I am trying to do my audio books, but I’m having trouble. First, the file size for a chapter is only 12KBand I know that is much too small. Then, I try to master in ACX requirements and once I start doing that it all sounds really bad and flangy or digitized. I don’t know what i’m doing wrong, but my time is running out and I just seem to be running in circles.
How can I set it up from the beginning to get the best possible sound? Can you also set it up TO ACX requirements to avoid a lot of post production editing?
Yep, 12K is way too small. I have a Chapter 1.wav here that is 128Mb. About 17 pages of a standard 5" x 7" paperback book.
Firstly - you need to save/export as a WAV file
and, setting up the ACX requirements is not dependant on Audacity, so much as your mic and equipment, and how quiet you can get your recording space.
Then with only minor tweeks, you get your sound right in Audacity.
None of it can be done in a hurry. Sorry
So, what equipment are you using?
How many words in a chapter? or standard pages perhaps.
What sort of “studio” have you set up.
I wouldn’t be surprising for that file to be quite small. It’s a common beginner mistake to thing that the .aup file is the whole project – it’s not. It only is a list of pointers to other files where all the actual sound data is (typically in a folder in the same location as the .aup).
PS: It’s “bytes”… “baud” has quite a different meaning.
As Robert suggests – export the project as a .wav and look how big that is.
And if you do have a terrifically small MP3 sound file, making a highly compressed MP3 file by accident will sound like what you got. Flangy and digitized.
When you finish a section of work, export as WAV (Microsoft) even though ACX AudioBook submission requirements are MP3. It’s good to save the archive of the work in perfect, uncompressed format. You can easily get from perfect WAV to any other format. Once you make a compressed, distorted MP3, the work will never get any better quality than that.
Are you following anybody’s audiobook instructions or demonstrations? ACX itself has good instructions and videos.
"my time is running out and I just seem to be running in circles." is a very bad sign. You are replacing a recording studio and production team with a microphone, laptop, headphones and script in the bedroom. We can help talk you through it—we have some custom tools for audiobook reading—but we are doing technical troubleshooting across multiple time zones and it’s not fast.
Let’s see if I can answer all these questions. First, I’m not new to voiceover work,or running sound boards. I am new to the ACX requirements, since I am not an engineer (but I do have engineer friends).
My equipment is M-Audio M-Track and Sterling SP50 Condenser mic with a pop filter on Windows 10.
Yes, I was looking at the .aup file. I have now saved to Wav and the file size is where it is supposed to be. Somehow my problem is that once I do Noise Reduction, that’s when I notice the flangy sound, especially during a breath into the first word after. I have the ACX Check and I have had to turn my gain all the way up to get as close to the requirements as possible. My RMS is still -23.8, but I think I can fix that in Mastering. I have a very quiet “studio” so very little room noise to start with, so maybe I don’t even need to do Noise Reduction, since it is not giving me the sound I need. I try to get everything right during recording (like I do in photography) so that there is less time spent mastering.
I have re-recorded the Intro and I’m going to send to my brother, who uses Pro Tools, so he can see where I’m at and if he can do anything I cannot. It sounds good, as long as I don’t apply any effects to it.
If you have any suggestions, they would be much appreciated.
Also, I thought I saw somewhere some notes on spacing between sentences, paragraphs,intros, etc, but don’t remember where I saw it. Do you have a breakdown of timing, so I can save it to a file to check?
Right. That’s what Noise Reduction does when you adjust it wrong or give it a bad Sample or Profile.
The object is to use the smallest correction possible.
One day as an experiment, I recorded an audiobook compliant sound clip in my quiet bedroom with simple equipment. The difference between the shoot and the “submission” was making it a little louder. Full stop. So don’t automatically assume Mastering or other extensive list of corrections, filters and effects. ACX has a sound quality failure they use if they catch you doing that.
Since you already know how to run a board, you adjust the microphone device for peaks at about -6dB on the Audacity sound meters. A perfect sound shoot should look about like this.
You can wander a bit lower, but probably shouldn’t go much higher. You may not push the meter all the way up so it turns red. That will produce permanent distortion. -6dB on the meters works out to 50% (0.5) on the blue waves
The Audacity sound meters are malleable and moveable. You can change their size, orientation and position for convenient use. Usually big is good. You have to keep track of them while you’re reading. Remember, no sound guy/lady.
You may find, as hundreds of people have, that you can’t get that loud. Modern microphone and microphone systems are “quiet” to keep people out of overload. Unfortunately, if everybody does that, you can’t make good level any more. See how you do.
When you think you get close, post a sample sound clip. Post WAV and post it raw. Zero corrections.
That sound device has a sound meter built-in. See how it compares with the Audacity meters. I would start out life with the bottom green light on, the next one up flashing and the red light never. Some makers cheat. They may display the red light before actual sound damage and an occasion red flicker is OK.
Okay, so last night, I had to turn the gain all the way up, to get to ACX requirements. My RMS was off by .6dB. I sent the file to my brother, so he can put it in Pro Tools and see if he can get everything right. As long as I do nothing to the file, it sounds great. As soon as I apply even one effect, then I get the issue. I’ll let you know what happens when my brother returns the file to me.
The usual recommended drill is to record such that your peak level falls between -6 and -10 dBFS.
Then in post processing step A is to “normalize” the clip and put the peaks at -3.2dB (a little margin over the -3 max to allow for processing).
If that leaves your RMS level too low then gentle application of a compressor application is the usual fix.
If after normalization your RMS level is too high then simply reduce the gain (or re-normalize to a lower value).
If after normalization your noise floor is too high then first attempt to remove the noise source from your studio. Failing that noise-reduction and /or some frequency equalization (such as the “high-pass for voice” filter that’s been offered up many times) can help.
Adjusting RMS or loudness is rough. If your basic, starting-out quality is a little off, you can chase your tail.
Changing loudness is basically moving the volume control up and down. Of course, it’s not that easy because that brings the peaks and noise (see that reference page) up and down, too.
To change one value without changing the others requires special effects and filters. You are creating beneficial, desirable distortion.
Rather than forum ping-pong, we can give you a push if you post a test clip on the forum. You can’t go over 20 seconds and Do Not apply corrections. Just a clean voice with the new louder level. And I’m not shocked that you need to run the microphone controls all the way up.