d:notes by Derek Leverington - Web 2.0, Technology & Advertising

« Creative Zen Micro - Near Miss | Main | Ask Google? »

April 7, 2005

Podcasting - Usability, Technical & Audio Production Suggestions

With every new channel comes the challenge of making it easy to use.

With podcasting being something of a new phenomenon and often being undertaken by those without training or experience in the area of audio production or experience in usability design, I thought it worthwhile to give some thought to some suggestions for making the content as consumable as possible.

Paying attention to some basic considerations can go a long way to improving the audio quality of a spoken word recording.

Appropriate Use of Meta Attributes
(with the common implementation for many being ID3 Tags in the mp3 file format)

Ensure you use the ID3 tag. In the absence of this, the player will only display the file name which may not be particularly communicative of the content of the podcast and may be named around entirely different considerations than its use as an plain-English description.

Recognize the character limitations of the playback devices and avoid unnecessarily long titles. In many players, longer titles will need to scroll like a horizontal marquis to reveal the entire length of the title, and at that, only when selected in the menu. So, as users will likely view several podcasts in a list, ensure the differentiating part of the podcast name (the individual program name) compared to others in the series appears relatively early in the title - saying within 20 or so characters. As a result, users can scroll through the list quickly without having to select a specific podcast name and wait the see the entire title scroll through the display to read the podcast name.

Not to pick on the For Immediate Release podcast produced by Shel Holtz & Nevill Hobson, whose program I enjoy very much and am a faithful listener of, but for the purposes of illustration and accessibility of the example to many, I'll use the For Immediate Release naming conventions as an example.

They typically are titled according to this naming convention:

For Immediate Release -- #19 -- March 28, 2004
For Immediate Release -- #20 -- March 31, 2004
For Immediate Release -- #21 -- April 4, 2004

In my player (Creative Zen Micro 5 GB) the screen cuts off after the '#' character, making it difficult to scroll through the list quickly and differentiate one podcast in the series from the rest.

The recommendation I would make is to choose a succinct podcast name that affords some characters left over to describe the individual program. This is a matter of design for the constraints and realities of the medium.

Ensure Absolute Consistency In Naming
Be sure that your podcast names are consistent to the character from one podcast to the next. As they will appear in a list in the player, any aberrations are quite noticeable and detract from usability and the visceral sense of quality of your production.

Audio Considerations

One of the benefits of mainstream radio is the level of production quality that we've become used to. Between the voice talent, writing (though not always), production proficiency and quality of equipment not to mention experience, the bar is far higher that what I would expect from many podcasters.

I expect that to hold true at this stage in the life of podcasting where the impetus for one to podcast likely has more to do with the desire to communicate than possessing technical proficiency and access to quality audio equipment. With many mp3 players having onboard microphones that can write directly to a digital file that can easily be posted as a podcast, the barrier to the lowest point of entry is quite attainable. That said, achieving a technically satisfying production and desirable content is another matter entirely.

Avoid Problematic Locations To Record
The key to being effective in this endeavour from a technical standpoint is clear and intelligible audio. Anything other than the sound of your voice works against clarity and ease of listening(unless in some instances the ambient noise of your surrounding is an important element of the content or creates a mood).

Avoid locations where there are surfaces nearby that will reflect sound back to the microphone. Flat, painted or shiny surfaces are the worst offenders for this. They will reflect sound back and give a few-millisecond echo throughout the whole recording. The use of soft materials can dramatically improve the acoustic characteristics of a space, so don't discount using bedding or blankets to improve the acoustic qualities of where you record. I may not be typical of every user, but I find a close-reflection echo quite annoying and distracting and it's a tell-tale sign of amateur production.

What's worse is that audio compression (discussed further down) which normally has an enhancing effect on a production only accentuates the presence of this type of audio problem.

Use Headphones When Recording
If you using discrete audio components i.e. not simply speaking into an mp3 player, consider wearing headphones so you can monitor what you record. You will inherently tend to speak in a manner that is more pleasing to the microphone and you will become aware of any obvious technical problem like the aforementioned slap-back echo and hopefully correct it.

Invest the Time To Experiment
Experiment with recording locations, equalization, microphone position, pace of the dialogue and your own technique for projecting your voice. A few minutes invested in this will absolutely make a difference in the quality of your recording. A few cumulative improvements will add up to a better overall production.

Maintain a consistent volume
Dramatic changes in the level of amplitude is something to be cautious of as many people listen to podcasts in their headphones. As such, dramatic amplitude changes are exaggerated and can be quite jarring for the listener.

If there is more than one person speaking, attempt for a balanced audio level between the two. Also, try the match as closely as possible the tone or tambre between the people speaking through equalization as recorded to tape or in post-production. If there are dramatic differences in tone or volume that result in one voice being louder or more easily heard, the user has to make a compromise when setting the volume of their player. If a user has to have the volume set higher due to the presence of ambient noise in their own surroundings, the level that they have to set it at to hear the quieter of the two people speaking may make the louder voice uncomfortably loud or annoying.

Compression
These days, almost all recordings have a appreciable amount of audio compression applied to them to. Among other considerations, this is done to enhance the recording to make it sound a rich and loud as possible on playback. FM stations also dramatically compress when broadcasting.

For those not familiar with audio compression, it's the process of reducing the amplitude (volume) of the peak levels in a recording. As a result, the overall level can be boosted resulting in the average volume of the recording being louder. If you listen to a compressed and an uncompressed piece of audio, even though the peak levels may be the same, the compressed audio can sound much louder than the uncompressed audio.

One of the benefits of this as a listener is that is provides a consistent volume level with no spikes or drops in volume.

If you have access to audio editing software, look and see if you can find a compressor to apply to your recording. I would suggest no more than 2 - 3db of gain reduction in most cases. After a certain point the voice no longer sounds natural.

A note of caution: using compression will elevate the level of background noise, mouth sounds and other ambient nosies that are not always desirable.

Other Potentially Useful Audio Processes
De-esser - to reduce sibilance - those ripping 's' sounds that again can be quite dramatic in ear buds
High-pass filter - this will filter out low frequencies - try a setting of 120 Hz to start to reduce popping p's (a pop filter or wind screen will also help with reducing popping on more explosive consonants . . . as does practise speaking on-mic)

And of course, it never hurts to have something worth listening to - this just makes it easier.

This is by no means comprehensive as I expect to add to this as I give this further consideration(although I didn't expect it would be this long to begin with on a first draft). I believe if these things were followed and podcasters were armed with some basic knowledge of audio production, we would be able to see the technical quality of podcast improved quite dramatically.

Many of these things listed are not difficult to accommodate. Where software is required to audio processing, it is not particularly expensive either. Any modern audio editor comes with standard audio tools to accomplish equalization, compression, de-essing, etc.

Posted by Derek Leverington at April 7, 2005 12:28 AM

Comments

Posted by: Lee Hopkins at May 3, 2005 5:02 PM

Posted by: Derek Leverington at May 5, 2005 12:46 AM