2nd Opinion – An Audiobook Checkup Tool

2ndOpinionWhat does it do?

This program is meant to be run once you have mastered your audio but before you send it off for publication. It is not meant to be used in place of an actual engineer. As it’s name implies, this program is meant to give you a “second opinion” about the state of your audio files.

  • This program checks/fixes the following:
  • Does each file peak at or below -3.0dB?
  • Does each file have an Average RMS between -18.0dB and -23.0dB?
  • Was each file recorded in mono (optional)?
  • Was each file recorded at a sample rate of 44.1k?
  • Does each file have between ½ and ¾ of a second of silence (-60.0dB or less) at the head?
  • Does each file have between 3 and 5 seconds of silence (-60.0dB or less) at the tail?
  • Does the retail sample have a length of 5 minutes or less?
  • This program also reports the total length of the audio for all files rounded down to the last full minute (as required for direct submission to Audible).

When this program locates an issue, it is labelled as either an ALERT or an ERROR.

ALERT – This piece of audio does not meet one of the specifications outlined above. The program will attempt to repair it for you. Please review any file changed by this program before submitting to a publisher.

ERROR – This piece of audio does not meet one of the specifications outlined above, but this program cannot repair this kind of issue. A human ear with engineering experience is required.

If your audio did not pass inspection, repair your audio and re-run this program.

How do I use it?

  1. Start the application by double clicking on it
  2. Click on Choose Directory
  3. Locate a folder containing WAV, FLAC, or AIFF files
  4. Click Choose
  5. Optionally, toggle optional settings *
  6. Click Analyse

* Optional settings include Check & Fix or Check Only, Limit audio above -3dB or Normalize down to -3dB, Limit output to Mono or Output same as Original.

This program expects the retail sample’s name to end _sample.filetype
(Example: Walden_sample.flac, Walden_sample.wav, etc). It treats the sample slightly differently than other files. If you have included one, please follow this naming convention.

The program writes its findings both to the window and to a file inside the folder called Report-2ndOpinion.txt. If it finds errors, once you have corrected them and you run the program for a second time, it will append its new findings to the end of the same Report-2ndOpinion.txt.

Because people sometimes forget to make backups, before examining your files, the program will backup your originals to a folder named OriginalFiles-date-time. The program will make a new set of backups each time it is run.

If the program finds no errors while it runs, it will let you know. 🙂

How much does it cost?

The program is free to use, but I always appreciate…

A Cup of Coffee

A ridiculous amount of caffeine was consumed while researching all of this stuff.
Add some fuel if you would like to help keep me going!
$5$10$15$20$25Other

Which file should I download?

Version 1.0 released October 7, 2015
Mac OSX (tested on 10.10 & 10.11): 2ndOpinion-mac.zip
Windows (tested on 7 & 10): 2ndOpinion-win.zip

How do I install it?

Mac OSX

  • Unzip 2ndOpinion-mac.zip
  • Move the application to the Applications folder (recommended)
  • Open System Preferences and go to Security & Privacy.
  • Set Allow apps downloaded from: to Anywhere

OSX Install Security

Windows

  • Unzip 2ndOpinion-win.zip
  • Move the application inside of C:Program Files (recommended)

What does it look like?

2ndOpinion UI

Frequently Asked Questions (FAQ)

What does it actually fix? What won’t it fix?

Peaks at 0.0dB or higher

WON’T FIX (but will alert you) – The audio has clipped. There is data missing from the file. Ideally, this audio should be re-recorded, not repaired. You don’t want to deliver broken audio to a client, do you?

Peaks above -3.0dB but below 0.0dB

WILL FIX – A Brick Wall Limiter is applied at -3.0dB. Since this program assumes that the file has already been mastered, at worst, there should only be a few peaks above -3.0dB. In this case, this kind of limiter will most likely make the smallest possible change to your audio while bringing it back into spec.

Optionally, you can choose to use Normalization instead reducing overall peaks to -3dB. Be aware that this will affect the entire file and not just the bits of audio that exceed -3dB.

Average RMS not between -18.0dB and -23.0dB

WON’T FIX (but will alert you) – I tried a few different methods of generic repair and was not satisfied with any of them. There are simply too many variables at play if your overall audio is either too loud or too soft.

All files not recorded to a single channel (mono)

WILL FIX – By default, any chapter recorded to 2 (stereo) or more channels will be mixed down to a single channel. This setting can be turned off before processing.

Sample Rate above 44.1k

WILL FIX – The audio will be downsampled to 44.1k.

Sample Rate below 44.1k

WON’T FIX (but will alert you) – Upsampling is not a good idea. Essentially, it amounts to asking software to “blur” your audio to fill in gaps. You don’t want blurry audio, do you?

Head is more than ¾ of a second

WILL FIX – The audio will be trimmed to 0.6 seconds.

Head is less than ½ of a second

WON’T FIX (but will alert you) – A short head could just be short, or it could have a stray noise in the middle of it that makes it seem too short. The software has no way of knowing if you started talking too early or if you simply bumped something.

Tail is more than 5 seconds

WILL FIX – The audio will be trimmed to 4.5 seconds.

Tail is less than 3 seconds

WON’T FIX (but will alert you) – A short tail could just be short, or it could have a stray noise in the middle of it that makes it seem too short. The software has no way of knowing if you turned off recording too soon or if you bumped something.

The retail sample is longer than 5 minutes in length

WON’T FIX (but will alert you) – Since the software can’t really listen to your performance, it has no way of assessing how it could shorten your sample.

Why doesn’t this software take MP3s as input?

This program is designed to work with lossless audio (like WAV, FLAC, or AIFF) rendered straight from your DAW (Digital Audio Workstation). It is best practice to do all edits in a lossless format before converting to a lossy format like MP3.

Why doesn’t this software convert the audio to MP3 automatically upon successful completion?

During development, an earlier version of this software did just that. When it became clear that distributing it with the MP3 feature intact would mean dealing with a licensing headache, the feature was removed.

In what circumstances would I want to disable the Check & Fix?

If you’d like to run the tests but do all repairs on your own, choose Check Only from the dropdown box.

In what circumstances would I want to switch from Limit to Normalize?

If a lot of your audio peaks above -3dB then you might want to switch to Normalize, which will reduce the volume of the overall file. Be aware that this will reduce your Average RMS more than using a Limiter. If your audio was well mastered beforehand, the Limiter will typically be a better solution.

In what circumstances would I want to disable the Mono Check?

To enhance the sound stage, audio books involving music, multiple narrators, or possibly a full cast audio drama, might want a final product to be rendered in stereo.

Can I change any of the other settings?

For now, no, but I am thinking about releasing an UberGeek Edition of this software where all of the parameters can be changed. If you’d be interested, let me know.

It crashes when I try to run it. What’s going on?

Permission-based issues on both Windows and Mac OS X can cause problems. Try moving the program itself to your desktop and running it from there. Make certain that the folder you have selected with the audio files that you intend to analyze is writable for both you and the program itself. And, be certain that the audio files are not currently open by another program when you are attempting to analyze them.

Issues have also been reported on Mac OS X 10.6 and older. I need someone running these older OSes to volunteer some time to test a custom build that will address this issue. Contact me if you are willing to lend a hand.

I have an idea for a new feature! Will you implement it?

Maybe. Contact me. Let’s talk.

Licenses, Notices, & Information

This software is distributed under an Attribution-NoDerivatives 4.0 International (CC BY-ND 4.0) license.It is distributed in the hope that it will be useful, but without any warranty. It is provided “as is” without warranty of any kind, either expressed or implied, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. The entire risk as to the quality and performance of the program is with you. Should the program prove defective, you assume the cost of all necessary servicing, repair or correction.

In no event unless required by applicable law the author will be liable to you for damages, including any general, special, incidental or consequential damages arising out of the use or inability to use the program (including but not limited to loss of data or data being rendered inaccurate or losses sustained by you or third parties or a failure of the program to operate with any other programs), even if the author has been advised of the possibility of such damages.

This software is written in TCL/TK and is used under a BSD-style license.
This software relies upon the SOX Audio Processing Library which is included under the LGPL License (LGPLv2).
The icon is derived from the Oxygen Icon Set which is included under the LGPL License (LGPL v2).