Fundamentals of Digital AUdio Podcast Episode 35 - Recording at 24 bit

You are missing some Flash content that should appear here! Perhaps your browser cannot display it, or maybe it did not initialise correctly.

Remember the Fundamentals of Digital Audio is part of the Home Recording Network. New shows in the network include Record Better Audio and Making Music with Mac OSX which is an extension of a book.

Recording speech at 24 Bit

Why CDs are 16 bit? Isn't that goood enough . Well for output yes. For recording stepping up to 24 bi has advantages

Levels and headroom

What happens when converting an anolog audio signal to digital

First, sample rate - how many smaples are taken per second 44,100 is the standard for cds and most audio Film and video sound often at 48,000 hz.

Resolution how many ones and zeroes assigned to each sample 16 bits or 16 ones or zeroes for each sample. 16 bit gives us over 65,000 potential values for each sample. 24 bit resoultion gives us millions of potential values for each sample. So the first advantage is a more accurate sample

The second advantage is more practical in many ways levels

Each bit gives us 6db of dynamic rane

16 bit 96db of dyanmic range

24 bit - 144db dynamic range (difference between the highest and lowest values)

Cell phone audio typically 8 bit I believe - less accurate.

Remember we want to give our levels a boost either with a voice processor on the front end or in software later. This gets our signal above the ambienbt noise the listner may be in while listening. But, each 6db we boost throws away 1 bit of resoluton and gets us closer to to cell phone sounding audio

24 bit allows us to record under "with headroom". Shoot for peak levels of appprox. -8 to -12db. This allows us to be safe and avoid clipping or exceeding 0db at all times which is a must. If not, you'll get terrbilble-sounding digital distortion

We can boost by 6 or 12 db or more - actually we could boost by 48db and still have as much resolution as a 16 bit recording. Now, boosting by 48db would be too much and clip the audio but the point is it gives plenty of headroom for bossting levels and compression. It makes post proceesing much easier

Our final output can be downsampled to 16 bit and output to mp3, mpeg 4 audio for web distrubtion

What if my hardware or software doesn't support 24bit

Easy part first software. Some software such as GarageBand only supports 16bit. You can simply get software that does including the free Audacity.

Hardware - if your hardware doesn't record at 24 bit, the best thing to do is put a hardware compressor in front of your recorder to get the levels boost on the front end. You still need to be safe and make sure you don't clip. Utlimately you may have to sacrifice 1-2 bits of resolution which isn't fatal for speech and podcasting - haeder to live with for music

Question feed back - a bit technical this time. If you'd like more explanation email me at digiaudio.podcast@gmail.com