How do I get rid of peaks/spikes in an Audio File before using Match Loudness?

Explorer ,
Dec 29, 2020 Dec 29, 2020

Copy link to clipboard

Copied

I am just starting to use Audition, so I'm sorry if I am asking something that doesn't make sense. I did a 15 minute voice recording to upload on Youtbue and when I listen to it most of the volume is too low. So I found online that Match Loudness is one way to make your audio file volume go up. I want to do this but is there anyway to get rid of all these spikes/peaks on the audio file before I boost up the volume for the whole audio file so that the volume is more consistent throughout the recodring?

Here is a screenshot of how the original audio file looks:

audio-file-spikes-peaks.png

TOPICS
How to

Views

2.2K

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines

correct answers 1 Correct answer

Adobe Community Professional , Dec 29, 2020 Dec 29, 2020

Wrong tool - that's not really what Match Loudness is for. You certainly don't need it for a single file; it's really a comparison tool to get multiple files to a common loudness setting, and with material like yours (high peak to average ratio) it's not going to work very well anyway.

 

You are quite correct in wanting to get rid of those peaks, though, although because one or two of them have hit 0dB, I'd suggest that you might want to reduce your recording level by at least 5dB just so they d

...

Likes

Translate

Translate
Adobe Community Professional ,
Dec 29, 2020 Dec 29, 2020

Copy link to clipboard

Copied

Wrong tool - that's not really what Match Loudness is for. You certainly don't need it for a single file; it's really a comparison tool to get multiple files to a common loudness setting, and with material like yours (high peak to average ratio) it's not going to work very well anyway.

 

You are quite correct in wanting to get rid of those peaks, though, although because one or two of them have hit 0dB, I'd suggest that you might want to reduce your recording level by at least 5dB just so they don't overload initially. Audition has at least three tools you can use to achieve a more levelled result, and you might want to try more than one - the results can vary a little with the material - there's no 'one size fits all' to this. There is one bit of common ground though - all of the processes are based around peaks at 0dB, and you should normalize your file to this level before you start. In your case it isn't going to alter very much, but you should remember this anyway - especially if you record at a lower level to avoid 0dB initially.

 

The oldest tool (it's been there since the beginning of Audition's predecessor) is Effects>Amplitude and Comression>Dynamics Processing. At the bottom of the presets list there's one called Vocal Limiter. In its initial state it probably won't take enough of the peaks off, but because you can see visually what it's doing, it is easy to make adjustments to suit your specific requirements.

 

Also, there's the Hard Limiter. This could possibly achieve everything you need in one go. Once again, it's fairly obvious what is happening, and also, experimenting is required. If you set the Maximum Amplitude to about -1dB, then the Input Boost will set up the amount of limiting. At a guess, from looking at your waveform, you probably wouldn't need to set that to more than about 9dB, as it will get harder to listen to - it's very easy to overdo this.

 

At a pinch, you could use the Speech Volume Leveler, but you'd have to do quite a bit of experimenting to get a good sound out of it, although it's possible.

 

With all of these, though, you have to be aware that if you limit the peaks and bring up the average level, inevitably the noise floor is going to rise. Personally my approach to dealing with that is not to worry about it whilst getting the sound right, and then using Audition's Noise Reduction (the process version, with a noise sample) and reduce the background this way. This avoids the other technique, that involves downward expansion of your signal, which can make it sound weird by losing the sense that it's 'real'.

 

That should get you started, anyway.

 

 

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Explorer ,
Dec 29, 2020 Dec 29, 2020

Copy link to clipboard

Copied

LATEST

Wow Steve I really appreciate your response to my post. I certainly didnt expect such a thorough and detailed answer. It looks like you've given me a solid starting point for me to start work with this recording. Thank you so much for your time and knowledge. And Happy Holidays!

Likes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines