Advice/Review on improving sound quality please

Hi guys. Sorry if this is in the wrong section, but it seemed the nearest to what I was after.

I’m doing some voice over work with Reaper DAW but I thought you guys could still give advice.

Could you let me know how I can improve the attached audio please?

I realize that part of the problem might be with my voice, but any tweaks or settings that help are appreciated.
It sounds slightly crackly (might be my voice also) and a little processed (I’m playing with settings to try and fix this).

I’m using Reaper with the following:

NS1 from Waves (Background noise remover)
Volume Adjustment

I also then use RX Elements and run:

DeClip. (Running Declick as I noticed my jaw gives slight clicks on certain words naturally and this seemed to help a bit)

Microphone is Rode NT1A
Sound Treated room (Tested noise floor etc a while back on audacity and it passed)
Scarlett 2i2 Audio Interface.

Attached an MP3. The Wav file was too large (4.61 MB) Not sure how to send this.

Appreciate any help.

It’s like Brian Blessed & David Attenborough had a baby.

You’ve got that ~130Hz booth-resonance thing happening

The free version of TDR Nova can fix it …

Lol - David Blessed perhaps?

Thanks for the links!

In general we request that you post “clean” unprocessed work. If you post a voice that has been through processing, we could be fighting the processing rather than the voice.

If you need a “declipping” tool, its a safe bet you’re recording too loud or have a volume setting wrong somewhere.

The forum will support 10 seconds stereo or 20 seconds mono at 44100, 16-bit.

A good goal is to record so your voice will occasionally hit blue waves of about half way (.5) and the bouncing sound meter about -6dB to -9dB.

It’s sometimes rough to get a home microphone to hit those values because home microphones tend to record low volume which can make them noisy (fffffff). This is a marketing decisions, not an electronic or audio one.

If no matter what you do or which settings you have you just can’t get high enough volume, you might try oblique microphone positioning (B).

That will let you increase your voice volume without increasing P-Popping and other mouth noises. You do have to be more careful about head movement. Plug your headphones into the 2i2 and listen in real time. That will help prevent volume wandering during performances.

Can you get the 2i2 volume knob to turn green?

Never blow into a microphone, but can you talk loud enough and turn it up to make the 2i2 knob turn red? That’s a good test that tells you how much volume range you have left. A red knob is the upper limit of the 2i2. You should never have production voice with a red knob, but anything up to that is fair game.


The NT1A is a side-fire microphone. You should be speaking into the grill just above the gold dot (with the microphone right-side up) The microphone doesn’t care if it’s upside down.


Thanks. I’ll post a raw file when a get a chance.

My recording room is small - square shape minus one corner if that makes sense. (Jutting out) I can just fit in it- So that probably doesn’t help things.

The declick (Rx elements) has been a bit of a life saver, as I mentioned my jaw clicks on certain words. I hope it doesn’t ruin/spoil the audio in doing so.

I’ve tried angling the mic slightly as well. I’ll try and post a couple of raw variations with different angles etc.

I think my best bet will be moving to another room (larger rectangle shape).
Just gotta find the time to retreat it etc.

I’ll keep you updated.


Every room, no matter what size, has resonant frequencies, even a >$3K booth.

The paid-for version of “TDR Nova GE” allegedly can detect & dynamically attenuate resonant frequencies
[ Can try-before-you-buy the sexist “Gentleman Edition” ].

Okay, I will try that Nova (Got me thinking of the Short Circuit movie now) :grin:

The spectrum plot is not affected by headphones/speakers/hearing
Atten test spectrum before-after.gif

You could learn from actual Attenborough recordings what his spectrum is.

my jaw clicks on certain words. I

There can be some very serious problems with applying layers of corrections to make your voice presentable. Even if you do get it all to work, you can’t change anything. The instant you change the microphone, studio, positioning, etc, you have to go through and reset all the patches. You also have to remember to apply everything and in the right order each time.

It’s strongly recommended that you correct fluffs, mistakes, and word-o’s in real time—right then rather than going back next week sometimes and record corrections. Off-time corrections are difficult for voice professionals to do and a new user with tons of corrections can be super difficult.

To be clear, if you make a fluff, pause with the recorder running and go back and re-read the whole sentence again. In post production, cut out the parts that didn’t work.

Never work on the original recording. Keep raw recordings in archive while you’re editing a copy.

Do everything in WAV, not MP3.

I think that’s most of the hints.


Thanks guys. Good advice.

Wasn’t sure if you needed the Catskill Farms Moo Crew posted in a specific place, but here’s the raw file. Thanks!

I was expecting David Attenborough, but it’s Karl Pilkington.

Just a static dip in EQ at 130Hz* could suffice to reduce the booth resonance,
but the optimum solution is a dynamic dip at 130Hz* which responds to changes in volume.
This can be done with the free version of TDR Nova …
compressing 130Hz in TDR nova.gif
Even after attenuating the 130Hz* still need to EQ and de-ess
A free auto-EQ plug-in worth having is Focusrite balance,
It has a M(ale) Vocal target, (the default target is “universal”).

[ * there are actually two bassy resonant frequencies: 129Hz & 259Hz, the first is the worst]

130Hz dynamic dip.xml (8.88 KB)

I’ve had a go at (dynamically) reducing the 2nd resonance at 259Hz,
(using same TDR Nova method as for 129Hz).
I’ve also used the “Universal” target, rather than “Vocal M” on focusrite balancer Auto-EQ plug-in

Pilkington Cows, before-after #2.gif

Thanks very much for the help Mr
Gervais! :wink:

I may have over/under corrected somewhat: I’ve got a touch of glue-ear at the moment, (TMI).
There is a wet/dry dial on the right of TDR Nova where you can adjust the correction to taste.
Similarly there is an an intensity slider on Focusrite balancer, (which essentially the same type of control).