Feedback request

Special Interest Groups Audiobook Production
1

I created this sample based on the instructions here: https://www.kozco.com/tech/audacity/TestClip/Record_A_Clip.html

I’ve created a few chapters for a book already, and when I use the documented techniques from this forum it all sounds okay to me, but wanted to get some feedback before I go any further. My setup is a 5 walls of blankets hung from the ceiling and a carpeted floor in a corner of my basement office, if that matters. I’m using a Behringer C-1 mic and UM-2 interface.

Thanks for any feedback!

Rick

2

Behringer C-1 mic and UM-2 interface.

I have one of those. One of each actually.

Sounds good to me. I applied Audiobook mastering and everything fell into place except the background noise (Room Tone) just squeaked past, so I applied gentle Noise Reduction of the Beast (6, 6, 6). Good to go.

Screen Shot 2020-08-04 at 8.38.29 AM.png
There is a defect in there. The first two seconds is supposed to be you holding your breath and freezing. I hear a lip smack at the one second point. That would seem to be unduly picky and it’s true it doesn’t make any difference in ten seconds, but when you start reading chapters, you are invited to leave longer room tone segments before and after. One way to get these longer segments is copy/paste a shorter segment. This would be a bad place to have a lip smack or clothing noises. We are reminded that -60dB noise means room tone has to be 1000 times quieter than your voice.

That’s the first third. I don’t hear any mouth noises or P-Popping, so that’s the second third. The third third will be more entertaining.

ACX is reluctant to consider any reading that doesn’t appear for purchase as either paper or ebook on Amazon. There is also a list of unacceptable books, such as instruction books, tech manuals, or public domain books.

Is your book conventionally published? If you’re not, this audiobook thing can be a lot more of a problem.

Koz

3

Thanks for the feedback, Koz. I have an ACX contract to do a couple of books that are already published, so that part is all good. This is my first book to narrate so I’m continually looking for tips, hints, and other advice.

As I’ve gone back and edited (thus far), I listen for extra clicks and smacks after I’ve applied the filter curve, rms normalization, et al and remove them as needed. I do need to work on breathing technique as I find myself having to edit those a fair bit.

Thanks again!

Rick

4

I’ve applied the filter curve, rms normalization, et al and remove them as needed.

The tools in Audiobook Mastering are for application to a whole chapter or segment after you get done editing and cleaning and intended to make the submission conform to ACX standards. They’re not cleanup tools. Filter curve is in the sense that it’s a rumble filter, but you should not be able to hear most of what it’s doing.

Are you doing the housekeeping? Export the raw work as a WAV (Microsoft) 16-bit sound file right after you get to the end of a chapter or segment, and then Export another one when you get to the submission stage. That’s your actual Edit Master. Only Then export the MP3 for submission. You can’t edit an MP3 without additional sound damage plus violating ACX compression quality requirements, and that first export keeps you from reading a chapter again if Audacity goes into the dirt during the edit/cleanup.

The voice is ‘suspiciously’ pleasant and well modulated. Do you read or perform—are you an actor?

Koz

5

I acted when I was in high school (lo many moons ago), and I help direct and tech direct the local high school drama club. If I didn’t have to buy food, pay a mortgage, etc, I would love to be an actor. I’m looking at audiobook production as an outlet for that.

Question about the WAV file: I’ve been saving my wip as AUP files. What is the benefit of doing that as WAV instead? Just saving drive space and reducing the sheer number of files involved? I’ve only been doing an MP3 as the last step.

And I note your comment on the audiobook mastering. My process thus far is something like this:

  • Record the chapter (I use punch and roll)

  • When done with the chapter, listen to the entire chapter again, checking for missed/changed words, breaths, other noises. Edit as needed.

  • Apply Audiobook Mastering

  • Listen again and do a final QC - this is where sometimes I hear things I might have missed before*

  • Trim beginning/end

    • if I find these, is it your advice to go back to the pre-mastered file and fix them, then re-master?

I’m hoping that as I get more experienced, I can cut down on these steps but right now I’m being super careful about every step.

6

I’ve been saving my wip as AUP files.

I know it appears you’ve been saving a single sound file. Audacity Project are actually much more complex than that.

This is an Audacity Project.

The _DATA folder can have thousands of tiny sound snippets inside. The AUP file is just the menu of how to put the show together from all the snippets. Never separate the AUP file from the _DATA folder and don’t change any names.

The Project (its proper name) is saved in Audacity’s super high quality internal format (so that part is good) but that makes it much bigger than any common sound file.

There are any number of forum complaints from people whose little sound snippets didn’t all come back when they tried to open a saved show.

“What’s an ‘orphan file’ and why didn’t my show open?”

If you lose or damage the AUP file, that is quite probably the end of the world for the show, in spite of you still having all the little snippets right there in front of you.

The next version of Audacity will have a very greatly improved Project format missing a lot of these problems.

But, to bring this around, Exporting a WAV (Microsoft) 16-bit isn’t the worst way to avoid possible Project problems. I’m the poster child for this since I’ve never lost a paid show. In my world, the WAV file was the desirable product, so I had no reason to mess with Projects.

At the other end, you produce your 192 quality, constant bitrate, MP3 show for export and then what? You can open that up for listening and nothing else. If you try to edit or cut it, it will fall below the 192 quality specification. So the MP3 does not make a very good edit master. I bet you’re asking what ACX does to it. The product degeneration is built into their different products and services, and they are a business, so they have to pay storage. MP3 is lighter than WAV.

What do you do with the show right after you make the MP3? That’s the magic time to also export a WAV in addition to anything else.

MyFamousAudioBookEditMasterChapter3.wav

That’s also the obsessive engineer never putting spaces in filenames.

Koz

7

Missed one.

  • Record the chapter (I use punch and roll)

  • When done with the chapter, listen to the entire chapter again, checking for missed/changed words, breaths, other noises.

  • Edit as needed.

  • Apply Audiobook Mastering

  • Listen again and do a final QC - this is where sometimes I hear things I might have missed before*

  • Trim beginning/end

Works for me. You may already have the minimum number of steps.

There is a trick here. If you apply Loudness Normalization and Limiter and they’re not needed, they don’t do anything. You probably shouldn’t apply Filter Curve more than once. You might be able to hear that. So that is the sole exception to the law of always apply all three and in order.

Koz

8

The room sounds dead: that’s a good thing, (no reverb or echo).

Free tools (plugins) worth the effort of download which work in Audacity on Windows …

Couture expander, (free version): attenuates the signal when you are not speaking.
G-Multi: muti-band compressor, attenuates momentary instances of extreme sibilance, (de-essing),
& attenuates momentary instances of extreme bass.

Currently only 32 bit versions of plugins work in Audcaity on Windows, (even if your computer is 64 bit)

9

Trebor - thank you for the analysis. I will definitely research the plugins you used!

Rick

10

ACX expects everything to match. So the beginning and end of each chapter, the chapters to each other, and the beginning and end of the book. It’s not unusual for new readers to get to the end of a book a “seasoned professional” and are horrified at the quality of the first few chapters.

Also, it’s good to remember that whatever you do to the work has to be done forever until he book is done. So no production missteps or leaving out a tool. If you find yourself getting lost in the mind-numbing process of reading out loud and editing (that never happens to me) you can trash the edit and go back to the Raw Reading you saved as either a Project Safety Copy or WAV file.

Or any of the “Save As” copies you made in the middle.

If you didn’t get too lost, you can also Edit > UNDO until you find your place.

There is one other problem with Projects. It’s tempting to keep updating one single project as you read, edit, and master. If that one Project goes into the dirt, that’s you going all the way back to reading the chapter again.

Koz

11

Koz - you’ve sold me! LOL

I did leave out the step that I have actually pasted in the last few seconds of the prior chapter to the current file to make sure everything sounds the same - tone, spacing - then I undo that. Haven’t had any issues so far.

I think I’m somewhat stuck in a learning loop. Since this is the beginning of the journey and I’m learning so much so quickly, I keep finding new advice and tips and tools - both narrating-wise as well as technical - and I want to keep applying them. But if I don’t just pick a time or spot on this curve to just get the whole book done, I won’t deliver it on time.

From now on I’ll definitely use the WAV approach.

You guys are great! Thanks for all the help!

Rick