getting home audiobook recording up to snuff

Dueling posts. I need to stop and read your last post.

I’m still managing Sunday.

You can put a hard-back book between the microphone stand and the towels for added stability. If it’s floor vibrations, that should reduce them to zero. At work I got fancy and cut plywood panels and painted them black.

The final looks nothing like that, but that exact configuration would work if I plugged it in.

Oop. Here it is. That’s not a towel any more, that’s carefully folded black Duvetyne and the paint is high temperature black because it’s bulletproof.

I can’t listen right now.


I got the clip from your server.

Maybe you do need to see the Audacity sound meters while you’re recording. You punched “It’s Admirable” and the blue waves went almost all the way up to overload damage.

There’s ways to fix that, but the loudness imbalance is too far for the regular tools and the Hard Limiter tool that I would like to try doesn’t seem to do the right things and I’ve never used it before.

By the way, the extra long USB cable has given you Frying Mosquitoes in the background. Yours isn’t this bad, but turn up the volume after I stop talking.

Now turn up the playback volume on the beginning of the clip you posted.

So there’s an extra step in there to make sure that goes away.

I need to get the explanation on the hard limiter tool. I don’t think there’s any other way to tame this.


Unfortunately, there’s just no physical way I can have the laptop within sight; the fan is far too loud, even to set up right outside the stall without padding between us. I can try harder to control my levels, and I’m interested in experimenting with the Hard Limiter you mention.

I was afraid of the extra USB adding noise. I’ll check with the local Free Geek this afternoon and see if I can get an extra-long USB Type B cable like the Snowball uses. Here’s hoping the noise is being introduced by the USB extension cord, and not by the distance.

If you can find a USB extension that uses heavier gauge wire then it might be better, but in general it is the length that, or rather the length combined with the fact that the Snowball is getting it’s power from the USB bus that causes the issue. An externally powered USB hub installed at the microphone end of the extension might help as well.

There is something really odd about the rest of your background noise. If it is electronic in origin it doesn’t match up with any mechanisms that I’m familiar with. If you are in a large apartment building then maybe it’s the HVAC equipment up on the roof. Although I wouldn’t worry about it too much, once we filter out the low-frequency rumble part you should be able to easily get past the ACX requirement. It’s also different (somewhat better) in the “shower3” test, which also makes me think it is probably building HVAC equipment.

I’m interested in experimenting with the Hard Limiter you mention.

Audacity doesn’t do anything in real time. The limiters, compressors and processors are all post-production tools and depend on you getting a good basic recording at the performance stage.

Turns out I did run into a tool quality problem. I turned over a rock. We should be able to recover from this. What you did, while a little scary, has not created sound damage and is a normal part of recording.

So you’re a pioneer. Isn’t that cool?

No, I don’t think so, either. I need to go away for a bit.


I love being a pioneer! You guys are being so helpful, so I’m glad you’re getting some data out of it.

I checked, but apparently USB Type B cords only come in 4 feet or less (edit: I’m seeing 10- and 16-foot models on Amazon for $5, so I might grab one in a future order). I can try a powered USB hub if it’ll cut down on extra noise, but the only hub I own is kind of crap and will probably just make the problem worse. As for electronic noise, the only other thing I can think of, besides it being the noise of the mic itself, is the ceiling light, which I can go ahead and turn off, since I’ll be in my little recording cave.

Any other suggestions, or should I go ahead and re-set up the ‘booth’ and try another sample with these changes?

On the sound meters issue: What if Audacity had an option/plugin to beep or otherwise audibly indicate when levels go over a certain pre-set amount? I could wear headphones while in the ‘booth’ and re-take a line if I hear the beep. This would also help narrators/producers who are sight-impaired.

Real Time listening to yourself on headphones goes a long way to suppressing the urge to get irrationally theatrical. This is the theory behind “sidetone” on telephones. Remember telephones? A little of your own voice was intentionally fed back to yourself so you can have confirmation of the connection and an idea of how loud you were. You can’t easily do that in cellphones, so you get the constant “can you hear me OK?” Because there’s no way to know.

This also gives cellphones bad microphones. I heard an interview where they were interviewing someone on the street and the guest was using earbuds with the microphone in the lead. It was “communication” only in the loosest possible sense. The sound quality was garbage and very difficult to understand, but the guest, of course had no idea. All she saw was the insane convenience of not having to worry about a separate microphone. She was a very prominent person, so the show producer wasn’t about to complain about it. Everybody loses.

I think we may have cleared your sound problem, but I need to sit and pay attention to it a but more.

I wonder if we didn’t do the beeping overload thing already…


I’ve tried real-time listening in the past and couldn’t get past the ‘bluh bluh’ effect it has on my voice, where my brain is struggling to keep up with what I’m hearing. With practice I could probably master it, if it turns out to be that important.

I never knew that audible feedback on phones had a name – yay for having learned something today!

I’ll wait to hear from you on the sound, but it’s exciting to hear that we’re getting close.

couldn’t get past the ‘bluh bluh’ effect it has on my voice, where my brain is struggling to keep up with what I’m hearing.

That’s not actually the problem. If you’re experiencing the “finger on the record” slowdown while listening to yourself, you’re probably listening to the computer headphone connection which is late or delayed. That drives everybody crazy. Far better to put your headphones in the desk drawer and leave them there.

The Yeti microphone, for one example, has a headphone connection in its base. It doesn’t have that delay problem. A quick look at the musical overdubbing setup I published has the headphones plugged into the USB interface. That’s in crisp real-time, too.


I’m not recommending using Apple Earbuds to mix music. They were graphically convenient and I had them.

I’m getting groggy. Do we think Eden Shower 5 is the current test clip? I’m going to peel off the other ones.

Yes, #5 is the newest test clip I’ve made. A few hardware suggestions have been made since, so I have a few things to try – but if it will be easily possible to get #5 to pass guidelines, I’d love to get a look at that leg of the process.

There’s no rush, so don’t mess with it when you’re groggy! I have the day off tomorrow, so I’ll have plenty of time to go over any info you post. Thanks so much!

OK. That’s more better. I got Shower5 to pass with minimal pain. The problem is not that I can do it, it’s how do I tell you to do it. I used a custom application of a custom tool to tame your unruly phraseology. Worse, as I said before, I don’t have a touchy-feely with this tool yet.



Now I’m really curious about what ‘unruly phraseology’ is.

Anyway, what if you wrote a tutorial on how to install the tool (assuming it’s an Audacity plug-in) and what settings you used with it? Your other guides have been pretty clear.

When you emphasized those words in the story that caused them to exceed the range of our normal loudness tools.

Are you OK with shortform postings and settings? If I said your story was -71dB, -23.5dB, -3.2dB, would you get that those are noise, loudness, and peaks? Did you write down what the ACX limits were?

The new Effect > Noise Reduction tool lends itself to that kind abbreviation. There are three adjustments and a common run of settings is 6, 6, 6, the Noise Reduction of the Beast (you can’t make this stuff up).

Normalize: DC, 3.2. Those are the first two of the three settings in Effect > Normalize. etc.\

As we go. I need to see if my latest series of incantations works with different clips from you. If you hadn’t punched those words, I would never have known we needed to do this.


Speaking of, what’s your suggestion re: shotgun blasts? They’re written as ‘BLAM’, but I think I’ll sound stupid shouting that into the mic.

You might try raising your vocal pitch instead of getting louder.

I don’t know how ACX feels about sound effects.

I don’t either. I’m pretty sure they draw the line at “radio theater.” I want to go back to the idea of someone telling you a story over cups of hot tea. They’re not likely to have an orchestra or special effects technician in attendance.

But there is also the technique of getting louder without actually getting louder. Speak the bangs loudly so the stress in your voice goes up, but back away from the microphone so your volume doesn’t change. You have to have a very good feel for the technology, be able to see the volume meters and have worked in radio for 35 years.


This is one way to patch Eden Shower 5 so it passes compliance without sounding wounded. You have to install two custom tools.

This seems crazy until the second time you use it. The tool settings are sticky, so they don’t change from application to application. So it may be work for the first chapter, but after that, you just go down the list: Tool > OK, Tool > OK. etc. Some might even say I should be able to automate it.

I don’t remember where we were on testing. I know I never finished the web page or document because I had contradictory requirements.

When I do this list, I get good ACX conformance with the supplied clip.

And as we warn people constantly, save WAV copies of your work in a safe place. If the computer goes into the mud during this process it could take the work with it permanently.


LF_rolloff_for_speech is a custom equalization filter that must be installed.
Limiter is a custom sound processor that must be installed.

Select the whole clip by clicking just above MUTE.

Effect > Amplify > OK.

Effect > Equalizer: LF Rolloff for speech, 8191 Length > OK

Effect > Limiter: Limit -6, Hold 10 > OK.

Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK

Effect > Compressor: Thresh -20, Floor -50, Ratio 2:1, Attack 0.2, Release 1.0, [_]No Gain Makeup > OK

Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK

  • Test -**

Just so we’re on the same page, here’s where I found the plug-ins: LF_rolloff_for_speech and Limiter (version 1). Other than having to dig those up myself, your instructions were very easy to follow, and I’m glad to hear the settings will be retained.

Attached is the result, with the last sentence trimmed off so that we don’t have to mess with email attachments. To my untrained ear, it sounds very similar to the patched one you sent yesterday.

Does this pass ACX’ requirements (and if so, barely or by a comfortable margin)? If so, it sounds like the next action is for me to record the first few pages and see if the same settings will have the same results. (I’ll also finally get to practice editing out bloopers.)