Apple has recently published two podcast related podcasting best practice guides. One on Marketing Best Practices and the second on Podcasts Authoring Best Practices. In this article our very own radio, loudness and podcasting specialist, Mike Thornton takes you through the Authoring Best Practices and offers tips and tricks along the way.
The Apple Authoring Best Practices guidelines cover formats, levels, and loudness metadata to help users to choose the best possible quality and file size tradeoffs while making sure you provide a pleasant and distortion-free experience for your listeners. Mike unpacks what Apple cover in such a way that non technical users should be able to understand.
Formats
In this context the Formats section is all about audio data compression formats. Data compression is different to audio compression, which is designed to reduce the dynamic range (the loudest to the quietest sound) of an audio source. Data compression is all about reducing the file size of the audio to make it easier and quicker to download and play over the internet, especially on mobile devices. Examples of data compression formats are mp3 and AAC, which is the format that Apple prefers.
We recommend that you produce a final mix file as a WAV file. This is an uncompressed audio format. Use 16 bit preferably at 44.1k sample rate but 48k is OK. Then depending on how your podcast is being hosted, you will either upload the WAV file and your hosting service, like Buzzsprout, (which is the service we use for our Pro Tools Expert podcast) will produce the data compressed versions. If you are uploading to your own servers then you will need to convert the WAV file yourself, and for Apple podcasting we would recommend (as Apple does) to use the AAC format, rather the more commonly used mp3 format as the quality is better with an AAC file of the same file size.
Levels
It is really important that your audio levels throughout your podcast are consistent. This is especially important as a lot of people listen to their favourite podcast during the daily commute where the background noise can be high. If the sound levels are going up and down your listeners will be having to turn their player’s volume up and down and very soon they will get fed up and go and listen to someone else’s podcast.
Getting consistent sound levels is something to take into consideration at the recording, mixing and mastering stages.
As well as getting the sound levels consistent throughout podcast, you also need to set the loudness correctly. Audio levels and loudness, although related are different. Audio level is a measurement of the audio signal and is often displayed using a dB, dBu or dBFS scale, whereas loudness is a measurement of the perceived loudness of a sound file and is measured in LKFS or LUFS. Don’t worry LUFS (Loudness Units Full Scale), and LKFS (Loudness K Weighted Full Scale) are exactly the same.
Unlike broadcasting where the delivery specs for loudness are usually either -23 LUFS in Europe or -24 LKFS in the US when it comes to online content the loudness should be around -16 LKFS although different providers can specify slightly different numbers. In their guidelines Apple recommend -16 LUFS with a tolerance of 1 LU either way.
The other parameter is the maximum peak level. With digital recording the maximum signal level is 0dBFS. But as with so many things it’s not that simple. There is another measurement that we use these days which is dBTP, where TP stands for true peak. True peak represents the audio level once the digital audio has been turned back into analog ready for our analog ears to hear. The key thing to be aware of is that before you hit 0dBFS (this is when the meter turns red) you may well have gone above 0dBTP and so it is preferable to have a true peak meter as the true peak level can be up to 3 or even 6 dB higher than the normal sample peak level.
Apple recommend that you do not go beyond -1dBTP in their Podcast Authoring Best Practices guide. My advice, especially of you don’t have access to a true peak meter, is not to go above -3dBFS. The problem is that when your signal level gets close to the maximum it is possible for the codec (the software that converts the WAV file into an mp3 or AAC file) to distort so although there was no distortion on the WAV file, the audio in the mp3 or AAC file that your listener will hear can be distorted. By playing it a little safe with the maximum peak level you will make sure the version of your podcast that your listeners hear should be distortion free.
Loudness Metadata
Some devices can read metadata embedded in the mp3 or AAC file. Metadata is just a posh name for data about data and as well as the file name, sample rate etc it is possible under certain circumstances to embed the loudness level in the file. This can then be used by the player to adjust the volume so that podcasts with different loudness values will play back at the same loudness on players that support this metadata. Unfortunately Apple gives no indication on how this loudness metadata can be inserted into the file.
More Resources To Help You Make Better Podcasts
Here are some links to a range of helpful content including the 2 new Apple Podcast Best Practice Resources…
Understanding Loudness - For Everyone Working In Audio Post Production
Loudness Metering And Resources
Free Loudness Metering Plug-ins