Every month I record an audio newsletter podcast for Zen Internet. This podcast is an audio version of their monthly e-mail newsletter that they send out to the majority of their customers. During the recording process however, things don’t always go according to plan!

Unfortunately, for the last two months I have been unable to record the monthly podcast, as I have been stuck in bed with ill health. Hence, I have not got quite as many bloopers or out-takes this year, for your your seasonal amusement, as in previous years.

Added to this, Winnie (our usual voice) has recently had a little baby; so with her new family addition she was not able to attend every recording session. Thankfully, Sue stepped into the breach on a these occasions to save the day, which helps account for the different voice with the South African accent.

I hope you enjoy our bloopers and manage have a little laugh at our expense…

Podcast Bloopers 2011

The Internet Monthly – Podcast Bloopers 2011:

Download audio file (bloopers_11.mp3)

Or, download: bloopers_11.mp3 [2’ 14”, 1.1MB]

Thank You

As 2011 drifts into 2012 and we start the yearly cycle again, I would like to thank you for visiting my blog and listening to the podcast and hope to see you back here again in the New Year.

Merry Christmas and Happy New Year!

I record and produce a regular monthly podcast for Zen Internet and have recently started producing a podcast for a gardening website that I help run with a friend. I have also produced a guided audio tour for a local town (which is produced just like a podcast). All of these podcasts I produce using a selection of free software products that I wouldn’t do without.

Free vs. Commercial

Free software is usually quick to load and very focused at achieving its specific aims. Conversely, commercial software often incorporates masses of features, most of which you are never likely to use but have to pay for nevertheless. This inevitably makes these larger, commercial programs much slower to load.

While I predominantly use the free software outlined below for all of my podcast recording and production needs, I do occasionally use commercial software for other audio editing tasks. However, I find these programs a bit big and cumbersome for my regular podcasting requirements.

Acquiring the Audio

No matter what software you choose to help create and produce your podcasts, you will first have to get your audio into the computer. How you choose to do this is entirely up to you, but is likely to depend on what level of audio quality you want and your budget. At the most basic level you could use your computer’s in-built microphone, but for a reasonable degree of quality you will want a semi-decent external microphone.

1. Audacity – Audio Recorder and Editor

Once your audio is in the PC in the digital domain you will need a piece of software to record and save the audio. Audacity is perfect for this task. I use Audacity for capturing all of the audio that I use for podcasts and general voiceover work.

Once recorded, you will probably want to edit the audio to clean it up and prepare your final podcast. If you want to produce a more complex audio production, I wouldn’t use Audacity in preference, but something more like Cubase or ProTools. However, for your basic audio editing and podcast preparation, Audacity is absolutely perfect. I use Audacity in preference to the bigger, commercial audio editing programs for precision editing of audio to remove clicks and pops etc. I find Audacity excels at this and is just so easy to to use.

2. WinLAME – Audio Encoder

Once your audio has been recorded and edited (preferably in WAV format) you will need to convert it into a suitable podcast format. The universally accepted audio file format for podcasts is MP3. WinLAME is an excellent piece of software designed to convert WAV files into the podcast friendly MP3 format.

While you can actually save files in MP3 format from directly within Audacity, I find WinLAME offers a much finer degree of control of the MP3 encoding process and also allows you to batch process a load of audio files in a single go.

N.B. Do not be tempted to save audio files you are working on in MP3 format, as every time you open and resave them you will loose audio quality. Save your files in uncompressed WAV format and only convert them to a compressed format like MP3 at the very last moment.

3. Paint.NET – Image Editor

You might well be wondering why I have included a piece of image editing software on my list of top 5 pieces of software to produce podcasts. Well, every podcast should ideally include an album cover that will be displayed in your player of choice, be it the ubiquitous iPod, or your computer’s media player. To produce this artwork, or to reduce existing artwork to a reasonable size (I would recommend 300 x 300 pixels), you will require some image editing software and Paint.NET is a very powerful and capable piece of software for the price (Free).

4. Mp3Tag – ID3 Tag Editor

Now that you have your podcast MP3 audio file and some suitable artwork for its cover, you will need to embed the artwork into the MP3 file and add some additional tags that can be displayed by the listener’s playback device. These “meta” tags can be used to explain a little bit about your podcast, e.g. category, year of release, copyright notices, author, track title etc. All of this additional information is embedded in your MP3 file using ID3 tags. To accomplish this and give your podcast that extra professional touch you will need some tag editing software.

Again, while Audacity allows you to perform simple editing of ID3 tags when saving MP3 files, a dedicated tag editor, such as MP3TAG, is much easier to use, more comprehensive and will allow you to batch process files as well as supporting many different audio file and tag formats.

5. FileZilla – FTP Client

Now that you have successfully recorded and edited your audio, converted it to MP3 format and added some cover artwork and descriptive meta tags, the final thing remaining is to upload it to your website. For this you will require an FTP client; not a glamorous piece of software, but a fundamental utility to have when working in the world of the web. A good choice for this is FileZilla, again a remarkably powerful FTP client which won’t even break into a sweat uploading your latest monolithic podcasting creation.

Happy Podcasting

So there you have it, the top 5 pieces of software that I regularly use to record and produce my podcasts and ones that won’t break the bank:

1. Audacity – Audio Recorder & Editor
2. WinLAME – Audio Encoder
3. Paint.NET – Image Editor
4. MP3TAG – Audio Tag Editor
5. FileZilla – FTP Client

If you have any favourite pieces of software that you use for creating and producing podcasts, please leave a comment below, I’m always on the lookout to improve things.

On the face of it, laying an acoustically reflective surface as a flooring option would seem a particularly stupid idea if editing and recording audio in the room was going to be one of its key uses. However, as is so often the case in life, there were other factors competing in the equation that demanded consideration and therefore necessitated the now irksome wooden floor.

The Need for Change

The choice of “purple” decor in my study was that of the house’s previous owners; not mine. After living with it for a number of years, I decided it was time for a change. As I spend a significant period of time in my study “nerding”, I chose to go for a clean, modern look and feel.

Prior to commencing the project, I measured the area in question and drew a 3D model of its basic layout using Google’s amazingly cool free 3D modelling software: Google SketchUp (see image to right).

Once I had the basic shell constructed, I then played with a few colour schemes and flooring options. My study only has one small window, making it quite dark at the best of times, so the original purple wallpaper was ditched in favour of clean walls painted in a light colour.

Flooring Options

My study was previously carpeted with a dark purple carpet, which while it made the room quite dead sounding, also darkened an already dark room.

I quickly decided on laminated wood as my new flooring of choice for a number of reasons. Firstly, I liked the look of it and felt it would help give the study a contemporary feel and create a more pleasing environment to work in. Secondly, it was light coloured, which would help lift the gloomy impression of the dark room. Finally, it was more durable and easy to clean, which was an important factor, as unfortunately I am confined to a wheelchair following an accident in July 2000.

Despite my convictions that I had made the right choice to use wooden flooring, I remained nervous of its acoustic implications; potentially making the room very reverberant. On an encouraging note however, all of the recording studios and audio editing suites that I have visited in the past all had wooden flooring. So hopefully my decision wouldn’t come back to haunt me.

Secret Plan

Once I’d made the decision to use wooden flooring, I then started to consider the possibility of placing acoustic tiles on the wall directly behind me (when I was sitting at my desk). This was the largest wall surface in the study, so hopefully some large panels would be able to mop up some of the reverberant clutter.

Typical acoustic panels are not the prettiest things in the world, particularly when you’re considering using them in an essentially domestic environment. However, with a little research I discovered that you could get acoustic panels with artwork printed on acoustically transparent fabric. Something that would look good, plus be functional. Result!

The first company I came across (Auralex) had an impressive range of artwork options for their Sonic Print panels, but nothing that really “worked” for me. I eventually settled on a company called GIK Acoustics after having seen a sample image of a guitar spread across 3 panels on their ArtPanel web page. Perfect, particularly as I was a guitarist in a former life.

Another hiccup however was that their standard panels were  2 foot x 4 foot, which would have been a little too tall for the wall I planned to put them on (a dado rail was in the way). Fortunately the guys at GIK Acoustics were very accommodating, and after much deliberation we managed to get the artwork to fit in a manufacturable way on a 2 foot x 3.5 foot frame.

Acoustic Performance

The acoustic art panels contain 2 inch thick mineral wool as the acoustically dampening material. While the manufacturer didn’t have acoustic performance results for the art panels, they did for their 4” panels. The results are apparently very similar, but with the thinner 2” panel not being as effective at the lower frequencies, rolling off at about 250Hz as opposed to 70Hz for the 4” panels. This isn’t too much of an issue for me as most of the audio work I do is speech based, which doesn’t tend to contain a large amount of energy in the low frequencies.

The End Result

When my study decoration project was initially completed (without the acoustic panels) the reverberation times in the room were quite high, making any recordings sound rather echoey. However, when the acoustic panels were installed the difference in sound was amazing; initially it sounded almost like an anechoic chamber to my ears as I’d become so accustomed to the live sounding room.

The acoustic art panels have certainly been very effective in taming the room’s reverberation, probably now down a normal level or even a bit below. I’m sure for true professional standards more would be required, but they’ve done exactly what I wanted and look really cool too.

I think I’ll live with the room as it is for a while before I rush into any decisions regarding further acoustic treatment, but as things stand I can highly recommend acoustic art panels as an ideal solution for a domestic environment: nice artwork plus acoustically functional.

Once a month I record a podcast for Zen Internet that is dedicated to the latest news and developments in the internet world.  I subsequently syndicate this podcast on my blog with Zen’s kind permission as The Internet Monthly. However, things often don’t always go smoothly during the recording process!

As with most things in life, things don’t always go to plan and recording podcasts is no exception to the rule; most of the time recording sessions progress relatively uneventfully. However, every now and then Gremlins creep into the system and the resulting mistakes of out-takes and bloopers are kept for posterity and for your seasonal amusement. Enjoy…

Podcast Bloopers 2010

The Internet Monthly – Podcast Bloopers 2010:

Download audio file (bloopers_10.mp3)

Or, download: bloopers_10.mp3 [3’ 30”, 1.6MB]

Thanks

As 2010 fades into 2011 and another year begins, I’d like to thank you all for following my blog over the last year and listening to the monthly podcast.

Keep tuned for more of the same next year.

Merry Christmas and Happy New Year!

The Internet Monthly is a podcast dedicated to the latest news and developments in the internet, produced for Zen Internet by myself, and subsequently syndicated on my blog. However, things don’t always go according to plan!

Recording podcasts is just like anything else in life; there are good days and bad. During our monthly recording sessions for the podcast, mistakes are invariably made, some of which I kept for posterity. As such, I have put together a short collection of these out-takes and bloopers for your seasonal amusement…

Podcast Bloopers 2009

The Internet Monthly – Podcast Bloopers 2009:

Download audio file (bloopers_09.mp3)

Or, download: bloopers_09.mp3 [3’ 14”, 1.5MB]

Thanks

As the final days of 2009 fade away, I’d like to take this opportunity to say thanks very much to you all for following my blog over the last year.

Hopefully the New Year will provide me with further inspiration for new and interesting posts in the month’s to come… we can at least hope!

Merry Christmas and Happy New Year!

Album or cover artwork can easily be embedded within MP3 files via ID3 tags. However, the ID3 specification caters for over 20 different image categories. So which is the best image category to choose, and is it ever worth embedding more than one image into your MP3 files?

Since the release of version 2 of the ID3 tag specification over ten years ago it’s been possible to add album artwork to your collection of MP3 files. This feature of the ID3 specification is now pretty well supported. Most MP3 players with a screen capable of displaying graphics (be they software or hardware  based), now support this feature.

However, as usual with these sort of things, MP3 player manufacturers tend to implement the ID3 tag standard in subtly different ways. This inevitably leads to unpredictable results for the user, particularly when displaying album art.

Embedding Images into MP3s

If all this talk of ID3 tags leaves you a bit cold, then it may be worth pausing for a moment to read my previous post on ID3 tags in MP3 files before proceeding, as they are fundamental to our further discussions.

Embedding images into MP3 files is fairly straightforward and indeed might be accomplished automatically for you by your CD ripping software. If this is not the case, then you will need a special piece of software called a tag editor to help with the task. There are lots of tag  editors available for a host of operating systems, with some good free ones available. My personal favourite is Mp3Tag.

If you want to know more about adding cover art to your MP3 files, the best image format to use and what image size you should consider using, then check out my previous post on embedding album art in MP3 files.

ID3 Artwork Options

The present ID3 standard (version 2.3.0) lists 21 different images that can be embedded within an MP3 file. In practice only one image is ever usually embedded, although an MP3 track could technically have 20 images (only one of the icons #2 or #3 can be embedded at a time). However, the space taken up by the images in such a file would become quite significant compared to the audio data.

The images available to choose from are:

1. Other
2. 32×32 pixel icon (PNG only)
3. Other file icon
4. Front Cover
5. Rear Cover
6. Leaflet page
7. Media (e.g. label side of CD)
8. Lead artist, Lead performer or Soloist
9. Artist or Performer
10. Conductor
11. Band or Orchestra
12. Composer
13. Lyricist or Text writer
14. Recording Location
15. During Recording
16. During Performance
17. Movie or Video screen capture
18. A Bright coloured fish
19. Illustration
20. Band or Artist Logo
21. Publisher or Studio Logo

There are two notable exceptions in the above list that are a little different from the others:

• The second option (#2) only allows images in the PNG format and images which are sized 32 by 32 pixels for use as a small icon.
• The eighteenth option (#18) in the list; a bright coloured fish! Some loony tune obviously decided that this was a good idea for a formal specification. Need I say more?

So, for all essential practical purposes, the list really has 20 options (ignoring the brightly coloured fish), or if you are just considering images of unrestricted size, then it has 19 options.

Support for ID3 Images

MP3 Players

Not all MP3 players are created equal when it comes to displaying ID3 images. For example, the Creative Zen Vision: M, while having a large colour screen capable of displaying pictures, does not seem to display any of the ID3 images as far as I can tell; no matter which image tag is used.

The basic Creative Zen displays the ID3 image from the last track in a folder, but insists on using this same image for all tracks in the folder, no matter whether they have different artwork embedded within them or not.

Software Players

Windows media player seems to display any ID3 image correctly, but providing there’s only one image in a track. If a track contains more than one image, it will default to the Front Cover image. Quite how you’re supposed to see the other images, I really don’t know, but at least this seems a sensible fall back position.

Podcasts

A lot of podcasts seem to use the Other image tag (#1 from the above list) for their artwork instead of the Front Cover option. I’m not really sure why this is, other than that their software defaults to this option. My preference would be for the Front Cover tag in this circumstance, but I’d be interested to see what the general consensus is on this.

Test your Player

Having tried a few players and found them to be wanting in their handling of ID3 images, I decided to do a structured test on those I had available to me.

Test Images

In order to test the various ID3 image tags on a range of players, I produced 21 different images corresponding to each of the image tags. Each image is a 300 x 300 pixel jpeg (except the 32 x 32 pixel PNG icon):

If you’d like to use these images yourself for your own experiments, you can download all 21 as a zip file. Each individual image is approximately 25kB in size:

• ID3-Images.zip (246kB)

Test MP3 Tracks

In order to test these images on an MP3 player, they each needed to be embedded into an MP3 file with the appropriate ID3 image tag associated with the image.

My tag editing software doesn’t support the 32×32 icon image tag or the nonsensical fish tag, so I ended up creating 20 separate MP3 files: the 19 basic image tags plus one MP3 containing all 19 images in the one track.

The MP3 track including the 19 separate images in the one file ended up with a resultant file size of 911kB for a 26 second 128kbps MP3!

• 21 – ID3 Image – All Images.mp3 (911kB)

If you’d like to use these MP3 files for yourself to test your own MP3 player’s performance in this respect, you can download all 20 as a zip file:

• ID3-Image-Test-MP3s.zip (2.5MB)

Report Back

If you’ve found these files useful and end up discovering some quirks with your particular player, please leave a comment so that others can benefit from your findings.  You never know, it may spur a manufacturer into updating their firmware to fix the problems and make their players fully compliant for future users.

Whenever you play a podcast on your portable MP3 player or PC, you will invariably see a host of information about the track on the player’s screen. This information typically includes the cover art for the track, the album, artist and track name, its genre and track number.

Such information is stored in the MP3 file itself using ID3 tags. However, in addition to the tag information in the example above, the ID3 tag standard can support over 80  different tags. So, out of this vast array of tags what are the key tags that you should consider using in your podcasts?

What are ID3 Tags?

ID3 tags are a simple and convenient way of embedding descriptive data about an MP3 track within the track itself. This descriptive data is sometimes referred to as metadata, which is data describing other data. The ID3 tag data describes the audio data of the MP3, e.g. the track title, artist name etc..

For a fuller and more in-depth description of ID3 tags, check out my previous post: What are ID3 Tags in MP3 Files?

Essential Tags

Downloaded podcasts will already have ID3 tag data buried within them, but if you produce your own podcasts, you have free reign to add whatever ID3 tags you wish. However, you could fill up your tracks with 50 different tags or more, but who is this really going to benefit? Not you for certain; it’ll probably take you longer to add all of these superfluous tags than it did to record the podcast in the first place.

Brevity is the key. Just include the bare minimum that users are likely to need, any more and you’re wasting your own time and wasting space in the MP3 file storing useless junk.

The tags serve two purposes. The first is to provide general information about the podcast that the user is likely to want to know, while the second is to enable the user to search through a large collection of tracks at a later date and hopefully filter out your podcast via its tags.

Track Title

This is undoubtedly the most important tag, being the title of your podcast episode and is a fundamental tag that should not be missed.

By way of example, the title I use for my monthly podcast, is the generic title of the podcast with the month and year appended:

The Internet Monthly – June 2009

Cover Art

Another important tag to consider is the cover art tag. As the saying goes; a picture speaks a  thousand words, and this tag gives you the option to brand your podcast in glorious colour, forever etching it in the user’s memory.

However, while this tag will help to connect the user with your podcast, it’s unlikely to serve much use when trying to search for the podcast later on.

If you’d like to learn more about embedding images into your podcasts and what image types and sizes to use, then check out my previous post on: Embedding Album Art in MP3 Files.

Artist

This is the producer of the podcast. For example, I record a monthly podcast for Zen Internet to a script that they provide, so I add their name as the artist, as it’s essentially their creative work:

Zen Internet Limited

Album

While this tag is designed for use with music albums, where it refers to the album which an individual track belongs to, in podcast terms it is commonly used to refer to the series name of the podcast (assuming the podcast is not a one off, in which case this field would be left blank). For my monthly version of the Zen Internet podcast, I use the following album tag:

The Internet Monthly

Track

This isn’t often used in podcasts unless you have a concurrently running sequence of episodes, in which case you could use this field to denote the episode number.

Year

This is the year in which the podcast episode was first published. This field is probably not of much interest to the user under normal circumstances, but can prove useful later on when searching through an archive of tracks to narrow down the search:

2009

Genre

Again, this tag is not of much use other than for categorising the track for archiving and searching purposes. For podcasts, the de facto standard entry for this tag is unsurprisingly podcast, although sometimes you may come across alternatives of speech or vocal.

Interestingly podcast is not defined in the list of genres in the official ID3 tag specification, even though it is common practice to use this as the genre name for podcasts. Both speech and vocal are officially recognised as legitimate genres however.

I decided to go with the flow and hopefully the specifications will catch up with conventional wisdom:

Podcast

Comment

This field is often used for show notes and additional information pertaining to the podcast. Some people use this field for copyright notices, although if this is your intention, their is a specific copyright tag field dedicated to the purpose.

As my company (Silicon Bay Limited) produces the monthly podcast for Zen Internet, I use this field as a slight promotional opportunity:

Produced by Silicon Bay Limited

Copyright

Instead of using the comment field for your copyright information, this is the correct tag for the purpose:

Copyright (c) 2009 Zen Internet Limited

WWW

Invariably your podcast will be associated to your website, so why not embed your website’s URL into your podcast? This way if the podcast gets distributed via other means than your own website, you’ll have a way of driving traffic from interested listeners back to your site:

http://www.richardfarrar.com/

In Summary

The key ID3 tags that you should consider using in your podcasts, that will give the most benefits to your listeners without overloading them with information are:

• Title – The podcast title
• Cover Art – Thumbnail of the podcast art
• Artist – The artist that recorded the podcast
• Album – Which series the podcast belongs to (if applicable)
• Track – The track number from the podcast series (if applicable)
• Year – The year the podcast was published
• Genre – The type of track, e.g. podcast, speech, vocal
• Comment – General comments
• Copyright – Copyright notice by the copyright holder
• www – The website address for the podcast

Manufacturers like to develop and retain technologies that gives them a commercial and technological advantage over their competitors. To this end we as consumers have ended up with a collection of competing, but unfortunately technically incompatible, solutions to the same problem; that of storing large amounts of audio on portable audio players.

Thankfully the majority of manufacturers (but by no means all) of these portable audio players have eased our potential incompatibility problems by making sure that their players can play the majority of compressed audio formats that are in common usage today.

So what are these common compressed audio formats, why do we need them and what are the associated problems with having so many different formats?

Why Compress?

Back in the dark ages of computing, not so long ago, computer memory was not so plentiful, or cheap as it is today. As such, storing large amounts of data, such as audio files, was very costly.

As necessity is the mother of invention, the limited quantities of computer memory available prompted the white coated technical boffins to develop fiendishly cunning ways of throwing away most of the audio data from these files, while still retaining a high level of perceived audio quality.

How they do this is rather complicated, but they essentially save space by not storing the bits of the music that we can’t hear. During compression of a 128 kbits/s MP3 file, typically 90% of the audio data is discarded, leaving a file size of one tenth of the original size. This in turn means that your MP3 player is able to store ten times the amount of songs compared to if it had to store the original uncompressed CD data.

MP3

MP3 (MPEG-1 Audio Layer 3) is by far the most well known out of all of the audio compression systems available today, and while it may not be as technologically advanced as some of the newer systems, it is the most widely used and provides perfectly adequate results for moderate bit rates for the majority of people’s requirements.

The major benefit of encoding your audio files in MP3 format is that you are pretty much guaranteed you’ll be able to play your resultant files on almost any modern media player or device. It is for this reason that most podcasts choose to adopt the MP3 format as their preferred compression system.

While most people think that MP3 is a free and open source format, it’s not! There are numerous patents owned by a host of different organisations that are applicable to the MP3 technology. This can make official licensing of MP3 technology for use in encoding software or playback devices quite tortuous. Out of all of the organisations that lay claim to the technology underlying MP3s, the German Fraunhofer Society is probably the most notable.

AAC

AAC (Advanced Audio Coding) is actually part of the MPEG-2 and MPEG-4 specifications and is designed to be a superior technological replacement for the older MP3 format. AAC is designed to provide improved sonic quality and transparency compared to MP3 files encoded at the same bit rate, although this advantage only becomes dominant for lower bit rates (below 128 kbits/s).

AAC is the default format that Apple have adopted for their iPods, iPhone and iTunes and for their iTunes online music store.  Sony has also adopted AAC as the standard audio encoding system for their PlayStation 3 and has been incorporated into their PlayStation Portable and recent Sony Walkman systems. Other companies adopting this format include Nokia and Nintendo (Wii and DSi).

WMA

WMA (Windows Media Audio) is Microsoft’s proprietary audio compression system built into its Windows operating system and is the default compression option used for ripping CDs to your computer using Windows Media Player.

Whilst WMA is a proprietary Microsoft system, the technology has been licensed by numerous third parties meaning that if you’ve ripped your entire CD collection onto your computer’s hard drive in WMA format (as I have), then you have a good chance of being able to play these WMA encoded audio files on a host of compatible devices.

I have a Creative Zen MP3 player, which plays WMA files without any problems and I also have two Netgear MP101 media players at home linked to my home network that can stream music from my server in WMA format.

ATRAC

ATRAC (Adaptive Transform Acoustic Coding) is another proprietary audio compression format, this time developed by Sony, originally for their MiniDisc systems in the early 1990s, but has since been used in the company’s range of portable audio players.

Since its inception, ATRAC’s compression algorithm has been improved over the intervening years, with the newer versions offering improved perceived audio quality over earlier versions for similar bit rates.

Unfortunately, ATRAC encoded files are not widely supported by other manufacturers and as such, if you have a Sony branded audio player and have ripped all of your CD selection into ATRAC files, you’ll be extremely limited as to where else you can play your audio files.

Licence Free

Ogg Vorbis

While the Ogg Vorbis audio compression format is not hugely popular outside of the geeky computer nerd circles, it is worth a brief mention as a number of commercial “MP3  players” support the format. More importantly though is the Ogg Vorbis system is FREE!

There are no patents or licence issues to worry about if you choose to adopt this system, although compatibility with other media playback devices may prove to be an issue.

FLAC

FLAC (Free Lossless Audio Codec) is another audio compression system worth a mention, partly because it’s free (as its name suggests) and partly because it’s lossless.

All of the above compression systems loose some audio quality in their compression process. FLAC does not, but the downside is that the same sort of compression levels are not achievable.

FLAC files tend to be between 40 to 50% the size of the original audio file, which is not that impressive compared to a typical 10% figure of an MP3 file.

Like anything in life, it’s a balancing act. If you can afford the additional storage space required and want original quality audio, then FLAC is the way to go. For the majority of people however, these tiny improvements in audio quality compared to the space saving benefits of any of the above compression formats will not be worth the penalty of compatibility issues.

Summary

While MP3 is the most common and well known of all the audio compression systems, it’s by no means the best. There are a multitude of different audio compression systems available, with just a handful of the most common ones mentioned above.

The choice of compression system can depend on many factors, but for most people, providing the audio quality is reasonably good and the file sizes acceptably small the overriding issue will probably be compatibility.

Thanks to the advent of digital audio and the power of modern PCs, it’s now well within the capabilities of virtually anyone to produce and publish their own radio play using the same technologies as used in podcasting.

As a small kid I remember producing plays with my toys, an upturned stool for a theatre, a few torches and the tales of Peter Rabbit by Beatrix Potter as a script. Since then my interests graduated to music production; I played in numerous bands and have always been keen on recording bands as a hobby.

However, my recent visit to an audio post production film editing suite at Pinewood Studios, re-sparked my interest in plays (I’ve always been a fan of the BBC’s radio series: The Hitchhiker’s Guide to the Galaxy) and got me to thinking about using the same equipment that I use for podcasting, but pushing it’s use a bit further.

All I needed was some creative inspiration, but in it’s absence, I did the next best thing and borrowed it from someone else…

Inspiration not Perspiration

In the beginning was the word and the word was written by John Davison; the word was Bananas.

John has grandchildren, five in fact, but it’s his two youngest grandchildren that were the catalyst for this radio play.

John would often tell his grandchildren bedtime stories, which he usually made up as he went along. However, after a while, John began to formalise his bedtime stories around a young gorilla named Bananas as the central character.

Having developed a few adventures of Bananas, test driving them on his grandchildren first, John published some of his stories on his personal blog, with the first story published being Just Another Happy Day At The Farm!

To accompany the stories on his blog, John flexed his creative muscles even further, drawing the main characters of his stories for visual enhancement.

Some of the themes for his stories were actually suggested by his grandchildren, giving him a creative head start. The story that I used as the basis for my radio play was one of these; George Helps Bananas and Farmer Jones.

Story to Script

Having the story of Bananas was the starting point for the radio play. However, it was only after a few glasses of wine on a dark and damp winter’s day at John’s house that the reality of turning this story into a radio play began to take shape.

A few glasses of wine later and the story of Bananas and George had been turned into a script for our fledgling radio play.

Recording

The recording of the radio play was done in stages, depending on people’s availability, with the recording setup being essentially the same as I use for recording Zen Internet’s podcast every month.

Character Voices

The first voice to be recorded was Winnie’s. Winnie recorded the narration, but also recorded the voices of Bananas and Mrs Jones.

The second voice to be recorded a few weeks later was that of John himself, who played Farmer Jones. During this recording session, John and his wife Connie also assisted in recording a few sound effects for the play, namely the stirring and the drinking of the hot chocolate featured in the play and the pouring of the bird seed.

The final character voice for the play was provided by my son James, who played George.

The only other voice for the play was that of the announcer, provided by myself.

Technical Setup

To record all of the voices I used a Rode NT1-A low noise, wide diaphragm condenser studio microphone (shown left) with a pop shield to stop the microphone “popping” with loud sounds beginning with Ps and Bs.

The microphone was also mounted on a “cat’s cradle” shock mount to help prevent vibrations from the microphone stand passing to the microphone.

The voices were recorded into the PC via a Focusrite Saffire firewire interface. During this process a little gentle compression and equalisation was added to the voices via the Saffire’s on-board DSP.

For those of you NOT interested in the nerdy technical details, then please skip the next geeky bit:

Compression: A compression ratio of 2.5:1 was used with a 10ms attack time and 350ms release time. The threshold was set to -30dB to give a compression of up to 6dB during peaks.

Equalisation: A high-pass filter set to 90Hz was used to filter out low end bumps and pops. A –4dB high frequency shelf filter was used, set at 16kHz to take a slight harsh edge off the sound. 6dB of boost was added at 200Hz with a Q of 1.7 and 6dB of boost at 4kHz with a Q of 1.2.

All recordings were done at a sample rate of 44.1kHz (CD sample rate) and were recorded as uncompressed WAV files to retain their maximum quality.

Editing and Production

Recording the voices was only a small piece of the puzzle. Now the rest of the pieces had to be assembled; only they didn’t come with instructions!

To assemble all of the recorded pieces, a computer based Digital Audio Workstation (DAW) was used. My weapon of choice was Steinberg’s Cubase 4 LE that came bundled with my Focusrite Saffire audio interface; an amazingly powerful bit of software considering it was effectively given away free.

The picture below shows a screen shot of the main editing window in Cubase. Using this window I assembled the various audio pieces into their correct order and colour coded different groups of tracks for easy navigation. The intro and outro music pieces were coloured orange, voices were yellow, foley effects were pink, ambient and animal noises were purple and all other sound effects were in green.

With all of the tracks in place, additional audio processing was performed with some compression, equalisation, de-essing and noise gating.

The final mix was then created using the software’s powerful mix automation features to remember the fader (volume) positions, pan (left and right stereo positions) and reverb settings throughout the mix.

With the final mix completed, the resultant WAV file was converted into an MP3 file encoded at 128kbps, joint stereo to produce a fairly compact file for downloading, while retaining a reasonable level of quality.

The final tweaks to the MP3 file were the addition of various ID3 tags to include the track title, artist etc. and the embedding of the cover art of Bananas into the file itself.

The Play Itself

So, you’ve heard how it was produced, but what about the final result?

Well, judge for yourself, here it is:

Bananas Helps George:

Download audio file (bananas-helps-george.mp3)

Or, download: bananas-helps-george.mp3 [5’ 21”, 5MB]

By now you’ve probably heard of the term “bit rate” used in connection with MP3 tracks and podcasts, you may even know that higher bit rates give better quality audio. However, like most things in life, nothing is simple and we can further complicate the whole bit rate discussion with variable and constant bit rates. So, what are the pros and cons of these constant and variable bit rates?

Choosing a bit rate for a podcast, or an MP3 track, is a bit of a balancing act at the best of times. On the one hand, higher bit rates (variable or constant) give improved audio quality, whilst on the other they lead to bigger file sizes.

Introducing yet another variable into the equation, that of constant or variable bit rates, only serves to confuse the situation even further. What then are the relative pros and cons that may affect your choice of type of bit rate to use?

Balancing Bit Rates

If you’re not familiar with bit rates, then my previous post on What are bit rates? should point you in the right general direction and help to get you started.

In essence, higher bit rates give improved audio quality, no matter whether variable or constant bit rates are used. However, higher bit rates will always lead to larger file sizes, which in turn lead directly to longer download times and require more storage space.

When choosing which bit rate to use for your podcast or MP3 track, you really need to go for the highest bit rate possible, while retaining reasonable file sizes and practical download times.

Accepting the fact that you need to choose the highest practical bit rate you can, which option should you go for next (assuming your encoding software gives you the option) constant or variable bit rates?

Constant Bit Rates

Constant Bit Rates (CBR) are pretty much as their name suggests, where one fixed, or constant bit rate is used for the entire file or track.

However, constant bit rates are not optimal as they waste valuable data bits encoding simple sections of audio, and conversely may not use sufficient data bits on more complex audio passages, resulting in poorer audio quality during these sections.

The more complex audio sections of tracks encoded with a constant bit rate will be of a lower audio quality compared to the more simple sections of audio within the same track, so a suitable bit rate must be chosen that provides sufficient audio quality for the most demanding sections of the audio being encoded.

Constant bit rates are easier and therefore faster to encode compared to variable bit rates and lead to consistent and predictable file sizes. Being an older format, constant bit rates are more compatible with older MP3 players.

Variable Bit Rates

Variable Bit Rates (VBR) are likewise pretty much as their name suggests, where the bit rate may change throughout the file or track being encoded.

Variable bit rate encoding analyses the audio file, adapting the bit rate as it goes to match the complexity of the audio it encounters.

For simple audio passages, the system automatically turn its bit rate down, saving valuable data bits, while retaining the perceived audio quality. Conversely, during more complex passages, the system automatically turns its bit rate up, using more data bits to retain the same level of audio quality.

Using this method, the average audio quality of a file can be increased as the available data bits are used more flexibly to encode the sound data with increased overall accuracy.

When encoding an MP3 track or podcast with the VBR method, the encoding software usually allows you to decide on the overall quality of the resulting track that you desire, after which the system encoder uses the bits required to achieve this defined quality level. Higher quality levels will require more bits.

Compared to constant bit rates, variable bit rates produce smaller file sizes for a given audio quality. However, as the bit rates are variable, it’s difficult to predict the size off these files in advance.

Variable bit rates are harder and therefore slower to encode compared to constant bit rates, as the adaptive encoding process is more complex an processor intensive.

Being a more recent encoding method, some older MP3 players or computer software may not be compatible with variable bit rate encoded MP3 tracks or podcasts.

Constrained Variable Bit Rates

A Constrained Variable Bit Rate (CVBR) is a flavour of variable bit rate, where the maximum bit rate allowed, or the Average Bit Rate (ABR) are predefined.

For example, encoding an audio track with the bit rate set to be an average of 128kbps should produce a resultant audio file of higher perceived quality than the same track of encoded at 128kbps CBR, as the ABR file can use data bits it saved during the simpler passages to add extra data bits (above the 128kbps average) to the more complex passages to improve the overall audio quality.

Which to Use and When

Compatibility

Generally speaking, encoding an MP3 track or podcast using a constant bit rate will be better for the sake of compatibility compared to using a variable bit rate, as some older MP3 players don’t support variable bit rates.

By way of example, in their guidelines for broadcast media, the BBC recommends the use of constant bit rates for their all of their podcasts and downloadable audio files.

Quality

If compatibility isn’t a significant issue for you, then you will probably be better off using a form of variable bit rate of encoding to squeeze the best audio quality that you can out of the available bits.

Whether you choose to use a constrained variable bit rate, so that you can keep a handle on the size of file you’re likely to get, or go for a fully variable bit rate, is entirely up to you.

However, before fixing which method you’re going to use, it may be best to do a few test recordings using different bit rates and bit rate types, before deciding on which suits your purposes best.

In Summary