I really, honestly do not know the answer to this question. What is the best bit rate to encode a podcast at? Also does that answer depend upon the fact that you are a listener or a podcaster or hosting service?
I do the post production work for the Global Geek Podcast. Before moving to TPN I always encoded the podcast at 44khz and 96kbps. That works out at about 35 – 40MB per show (depending on length between 40 minutes to an hour). We have what I think is great audio quality, but am I spoiling ourselves and our listeners and potentially excluding others?
We have never had a complaint about the file size of the show. No-one has ever said it was too big. People have commented on the quality and said it is great and we have worked hard to get it that way. But I now question if that is over kill. So I tried to figure out what bit rate is the most common. I did a very small survey of the podcasts I have on the computer. I only have nine on it at the moment – most of them are on the MP3 Player (where they should be).
Anyway I got the following breakdown:
Total of 9 Podcasts:
- 2 encoded at 96kbps
- 4 encoded at 64kbps
- 3 encoded at 48kbps
A conclusive survey that does not make. But maybe I am aiming too high. What quality do listeners expect of a podcast? Do they want a small file and lower quality so that they get the content without the bandwidth. Or do they want great quality and a larger file size? With the size of MP3 players now the storage is not an issue I don’t think. But I know in Australia the cost of bandwidth might be. The cost of faster connections is expensive and so many users are on a maximum of 256/64 or 512/128. So does a larger file size deter them from listening to our show? Could we have a bigger audience if we made it smaller and if that is the case what size is acceptable?
With the uptake of broadband technology there is a step towards encoding at a larger bit rate but what should it be? Perhaps 64kbps is a good place. I listen to quite a few podcasts that are recorded at 64kbps and they sound good. A one hour podcast encoded at 64kbps is about 28MB (voice only). Is this a big difference to 96kbps? Well it is between 10 and 15 MB. Will that mean the difference between more listeners and a balance between keeping your existing ones because of what they expect? Will you loose listeners by lowering the bit rate dramatically and will it matter because of the number you pick up. To me it does anyway, I care that we keep the listeners we have.
The other big consideration here is the hosting cost. I know that I had to go to the plan one up from the basic plan in order to have the podcast encoded at such a high bit rate. So that privilege cost me $10US/month instead of $5US/month. That was a cost that I thought was worth it. Also what if your podcast is being hosted by a network, what file size is reasonable for them to host? Is it acceptable that you have a higher bit rate than the other shows that are hosted there and is it necessary? Personally, I would like to find a happy medium between file size, bit rate and quality. I want the best quality at a reasonable file size. I don’t want my hosting provider to get pissed off that the show is too large. In addition to that fact; the network wants as many people to listen to as many shows as possible. If it is possible that people are “turned off” by a large file size, then that is not for the benefit of the network and I would not do it. In that instance the file size should be smaller at the sacrifice of quality for the benefit of the network and I need to accept that.
As a listener I do not care what size a file is. I have a fast Internet connection and it really does not bother me. I like high quality podcasts but I listen to some that are not of a high quality as far as bit rate because the content is good. So is good quality a cover for shit content? If it is; it is not sustainable long term. So as a listener of podcasts I don’t search for podcasts based on audio quality or file size, and maybe I have just answered my question in part.
Having made these points I will say that some basic editing will improve quality out of sight. I have turned off podcasts because they have not bothered to do this basic editing. They were unlistenable and total shit and they should have thought the same! I wonder if some podcasters even listen to it after they have recorded it. So what do I mean by “basic editing?”
Basic editing in my opinion is:
- Setting levels before you start, especially if you are recording Skype using a software application. This means setting your levels with enough “headroom” to get loud during a podcast so that you don’t “clip” the recording. And not so soft that you have to amplify it dramatically to get something to work with.
- Don’t edit the podcast as an MP3, MP3 is a “lossy” format and gets worse and worse in quality every time you re-encode it or open it and save it.
- Run a compressor on the audio to “smooth” the audio. That is take out the high’s and bring up the lows.
- Run the compressor a few more times.
- “Normalise” the audio, basically set the zero level. Makes the podcast the same volume and means that the listener isn’t constantly turning their volume up and down.
- You may need to “amplify” the whole audio after using the compressor and normalising the audio. You don’t want the listener running out of volume because it is too soft!
- Any added or imported audio needs the above steps.
Believe it or not the above takes the least amount of time in my editing but makes the biggest difference. I do go a step further and edit the actual audio and take out the umms and errs and we always stuff things up and say well we will edit that out. The time is also in the transitions and the mixing of the imported audio, making it all work together (the best that I can). So maybe you can see why as a podcaster I want it to sound as good as I can, I put a lot of effort into both the pre and post production. But is that at the neglect of other issues? Is this basic and advanced editing enough to make it a “quality” podcast?
Please leave a comment and tell me what you think. Tell me if you are a listener or a podcaster. Podcasters, tell me what you encode your podcast at and why. Listeners please answer my questions for me. As I said at the start of this post I really do not know what the right answer is, that’s why I have posed lots of questions. It would be great to get some answers, although I am not sure there is one.
Odeo
I know that I rant about the uselessness of Internet Explorer and the virtues of Firefox. That has not changed but it would seem that Internet Explorer is more broken now than it ever has been.
Rooster's Rail 
Secret Mikons Business
October 29, 2006 — The RoosterMikons that I wrote about some time ago now is a site that allows you to create and share what they call “mikons”. Mikons for everyone else could be called avatars or a logo. Indeed you could use their “mikon machine” to produce a personalised logo just for you or your blog or anything really. Read all about Mikons either on my post dedicated to the site or take a visit, register and start creating, sharing and now commenting on your own mikons!
This is another piece at mikons that will make for a more enjoyable user experience. The development team are working hard to give users a solid platform. Not a gigantic leap for them but a step and I really believe in mikons as being a unique site offering something that people really want. I have created a number of logos on the site and I think that is a powerful direction they could go. One of them is this blog’s mascot the beloved Rooster! Well I like him.
Spsst… Hey you, yes you… There is something very exciting about to happen at mikons.com. Sorry but I am not allowed to say anything yet. But it is big, yup really big! As soon as I can I will let you know. Hopefully you will hear it here first. It is even better than being able to comment. I hate it when I have to keep secrets, but then again I could just drop a hint… no no no, damn it. More soon!
What would make your mikons better?