Difference between revisions of "MTD2 class 9"
(→Aliasing) |
(→Quantization) |
||
Line 9: | Line 9: | ||
* Quiz Three Next | * Quiz Three Next | ||
** The quiz will cover [[week digital theory 1]][[Audio Digitization 2]] and chapter 3 from Sound Design for Interactive Multimedia | ** The quiz will cover [[week digital theory 1]][[Audio Digitization 2]] and chapter 3 from Sound Design for Interactive Multimedia | ||
+ | |||
+ | ==Three types of digital errors== | ||
+ | |||
+ | * Quantization Error - Difference between actual analog value and quantized digital value | ||
+ | * Aliasing - Undersampling | ||
+ | * Jitter - Time errors | ||
+ | |||
==Quantization== | ==Quantization== |
Revision as of 15:50, 28 March 2006
Contents
In class
- Review and discuss rough movie edits.
- Digital Theory 2
- Listen to Beethoven exmaples (Sample Rate and Bit Depth)
- Discuss Digital Errors
- Look at sampling circuit
- Quiz Three Next
- The quiz will cover week digital theory 1Audio Digitization 2 and chapter 3 from Sound Design for Interactive Multimedia
Three types of digital errors
* Quantization Error - Difference between actual analog value and quantized digital value * Aliasing - Undersampling * Jitter - Time errors
Quantization
Quantization Signal Processing
Discrete levels based on the number of bits
Analogy
Analog is like a bucket of water while digital is like a bucket of ball bearings. Analog has an infinite resolution while a digital system can only resolve thins to finite values.
"...if sheets of typing paper were stacked 22 feet, a single sheet of paper would represent one quatization level in a 16-bit system. In a 20-bit system, the stack would reach 349 feet. In a 24-bit system, the stack would tower 5592 feet in height - over a mile high." -Principals of Digital Audio, Ken C. Pohlmann McGraw Hill
- The more bits you have the more dynamic range you have. Each bit is equal to aporximately 6 more dB. 16 dB s about 120 dB of dynamic range.
Quantization Error
Quantization error = the difference between the actual analog value and the quantized digital value.
The less bits that you have to quantize a signal the larger the quantization error will be. More Bits = Less Quantization Error
Quantization Error is alos known as Signal
Dither
Dither = noise added to signal
addition of a small amount of white noise to the signal to reduce signal-to-error distortion ...also aid in processing and bit depth conversion... see dither article
Dither, Article Sampling Frequency/Aliasing
Get smapling gif sampling1.gif (39564 bytes)
Nyquest Theorem -- The sampling frequency determines the high frequency limit.
Sampling frequency = discrete time frequency of the sampler
44.1 kHz = CD audio
square wave @ 20KHz won't make it w/ a 44.1 kHz sampler
Aliasing
Aliasing = error caused by under sampling -- cars wheels in movies
Need a brick wall filter at the imput of the samplig circuit with a cut off frequency of 1/2 the smapling rate
Jitter
Error due to clock synchronization...More on this next time
Signal Flow
low pass filter (ideally a brick wall filter) --> AD Converter --> DA converter --> low pass filter
pcm.gif (23735 bytes)
MORE
READ dither Article
http://www.digido.com/ditheressay.html
READ Jitter Article
http://www.digido.com/jitteressay.html
Digitizer
Premiere Audio Demo
How to make Sound forge default editor for Premiere
Flash Tracing Demo
How to trace still images in flash
Homework
- Work on Audio for Storyboard movie
- Post another version of the exported movie
- Quiz Three Next week