I’m considering dipping my feet into video and making one. I’m imagining my “workflow” will be VCV Recorder -> Some video editor -> youtube.
Any suggestions of best practices? Esp which codecs to use through the pipeline, and output settings for you tube to get best audio quality.
Thanks!
I’ve been using Camtasia, because I had already bought it for doing product demos for the day job,
It is pretty slick, easy to add annotations, compile lots of video and audio into a single clip - even a direct push to Youtube and other platforms.
One HUGE downside. It can’t record system audio from ASIO drivers 
Looking on their site, people have been asking for this for years…
But, if you use VCV recorder, it is real easy to import the audio from that into the video you made.
My day job is influencing me to other video editors ;-). With your editor do you specify any audio encoding output settings, or is the built-in youtube preset good enough?
My day job is writing software to make people drink more beer, so… 
But for Camtasia, the output settings are incredibly configurable - for codex types, resolution (audio and video).
I use VCV Recorder for quick stuff that I don’t want to tweak. But, for Youtube, I use OBS and then Shotcut. My problem with VCV Rec is that it doesn’t format to YT’s desired resolution of 1920x1080. With Shotcut, I can tweak the gain and fade in/outs and it has a preset for Youtube.
what resolution does VCV rec output? Whatever the actual window size is?
yes. This is not a problem for YT, per se. But, it causes problems if you want to format the end tags.
What does that mean?
Those little boxes that YT can place at the end of your video to refer viewers to previous videos. An audience retention thing.
OBS or VCV Recorder, sync audio if needed and edit video in DaVinci Resolve, which is also a crazy good audio mixer that can run VST plugins. All of them are free. I use the Youtube export preset, that’s it.
And never clip the audio, because these video formats are cancer to audio and create horrible artifacts. Better to record 24 bit with headroom and use a final limiter in Resolve afterwards. I use Host FX with a Waves limiter a lot for things that don’t need editing. OBS is weird, it clips at -5db, which seems to be the 0 point.
Being on Windows, I use the Voicemeeter driver to be able to intercept the ASIO stream in OBS studio.
Recorder is good for video but lossless, so it results in a huge file I have to convert. And once in a rare while I stream performances on Twitch, so I have an OBS setup. OBS is nice because it encodes directly to MP4.
Given the scale of my patches, I generally always a Recorder instance off-screen capturing FLAC even if it doesn’t always capture video. It’s a good insurances in case the overhead of recording video causes audio glitches, since Record doesn’t operate in real-time. (+ in case my jam accidentally turns out to be a legitimately good song)
To edit the video, I use Shotcut for editing when the job is simple, DaVinci Resolve when it’s more complicated, and Blender the days I hate myself.
Sometimes I post excerpts of jams for Twitter friends and the full version on Youtube, so I use LosslessCut to trim MP4 to fit its 2:20 time limit, as it’s incredibly fast without any re-encode or quality degradation.
On those days I use ffmpeg to trim / split / join / resize / add subtitles / adjust the audio / whatever.
ffmpeg can do literally anything, but by the time you’ve figured out the syntax you might as well have edited the individual bits yourself in a hex editor.
My workflow:
The main problem I’ve had is with the audio artifacts introduced during final rendering . It’s not a problem I’ve managed to solve completely but setting the audio bitrate higher mitigates the problem.
Windows here, too. tx for the ideas.
Somewhere out there in the world is some software from a big company that internally shells out to ffmepeg for all audio mixing… Or so I’ve been told 
Obs for recording using gpu power, asio link pro for proper audio routing for my system (doesn’t need for you probably) then DaVinci for editing and export in MP4 or MOV and AAC for youtube. The thing I don’t understand yet - why Obs file size is 500mb but after DaVinci 2.5 gb? But I don’t care for sure because the whole chain is pretty handy and doesn’t require everyday settings check. Also Obs can do multichannel recording so I can easily record a voice and VCV patch simultaneously and then add some separate dynamics processing in DaVinci for the recorded voice.
It’s probably because it renders at higher quality than the source? You can convert an mp3 to WAV and make it bigger too. Oh well, at least it doesn’t degrade it.
maybe it transcodes from one lossless codec to another, and it’s favored native one isn’t very small? But could be anything, right? Video is so confu sing - I worked on video encoding for a year or two and man was that baffling!
as Russians say about a paper - youtube will endure anything (I mean gigabytes 
But the speed of upload is also important sometimes when you’re on the run. But hopefully Davinci process to youtube standard setting while OBS. Anyway there is no way in OBS to make nice sidechaining from a voice to video (or maybe yes already, there are a ton modules for obs laying around - still looking for zooming lens for example - seen one time from one streamer)
maybe. Youtube has some standards I believe. But it also transcodes to the same filesize when set in 1080p master. So who knows, I think this is something inside OBS too because I set it to GPU render and files are always smaller than at CPU render.