Podcasting Basics, Part 5: Loudness for Podcasts vs. Radio

podcastloudness_Featured

This article evolved from an email to the AIR listserv back in March. Questions around levels for podcasts and radio kept coming up, so I sat down one night in the midst of preparing for a move from Minneapolis to Washington, DC, to hammer something out to address them. I ended up with a quick-and-dirty explainer, along with a couple of step-by-step examples for various platforms. What follows is certainly not an exhaustive explanation, but instead an attempt to give those who are curious a little more information.

TL; DR

Radio and podcasts need two different loudness targets: -24 LUFS for radio and approximately -18 LUFS for podcasts. It’s best to do all of your production at the lower “radio” level, even if your final product is to be at the higher “podcast” level. To get from -24 LUFS to -18 LUFS, use the steps outlined at the end of this document or use software specifically made for the purpose.

First, Some Context

The loudness standards for public radio distribution are welcome in that they allow us to define a single target for moving content around the various public radio systems, regardless of its production style. My previous article at Transom, The Audio Producer’s Guide to Loudness, addresses how to achieve the -24 LUFS target used in radio production.

The radio standard doesn’t address a loudness target for podcasts… and it’s not likely to. Those of us involved in making that standard were primarily concerned with passing content amongst the distribution points in the public radio system.

Podcasts are addressed in a recommendation by the Audio Engineering Society. It’s long and technical (I know because I helped write it!) so I’ll spare you the details: it recommends that podcasts target somewhere between -20 LUFS and -16 LUFS depending on the style of content. Most of you reading this probably produce talk content, and that means -18 LUFS, the very middle of the recommended range, will suit your needs nicely. (If you are already producing podcast content at -16 LUFS, a number that was often suggested until the AES recommendation was born, don’t sweat it, you can keep doing what you are doing — or move down a little to -18 LUFS.)

So now we’ve got two loudness targets depending on your destination:

  • Traditional public radio: -24 LUFS
  • Podcasts: -18 LUFS

YES, YES BUT WHY THE HECK ARE THERE TWO STANDARDS?

The lower -24 LUFS radio spec allows for a wide dynamic range. That means audio producers can create programming with soft, low-level peaks and louder, hotter-level peaks and everything in between (within good production practices) as long as the overall average loudness is -24 LUFS. Talk radio might not take full advantage of this wide dynamic range, but music programming certainly does, so the folks who made the radio standard had to keep a wide variety of content in mind when they created it. This dynamic range is what allows one producer to have a very compressed and punchy sound while another has a more natural and open sound. As long as they target the same loudness value, in general the two programs will sound equally loud over time.

Podcasts (and podcast-like content), on the other hand, are frequently enjoyed on mobile devices. Unfortunately phones and tablets don’t have the highest-quality audio playback capabilities. They do just fine with louder content like pop music but they have a hard time dealing with lower level content like that produced at -24 LUFS. You can certainly turn up the listening level on the phone, but it may struggle to raise the -24 LUFS content to a consistently comfortable listening volume. If you were listening to your stereo, you’d turn the volume knob up to a comfortable listening level and you’d probably still have a looooong way to go before the knob hit “max” volume. In fact, if you turned the stereo up to max, your cat would freak out, your landlord would evict you, and you’d probably suffer from some severe hearing loss (you poor sap). Your phone, however, just can’t compete with this due to the limitations of its electronics. It doesn’t have the capability to significantly boost lower level audio like your stereo does. (Frankly, to protect your hearing, you don’t want it to, otherwise you could accidentally hurt yourself when listening on headphones.)

Compounding this problem: when folks listen to podcasts on a mobile device, they are likely on-the-go. Commuting via metro (ahem… subway… I’m a DC boy at heart it seems), bus, walking, whatever. Or they are listening while doing the dishes or driving in the car. All of these scenarios will have loud background noise that competes with the audio.

Louder listening environments, combined with a mobile device’s inability to pump up the volume, means we need a higher loudness target for podcasts.

Let’s take a Nick Quah-style GIF break, shall we?

LoudnessGIF1

Another Thing

The possibility for a wider dynamic range in the radio spec works well for radio because every station includes processing before the signal is broadcast over the air. Processing like compressors, limiters, companders — equipment that squashes down some of the peaks, brings up the lower levels, and reduces the dynamic range. This processing helps the signal survive over-the-air transmission, helps it be more audible in environments like the car or the kitchen, and shapes the sound of the station. Each station wants to tailor its sound based on its own tastes and content, and they do that by processing their signal differently than the other stations in their market. One station may prefer a more natural sound (often the choice most public radio stations make), while another station may prefer a more compressed and aggressive sound (typical of Top 40 or commercial talk radio).

The public radio loudness standard ensures that each station will receive every program from the distribution system at a consistent loudness. That consistency allows them to fine-tune their processing to their liking and create the sound they want.

If the processors are squashing the dynamic range anyway, why don’t we just use the -18 LUFS podcast target for radio production? It would be easier to have one number for everything.

Well, a higher loudness target means lower dynamic range and that effectively steals the dynamic range from those producers who want it. There are certainly producers out there (like yours truly) who prefer to have the option of a wider dynamic range when they need it. Music productions especially want that wider dynamic range, but so do some more traditional talk programs like many of the daily news magazines you know and love.

Another reason not to produce everything at -18 LUFS is a logistical one. The higher peaks associated with a higher loudness target make it really difficult to produce much of anything without some kind of processing… especially when you are producing live radio! Case in point: if you sat down with your portable recorder and shotgun mic right now and tried to record your voice at the equivalent of -18 LUFS, you’d distort the recorder. To prevent the distortion you’d need compression and limiting in line before the signal reaches the recorder, and that’s before you’ve even gotten a chance to listen to your audio. You’d have to process the sound before it was recorded. That’s not best practice and it would be expensive… no good.

 

To sum up…

 

  • Traditional public radio: -24 LUFS
    • Allows for a wider dynamic range for those who want it.
    • Is at a level that is low enough for everyday production without needing compressors and limiters.
    • Ensures stations receive content that is consistent from program to program.

 

  • Podcasts: -18 LUFS
      • Overcomes the limitations of mobile devices.
      • Overcomes the limitations of the environments where our listeners find themselves.
      • Requires reducing dynamic range.

     

 

Getting From -24 LUFS for Radio to -18 LUFS for Podcast

You’ve read The Audio Producer’s Guide to Loudness so you have the tools and know-how to produce content at -24 LUFS (if you need more help, let me know.). That’s great, because as stated before it’s much easier to record and mix audio at -24 LUFS. I recommend doing all of your production (recording, editing, and mixing) at -24 LUFS. Even if your content will never air on the radio, produce it at -24 LUFS first. Once everything is sounding amazing at -24 LUFS and you feel like the mix is done you can convert it to the -18 LUFS podcast target.

(Yes, there will be opinions raised against this idea. It’s an extra step (read: some extra mouse clicks). Also noise floor concerns are sometimes cited… but gear these days is of such quality that noise floor is not a major concern. And, frankly, distorting your recording device — which is what would happen if you tried to record at a super high loudness like -18 LUFS — is way worse than any potential noise floor problem.)

To convert from the lower -24 LUFS mix to the -18 LUFS podcast target we are going to reduce the dynamic range — reduce the peaks — so that we create room to raise the overall loudness of the file. If we tried to raise the loudness of the file without reducing the peaks, we’d end up with distortion, as many of the peaks would push to 0 dBFS. That’s bad. By managing the peaks first, we give ourselves more headroom — the space between the highest peak in the file and 0 dBFS. We’ll first use a limiter (either manually or via an automated workflow) to reduce the peaks and increase the headroom, then we’ll raise the gain of the entire file (again, either manually or via an automated workflow).

  • Mix to -24 LUFS average.
  • Limit to manage peaks.
  • Increase gain.

(This method, by the way, was introduced by Thomas Lund in his 2013 paper titled Audio for Mobile TV, iPad, and iPod.)

There are a few ways to do this, but I’ll outline some straightforward methods here using commonly available tools.

Pro Tools

    1. Bounce or export your -24 LUFS mix to a file. Be sure to label it appropriately so you don’t get confused later, because we are going to end up with two files, one at -24 LUFS and the other at -18 LUFS. I typically put “neg24” in the filename of this file.
    2. Import your file to a new session or track.
    3. Open the AudioSuite plugin called “Maxim” (see above) (Audiosuite -> Dynamics -> Maxim). Maxim is not a great limiter, it’s a free one that comes with Pro Tools. If you have access to a nicer limiter, such as the Waves L1, or similar, please use it instead! You can probably use the same settings as we indicate in the graphic here, but you may need to experiment a little with the specific plug-in you’re using.

      photo of Maxim plugin
      Maxim Plugin
    4. Adjust the parameters to look like so
    5. Highlight the entire file.
    6. Click “Render”.
    7. Assuming you started with a -24 LUFS mix, you now have an approximately -18 LUFS file!

AudioSuite

Hindenburg

The folks at Hindenburg make this easy. You can simply use an export feature to convert your mix to -16 LUFS. It’s not -18 LUFS, but it’s pretty darn close. Follow this guide.

(A quick aside — Hindenburg also includes a cool Auto Level preset called US Public Radio to help you make the -24 LUFS mix.)

Adobe Audition

Adobe Audition has a nice feature called “Match Volume” that will do the conversion to -18 LUFS. It’s not perfect — it doesn’t let you fine-tune the peaks, but it does a decent job.

    1. Open the Match Volume window.

      photo of Match Volume adjustment
      Match Volume
    2. Drag your file from the File window to the Match Volume window.
    3. Click on Match Volume Settings. In the “Match To” field choose “ITU-R BS.1770-2 Loudness”
    4. photo of Match Volume settings
      Match Volume settings
    5. Select the “Use Limiter” box and make sure the other settings are set like this:

      photo of Use Limiting setting
      Use LImiting
    6. Click Run!

(Insider info — I’m told that a future version of Audition will likely include a preset for US Public Radio to help get mixes normalized to -24 LUFS.)

Voilà. You bumped your -24 LUFS radio mix to -18 LUFS for your podcast… AND you didn’t have to change any of your normal mixing techniques to get there.

Adobe Update

[Editor’s Note: Added May 27, 2016]

A quick update. It turns out the wonderful folks at Adobe have updated Audition CC since I wrote this article. They’ve changed the name of the tool from “Match Volume” to “Match Loudness” AND added a nice adjustable True Peak limiter. It’s perfect for the needs of anyone needing to get a -24 LUFS radio mix to -18 LUFS for podcast distribution!

If you have the newer version, here are the steps you should follow:

  1. Open the Match Loudness window. 
  2. Open Match Loudness
    Open Match Loudness
  3. Drag your file from the File window to the Match Loudness window.
  4. Click on Match Loudness Settings to open the setting window and choose “ITU-R BS.1770-3 Loudness” from the dropdown.
  5. Choose ITU
    Choose ITU
  6. Select the “Use True Peak Limiting” box if it’s not already selected.
  7. Make sure the other settings are as so:
  8. Adjust Settings
    Adjust Settings
  9. Click Run!

Thanks to Paul Figgiani for pointing me to the new version.

There are other tools recommended in one of Jeff Towne’s podcasting articles at Transom, like Auphonic, which is definitely worth a look. Many of these tools replicate the exact same steps we just did — but in fewer mouse clicks. Also, if you can afford it, iZotope offers a couple of products that will do this step for you very easily, iZotope RX and RX Loudness Control, and their processors sound great. I highly recommend them.

I’ll leave you with one more GIF:

Loudness GIF2

Comments

Your email address will not be published. Required fields are marked *

*
*
Website

  • Paul Figgiani

    5.24.16

    Reply

    Rob,

    Good article. I’d like to share a bit of subjective insight:

    Well produced audio that may have been previously mixed or Loudness Normalized to -24.0 LUFS may not be well suited for significant added gain offsets. It depends on the inherent dynamics. For example, an -18.0 LUFS target will require a hefty amount of gain when the source is -24.0 LUFS. If the audio is highly dynamic, limiting will be significant. Even more so when targeting -16.0 LUFS. In some instances dynamic range reduction *before* the final gain offset (regardless of how it is applied) will optimize the the audio resulting in less aggressive limiting. In fact, in my view Spoken Word audio slated for mobile device consumption (Podcasts, etc.) will exhibit improved intelligability when the dynamics are carefully managed. I’m surely not suggesting over-compressing. The documented steps above can sometimes result in hard limiting to maintain your suggested maximum Peak Ceiling. In my opinion this should be avoided. OTOH controlled dynamics before the final gain offset will most certainly reduce the amount of limiting that will be necessary to meet the suggested target(s).

    -paul.
    @produceNewMedia

  • Flawn Williams

    5.25.16

    Reply

    I agree with Paul that hard limiting is not a complete solution here. Some moderate compression of individual elements can improve the results. I tend towards low ratios and low thresholds for those compressors, as opposed to just smashing the peaks.

    One possible factor for why there would be a need for separate LUFS standards for public radio broadcast: almost all radio stations have some kind of signal processing (including dynamic range reduction) between the audio program source and the transmitter. But file-based deliveries such as podcasts don’t have that intermediary between the producers and their listeners. And radio station staff should note that, when they are preparing online files for their station’s site, they need to know if the audio they are posting has or hasn’t gone through their station’s ‘air chain’.

    • Rob Byers

      5.25.16

      Reply

      Flawn,

      I agree with you on both counts! The workflow proposed above assumes one has already mixed their piece to -24 LUFS using good audio practices (see: http://transom.org/2015/the-audio-producers-guide-to-loudness/ ). That would include managing excessive peaks with either level automation (rubber band/volume automation) or compression. Managing levels manually in this way will yield the best sounding mix.

      The hope is that producers can use the above steps to simply get a mix that is already mixed to -24 LUFS (and uses good practices) bumped up to -18 LUFS for podcast use. The step of limiting outlined here *should not* be used in place of controlling the dynamics of individual elements within a mix. (Paul, I think this answers your concern as well.)

      Yes, yes, yes – stations needing to be aware of air chains and workflow when choosing a pickoff point for this process! Also, I hope I addressed your comment re: broadcast processing in the article.

      Thanks,

      Rob

  • Flawn Williams

    5.26.16

    Reply

    Yes, Rob, you addressed that nicely. I somehow missed it in my first scan of the article.

    Thanks to you for this piece, and to Jeff Towne for the rest of the series, and to everyone who pushed through the difficult process of getting these standards created over the past few years to tame the ‘Wild West’ conditions of earlier times.

    I’m now working on production for directly-distributed podcasts, as well as mixing for some public radio shows, and it’s a mind-bending (and monitor-bending!) effort to develop the two different sets of production techniques to satisfy both camps.

    For people venturing into using compression on individual tracks, it’s good to be aware of where the compressor fits in the flow of audio. In Pro Tools, for instance, the ‘Clip Gain’ volume controls affect the audio before the insert compressor plug-ins, while the track volume automation timeline changes the volume after the compressor. Each can be useful in achieving a consistent mix, but the effect on the sound is different. In Adobe Audition you can select pre- or post-volume insert points for the compressor.

    Also, Rob, welcome back to NPR! Minnesota’s loss is DC’s gain. But your work on this topic while working for each of those organizations has benefitted both of them, and the wider audio community as well.

  • Scot

    9.09.16

    Reply

    Fabulous article. Thank you very much.

Your email address will not be published. Required fields are marked *

*
*