How do I filter out sounds over certain amplitude/loudness?

Audacity 3 Help Forum Windows
1

Hello everyone. I’m a complete newbie to all things related to sound recording, other than having a very basic understanding of bitrates, sample rates, and basic acoustic sound principles. Other than that, I do not have a simple understanding of the digital audio world, so please forgive me and bear with me.

When I record sound through my phone, I set it on the table, and leave it on recording mode. But when someone sets a particularly heavy book on the table by just dropping it on the table, or when I accidentally drop the phone onto the table (even from just half an inch from the table), it creates this really painful noise when listening to playback–especially if my subject sounds very quiet in the recording and I max out the volume just to hear what he/she is saying. Is it possible–and can someone provide me with basic instructions on how–to basically use Audacity to create a certain “amplitude ceiling/floor” so that those really loud sounds get their amplitudes cut in half or more? Or if that’s not possible, or simply doesn’t work, is there any way I can use some sort of Graphical User Interface to completely remove those spikes from the waveform by just circling them and then pressing a delete/erase button? I don’t know how negatively that will affect the resulting audio if I do that, but I think it would be cool if it were possible.

I have uploaded a screenshot of one of my sound files in Audacity–and have circled the spikes/needles of loud sounds that pierce my ears whenever I listen to them–just to try to give you a stronger understanding of my situation.

None of the sound recordings I have are of songs, they are of a person speaking normally. The sound files I recorded are in WAVE format, 44100 Hz sample rate, 16-bit PCM (in case this information is necessary for the above). Incidentally, can I have some opinions on what the best compression format is for anyone? I’ve merely heard some opinions on the internet that OGG vorbis is better for lower bitrates and mp3 or some other format (AAC?) is better for higher bitrates. Then, FLAC for lossless, which I think I might need a little better understanding of to compress my files into and discard the original WAVE formats. Could anyone find it in their hearts to enlighten me a little? I know this is vague, but I prefer my recordings to be (at least) “clear” while being as small in size as possible. I am accustomed to hearing MP3’s at 320 kbps, and I regard them as acceptable in terms of clarity (or “real-life-likeness” or “I feel like I am actually listening to the singer sing into both my ears, not a reproduction of it”), but maybe there’s something even better out there.

I appreciate any kind of help, instructions, and opinions on compression formats. Thank you.
Audacity Audio Spikes Screenshot 2nd ver.JPG

2

There’s a plug-in effect called “PopMute” avaiable here: https://forum.audacityteam.org/t/new-plug-in-effect-popmute/17499/1
(be sure to click on the link that says “Current Version” rather than either of the obsolete versions).

Instructions for how to install: http://wiki.audacityteam.org/wiki/Download_Nyquist_Plug-ins#Installing_Plug-ins

There are “help” instructions included in the effect, but if you need further help please feel free to ask.

3

Hi steve, thank you for your reply. I tried out pop mute the night you posted this effect, but I couldn’t figure it out. I looked at the view help box several times, and to be honest, I couldn’t really understand it… I tried your default threshold and mute level of -6.0 and -24, but they didn’t really do anything. Then I tried (if you look at the screenshot, you might understand where I’m coming from) a threshold of -.02 and trying a bigger mute level, but it didn’t do anything noticeable, either. I mean, I listened to my recording after I applied the popmute effect, but it sounded pretty much like my recording before I applied popmute, and the waveform didn’t look different, either. How do I determine what threshold level I want popmute to use, and how do I determine what the mute level to be? Am I correct in thinking that the mute level is how much I want every sound above the threshold level to be muted/chopped off by, in terms of dB?

4

If you could post a short audio sample (preferably in WAV format), I’ll have a look and see if there are settings that will work for this job.
The audio sample only needs to be a couple of seconds - just as long as it illustrates the problem that you are trying to fix.
See here for how to post an audio file: https://forum.audacityteam.org/t/how-to-attach-files-to-forum-posts/24026/1

5

Thank you steve. I’ve uploaded 4 seconds of the audio recording of the screenshot I uploaded here a few days ago. At first, it might not sound like there’s any voice at all, but if you crank up the volume, you’ll hear that it is there, but just very quiet. But, I wouldn’t do that if I were you, because 2 seconds into the clip, there’s a loud “TICK” that might hurt your ears or equipment…In addition to that TICK noise, there are loud coughs and chairs scraping the ground that I want to suppress that I’ll try to upload, too.

6

That example is so extreme that I would probably run Pop Mute twice.

The important thing to remember about Pop Mute is that affects audio that is above the set threshold.
You will notice that the threshold setting is in “dB” and has the range of -24 to 0.
“0 dB” is when the waveform is the full height of the audio track. -24 dB is a pretty quiet level.

There is a little “trick” that allows you to accurately find out the level (amplitude) of the waveform, or part of the waveform. Let’s take that audio clip that you posted as an example:

Select the entire audio clip.
Edit menu > Select the Amplify effect.
Note that the Amplify effect shows “Amplify by 6.7 dB” and the “New Peak Amplitude” will then be 0.0 dB. This tells us that the current amplitude must be -6.7 dB (minus 6.7), because we know that amplifying by +6.7 will bring the level up to 0dB.
Click OK to amplify up to 0 dB.

We can clearly see a huge “click” at about 2 seconds. The audio just before the click looks like it is about the next loudest, so what we will do is to use that find the “Threshold” settings. We will set the Threshold a little above the level of that part so that we only affect sound that is above that level (the click).

So, to find the level of the audio just before the click, select it, then call up the Amplify effect, but do not apply it.
findlevel.png
“Amplify” is offering to amplify by 28.2 dB to bring the level of the selection up to 0dB. This tells us that the peak level of the selection is -28.4 dB (minus 28.4).
We do not want apply the Amplify effect now, so press the Cancel button.

So we want to set the Pop Mute “Threshold” above -28.4 dB.

Because we are dealing with negative numbers, “above” means that we want a “less negative” number, so -24 dB will be perfect - it’s “above” -28.4 but not too much above.

Select the entire audio clip and bring up the Pop Mute effect.
Set the “Threshold” to -24 dB (the level that we have just worked out to be a bit above the audio that we want to keep.

Now we need to decide how much to reduce the level of that click.
If we reduce it all the way to silence, then it will leave a noticeable “hole” in the sound, so we probably don’t want to reduce it that low.
We probably want to reduce it so that it is a bit lower than the rest of the audio.
Now we know that the audio we are keeping goes up to -28.2 dB, so we want to be “lower” (more negative) than that. Let’s try -40 dB as that is about 12 dB below our “-28.4” level. So, set the “Mute Level” to -40 dB.

The other two settings can normally be left at the defaults (10 milliseconds).

Click the “OK” button and the click will magically disappear.

Now apply the Amplify effect (default settings). This will make the voice much louder so that we can hear it properly now.

End of stage 1.

Stage 2 we will deal with the cough and chair scrape.
The method is pretty much the same, we just need to use different settings. Muting these noises too much will leave an obvious “hole” in the sound so we will not remove these noises, we will just make them a bit quieter.

First we pick out (by eye) the highest (loudest) part that we want to keep and “measure” it as before using the “Amplify” effect.
(note there is a click just after 1 second, so we want to avoid that when we make the measurement. I make the level about -7 dB, so we will set the Threshold a little above that, say -6 dB.

We want to avoid making a “hole” in the sound, so we will set the “Mute Level” just a bit below the threshold, say about -10 dB.

We can leave the “Lookahead” and “Release time” at the default settings, or because we are wanting to affect quite “long” sounds, we can tweak these settings to make the transition a bit more gradual - try setting these both to about 60 milliseconds.

That’s it.

Note that when you use this method on the whole track, the settings may be a bit different, but follow the procedure, get the “measurements” by using the “Amplify” effect, and remember that “less negative” is a “higher” level.

7

Thank you so much, steve! Your detailed instructions taught me more than just how to suppress painful noises and really helped make my recordings so much easier to listen to! Sorry it took me 4 days to get back to you. I was trying out pop mute, and I find that running it 3 or 4 times using like a -10 to -20 mute level helps to make my recordings not have gaping holes in them when I’ve got loud noises that have huge differences in amplitude. Thanks again!