Meeting ACX Check makes my audio sound overdriven

I record and correct.
I follow the steps of:
Noise Reduction
Filter Curve
RMS Normalize (introduces lots of red peaks)
Limiter (clears the peaks)
ACX Check. (It passes)

But the audio sounds blown out.
I need help.
Thanks in advance -Frank

Here is a raw audio sample

Here is the ACX Check - Passed
Noise Reduction, Filter Curve, RMS, Limiter sample


Did you use the recommended ACX Mastering procedure documented on this forum? I ask because performing that procedure on your original sample produces a result which, for me, doesn’t pass ACX maximum noise level requirement. Your processed result does pass the ACX check though.

To me the processed results, both yours and mine, have peaks nowhere near 0dB and they sound OK to me.


You have a vocal oddity. Your voice is profoundly non-symmetrical. The “up” parts of your blue waves are double the volume of your “down” waves. That’s not fatal or anything; I know announcers with that kind of voice. But that can cause some interesting affects in the mastering tools. RMS Normalize doesn’t care a fig where the waves are. It corrects for overall blue wave energy (that’s what RMS means) and then goes home.

Limiter cares a lot where the waves are and it goes through smashing and slashing (gently) anything that will fail ACX Peak.

But before we indict your father’s genes, I noticed something else. Your presentation is way noisier than it should be. The noise seems to be well-behaved, normal background noise (spring rain in the trees “fffffffff”), just way too much of it. Your Raw sample is close to perfect voice volume, so we have even more mysteries than is apparent.

Which microphone are you using? There is a condenser microphone on the market which claims to be able to plug straight into your computer soundcard with a simple adapter cable (included). That’s according to Promotion and Publicity. Meanwhile, Engineering is in the background screaming about plugging it into a proper 48v phantom-power microphone preamplifier for good sound. I think it says that in the instructions somewhere.

I think what you got is what happens if you use the simple adapter cable. Noisy, distorted sound. It does work as long as you don’t try to do anything serious with the sound. Most podcast producers probably wouldn’t notice.


I think you’ve used the limiter on “hard-clip” setting, when it should be on “soft limit”.
Hard-clip looks & sounds like overdrive: flat tops …

Using the tools wrong is possible, too. This is from the Mastering instructions. Note the limiter settings.

Screen Shot 2020-01-14 at 1.26.20.png
Still, you should not need stiff Noise Reduction to pass. There is still a mystery.


You leave my Father’s genetics out of this :wink:
The microphone is a Zingyou ZY-007 connected to a Focusrite Solo 3rd Gen with 48v phantom power.
I have the microphone gain set around 70% and monitor set to about 40%
I expanded the Audacity Meter as wide as possible to watch for it jumping to the yellow and red.
I am using a soft limit and those exact settings from your instructions. (THANKS!!!)

Because of the high level of background noise, I can’t use the 6,6,6 Noise Reduction without failing Noise on the ACX Check.

The noise doesn’t seem to be coming from the lighting, computer, or microphone & USB feedback, but it might be the furnace circulator, midwinter and it seems to be running all the time.

I tried again and moved a lot closer to the microphone and pop filter to get the meters hitting yellow most of the time.
Your comment about the -60db floor meaning that my voice needed to be MUCH louder than the background noise made me think I was just too quiet to start with.
The last attempt resulted in much wider blues and less red after RMS normalize. Sooooo, I think you solved the first part.
Now I need to figure out “quiet space”.
Thanks -Frank

I can’t use the 6,6,6 Noise Reduction without failing Noise on the ACX Check.

Remembering that the goal is to produce an acceptable recording with no noise reduction.

Ding! This is the magic microphone.

The package supplies an outboard “sound card” and adapter cable with soundcard plug on one end and XLR Female on the other.

Screen Shot 2020-01-17 at 14.56.22.png
Right out of the packing, they expect the microphone to work on the 5v coming from the computer.

I think the ffffff noise is coming from the microphone and is normal for this unit. I can see a little wall power interference in the noise (60hz, 120Hz, etc), but some of that is going to die in Audiobook Mastering. That’s one of the things “Low Rolloff” does.

That’s one of the Amazon packages for $34 USD. By my guess, the boom mount and accessories in that picture cost more than $34, so the microphone was free.

I think you got the best performance this microphone can deliver.


They’ve got the cheek to call it “Professional … Studio … Mic”. For USD$35 I don’t think so.

Before putting the “professional” mic in the circular file, check it is actually getting the 48 Volts …

Is the “48V” button on ?.

Yes I have the 48v turned on.
I am hoping the problem is room noise I can dampen.

Btw what causes the unbalanced voice curves?

I am hoping the problem is room noise I can dampen.

It’s possible, but unlikely. Room noises tend to have a signature or special “musical” tones. Air conditioners have some tones from the power system (60Hz in the US) and many computer fan noises have similarities. Yours doesn’t. It’s just spring-rain-in-the-trees shshshshshshshsh. That’s electronic noise and can be generated by not spending very much money on the electronics.

I’m not making that up. If you look at most sound mixer advertisements, they almost all promote their microphone preamplifiers. “Has Signal Spritzer Mix Preamps just like our Studio Consoles!!” This is why. Microphone electronics are hard to make.

I have the 48v turned on.

Do Not select Air and turn gain 2 all the way down.

what causes the unbalanced voice curves?

It’s possible your voice is naturally doing that. Mine does that a little, but it’s much more likely the free microphone. Can you record on anything else as a test like your laptop built-in? Nobody will give the built-in studio quality awards, but I bet the blue waves are better.


The constant rain-like hiss is generated by electronics, it’s not noise in the room.
All electronics generate it to some degree, but in your case it’s conspicuously loud.

https ://www

I am going to try some more microphones.
Mac Built-In Microphone,
Bose QC-15 Microphone,
and hoping to borrow a musician friend’s microphone.

I’ll load some “Catskill Farms” audio with each setup.

There’s a YouTube channel where voice-over microphones are compared,
( but I think the cheapest one is USD$50, & that’s 2nd-hand ).

My friend only had a USB microphone for his music recording. So I have ordered two new mics, MXL Mics 770 and the Audio-Technica AT2020, so I can test them in my basement “studio”.

MXL770 Microphone Audio Test

Wow, KOZ is a master. (The rest of you are also most amazing and fantastic, too)

I can pass the ACXCheck with only Floor, RMS, and Limit using the new mic.

With the clothes washing machine and dryer going I can still pass if I run Noise Reduction @ 6,6,6
For an inexpensive microphone, this works surprisingly well.

I’m nitpicking:
There’s a bassy resonance on “farm’s rich”, & a high-pitched twang on “Green”.
They could be caused by being too close to the mic, or the way the mic is mounted.

I listened and couldn’t hear the bassy & twang. I think you might be hearing the freakshow that is my natural voice :wink:

Not unless you’re the tin-man from the Wizard of Oz.
The bassy resonance is like something rigid & hollow the size of a small tin-can,
the twang is a bit like someone plucking the ring-pull on a can.

My bet is the body of the mic itself is resonating, and the twang is shock-mount or other nearby springy thing.

If the mic is connected to a table it could be picking up extraneous noises from that via conduction.

That’s after mastering, right?

the freakshow that is my natural voice

The Audiobook Mastering Suite is only going to get you past the ACX basic mechanical specifications. Are you loud enough, etc. We can’t test for theater. Are you pleasant to listen to? Do you make my ears hurt? That test happens at the ACX second pass Human Quality Control where an actual person listens to the work. This is where tongue ticks, gasping, and wet mouth noises live (if you have them).

Oddly, most tests don’t care if you have a regional twang or national tilt to your voice. I can listen to a story in “Texas” or “Irish” all day long as long as it isn’t so deep or fast the words get lost. There’s a YouTube show where a woman does her whole comedy standup in “southern.”