass filter and thus removing the hiss/noise from higher frequencies. Nevertheless, I’ve never heard of this release of OBS and O l would definitely check it out. Thanks for sharing it!

  2. The FFMPEG custom output option is so underutilized in OBS. I really hope they revamp it to be more intuitive. Something like encoder presets for more professional applications like this would go a long way. We're clearly seeing a rise in the use cases for OBS outside of straight gameplay capture and/or streaming. It'd be nice to see those other use cases (podcasting, DSLR mirroring capture, etc.) be more well exposed and developed for OBS.

  3. This would have been a great alternative to the mixpad program from nch software I've had to use. Does it works with the stream deck?

  4. GREAT !!
    Is there a way that you can do a video on ASIO devices ? Not necessarily audio interfaces but when there is a single USB ASIO audio interface (like a guitar processor – Boss GT-1 GT-100 GT-001 GT-1000 etc ..). I have trouble when DAW has ASIO and I also use USB mics. And how this could be done in OBS.
    BTW, I love the videos and support you with as many views as I can

  5. May I ask: how do you get the "Video Transcode" option (to remux mkv to MP4) when you right click on the file? 4:02

  6. I've been recording system audio and multiple channels from my audio interface into discrete tracks for years for my podcast. I started out using Jack on Linux with the jack_capture application and for the last two years have just used Ardour. Similar results can be had with Voicemeeter, but I personally think the JACK Audio Connection Kit and Ubuntu or Arch Linux is a far better solution.

  7. Reaper (DAW) handles video. I wonder if you could open up that multitrack MKV file in reaper and slice it up there. It exports video as well.

  8. Dude – I've been struggling to record both sides of a skype call with decent audio quality for ages, spent hours today going through different tutorial videos and articles with nothing working – This video finally gave me an acceptable solution. THANK YOU!!

    I had one issue and the solution I found might actually be a bit easier than your Premiere work around: The MKV file wasn't playing any audio when I opened it in media player (I don't know if that's normal, but whatever) and I found if I just dropped it in Reaper the audio would transfer fine and I can save it as an mp3 or wav or whatever suits from there – easy! Hopefully at least one person can use that tip and I can consider the favor paid forward.

  9. In the latest version there is no way to select the audio encoder that I can tell and also mkv isnt in ffmpeg 🙁

Leave a Reply

Your email address will not be published. Required fields are marked *