So, I want to record a stream with a microphone. I wanna know how can I improve voice quality considering it's a live record, meaning there's no after-edit.
I've been using Audacity for my Youtube videos, but I don't think this program can be used to live editing.
Please help.
There is fairly little you can do software-wise to improve the sound quality of your input "on the fly", especially if all you're using is a headset or a standalone cheap mic. Studio mics are more suitable for higher quality audio without any need for additional post processing, but even then you can't expect much if you're using nothing but intergrated audio in your PC.
An audio interface is your best bet if you've got the money to invest into your videos, as they can provide you with the neccessary software tweaks such as noise gates and a wide variety of effects. Also, it's always good to have a physical pop filter on your mic, at least those can be found pretty cheap.
If money is a problem, the only thing you can do is try and tweak around your audio settings (in Windows). But in that case I'd recommend just recording your audio uncompressed (yes, you can do that in Audacity, Ocenaudio, Sound Forge etc.) and then make the neccessary post processing fixes. It's a bit of a longer process but it's most certainly the cheapest solution.
I have one more question. Bandicam, which I use, provides an option to seperate audio and video track - very useful, because I can then edit voice in Audacity. However, there's a problem, because I cannot connect those tracks back since the audio quality changes regardless of the program I try to use (Windows Movie Maker, Camtasia). I thought it's a case of frequency, which value should I use for that? (Eventually, which software can I use to combine tracks that won't drop the audio quality?)
Further explanation: Audio track edited in Audacity plays well in the program or as a music file. Sadly, after many attempts I;m not able to keep good quality when combining video and audio track in Camtasia or WMM. I know that Camtasia itself has some problems in terms of frequency of the audio files, but Internet doesn't have a solution.
First of all, try using OBS or Xsplit to record - I'm pretty sure those programs can do the same thing as Bandicam.
As for audio, it's not all about frequency - for YouTube purposes it'll be fine if you use 44100 Hz at 16bit depth. Any audio editing program can convert these values, so as long as those values are the same between the audio clips, as well as the format in which the audio is recorded, the actual "quality" part i.e. the bitrate of the audio doesn't matter. As for cutting and merging audio streams in post processing, I prefer Sound Forge, but it's not free - Ocenaudio or Audacity should be just fine as long as you keep your audio uncompressed.
And here's another hint - if you already separate audio and video streams, make changes to both and compress them, then use a muxer to merge the two streams instead of compressing them further.
I know this sounds a bit confusing, so let me make it simpler:
1) Always record both audio and video uncompressed (high bitrate MP4 for video saves you additional rendering and compressing if you want HD on your channel - WAV is a great format for audio) with Xsplit or OBS
2) Edit your video the way you always do, save it without the audio stream in MP4 (I assume you use the H264 codec)
3) Edit your audio and save it as MP4 (alternatively, as either AAC or M4A; for YouTube 192kbps is fine)
4) Use a muxer (I recommend MKVToolnix, works with MP4s the same as MKVs) to "mux" (merge) streams together and you're done
Quote