Low RMS levels

“I thought that cards were a frame-up
I never would try.”

The good news is I got the raw performance to pass reasonably easily. It even sounds human. So there’s the first two poles of performance success (the third one is client satisfaction).

Screen Shot 2018-02-16 at 16.55.54.png
Apparently, you did, too, but I did it in one fewer steps. I did one noise reduction at 9, 6, 6. There’s an audiobook directive that states use the fewest corrections possible. ACX has a failure called “Overprocessing.” You can’t get there by beating your voice with a stick.

Even better, the overall raw performance volume comes in at about -9dB which is at the lower end of -6 to -9 presenting recommendation. So just pump the microphone volume knob up a bit and you should be good to go.

We should address your studio noise. I can easily make your noise reading much better by carefully filtering out a 120Hz hum tone you have somewhere in the room.

The 120Hz tone (low organ pedal) is probably coming from your air conditioning or heating or possibly a refrigerator or freezer. If you can identify that and turn it off during performances, it will make it easier for the other tools and filters to do their job.

Effect > Equalizer should not be flatten. It should be Low Rolloff for Speech.
Screen Shot 2018-02-16 at 17.17.56.png
And it should go first, not last. It’s job is to eliminate room and building noises before the other tools have to deal with them. Your raw clip has significant thunder/earthquake noises in it that vanish after application of Low Rolloff for Speech

I don’t hear any echoes or recording in a bathroom noises, so that’s a major advance right there.

What did the wine-glass/cellphone sound turn out to be? I don’t hear any remnants of it.


It is possible to get rid of the 120Hz hum with another mastering step. That’s not ideal but it does work.

Apply this step first, before you do anything else.

Effect > Notch Filter: 120Hz, Q2.

Then Equalizer: Low Rolloff…
and the rest of the steps.

And you may be done.

One bad effect of applying too much Noise Reduction is it makes the voice slightly crisper and brings out your mouth noises. All these tools interact.


The ping was an alert…problem almost solved, removed alert tones for everything in preferences…hopefully that will work. Usually when that happens I just re-record and delete the dingy part.

I think that hum may be the pump for the fountain outside, I’ll run a check on that. It is none of the other things you mentioned, but I have 3 options, the electric piano has a hum even though it is off, ???, ditto with the flat screen, or my Mac has a very slight hum, though I have done all I can to mitigate that.

Going to re-edit my already recorded 27 chapters, ugh, but OCD demands I do so. Thanks so much again.

P.S. Notch filter was not the answer, changed my voice…but I couldn’t pick up thunderer earthquake noises…puzzling

I can show you how to measure these tones, but I’m on the wrong machine, in the wrong place.

Though your face is charming,
it’s the wrong face

Then you won’t get along with Low Rolloff for Speech, either. It changes some audible vocal tones. It’s usually OK because it solves a number of other problems.

Bottom line, if you sound OK (and you did to me), then you may be good to go. Final acceptance will be with the client.

You didn’t say how you got around the talking in a rain barrel effect. Remember, we’re collecting these to help other people. It’s a Forum, not a help desk.

Final note: Equalizer: Flatten doesn’t do anything. Leave that step out.


I only use the “flatten” in equalizer when I am dropping in a re-take and it is at a different volume than the rest of the recording, I just nudge it up or down a bit, and it works.

"I’'m singing in the rain, just singing in the rain, what a glorious feeling, I 'm recording again!)

I have no idea how I swam out of the rain barrel, somehow you threw me a life ring and I used it. In going through the first 27 chapters (!!!) and “re-mastering” (trying to only remove a click, hum or slurp, but also having to replace entire personality voices ((need next time to have better pre-record communication with the author, either that or quit giving her/him options they like better than the first take))…this particular book has been a vast learning experience for me, 5 hours of on file recording, 10 chapters to go, and what with my challenges, I believe I have been at it for at least 50 hours. I am working for free here.)

But I get to meet neat people like you…so it is worth it. Picking your brain for free, I almost feel guilty.

Do you work for hire?


Picking your brain for free, I almost feel guilty.

The longest posting on the forum was Ian in Hollywood (the neighborhood, not the metaphor). 39 chapters and over a year. He was trying to read for audiobooks in his apartment on La Brea. That’s it. That was the whole job.

Do you work for hire?

Just whatever I can get with a navy bandanna, black T, tight denim shorts and pink Vans on Santa Monica Blvd, although we try to wait until later in the year when evenings are warmer. Not everybody likes goose bumps.

If you’re interfacing directly with the client, then you won’t have all those nit-picky specifications to worry about, although we frequently say, if you can pass ACX, you’re generally good to go with anybody else.

somehow you threw me a life ring and I used it.

I wish I could remember which step that was. I know you claimed the one or two “usual suspects” I listed were not bad when you tested them.


From here it’s basic bookkeeping. As I’m sure you’re already finding out. Keeping track of the takes, corrections, versions etc. is not trivial. And I would stress very strongly be able to point to two different locations holding your original WAV performances and any other backups you might need.

“Oh, darn. My MacBook won’t start.”

There is one oddity I should warn you about. Keep filenames dead simple. Upper, lower, numbers, -dash- and underscore. Macs can deal with fancy punctuation, but you may get yourself in trouble if you send a sound file to someone who can’t. Worse yet, your file transfer service might start to lose things at random.

20180218-1915_TestVoiceOver.wav That’s today’s sound file from 7:15PM in acceptable format to everyone. Being obsessive, I don’t use spaces, either.

Never give a right-brain creative a list of options.

Just don’t. That never comes out well.


“Just whatever I can get with a navy bandanna, black T, tight denim shorts and pink Vans on Santa Monica Blvd, although we try to wait until later in the year when evenings are warmer. Not everybody likes goose bumps.”

Smiling, have a visual, …roller skates seem right.

I have written and deleted paragraphs…the world is NOT ready for me yet…or at least so far…


One martini, two martini, 3 (no!) 4 (glasses of wine), lamb chops, crisp salad, baked spud, dark chocolate, and press “send” into the night. Let’s hear it for imagined anonymity. Hah!

roller skates seem right.

Chicago 9s. Black, high-top.

Unfortunately, the mode is rollerblades.

My sole talent on rollerblades is failing to kill myself.

I have written and deleted paragraphs

I thought you had a client.


the paragraphs written and deleted were martini’d attempts at being witty

omg, of course, blades, age is showing

“Ha, Ha, Ha, Ha,
Stayin’ Alive, Stayin’ Alive.”


Did you know that the beat of that song is exactly (or so) the same you should hum to yourself when doing CPR? True fact, learned it getting certified.

Okay, back to an answer to your question for the forum. You asked how I got out of the wine glass/rain barrel.

I just finished editing/mastering (she said, understanding the utter foolishness of an inexperienced newbie even presuming to use those words), a chapter that had to be piece-mealed together from multiple recordings, (hard drive wiped clean, files lost) ergo, a variety of volumes, levels, and sounds. I believe I was successful because of the settings you gave me to use, particularly the RMS Normalized.

Each individual recording section passed the ACX tesst, but they were not cohesive when strung together. I took each disparate area, put it through the RMS Normalize (-19) (when the volumes were too low), followed by Limiter. and finally Noise Reduction (9, 6, 6). Made an excruciatingly long final review using High Pass Filter to soften the far too numerous p, b, and some t pops, and then, if necessary, expanded the sound waves to a point were I could isolate the worst culprits and take out one Hertz? Anyway, 21 minutes that pass even my OCD tests. Even getting better at re-recording a word or a sentence and sliding it into the original sound track almost unnoticeably.

Author happy, (low standards), Narrator happy, (standards getting very high), Tutor happy?

Author happy

Works for us.

passed the ACX tesst, but they were not cohesive when strung together

ACX Check is a technical standard. It is designed to help you pass the ACX “Robot,” generally the first inspection your submission hits when it arrives.

This process helps filter out people who submit technical trash. To their credit, ACX will usually try to suggest corrections along with the rejection.

But after that, the submission goes to Human Quality Control [Dum Dum Dummmmmmm]. There, the work has to be entertaining and theatrically correct. This is where unmatching chapters, sibilance city, echoes and cellphone voices go to die. HQC’s standard is listening to somebody telling a fascinating story in real life over cups of tea. P-Popping and other distortions need not apply.

You do have to pay attention. One poster got bounced when she submitted work with the wrong amount of silence in the chapters. Fortunately, that’s between you and the client.