Noise Ceiling


I’m new to the Audacity forum and am having problems with the noise ceiling. This recording was done at home. Please analyze it, and tell me what is needed. I also recorded another part of the novel in the studio of a musician, which I’ll post next. I just edited the audio for time, and nothing else.



Hello again,

I had to cut-off a second of this audio because it was to long to attach. This was recorded an a musicians studio with the noise floor well below -60. However parts of the “noise” are above -60 according to the ACX. Is it possible that the check is considering part of my voice as being noise? I am surprised at these standards because I am sure that the studios at the two radio stations I worked at in Nashville would need additional sound proofing in order to pass ACX standards. However, I’m willing to do what it takes to conform to the standards because it’s my novel I’m reading, and my publisher wants me to record an audio book.

Many thanks for the help.


Yes, ACX check is not particularly “smart”. It is merely reporting the quietest contiguous 1/2 second section of audio it can find in the recording. In this test I don’t think there is a gap of that sort, I can hear you breathing through out the quiet bits. Hold your breath for 2-3 seconds (and don’t move either) to get a good test.

There is a fair amount of sub-sonic noise, and I can hear a hum in the background that I’m guessing is a motor of some sort (HVAC unit, fridge, etc).

In any case I was able to get you clip to pass ACX test with the following steps:

Normalize to -3.1db peak
Apply the “LF Rolloff for speech” equalization curve.

I did it with AudioBook Mastering 4 and the addition of a pitch specific filter tuned to your motor hum.

Plus Effect > Notch Filter: 120Hz Q3 > OK.

You’re close in the raw reading, so several different mastering processes should work.

Screen Shot 2018-08-01 at 10.49.07.png

When you choose a mastering technique, you need to keep with it for the whole book. ACX doesn’t like reading sound changing chapter to chapter.


Thank you all very much. You all have been a great deal of help. One more question. Which audio studio seems the best for ACX recording, Dave’s studio, or Robert’s studio? I need to choose between the two. I also understand being consistent with mic placement, ambience etc. is more important for book recording than broadcast, Thanks again.


If someone forced me to pick one, I would pick Robert. That sound clip has very slightly less noise than Dave’s clip. The difference is very slight and it’s going to be all up to you which one you like listening to. That sends you back to your high quality headphones or good quality speaker system for vocal analysis.

Both have that hum tone in there (mmmmmmm) and about the same volume. All things being equal, I would be getting rid of that first. It will make the presentation sound better without that extra notch filter and it will reduce the processing steps by one.

I boosted that super-quiet room tone two seconds at the beginning to illustrate the hum, but I couldn’t do it because you were making so much noise.

Gargl, smack, tick, tick, Gasp.

That two seconds is harder than everybody thinks.


OK Koz, many thanks. What I plan to do is search out the noise source and shut it off. Then next week I’ll sent you another sample sans breathing and mouth noises at the beginning. I’ll do it in my studio. I think it’s best to get all of the technical problems ironed out before I attempt to submit chapters of the book to ACX. Thanks again.


I shut off all electronics in the room except for the laptop, and placed woolen sound insulation over all of the windows. The jack on my good cans broke, so I couldn’t hear any hum if I had it. (Could you suggest a good pair of headphones?) The first few seconds are the room tone without me breathing. Then I took a breath and talked. I used an AT2020 USB mic and mastered it in Reaper, but didn’t process it. It was recorded in my home studio. Thanks much for the help.


I can make that pass with just the application of the “LF Rolloff for speech” EQ curve and normalizing to -3.1 db peak.

I don’t hear any hum or other annoying background noise, only a low rumble that might be the HVAC or it might just be the natural noise of the microphone.

If this setup is repeatable, I would say you are in good shape.

OK thanks, I’ll go with this setup. I appreciate the help very much, because I was starting to get really frustrated. I have worked in radio broadcasting on-and-off for decades, but have never been up against such stringent requirements. But, one lives and learns. Thanks again. :smiley:

have never been up against such stringent requirements.

by yourself.

You are the recording engineer and mastering producer, probably for the first time.

In radio broadcasting your work is followed by the transmitter processors up on the hill. You can be remarkably sloppy in the studio and the transmitter will even everything out for you.

I listen to a talk show over the air and I recorded it that way for a long time for application in hiking. I started to use the much better, clearer, posted podcast except volume variations made the work almost unlistenable. One performer mumbled in his beer and the other had a thermonuclear laugh.

I had to come up with a way to simulate transmitter processing. Nobody at the studio ever worried about it.