I’ve been podcasting for about 11 years. During that time, I’ve gone through much trial and error, have done a ton of research, and eventually settled on a formula for producing podcasts. No, this isn’t about marketing your podcast or how to get rich or even make money. This is about the technical aspects of your podcast file itself and what every podcaster should do when they produce that final file that every subscriber and listener will play on their podcast app of choice on their phone, or even via the web on their computer.
In this post I’m going to zero in on two elements; loudness and file size (file size will be directly linked to bitrate, stereo versus mono, etc.)
What determines the final size of your podcast episode relies on a few things. One, of course, is the length of the podcast episode itself. A two- or three-hour long podcast is going to have a larger file size than a 30 minute or 1 hour podcast. The Sample Rate and the “bit-ness” are the other two factors.
It’s common knowledge from the early days of MP3 files (specifically as they relate to music) that 128k (kilobit) stereo is considered “CD Quality” for playing music from an MP3 file. The sample rate is usually 44100 Hz (or 44.1 kHz as it’s usually written).
For example, a 10 minute 23 seconds long DDP episode (as a test) that is 128 Kbps at 44.1Khz Stereo is 9.52 MB in size. That’s not a bad size but the episode is only a little over 10 minutes.
Now, let’s use the latest episode of PX3 from Justin Rober Young. Justin produces his episodes in 192 Kbps, 44.1 Khz 32 bit Stereo. Episode 242 is 72.9 MB in size and is 50 minutes and 25 seconds long. That’s a pretty large file for a 50-minute episode. Personally, I don’t understand why Justin decided to put his episodes out in such a high bitrate. It’s not needed and it’s wasteful.
For the longest time, the recommended MP3 specs for your episode were 64 Kbps, 44.1 Khz 16 bit Mono. I personally like the sound of Stereo, but let’s see what file size we wind up with for Justin’s episode at 64 Kbps Mono. File size says — 26.6 MB (27,319 KB). Sound quality? Again, not that noticeable.
But what about Stereo? The best course of action to maintain good quality sound for a file that is 64 Kbps and Stereo is to change the sample rate by half to 22050 Hz. With rounding the way Windows does file sizes, the file is the same — 26.6 MB (27,320 KB). However, the sound is better because now you have stereo instead of mono. I’ll share links to the different file formats so you can listen for yourself*.
PX3 192 Kbps, 44.1Khz 32-bit Stereo (72.9 MB)
PX3 128 Kbps, 44.1Khz 16-bit Stereo (49.7 MB)
PX3 96 Kbps, 44.1Khz 16-bit Stereo (38.2 MB)
PX3 64 Kbps, 44.1Khz 16-bit Mono (26.6 MB)
PX3 64 Kbps, 22050Hz 16-bit Stereo (26.6 MB)
There have been numerous articles written on this subject, but I’m going to do my duty and discuss it once again. Loudness or “perceived loudness” is measured in LUFS. What is LUFS? Here’s the full definition:
LU is an abbreviation for the term Loudness Unit. Loudness Units differ from decibels in that decibels measure the level of air pressure generated by the sound, whereas Loudness Units are an arbitrarily established unit for audio work which allows a person to control the resulting output volume of a piece of audio. In practical terms, one Loudness Unit is equal decibel of air pressure generated by the audio. LUFS is a term for Loudness Unit Full Scale, which allows the measurement of loudness of a piece of audio without a reference, whereas a decibel measurement requires the reference of standard air pressure, against which to measure the air pressure generated by the audio.
(Courtesy of https://backtracks.fm/resources/podcast-dictionary/lufs)
In simple terms, it’s how loud your audio sounds to the listener. Many audio production software have a way of processing your audio to a certain LUFS specification. The going consensus is your audio should be at -16 LUFS for Stereo, and -19 LUFS for Mono (remember in audio the level is express in a negative number, meaning in this case that -16 is louder than -19). The reason for -19 LUFS for Mono is because when a mix is exported from Stereo to Mono, 3db is usually added to the overall loudness of the resulting audio. As an aside, YouTube recommends -14 LUFS for YouTube videos.
Why is this important? Imagine you’re in the car and you’re listening to the latest episode of one of your favorite podcasts. You turn up the audio so you can hear it comfortably, without straining your ears. The episode finishes and the next podcast episode from a different producer is queued up. When it starts playing, the audio is so loud that you race to your radio or device to turn it down. Now, once that episode has finished, you have another podcast in the queue and when it starts playing, it’s so low you can barely make out what anyone is saying, so you turn it back up. This is the problem.
LUFS isn’t new. It’s been used in radio basically forever (though the levels are different — even between countries).
It’s important for the sake of your listeners to abide by these rules and try your best to have your audio at the correct LUFS level. Fortunately, many audio software editors have this capability built-in; I use Adobe Audition and it has a function called Match Loudness.
So, for the best listener experience from one podcast to another, it’s best if all podcasters adhere to these guidelines for loudness.
I’m not an authority on podcasting, but I have been doing it for a long time. I’m a techie at heart (I own my own IT Consulting business), so the technical details are very important to me. I spent years testing various bit rates, sample rates, etc. to get the best sound with the smallest file size. Granted, many of us now have large Internet pipes at our homes or businesses, but we must always consider those less fortunate around the world that don’t have unlimited bandwidth on their devices. Balancing quality audio with the smallest file size we can get without compromising the overall listenability of our podcast should always be on our list of goals for our podcasts.
*Please be aware that I took Justin’s original PX3 Episode 242, which was already in MP3 format, and re-saved it in the various kilobits and sample rates; it’s better to do this with the original uncompressed .WAV file because otherwise, as with what I did, I’m recompressing an already compressed audio file each time I re-sampled or changed the bit rate.
Also, I’m specifically talking about file size as it relates to MP3 files. You can produce episodes in AAC, which will give smaller sized files with good quality audio. Opus is another good codec that provides much smaller file sizes with good quality audio. However, MP3 is still pretty much the standard.
Simple southern dude with a background in technology.