I use Audacity 3.5.1 to edit audio files. The main goal is to remove constant background noise (such as hissing, humming of air conditioners or computer fans) from the recorded files, so that the quality of the original sound is minimally damaged. Despite using the standard Audacity Noise Reduction tool, the results are always slightly artificial or accompanied by “watering/gargling sounds”. Steps I have taken so far:
1- Accurate profiling of the pure noise section.
2- Using the default settings of the Noise Reduction tool.
3- Gradually reducing the “Noise Reduction (dB)” and “Sensitivity” parameters.
Do any expert friends have experience or more advanced methods to achieve clearer results? For example:
Is it recommended to use external (VST) plugins like (Noise Reduction VST)?
Are there specific optimal settings for different noise profiles (like fan noise vs. hum noise) that I should follow?
Is it recommended to use a combination of tools like Equalization (EQ) to reduce noise frequencies before or after Noise Reduction?
Thank you in advance for your valuable advice and experiences.
If you’re using compressor - especially including “make up gain” to make the files louder - or any other tool to enhance the volume, you need to remove the noise FIRST. You may not be able to hear the noise but you can see it on the waveform. Then after compressor you will find more noise and will need to remove it again. Unless the files are terribly poor quality, this usually works good.
I edit my church’s sermon recordings to post to our website. We are in the “dark ages” and they still record them on an audio CD (and the entire sound system is still analog). Our volunteers are not professional audio engineers, so these recordings are not always the greatest. A couple of months ago we had one recording which was very low volume but had lots of noise. Using the Audacity noise reduction tools caused a lot of distortion in the main audio, so I tried Adobe podcast and it was awesome! The noise removal was wonderful and the speech sounded great! I’ve been using it ever since. It also does a great job with plosives and other microphone pops! Adobe’s fee plan currently allows files only 30 minutes long and I think there is a one hour per day limit.
Most of our sermons are a little longer than 30 minutes so I break them up into two files, run each file through Adobe podcast, then put the files back together. I still then edit the file in Audacity because I’m not satisfied with the loudness of the files and use the compressor function (including “make up gain”); however, I use less compression than I did on the non-Adobe-enhanced files. I’ve noticed the enhanced files then have lots of very high-pitched “s”es so I need de-s using a low-pass filter in a lot of places. Occasionally there is an “s” which I can’t hear and I have to amplify it. There are also still breaths to remove, but nothing like there were in the files that I did not run through Adobe podcast. I also take out a lot of loud hard consonants. So I’m not sure if using Adobe is saving me any time, but overall I think the quality is better because of the noise removal and the plosives and other microphone pops and background thumps.
