These are my notes from a presentation by Will Birchett of the University of Oklahoma on “Digital audio alternatives” at the ODLA 2007 conference.

What is digital audio
- bits and bytes
- not analog
- when an analog audio signal is converted into electrical on/off pulses

everything in analog audio is voltages
- digital system takes those sound waves and does an approximation to figure out how to reproduce that with bits and bytes

with digital, there are two standards agencies

ITU-T
- G.711
- G.722
- G.728
- G.729

ISO/IEC
- MPEG-1 Layer 3 (this is commonly called mp3 format)
- MPEG-2 Layer 3
- MPEG-4 AAC

standard mp3 player can’t play dolby surround
- mpeg-2 layer 3 does have 6 channels
- most digital audio now going to MPEG-4 AAC, allows for up to 64 channels

mpeg-1 layer 3 is 2 channel
mpeg-2 layer 3 is 6 channel
mpeg-4 AAC is 64 channels

how is the conversion done
- a little black box!
- an A2D chip: analog to digital
- converts voltage to binary data at thousands of ___ per sec

digital is an approximation of an analog wave
- there is no exact representation
- conversion: tries to “best guess”

analog to digital hardware: laptops, laptop USB devices, USB microphones

has a box that takes 14 channels into a firewire interface

I have a 32 channel recording studio

when you are doing musical recordings, a digital recording, you don’t mix your musicians
- you want to record each musician separately
- this lets you post-process on your computer
- don’t have to get the whole band sounding good at the same time
- can record at different times, and get the best performance from each individual, that can even be on different days
- this is taking out the best bits

audio recording software, can lay 5 separate takes side by side, and get the best from each
- with analog, time coding is essential, to time align, it pretty hard to do this
- recording studios used to do this with reel to reel tapes
- digital has taken recording and made it where more people can produce awesome content

3 basic concepts

1- bit depth
- referes to the number of bits that you have to capture audio
- 16 bit = 65,535
- 24 bit = 16,777,216

the more bits you have, the more information is there

a phone line is 8 bit
- all the phone providers boost that up to 64 bit, still an 8 bit line
- the more more bits you put in, the better picture you can get with digital audio

- those numbers graph out from the lowest bass to the highest treble

bit depth tells you how good of a recording you can get

laptops, many portable recorders are 16 bit

a 4 bit wave doesn’t have a lot of data
- look at the graph of where the approximation meets the actual analog wave

Going from 16 bit to 24 bit is more than a hundredfold increase in values that are sampled / quality
- you don’t see 24 bit in videoconferencing or television
- it is more for creating your MASTER audio file
- takes a LOT more room for storage

2- sample rate
- refers to the number of times the audio is measured or sampled per second
- how many times it takes a digital slice of that analog signal
- you take more samples more often, you get a little better
44.1 is the most popular sample rate, that is used on all CDs for music (44,100 samples per second)
88.2 is double that, done for video editing and motion pictures
HD audio is 96,000 samples per second

most interfaces don’t give you that info

[WHEN YOU GET INFO ABOUT A MP3 OR OTHER AUDIO FILE IN ITUNES, YOU CAN SEE THIS INFO]

3- bit rate

how much data per second is required to transfer the file
- bit rate = (bit depth) x (sample rate) x (# of channels)

examples
- telephone: 8 byte bit depth x 8 kHz = 64 kb bit rate
- CD audio 1.411 M bits/second (million bytes per second)
- MP3 could be 192 kilobit per second

CDs and MP3s are completely different in terms of bit rate

2 ways of recording a digital file
- pulse code modulation (PCM) – WAV files
- create bit data when there is no audio
- that is why WAV files get very big, very fast

MP3 takes same data, figure out when there is no sound, and then cut that out
- when there are higher and lower frequencies with no data, that is cut out

[SO THE BIT RATE IS WHAT WE COMMONLY REFER TO WHEN WE TALK ABOUT ENCODING A MP3 FILE: AT 128, 196, etc. Will's explanation of what a mp3 is and how it works

FM radio is 64 kilobits per second
- some of the high ends are gone, not as crisp
- compressed down, less bit depth

at 32 KBPS the audio REALLY sounds different

number of frequencies in the human voice is much less than music
- so most spoken podcasts are published as 32 kbps mono

so the comparsion of CD audio to mp3 is;
- representative mp3 bit rate: 128 kbps
- CD quality bit rate: 1,411 kbps

[I STILL NEED TO UNDERSTAND THE DIFFERENCE BETWEEN BITS AND BYTES. I KNOW THIS IS BASIC AND I'VE READ IT BEFORE BUT I DON'T REMEMBER THE DIFFERENCE AND CAN'T EXPLAIN IT TO SOMEONE AT THIS POINT.]

Will demonstrated p

bit rate is really chosen by your application

podcast just needs 32 kbps mono
- your voice is NOT stereo
- recording a speech: just need mono recording, and 32 kbps recording
- these other sample rates are for musical instruments

biggest misconception with digital audio: I’ll just choose a higher bit rate and my quality will increase
- that is just going to capture more frequencies, and is going to boost your

32 kbps mono you can get about 10 MB for 1 hour
- 80 minute CD is 720 MB

when you do digital audio, set your

most of the audio recording industry doesn’t capture 24 bit
- most captures at 16 bit

another side of the house says capture

downside to variable bit rate
- if you want to edit that
- VBR files are VERY hard to edit
- that is like taking a VCR, recording on extended play, then recording on standard play, then trying to import that into a video editor: hard to do

I have digital sound. now what?

storage types
- optical disk (CD, DVD, HD-DVD)
- digital audio tape (DAT)
- hard drive
- flash memory

storage options are a lot greater with digital
- analog: reel to reel and magnetic tape were the only options, and over time those media forms would self-corrupt

AES standard (EBU)
- 2 channels of audio, 24 bit depth, twisted pair wiring, XLR connections
- used in digital broadcasting

analog audio is voltage, so hums are caused by RF radio, RF mics, AC power, the power field from those overlaps itself onto the voltage of the analog audio, causing quality drops (hum)

S/PDIF
- Sony Phillips Digital Interface
- that is in all home audio gear
- 2 channels of audio
- consumer version of AES/EBU
- 20 bit resolution
- used in home theater

cost of RCA connections for S/PDIF is much less than XLR

ADAT
- used in digital recording studios
- 8 channels of audio
- on 1 cable
- 16 bit resolution
- fiber or DSUB-25
- used in pro audio

Now I have a CD with some tools on it
- 2 folders
- 1st is audio samples with varying bit rates
- I have hyperlinks to different reference sites on wikipedia about digital audio
- also some audio tools and test equipment
- standalone digital recording
- computer-based digital recording
- audio training
- CD software: all free software
– audio recording: audacity, lame mp3 encoder
switch software (is a free version and a pro version)

I use QuickTime pro for podcast recording instead of Audacity
- QT Pro is another tool for converstion

I record music at 192 kbps, but downsample audio to 32 kbps for a podcast

variable bit rate encoding WILL reduce your file size of your final produced podcast

also real-time analyzer software program (Windows only)
- used when designing a system, to see frequencies
- when you are recording something, you can watch what bit rates would be cut off
- you can visually see on the screen audio you are seeing (like bouncing LED bars on a stereo)
- can use that to make a better decision on where to cut your audio and still make it intelligible

last: signal generator
- is a professional noise generator
- people sell these as hardware, but there are also free virtual ones like this (windows and mac versions)

normally when you record digital audio there filters being used
- audacity gives you choices about selecting those

other software he has
- MOTU FireWire Console
- FireWire CueMix Consule
- AudioDesk 2.03
- Will’s waiting on the new version of Logic Pro

Technorati Tags:
, , , , , , ,


Check out Wesley's new ebook, "Mapping Media to the Common Core: Volume I." (2013) It's $15!

If you're trying to listen to a podcast episode and it's not working, check this status page. (Wes is migrating his podcasts to Amazon S3 for hosting.) Remember to follow Wesley Fryer on Twitter (@wfryer), Facebook and Google+. Also "like" Wesley's Facebook pages for "Speed of Creativity Learning" and his eBook, "Playing with Media." Don't miss Wesley's latest technology integration project, "Mapping Media to the Common Core / Curriculum."

On this day..

Share →
  • http://musicisnotforinsects.blogspot.com/ Ken Pendergrass

    Wesley- These are great notes! You take notes like me, especially the notes to yourself. Thanks for the links and a great resource for digital audio alternatives and some explanations about bits, compression and the whole mix of digital audio jargon.
    -Ken
    p.s. just discovered your “eyes right” blog also…

  • Robert

    Great Stuff! I learned a lot, and understand many things better, Thanks!

  • http://faculty-staff Will Birchett

    A bit is a single binary 1 or 0. A byte is a group of 1s and 0s. So looking at bit rate, you see bitrate = bit depth * sample rate * #or channels.

    for CD Audio

    1.411Mbps = 1411200 = 16 bit * 44,100 * 2 (stereo)

    http://en.wikipedia.org/wiki/Bytes
    http://en.wikipedia.org/wiki/Word_%28computing%29

    Hope this helps

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 Unported License.

Made with Love in Oklahoma City