Download Powerpoint notes.

Art 299
Sound, Audio, Acoustics as Creative
components
Dr. J. Parker
Fall 2009
Sound
What we hear as sound is caused by rapid
changes in air pressure!
It is thought of as a wave, but that can be
confusing since we are immersed in the
medium. A wave seems to be on the
surface.
Sound is 3D
Sound sources is at the centre of a sphere.
Pressure changes (higher, lower, higher …) move outwards
from the source.
Pressure and Amplitude
Loudness is a matter of how big the
pressure change is.
If we let normal ambient air pressure be 0,
then sound amplitude is +x to –x for some
way to measure pressure. Air pumps at
gas stations here use Pascals, but we can
use whatever we like.
PHYSICS ALERT !!
1 Pascal = 1 Newton per square metre
Standardized normal air pressure is:
P0 = 2 X 10-5 N/m2 = 0.00002 N/m2
= 20μ Pa.
Why do we care? Because a decibel, the
standard measure of sound level, is:
10 log10
where p is the rms pressure
of the sound.
Who cares?
All sound folks use decibels, or db. You get
to know (feel) how loud things are after a
while.
Threshold of Hearing (TOH)
0 dB
Rustling Leaves
10 dB
Busy Street Traffic
70 dB
Front Rows of Rock Concert
110 dB
Military Jet Takeoff
140 dB
Instant Perforation of Eardrum 160 dB
How loud is it?
The sound of leaves is 10 db.
The sound of a pencil dropping is 20 db.
How much louder is that ?
10 times!
Decibels are based on powers of 10.
If one sound is 10x times more intense than
another sound, then it has a sound level which is
10*x more decibels than the less intense sound.
Frequency
If you stand in one place, the pressure
waves will pass you. The number of peaks
that pass per second is the frequency.
Measured in Hertz (Hz), formerly cycles per
second.
60 Hz = low hum
440 Hz = ‘A’ on piano
4000 Hz = limit of telephone (voices)
Frequency
If we draw a curve that represents pressure
VS distance we can see that sound does
look like a wave.
Frequency
Since the speed of sound in air is a constant
(pretty much) then the number of peaks
that pass is a second is related to the
distance between them is a fixed way.
Also, the horizontal axis could be distance
just as easily as time.
Frequency
V = l/f where v is speed of
sound, f is frequency.
Amplitude
Frequency
Frequency is a precise way to specify pitch.
A 440
E above C
Summed
Oboe A
Oboe E
Phantom
B 3960
Frequency
Interesting sidebar
The speed of sound is different in different
media (air and water). When a sound
moves from one medium to another, part
of the sound is transmitted, part is
reflected back!
Sound can reflect
Multiple sounds
What we hear with multiple sounds is the
sum of all of them.
Sound Systems/Tech
A sound system is a collection of electronic
components designed to record and
display sound. Details depend on
applications:
Theatre: large output, many speakers
Studio: large input, many microphones
Microphone
Converts air pressure differences into
electrical voltage.
Microphone
Speakers
For display of audio.
Amplifier
Changes volume (voltage/current/power)
levels of a signal. Needed to get good
sound levels from a speaker.
Mixer
Control volumes of multiple inputs into
(multiple) outputs.
Complete Sound System
Recording
How can sound be recorded?
How can it be stored on a computer?
What tools do we use for this?
Where do the wires go?
Why am I doing this?
It is still sound if it is on a hard drive?
How much disk space do sounds need?
Recording
To record sound on a computer is very simple.
Recall that microphones change sound (pressure)
into voltage (electricity).
A computer ‘sound card’ can change voltages into
numbers (just as we do: 12 volts, for instance)
which can be stored in a computer memory.
A sound is therefore a sequence of numbers that
represents voltages (that represent pressure).
Digital Sound
How BIG the numbers can be dictate
accuracy of voltage samples ->volume or
amplitude.
8 Bits (binary digits) can store numbers
between 0 and 255 (256 different
frequencies)
16 bits has 32768 different frequencies.
Called depth or quantization.
Digital recording
How fast you play the samples back dictates
frequency.
Obviously one plays them back as fast as
they were recorded. However, more
samples per second means higher
frequencies can be recorded.
Sampling
Sampling
Quantization Distortion
Quantization is the process of selecting whole
numbers to represent the voltage level of each
sample. The A/D converter must select a whole
number that is closest to the signal level at the
instant it’s sampled. This produces small
rounding errors that cause distortion.
Quantization distortion increases at lower levels
because the signal is using a smaller portion of
the available dynamic range, so any errors are a
greater percentage of the signal.
Clipping
File Sizes
Sampling Rate x Resolution x Number of
Channels x Time in Seconds / 8 =
File Size (in Bytes)
44100 x 16 x 2 x 60/8=10,584,000 /minute
Sizes
Bytes per minute, uncompressed.
Sampling Rate Resolution Number of Channels
44,100
16
2
44,100
16
1
22.050
16
1
11.025
16
1
11.025
8
1
File Size
10,584,000
5,292,000
2,646,000
1,323,000
616,000
Recording sound – MAX OS X
http://audacity.sourceforge.net/download/
Recording Sound – Windows XP
Sound recorder
•To open Sound Recorder, click Start, point to All
Programs, point to Accessories, point to
Entertainment, and then click Sound
Recorder.
Audacity
Sound Forge
Sound Effects
http://www.pacdv.com/sounds/index.html
http://freesoundfiles.tintagel.net/Audio/
http://www.a1freesoundeffects.com/
http://www.pachd.com/sounds.html
http://www.soundhunter.com/
http://simplythebest.net/sounds/WAV/sound_effects_WA
V/index.html
http://www.acoustica.com/mp3-audio-mixer/sounds.htm
http://ljudo.com/default.asp?lang=tEnglish&do=it
Edit: To alter to fit a purpose
Sound files contain audio stored as
samples.
We can change the value of any sample or
group.
We can copy samples from one place (time)
to another.
We can delete samples.
We need an interface
Samples are averages of sound values
over time. Not a good abstraction.
Originally, editing was done using audio
tape. Tape was physically cut with a blade
and then taped back together.
Old Style
Tape Editing
Cut editing (also known as destructive editing) consists of physically cutting
magnetic recording tape, removing an unwanted portion and rejoining the
recording tape with a special adhesive backed splicing tape.
Monophonic and stereophonic 1/4 inch tapes were traditionally edited in this
way. Cassette tapes can be cut edited but it is difficult to locate the precise edit
point. The technique is sometimes useful for retrieving damaged cassettes.
Multitrack tapes (4 or more tracks) are not cut edited. .
Tape Editing
Removing a section of 1/4 inch tape
1. Play the tape until you hear the first cue,
that is, the sound at the beginning of the
section you intend to remove.
2. Put the tape recorder into edit mode.
(See the operator's manual.)
3. Rotate the tape spools by hand with the
tape against the replay head,
(usually the one on the right).
4. Using a chinagraph pencil, mark the
tape at the beginning of the section to be
removed.
5. Mark the tape at the second cue, e.g. at
the end of the section to be removed.
6. Pull the tape away from the replay head
and lay it in the guide of the editing
block, oxide side down.
7 Line up your chinagraph mark with one of the
diagonal slits in the editing block.
Choose the 450 (45 degree) slit for a stereo
recording, the 600 slit for a mono recording or
the 900 slit if you are removing clicks.
8 Cut the tape by running the corner of a singlesided razor blade along the slot.
The corner of the blade should run along the
bottom of the slot in the editing block.
It does not matter that the corner of the blade
becomes blunt because that is not the part of the
blade that slices the tape.
9 Now cut the tape at the other chinagraph mark
10 Remove the unwanted section of tape but keep it safe
in case you need to replace some or all of it
11 Slide the two ends of the tape together in the editing
block.
The two cut ends should just butt together. There should
be no overlap and no gap between the ends.
12. Cut about 2cm of adhesive splicing
tape. Touch one end of the splicing tape
on to the side of the razor blade, so that it
lightly sticks there.
13. Use the blade to transport the cut length
of splicing tape to the editing block.
14. Stick the length of splicing tape over the butt
join and press out any air bubbles with your
finger nail. .
15. Remove the magnetic tape from the splicing
block.
16. Play the tape at a good listening level and
listen to the edit point. As the edit passes the
replay head there should be
• no change in the level (loudness) or quality of
the foreground speech, music , e.t.c.
• no change in the level or quality of the
background noise
• no change in the rhythm of the speech or music
Anyone wish to try this??
So, how does this translate into Digital??
Let’s use GoldWave: it is free (evaluation version), and is relatively full featured.
http://www.goldwave.com/
The sound (tape) is in a window
Time advances from left to right; volume (voltage, intensity) is up/down. There
Are two channels here (stereo)
Time (in seconds)
The controls
Time advances from left to right; volume (voltage, intensity) is up/down. There
Are two channels here (stereo)
Paste
Scale
Into new window
Open
Save
Copy selection Delect
File
To Clipboard
New
Selected data
Which channel
File
Paste
The device controls
Time advances from left to right; volume (voltage, intensity) is up/down. There
Are two channels here (stereo)
Volume
User play
(looping)
Record
Balance
Speed
Right channel
Play
Rewind
Left Channel
Editing – Grab, cut, paste
Left Click on left side
Of sound to be copied/
deleted
Right Click on right side
Of sound to be copied/
deleted
Then select cut or copy or delete,
or use del or ctl-x or ctl-c
Editing – Sound Forge
Advantage – you can hear the sound as you move a cursor along the time line.
Editing – ‘Colors’
From the computer commercial on TV
Kira Willey - Colors.MP3
Colors – edit 1
Final edit
Recording – microphone work
Sound forge
Recording – Audacity
Available at http://audacity.sourceforge.net/
Open source!
Recording – Audacity
Goldwave
Sound Forge
Audacity