Noise Removal is on a list to be changed to “Noise Reduction,” because that’s much more what it does. You may have just overused it. It’s purpose is not to make your performance perfect. It can’t do that, but it can help with a single, annoying noise like fan or air conditioning. It can’t help much with “white” or “rain in the trees” noise.
When you take a profile of a noise, the tool generates filters to remove that particular noise from the show. The problem with white noise is the tool generates filters to remove everything. The only way a tool like that can work is to rely on another tool that tries to figure out where your voice is and do no noise reduction during your words.
This almost always leaves tails or other artifacts on the voice. You are urged strongly not to need noise reduction at all since it’s always a crutch and sometimes not a very good one.
The Smoothing value is the setting for how much voice makes it through without filtering. You can also use really tiny values of Noise Reduction (first value) and increase slowly until the trade-off between background noise reduction and increasing tinkly compression effects is optimal.
Or better, not use the tool at all. Get as close to the microphone and loud as you can without creating mouth distortion effects or overload. This hiss noise is almost certainly the noise generated by the microphone itself. It’s always there and it’s your job to make your show louder than the noise.
One of the disadvantages of an all-in-one USB microphone is you can’t change the preamp (and the preamp noise) like you can with pure microphones.
I’m not listening on a good sound system (it being 3am Pacific), but it’s possible you’re listening to your computer fan noise. The cure there is obvious.
If you listen to Test 2 (Without noise removal) how bad does the background noise sound to you?
The problem is the podcasts I’m creating I want to eventually sell, so I want it to sound as good as possible. I believe the Samson CO1U is quite a good mic, so surely it’s possible to get it better?
Worse than it could do given the equipment that you are using.
There is low frequency noise that can be easily reduced using the High Pass filter effect. Try setting it to a frequency of about 60 to 80 Hz and various “dB per Octave” settings to see which sounds best to you. Ideally you are aiming to reduce the “rumble” without affecting your voice sound.
There is quite a strong peak at 120 Hz, which can be removed easily with the Notch filter effect. Try it at 120 Hz, q = 20.
but probably the most significant improvement would be from getting closer to the microphone. You will need to use a “pop shield” so as to avoid “wind blast” effects from your breath on the microphone.
If you don’t have a pop shield, they can be purchased from about $20, or you can make one for next to nothing (examples on Google).
The strong presence of 120 Hz could be your microphone listening to your compact fluorescent lamps or other AC powered devices in the area. You should cure this at the beginning because removing 120Hz with a notch filter inside Audacity could affect the low pitch quality of your voice.
Try turning off all your lights and/or recording in bright sunlight once and see if the 120Hz Peak goes away. You may still be able to get tungsten lights for use while recording. Don’t dim them. That makes noise, too.
The rest of it is pure microphone hiss at about -45dB which is about right for this microphone. There’s no shortage of complaints of low volume and high hiss level with this mic in particular and the whole range of USB mics in general. The instant you remove the audio operator and mixer from the sound channel, you have to make compromises and one of the compromises is to lower the performance volume. Nobody can recover from overload/clipping (too loud), so most of these mics tend to be restrained in their vocal performance. The usual complaint is of playing a guitar and not getting a good recording because of noise.
Run the Noise Removal but don’t push it. Use the tool to reduce the noise 9dB to 12dB instead of trying to make a velvety “black” background. Mess with Smoothing until the voice sounds reasonable. That should pretty much leave the voice alone and give you a presentable product.
That’s assuming just getting closer with a pop and blast filter doesn’t work.
You are warned that post production filtering and rescue are going to be a stabbing pain in the neck if you get rolling on this. It doubles the production time on each show and increases the likelihood that you are going to post a noisy show by accident – or run out of time to post a show and miss a deadline.
This is your show with 9dB noise reduction and 300 smoothing. I used a serious noise chunk of the beginning of your clip as the profile. There is a snippet at the beginning and end with no removal to show the contrast.
How big is the room you’re recording in? This microphone has little or no response for sound arriving at the sides – it’s a supercardioid – but it does respond from the rear. You might want to put a quilt or blanket right behind the microphone to help suppress room echoes and tighten up your voice.
You might also want to put a quilt on the desk to eliminate vocal slap. Forget the rest of the equipment and pay attention to the doubled-over moving blanket on the desk. This was a radio broadcast that I shot.
Did you play that clip I posted? That’s your raw sound for the first second, then the corrections I made, then the bad sound again. If the sound in the middle was OK for you, then I know how to get that. If it’s still too noisy, then this may be painful to fix.
Noise removal works in two steps. The first step is the Profile step where you drag-select some bad audio to tell the tool the kind of sound that you don’t want – the noise. Then you run the tool again (it remembers the noise) and remove the noise according to the settings on the control panel.