Phys 1240: Sound and Music LAST: wind instruments TODAY: Jeff Merkel on “sound in the studio” NEXT: Wrapping up - wind instruments, voices, sonograms, …
Digital Audio
No new reading! No more assignments after CAPA #9 (tonight)
Recording, Mixing, and Delivery
Final Exam : Sat 7:30-10. See web! Chris and David’s office hours accordingly rescheduled: Chris Tuesday (usual time, 2-4) David Tuesday (usual time, 4-6)
• Microphone Transduces sound, turning acoustic pressure waves into an electrical signal in a very similar way as our ears • We now can record voltage as a function of time. • We’ve seen this voltage vs. time many times now, which is a waveform. • But how do we actually record?
Recording Lots of different ways to record digital data, but the most common full-fidelity is PCM, or Pulse Code Modulation Audio CDs, .wav, .aiff A/D (Analog to Digital) conversion Basically “connect the dots” on a grid system Amplitude
“Capturing” audio waves with microphones
Time
1
Digital Sampling We now need to decide how often to “sample” the voltage signal.
time
Concept Question 1 I have a 220 Hz wave and a 1000 Hz wave. Which one needs a higher (more samples per second) sample rate to be resolved? A)220 Hz B)1000 Hz C)They use the same
Nyquist Theorem Nyquist’s theorem states that an analog signal wave form can be reconstructed without error from samples taken at equal time intervals. The sampling rate must be greater than or equal to twice the highest frequency component of the analog signal.
Concept Question 2 If humans can generally hear up to 20,000 Hz, what rate must the sound be sampled at to achieve the full frequency range? A)10,000 Hz (10 kHz) B)20,000 Hz (20 kHz) C)40,000 Hz (40 kHz) D)No freakin’ idea
2
Species
Approximate Range (Hz)
human
64-23,000
dog
67-45,000
cat
45-64,000
cow
23-35,000
horse
55-33,500
sheep
100-30,000
rabbit
360-42,000
rat
200-76,000
mouse
1,000-91,000
gerbil
100-60,000
guinea pig
54-50,000
hedgehog
250-45,000
raccoon
100-40,000
ferret
16-44,000
opossum
500-64,000
chinchilla
90-22,800
bat
2,000-110,000
beluga whale
1,000-123,000
elephant
16-12,000
porpoise
75-150,000
goldfish
20-3,000
catfish
50-4,000
tuna
50-1,100
bullfrog
100-3,000
tree frog
50-4,000
canary
250-8,000
parakeet
200-8,500
cockatiel
250-8,000
owl
200-12,000
chicken
125-2,000
Concept Question 4 Common CD’s have a sample rate of 44,100 Hz. What is the highest frequency that a CD can accurately play back?
Concept Question 3 For a dog that can hear up to 45,000 Hz, how high must the sample rate be for it to hear full fidelity? A)22,500 Hz B)45,000 Hz C)90,000 Hz D)No idea
Bit Depth We now need to work on the y-axis of our grid
A)22,050 Hz B)44,100 Hz C)88,200 Hz D)Not a clue
3
The Binary System In the digital world there’s only 1’s and 0’s The Binary series: 0, 1, 2, 4, 6, 16, 32, 64, 128, 256,… 1-bit can be 0 or 1 so there are 21 combinations. 2-bit
00 01 10 11
0 1 2 3
4 combos, 22
3-bit
000 001 010 011 100 101 110 111
0 1 2 3 4 5 6 7
8 combos, 23
16-bit = 216 = 65,536 combinations 24-bit = 224 = 16,777,216 combinations
This now gives us a number association with the y-axis
32,768 2n
Concept Question 5 n = number of bits
0
-32,768 0 Because waveform amplitudes are analogous to displacement from zero, let’s adjust the scale so that the center is 0, and the outer bounds are + and – of half the number of combinations. So for a 16-bit system, there are 65,536 combinations. The upper bound corresponds to 32,768 and lower to -32,768
A 24-bit digital audio system has 16,777,216 combinations. About how many steps are there between the zero point and one of the outer bounds? A)70,000 B)8,400,000 C)17,000,000 D)34,000,000 E)Way too many to count So a 24-bit system represents a 256x improvement over a 16-bit system!
4
What happens if you go beyond the maximums? There simply isn’t data outside the outer bounds. Clipping of the wave occurs.
What if the wave doesn’t exactly correspond to a number on the y-axis? This is called Quantization Error. Because the wave is slightly distorted from the original, fidelity is lost.
Results in nasty sounding clipping in the form of severe harmonic distortion. Also, any high frequency data on the crests of the waves is lost.
We now have a “plot” of our x and y axis. How much data are we using up? For CD quality audio, which is 16-bit, 44.1 kHz: 16 bits x 2 (stereo) x 44,100 Hz x 1byte/8bits = 176,400 bytes/sec
176,400 bytes/sec x 60 sec/min =
Dithering introduces randomness when lowering bit depth which actually increases perceived smoothness.
Mixing Each instrument/mic is recorded to it’s own track. The streams of audio are then processed and combined. Mixing multiple tracks down to a stereo two track compilation is called “mixing down”
10,584,000 bytes/min, or 10.6 Mb/min
5
Time to make a CD! CD’s are
Mastering Mastering is the art of compiling all the songs into a single project and creating the actual master that will be mass replicated. Make the music sound good on many different playback systems. Listening to music in terms of frequency. Quality control issues.
Let’s make CD’s! Compact Discs work on the same binary principals as regularly stored data, but done in an optical fashion. Pits diffuse the laser while nonpits reflect.
Compressed Data Audio MP3, Ogg Vorbis, WMA, etc. Work on the psychoacoustic phenomenon that there’s a lot of stuff you can’t hear that can be removed. Masking is the main principal behind it. Masking in time and frequency.
6
Surround Formats Dolby Digital DTS SDDS
5.1 5.1 7.1
THX – a quality standard developed by Tomlinson Hollman/George Lucas.
7