Please review my first audio book recording

Hi all!

I’ve been in podcast production for the past two years and just now getting into audiobook production. So I know my way around Audition and now starting to learn to use audacity, however, I’m not a professional sounds engineer and primarily self-taught, which is why I would appreciate some insight and tricks of the trade from everyone here on this current project.

I have two questions:

  1. When recording, even though it sounds very loud thru my headset, the actual recorded volume is very low. We use a Beringer XR18 in the studio and I dialed the mic way up. I have to boost the sound in audition/audacity by 10db to reach the RMS standard with ACX analysis. Is there anything I can do with particular settings to create a better recording? I’m attaching a sample containing both original recording and final result so you can see what I mean.

  1. I send my sample thru ACX audio lab a few times. Of course the RMS started out way too low. I kept boosting it by a little each time and eventually got the “no issue found” report, hence the final sample. But the trouble is I get this result when I ran the analysis in Audacity:

Peak level: -3.20 dB Pass

RMS level: -25.06 dB Fail (too quiet - RMS must be in range -23 to -18 dB)

Noise floor: -70.44 dB Pass

How do I go about resolving this? Any instruction is much appreciated!

I have a couple of suggestions at the reading step, but I applied some Audacity tools and got it to pass ACX Specifications.

First I clipped off the last five seconds of the work leaving everything the same volume.

Then I applied Audiobook Mastering. I cheated and did it with the macro in one step rather than all three steps one after the other. Both work. Then I applied Audacity ACX-Check and found the background noise is too loud (I’ll get back to this later).

I applied Noise Reduction of the Beast (6, 6, 6) and got this illustration. That sample sound clip should pass both Audacity ACX-Check and ACX Audiolab.

Screen Shot 2022-11-09 at 3.50.57 PM.png

That performance is too noisy. It should be possible to hit the noise specification without the extra Noise Removal step. But even with Noise Removal, the noise is still right on the edge. I use -65dB as loudest permissible noise (quieter) rather than -60dB which is the actual specification. I do that because your work is going to go through additional processing and manipulation at ACX and if you start with -60dB, it’s easy to hit -59dB by accident and the quality control bells are going to ring and reject you.

Back to the studio. Which microphone, how far away are you? Can you get Audacity blue waves at least half-way up on occasion, roughly like this?

0.5 or half-way up the blue waves and -6dB up in the bouncing sound meters is the same number.


Rather than use Noise-Reduction, (which will damage the voice to some degree),
I used the free version of Couture to push down the noise floor,
(~9dB is about as much as you can get away with without people noticing an effect has been applied).
Couture also reduces the room-reverb a little bit, see … https ://

We may not need any of this if we can get more moxie (technical term) into the voice recording.

You can do amazing things with oblique positioning (B) instead of trying to announce head-on (A). A firm, louder voice with little or no P-Popping or breath noises. Most wind noises go straight in front of your lips.

Home microphones tend toward weak, quiet, gentle voices because of corporate marketing and publicity, not production value. Quiet microphones are less likely to get returned.


I suspect that the XR18 may be overkill for your project. While it can be extremely flexible in its capabilities, the downside is it extremely complicated and confusing to operate. Your question may be better directed at the XR18 specialists (try their forum).

Having said that, run the X-Air Windows controller program for the XR18 and establish a connection. Then click on In/Out (next to your Setup icon), then select USB Sends. You may find you are able to choose output from any of “Analog”, “Input”, “Pre-EQ”, “Post-EQ”, “Pre-Fader”, and “Post-Fader”, and various other selected levels to each USB channel. Perhaps you have selected “Analog”. You may need to change this to “Input” or to “Post Fader”. You may also choose to output the Main L or R to USB port 1, for example.

I’m going to take some time to properly study your comments but I want to first thank you for your time and insight!

It worked!

I used the ACX analytics tool and it passed that as well. However, I did have to use the “Insane” setting to export otherwise ACX tells me it’s bit rate is too low.

Does this mean I can proceed to recording the next chapter? What are some the potential risks for it to be rejected by ACX? Please give it a listen and I’m here to learn!

I will definitely reposition the mic in the future, but unfortunately the space I’m renting to use uses XR-18 and there is no alternative options. I’m going to look up specific instructions on their forum.

Thank you soooo much for your help!
Screen Shot 2022-11-14 at 11.20.18 AM.png
Screen Shot 2022-11-14 at 11.15.26 AM.png
Screen Shot 2022-11-14 at 11.05.35 AM.png

I suspect that the XR18 may be overkill for your project.

I suspect he’s right.

It’s critically important that you control and monitor your voice volume at the beginning of the chain—at the microphone electronics in the mixer—and not wait until you have a problem later with the digital sound. I don’t see a sound meter anywhere on this mixer. You’re supposed to see that your voice level is too low before the computer gets to it.

The product information says the mixer has a convenient integrated Wifi module for wireless control. Are you supposed to adjust and monitor your voice level using a separate controller and sound meters? I’m lost.

It feels like this mixer is intended to be part of a much larger and complex sound system and not something to capture your voice for a book. Was this a recommendation or something you already had on hand?

What’s the microphone?


What are some the potential risks for it to be rejected by ACX?

Unless they changed it, they won’t accept a partial or test submission any more like when I did it. They will only accept the whole book. They offer ACX Labs which will tell you the same things that Audacity ACX Check does, except they don’t measure noise.

Nowhere does anybody evaluate theater. That takes a human. Do you read slow, stutter, or have mouth noises? Those failings will wait until they reject the whole book. What fun.

A note about submission format. You should be doing all your work in perfect quality WAV format. Only when you have a perfect performance do you produce the Edit Master WAV archive and only then the ACX MP3 submission file at 192 quality or better.

You can’t edit or change an MP3 file without the quality suffering. Never use MP3 in the middle of production or editing.

They recommend, but don’t insist on mono—single channel (one blue wave)—presentation. They will accept stereo (two channel) but then they insist you have to produce the whole book that way.

They insist everything matches including the beginning and end of the book. This is where first time readers get into trouble. You read the first chapter as a rank beginner and the last one as a seasoned professional. And then think about re-reading the first few chapters over again.


Another note. It’s Constant Rate MP3. That one lets you select the rate value.

Screen Shot 2022-11-14 at 9.10.04 AM.png

I will be in the studio tomorrow and I will take a closer look at the current setup. I rent the space and equipment from someone else so unfortunately there isn’t an option to change them. I will take a look at the microphone brand tomorrow as well. Thank you!

Yes, I am thinking in USB 1 - Channel 01, for example, you might want to change the red circle=Analog to a cyan circle = Post Fader. Or ouput Main L or R to USB 1. Good Luck. If It doesn’t look easy that is because it is not. :wink:


You mentioned that you did the initial mastering with a macro rather than the individual tools in sequence-

I’m a bright-green narrator. I have a lot of stage performance experience, and enough coding experience to get into trouble, but my mastering process is currently brutally inefficient and probably negatively impacting my quality; e.g. bouncing back and forth between amplification and limiting. I just read another post of yours in which you mentioned how new readers tend to over-edit, which I have certainly been guilty of.

I wonder if you would share your macro, or more likely, point me to the thread where you already have?

(For context, I just submitted my first completed audiobook and it is currently in it’s ACX approval stage, and I am doing the recording for my second audiobook, and I’d like to mess it up as little as possible in this learning curve.)

Thanks for all your excellent advice on here. I would have messaged you directly with this question, but as a brand-new account it would not let me.


You mentioned that you did the initial mastering with a macro rather than the individual tools in sequence-

There is a formal posting for Mastering and I can post the Macro here.

I will be in the studio tomorrow

If it seems like we have been operating cross-purposes through much of this, it’s because we have. You have been using a formal Sound Studio and have been accepting their work for further processing at home and then submission. When most forum posters refer to their “studio”, they’re talking about the spare bedroom in the back of the house.

So you don’t have to worry about the microphone, preamplifier, sound desk, mixing console, soundproofing, environment suppression, recorder, or any of that. That’s being taken care of by the Studio. You might keep that in mind if you decide to actually use your spare bedroom to record. It’s not as easy as it looks. I got supersonically lucky. The original owners of my house had a boy who played drums and the tiny third bedroom is soundproofed. I also have a garage jam-packed with boxes and junk which makes a terrific soundstudio.

This is the exhaustive, Encyclopedia Britannica version of mastering.

This is the TLDR version.

And if that’s too much work, I created a Macro which will do the whole thing in one swoop.

Audiobook-Mastering-Macro.txt (498 Bytes)
This is a picture of the programming. Computer programs don’t need to worry about the length of text lines, so the programmer can write them as long as they want.

If your screen is large enough, you can look and see the individual steps called out from when you were doing it by hand.

I’m a couple of versions behind, but I believe you can install it like this:

Audacity > Tools > Macros > Import > Point to the macro file > Open.

It should appear in the macro list if you have one.

Select the audio to be mastered. Tools > Apply Macro > Audiobook-Mastering-Macro. There is no OK. It just does it. It is a little strange watching all the steps blasting by.

We should note that the Mastering Macro guarantees Peak and RMS (Loudness). You can throw trash in there and it will come out passing ACX inspection. It may not pass noise, however. Noise kills most home performers. You don’t have to worry about that because that’s up to your Studio.

Another note. The Macro has no intelligence at all and doesn’t use compression. So if you spend the first half of a chapter whispering and the last half shouting, you will get a theatrical disaster that will technically pass ACX-Check.

It will probably not pass ACX Theatrical Quality Control. Some human is going to notice that your voice volume in that chapter is wacko.


I would have messaged you directly with this question

That’s not recommended.

It’s best to keep all the questions and answers in the open forum so the most readers possible can benefit from the conversation.

Also, Do Not go back and change an old message or posting. Nobody will see it and it may result in several different conversations in the same message thread.


That is incredible information and thank you so much for helping us rookies onto the right path.

The studio I rent uses audio technica microphones, and although they’ve done quite a bit of sound proofing, the building they use has very noisy pipes and AC. I turn off the AC when I’m in there but I feel like i can still hear some humming in the backgroupd once I apply loudness normalizer. Somehow it passes the ACX check, but I’m concerned that something will cause it to be rejected by the human inspector.

I beg to differ: it does not mention dynamic-range-expansion,
e.g. expanders like Couture ([u]free[/u] edition) can squash down the noise-floor when there is no speech,
and can even reduce room-reverb a little

Don’t overdose on Couture, it will then just be a conspicuous noise-gate. Typically -6db, -9db at most.
(-6db could be the difference between passing & failing ACX )