2: Signalling Flashcards
why are digitised signals possible?
information can now be coded into binary digits which can be transmitted in multiple ways.
What are digital signals represented by?
Binary numbers (digits)
What are the advantages of digital over analogue signals
1) Digital signals can often be sent, received and reproduced more easily than analogue signals because they can only take a limited number of values
2) faster transmission than analogue signals
3) Digital files can be compressed to reduce their size, and manipulated easily for artistic effect
4) Noise resistant
What are the disadvantages of digital over analogue signals
2) Because digital signals can be copied more easily, digital information (films etc.) can be reproduced illegally.
3) Confidential information, such as personal data and photos, may be stolen without the owner’s knowledge or consent, more easily by hackers, infected networks or malicious websites
4) digital signals easily scrambled
Are Analogue signals limited in the values they can take?
No. They also vary continuously from one value to the next.
how can a signal get weak and what can you do about it?
what is noise in the context of signals
When you transmit an electronic signal, it will pick up noise. From what does it pick up noise?
what is a problem with apmifying a signal with noise
1) if it’s distorted, you can amplify it
2) random variation on the signal
3) Electrical disturbances or other signals
4) the noise is amplified as well
Why is it easier to reconstruct the original signal from the noisy signal with digital signals? Why is it important that it’s easy to reconstruct the signal
The number of values, a digital signal can take is limited
You need to get an accurate representation of what was sent
Can Analogue signals be digitised?
what is the process called, define it
Yes
the process is sampling, its the process in which the displacement of a continuous analog signal is measured at small time intervals and converted into a string of digital binary numbers (samples)
How do you digitise a signal, where each sample taken is coded with 3 bits, potential problem?
a sample with 3 bits means that there are N=23 levels called quantisation levels to represent the signals value.
You take the value of the signal at regular time intervals then find the nearest quantisation level.
Each quantisation level is represented by a binary number so you can convert the analogue values to binary numbers
The digital signal you end up with won’t be exactly the same as the analogue signal if the nearest quantisation level doesn’t match the signal when a sample is taken
What 2 factors affect how well a digitised signal matches the original?
The difference between the possible quantisation levels
the time from one sample to the next (sampling rate)
What happens if a signal is digitised using only a few, widely spaced samples?
It’s likely that a lot of the analogue values sampled will be far from the nearest digital value.
what is quantisation error?
the difference between the signal value and the closest quantisation level
What is the advantage of increasing quantisation levels?
The more closely the digitised signal will match the original
What is resolution?
equation for resolution
a signal is detected over a 12V range, 8 bit sample of this signal is produced, calculate resolution of this sample
the smallest change in potential difference that can be determined.
resolution = potential difference range / number of quantisation levels
12v range / 28(256) = 0.047V
What is the advantage of using a lower resolution?
reduces the demand on data storage and transmission speeds
What limits the number of bits that can be used for sampling, and so limits the number of quantisation levels
Noise. If the original signal contains noise then you would sampling the noise to greater detail rather than ignoring it. you should not have a smaller gap between quantisation levels than the size in noise variation
how to calculate maximum number of useful quantisation levels?
max useful levels = total pd of noisy signal variation / pd of noise variation
Vtotal / Vnoise
number of bits required for maximum useful quantisation levels
Q: a signal has a max total variation of 200mV. the noise variation is 5mV. calcualte the largest number of bits per sample worth using to encode the variation
b = log2 Vtotal / Vnoise
log2 0.2V / 0.005V = 5.3, so this goes up to 6 bits
we go up because 5 bits gives 32 levels, whereas the Vtotal / Vnoise gives 40 max useful quantisation levels, so we go one up
2 conditions to be met for accurately reconstructing og signal
signal cannot contain any frequency above a certain max
minimum sampling rate must be twice the highest frequency component
Minimum sampling rate =
Twice the highest frequency component
Why does the sampling rate have to be high?
To record all the high frequency detail of the signal
To avoid the creation of aliases
What are aliases?
Low frequency signals that are created from having a too low sampling rate
what is the sampling rate for music, how do you ensure aliasing doesn’t occur?
the human ear can’t detect frequencies above 20KHz, so for music to be sampled accurately, it is sampled at a frequency of 44.1KHz. filters remove frequencies above 20kHz in the original signal so aliasing doesn’t occur
what can be the result of not sampling frequently enough and not filtering out frequencies above a maximum
you lose high frequency details
you can also create false low frequencies which is known as aliasing from frequencies above the max limit
define bit rate
rate of transmission of digital information
equation for bit rate units
a cd uses 16 bits per sample at a sample rate of 44.1kHz for 2 channels. calculate combined bit rate for 2 channels
bit rate = samples per second (sample rate(Hz or KHz)) * bits per sample
(16 * 44100) * 2 = 1.4 * 106 bits s-1 = 1.4MHz
How can you speed up the process of uploading images?
Compress your image to reduces the file size
Also uses less storage
how to calculate duration of signal
a song occupies 28Mbytes of memory on a phone. if the bit rate is 1.4MHz for the system, calculate the duration of signal
number of bits in signal / bit rate
(28 * 106) * 8 / 1.4 * 106 = 160s = 2 mins 40 sec
how to calculate number of bits required for storing a recording that lasts 3.5 minutes with a sample rate of 40KHz where there are 16 bits per sample
bits = number of seconds * sample rate (samples per second) * bits per each sample
seconds = 210 Seconds, there are 40000 samples per second each with 16 bits
210 * 40000 * 16 = 134400000
how to find sample rate from a graph
find number of samples taken in a second
what does it mean if an image has a 6 bit resolution
the image uses 6 bits per pixel
what is meant by speed of transmission
speed at which the signal travels
what is a signal
the variation carrying the useful information being transmitted
