Matching Inserted Audio

I am attempting to record an audiobook on Audacity. Of course, I make a few mistakes when reading some sentences, then I am attempting to correct the mistakes by rerecording the sentences at the end of the recorded chapter. The problem is that when I apply Effects to the entire chapter (such as EQ and Noise Reduction) to make both the original and inserted rerecorded segments sound the same, the effects do not work to match up everything. When I attempt to apply effects to the inserted segments, I cannot match them up with the original segments. How can I overcome this without rerecording the entire chapter?

You can try real time patch generation.

When you make a mistake, pause for short time, back up to the next even sentence or phrase and read the correction—and then just keep going. This has the advantage of keeping you exactly matched with the rhythm, emphasis, pitch, and tone of the show. Some readers ring a bell or make a loud noise (clap) so it’s easier to find it later. You should be able to just edit off the broken bits leaving a smooth reading.

It is strongly recommended that you Export a protection WAV file as you finish reading before you start chopping against editing errors (or the power fails damaging the reading).

If everything fails, then you will need to either read the whole thing again, or make the correction at the end of the reading and hope for the best.


You should not be applying custom corrections to each chapter. You can force that to work, but it’s not optimal.

Ideally, you read a chapter, Export a WAV protection copy, do the corrections, mouth noises, fluffs, errors, etc. Play it and make sure the presentation is smooth and even. Apply the Audiobook-Mastering-Macro and check it with ACX-Check. Maybe apply gentle noise reduction. Maybe not.

What kind of effects are you using?


Koz: Thank you for your responses! I have been applying Loudness Normalization, EQ, Limiter, Compression and Noise Reduction to each file. I had watched a video on YouTube that recommended I do so to meet ACX standards. Are you saying I do not need to apply these to meet ACX standards?


I left out a step in the fluff correction. When you make a mistake, leave the recorder running, look back to the last even sentence or phrase and just starting reading again, this time with the correct words. It’s a New User mistake to stop the recorder and try to do a live edit to correct the mistake. That’s amazingly hard to do and it throws off the swing and rhythm of your reading. It’s much better to do it in post production editing. This also allows you to do the edit at a sentence break which it’s much less noticeable.

I really should record a mistake as an illustration.

I’m saying you may not have to work that hard. The Audiobook Mastering Macro has Equalization to get rid of thumps, bumps, rumble, earthquakes, thunder, and some P pops. It preconditions the work for Loudness Normalization which places the work in the middle of the ACX loudness range. This overall volume change may give you tips and peaks louder than the -3dB limit, so we follow that with gentle peak limiter. You would think that would change the loudness and it does, but so slightly that the chapter stays in loudness range.

All with one key click.

It doesn’t affect noise, but if you recorded your performance in a quiet, echo-free room, you may not need it.

If you do need it, we have a recommendation for that, too. Noise Reduction of the Beast. Run noise reduction with 6, 6, 6 settings. It gives you half the noise you had before and nobody can tell what you did. If you have noise values worse than that, it may be valuable to find where noise is coming from and fix it there.

Let’s see if we can get a forum sound sample to post.

That was written for the old forum software. You would be using the Upload arrow symbol at the top of the forum text window.


Do you use ACX-Check? That will tell you your three sound readings and is an approximate match to the ACX Audiolab

We measure noise and last check, they don’t.



Pardon me for bothering you again, but I cannot locate the Audiobook Mastering Macro plugin. I downloaded a file from the Audacity/Nyquist site, but it is just a txt file. Can you provide me with a link for the download?

Thanks again,


If you have the right file, it is, in fact a text file. Much of it isn’t English words and the layout of the file is wacky.

This is what happens if you open the file in a text editor and stretch it out. It really is four steps.

Right. It doesn’t fit, but Audacity doesn’t care.

This forum message (scroll down) tells how to install and use it.

I can’t find the forum tools to let me point to one part of the message.

Were you able to record that ten-second voice test? Sometimes we can immediately work around problems by just hearing your voice.

Let us know if the forum refuses to accept the WAV sound file.


Koz: I have uploaded the 10 sec. audio test, and I have successfully installed the Macro in my program, and I have successfully utilized the Punch-in Repair tool. Looks like I am on my way to success with the program.


Did the forum complain or send you messages? I don’t see your sound file. Could you see it after you uploaded?

I’m only beating this up because we think this part of the forum may be broken.

You are the test case.


WAV upload works for me …

(but the name of the WAV file is not displayed in the post).

Koz: I could not find an attachment command on the Forum page, so I used the Upload file command at the bottom of the page. That may be why you could not find the file I sent. If you can help me find the Attachment command, I will resend.



Trebor: Thanks for IDing the upload link for me. But now Audacity says that I am a new user, so it will not accept my file. :frowning:


The old forum used to do that, too, but in the old forum, we knew what the rules were.

It’s not that important. It’s just sometimes the sound test can point to problems you didn’t know you had. There were two forum posters who were trying to read for audiobooks and they both had trouble speaking clearly. In my very own personal case, my submitted sample reading failed because of mouth noises. Software can’t catch any of those.

I think it’s still true that ACX will not accept any reading unless it’s also for sale on Amazon. Can I buy your book right now?


This topic was automatically closed after 30 days. New replies are no longer allowed.