Podcasts as any other media resources could potentially exclude several users from learning, if they are inaccessible.
TechDis, the JISC funded UK educational advisory service, working in the fields of accessibility and inclusion, have produced a very comprehensive list of things to consider when using podcasts for educational purposes:
http://www.techdis.ac.uk/index.php?p=3_10_13_3
Give people as much choice as possible in the ways they can access the podcasts e.g. via RSS, direct link to MP3 on web page, Itunes, all the main podcast directories.
Make sure the page that offers information and links to the podcast conforms to W3C guidelines http://www.w3.org/WAI/
use the best quality recording equipment available and maximum quality recording settings during production. Keep background noise to a minimum during any 'vox pop' type interviews and avoid background music when people are speaking. Never allow people to talk over each other in group discussions. Keep the final mix pretty central to both left/right channels.
always produce a transcript from the final version of the podcast and post it on the website.
If possible, offer various quality versions (i.e. file sizes) of the podcast (low, medium, high).
Wherever practical, consider producing both video and audio podcasts, with the video podcast including subtitles and use of images and diagrams to help convey complex information.
The above 3 sections have been recently circulated by Dr. Ian Francis (University of Dundee)
If you pre-script your audio, then transcripts are already ready for your podcasts.
Alternatively, you may need to manually produce transcripts. However, it has also been suggested, that a succinct summary of the digital audio or video file is more useful for the listener/viewer than a detailed transcript. So, when your segments are short and used as stimulus for further discussion or thought, you could instead include the main points discussed in the text area of your message in LUVLE.
If you prefer to use voice recognition software to produce transcripts for podcasts that have only one speaker, Dragon Naturally Speaking is an option. Please note that the Dragon software needs to be trained beforehand in order to pick up your accent. Although it is very accurate most of the time, you may need to allocate time checking the transcript and correcting any errors.
If you want to add captions to your video file, you could use software such as MAGpie (free) or Hi-Caption (if you want the final video file to be streamed). However, neither of them support captioning mp4 files. The most common supported files are: avi, wav, rm, wmv, mpeg, mov. The exported files can be played by either Windows Media Player, Real Player or Quicktime. For more information please read:
Captioning with MAGpie 1.0: http://www.webaim.org/techniques/captions/magpie/version1/
Captioning with Hi-Caption: http://www.webaim.org/techniques/captions/hicaption/
Camtasia (a rich media presentation software) can export your video file with captions in m4v format which is playable by Real Player, iPod, iTunes, Quicktime, etc.