MP3s For Podcast

[marindakim]

Hi,

One of my clients has asked me to create MP3s for a podcast. I work primarily as a composer and in audio post for video so my experience with MP3s is limited and doesn't extend to podcasting. In addition to any general observations you might have, I have a couple of questions I hope you guys might be able to comment on.

I expect to work with mono files recorded in the field at 44.1. When using the MP3 Option what bit rate/ sample rate should I select? Also, is there a maximum or 'safe' limit as to how large MP3s for podcasting should be? For example, a mono MP3 of a 30 or even 60 minute speech? And do you recommend normalizing the files to 0dbfs prior to exporting them or should I normalize to some other value?

Finally, have you had any experiences with ID Tag Typing that you can pass along?

Thanks - I'd appreciate any recommendations you might have.

Marindakim

[Eric L]

Hi Marindakim,

First off, what type of material will be recorded for the podcast? This will determine what bit rate and number of channels you may want to use. You say in your post you expect to use mono files. So let's assume we are talking about human speech only and not music.

If this material is only for true podcasting that will be listened to on an Apple Ipod, then you should use the ACC codec as you will get better results at lower bit rates than the traditional mp3 codec. The only drawback with this is that non Ipod players can not listen to the pod cast.

Since a pod cast is technically not a realtime broadcast, size limits are not essential. The user downloads the pod cast file to there ipod and plays it back locally, not from a server. So bandwith is not a major concern. That being said, you do not want your download to take forever or no one will want to wait for it.

Knowing all that, Mono, 44.1, 128k AAC is a decent set up for human speech without it sounding all alien like. Yes, you should normalize to the top (-1dbfs) as this is like radio, not tv or film. You should definitely use the ID tag info to add copyright information and any other details you would want the listener to know. This is easy and can be done in Itunes. As a matter of fact, Itunes is a great application for all this and can make standard mp3 files as well. The same specs for ACC should work just as well as for mp3. The AAC will just sound better. If you are adding any music to the pod cast I would use a minimum of stereo, 44.1, 192k, AAC. Do some tests for yourself. The difference between bitrates and codecs is very noticable. good luck.

[marindakim]

Eric,

Thanks so much! That's a big big help. I'm sure non IPOD users will also be downloading these files so I guess I'll stick to using the MP3 Option unless there are other reasons you know of that mitigate against using it; I gather from your answer that MP3s created there will play across more platforms. And I'll run some tests comparing bit rates from that application. I'll look for the best compromise between downloading speed and sound quality as I anticipate some rather long speeches.

Thanks Again!

Marindakim

[pneyrinck]

Regarding ID3 tags, you need to decide between you and your client what exactly you want to be displayed for Name, Artist, Album, Composer, Genre, and Comments fields in an MP3 player. Because podcasts are often not music, those fields can be used different ways. You should look at various podcasts that you want to possibly emulate and see what they do. And after you encode, you should verify it looks like you intended in iTunes and Windows Media Player. Finally, be sure to keep the person who will create the RSS feed in the loop. The RSS feed is XML text that has fields in it that should match the ID3 tags you encode.

Paul Neyrinck
www.neyrinck.com

[marindakim]

Hi Paul,

Thanks for your suggestions. This is great info I can and will use!

Marindakim