Matching sound quality of edited sentence in an audiobook narration

Hello,

I’m using Audacity version 3.1.3, and Microsoft Windows 11 Home to narrate an audiobook using Yeti Blue USB Microphone, Knoxgear Pop Filter, and Knoxgear TX100 Closed-back Studio Monitor Headphones.

I made an error on a sentence of an audiobook I am narrating, so I edited the sentence and replaced the incorrect sentence with the correct sentence. When I played the audio, the sound quality of the edit was different when compared to the previous and subsequent recorded parts of the chapter. Is there a way to make sure the audio sound quality of the corrected sentence matches a pre-recorded audiobook recording?

Thank you in advance for your assistance.

Can you post the error? There’s nothing quite like offering detailed technical advice blind…or deaf in this case.

Drag-select a portion before and then going into the edit. A portion that illustrates the quality shift, say three to five seconds before and three to five seconds after. Export as a 192 quality, constant MP3 and post it here. You will note 192 quality is the ACX/Audible submission quality.

Scroll down from a forum text window > Attachments > Add Files.

If you’re recording in Windows 11, there is a festival of problems you can have by Windows trying to “help you.” You can also have sound problems by Zoom, Skype, Meetings, and other programs trying to “help you.”

And finally, the microphone itself can have driver and software packages trying to “help you.”

You picked the worst possible repair condition. Come back next week and try to patch it.

If you know you’re making a fluff or mistake, the prevailing advice is stop reading, go back at least one full sentence and read it all again. Do Not try to patch one word or other short phrase and Do Not stop the recording. Later in editing, you go in and slice off the mistake and the new and old sentences should fit together. You get the swing of how to do this as you go. This technique closely matches the mistake and the voice quality, presentation, and emphasis of the patch because they’re seconds apart.

Do you listen to your live voice as you go? It should be possible to plug your headphones into the Yeti and listen in real time. That goes a very long way to making your chapters match because you can hear in real time when your presentation, emphasis, and rhythm start drifting off.

That also gives you the ability, absent all the other computer problems, to do a patch by using Music Overdubbing techniques. Play the work up until the patch and then start announcing and match the voice quality in your headphones in real time. That’s not easy, but it’s a lot easier than trying to announce a patch cold.

Koz

Thank you so much for your response. I’m not sure how to “Export as a 192 quality, constant MP3 and post it here. You will note 192 quality is the ACX/Audible submission quality.” Can you please direct me to any available guidance on how to accomplish this please?

Try File > Export > Export as MP3, then near the bottom of the pop-up window, set “Bit Rate Mode” and “Quality”

When you are making a post on the forum, look below the box you are typing in. There is an “Attachments” tab.

A couple of useful links relating to jademan’s reply:
https://manual.audacityteam.org/man/mp3_export_options.html
https://forum.audacityteam.org/t/how-to-attach-files-to-forum-posts/24026/1

Thank you Steve

Here is a test MP3 file that illustrates the problem I need assistance in resolving. Thank you for your assistance.

It sounds like they were recorded in different locations:
the first part has much more reverb(eration), noticeable on the esses …

because of that, IMO, it’s going to be impossible to make the audio match so no-one can spot the edit.

Ok. Thanks for your assistance. My next question is how do I create space within the track to insert a missed sentence?

It’s common to think you need to jam everything onto one track. You can do that, but it’s difficult, confusing, and time consuming to do it that way.

Another way is to split the chapter up into before the error (top) and after the error (middle). Then you can record the correction as a third track (bottom).

Screen Shot 2022-02-23 at 6.24.22 PM.png
You can slide the tracks left and right (sooner and later) so they all meet. Audacity will try to play them one after the other left-to-right. It gives you the ability to apply filters, effects, and corrections to one track without affecting the others.

Later, when you export the performance to a sound file, it will warn you that it is going to melt everything together into one continuous, single track.

~~

I did wonder when I listed to your sample, how you produced the “good” sample. It has some tonal distortions (from Trebor, above) and some damage that comes from applying noise reduction too strongly. The patch doesn’t have any of that distortion. Did you produce the original voice work? Are you at home trying to replicate or correct a studio performance?

Koz

Thanks for the assistance. Yes, it’s my voice producing the original narration.

Yes, it’s my voice producing the original narration.

???

You can get an effect like that if you record your voice with Skype, Zoom, or Meetings (for example) sleeping in the background of your machine rather than completely closed. Those applications apply very serious processing to keep all the voices, computers, and internet locations performing OK.

Then the correction sounds like it might be a simple recording without all that “help” going on. So it’s no shock that it doesn’t match.

I don’t begin to know how to fix that.

Koz

Ok. Thanks