Hello everyone, first time poster. I am starting to record audio for a YouTube project I want to start, and while I have had others record me, I've never particularly recorded or edited myself before. I'm not really striving to be a perfectionist, but I'd like something that at least sounds decent for my first attempt. For background, I have an Elgato Wave 3 with some some basic plugins for Noise Removal and EQ (though I won't claim to fully understand what I'm doing on the EQ side).
I recorded my first voiceover, and the thing that jumped out at me was, at least to me, how harsh my pronounciation of "S" came off as on the recording. Of course I am aware of the existence of the DeEsser, and immediately went to YouTube to try to understand how to troubleshoot and solve the issue. The video that to me had the clearest explanation was this one:
https://www.youtube.com/watch?v=hrSVTJmMpbw&t=252s
And so I attempted to apply this to my own case and found that...at least to me...nothing seemed to hit the right balance between actually making a discernable difference and not just hit the gain on EVERYTHING in the clip. I'm wondering at this point if the problem is actually in my own head/expectations, or if I'm actually doing something wrong.
Here is a link to my Audition project file/.wav file of the raw recording without any treatment:
https://drive.google.com/drive/folders/1l2IIy8Ws6oEw8X-rR4WWImz7uyuNr3Dq?usp=sharing
My first step was identifying an "S" sound using the Waveform and Spectral Frequency. The very first example is when I say the "S" in "Scrolling Down the Belt", very close to the beginning. I isolated the "S" sound using Spectral Frequency and the Marquee Selector Tool to where I could hear only it. And if I understood the video correctly, I was really looking for parts that showed up as orange on the heat map. While it maybe wasn't as obvious to me as it was on the example in the video, I think I managed to locate such a spot:
With that identified, I opened up Frequency Analysis, looped the clip, did a hold, and got what you also see in the screenshot above. If I hover over the highest red peak there, I get a value of around 3747 Hz.
With this, I opened up DeEsser and entered 3747 as my CenterFrequency, with a default value of 3179 Hz as my Bandwidth. Given that the spike in the last screenshot was somewhere between -35 and -40 dB, and I seemed to have a similar spike within my range on the waveform in the DeEsser, I figured I had the range about right. This is what my waveform looked like in DeEsser with just the highlighted clip playing:
From here, with the "S" clip still selected, I slide the Threshold from 0 dB downward until my Gain Reduction finally started having a significant value appear on it. It took me until about -33.4 dB before I got even a single line appearing on the Gain Reduction. -35.3 dB got me two lines, -37.6 dB got me three lines, etc.
So I decided to start with the value of -37.6 dB. I then de-selected my highlighted "S" clip and listened to the entire thing with the DeEsser turned on. From here is where I question my own sanity/expectations because...it seems either to be doing basically nothing different, or the Gain Reduction is firing all the time. And according to everything I've seen, you do NOT want the Gain Reduction firing all the time. So I'm at a bit of a loss at this point, and I would love for someone to tell me what they feel like a reasonable level is for this particular example and how they came to that decision if possible...just because at this point I feel like I'm never really going to land on it.
Thanks a ton for any help that anyone may be able to provide!
AdChoices