For the final assessment for my Audio Engineering unit, I was given a 2 page assessment which consisted of a variety of Pass, Merit, and Distinction level questions (marked with either a P,M,D next to each one bellow.) My answers where as follows.
Define the following terms: (P)
Signal to Noise Ratio:
The Signal to Noise Ratio describes the balance between a desired signal and the background or system noise that is recorded with it. No matter what you record with you will always get some element of background noise, which might take on the form of system hum or air conditioning, and therefore will be on an audio recording whether you are aware of it or not.
The key is to have a signal to noise ratio that favours the signal, as this will give a clearer recording, that’s why it is important to have good gain staging, as a weaker signal will have a more noise, meaning that if you turn the volume up, you are turning both your signal up and the noise.
Using an “Alias” means to take on a name that differs from their own, so therefore in audio it occurs by the Sample Rate not reading the waveform correctly, mistaking a frequency for a different one, resulting in a distorted signal.
Dither is essentially shaped noise which is used to aid Bit Depth transitions, with lower the bit depth requiring a greater dither. where Bits have been removed, it works out where to add the noise to smooth out the waveform.
Bit Depth refers to the amount of times a signal is read each second it is sampled, with the higher the bit depth meaning the more times its it read and therefore gives a closer representation to the actual sound.
Binary is a numerical system where the outcome can be 0 or 1, a computer uses binary, where the two states represent off and on, and therefore by adding more “bits” you increase the amount outcomes.
Sample Rate dictates how many times a signal is sampled per second, such as 44.1Khz, so therefore when it plays back, it will play 44,100 samples per second.
Headroom refers to the loudest your signal can go before the signal clips, on analogue equipment (Measured in dBVU) the signal may go over the threshold, but wont clip, it just won’t be as loud, while a when a digital signal (measured in dBFS) will hit the ceiling and clip, causing the signal to be distorted and extra harmonics being added.
dBFS (decibels full scale) is a digital unit of measurement where the defined maximum level is 0dBFS, so will not go into a positive decibel figure, as well as these systems having a reduced dynamic range compared to their analogue counterparts.
Explaining the Advantages and Disadvantages recording at higher qualities compared to the normal CD quality standard of 16 bit/44.1KHz. (M)
When setting up a session, many DAW’s give you the option of what to set both your sample rate and bit depth, each which comes with its benefits and limitations:
For instance if you set your sample rate at a higher quality, such as 48Khz or 96Khz, your data is being quantized at a quicker rate per second, meaning that when the signal is being played back, it will give you a closer representation of the incoming signal.
However once you start getting past 96Khz, such as up to 192Khz, you may be gaining a larger bandwidth, but within the 20Hz-20Khz range jitter becomes more and more of an issue, since you are sampling more times per second, resulting in the chance of aliasing becoming more likely.
An example of this is that it isn’t uncommon that microphones with a high frequency response when driven to hard start to cause aliasing problems, resulting in a gritty, harsh sounding top end. This is due to the anti-aliasing filter rounding down source harmonics, such as 23Hz to 21.3Hz, which will sound off. But if you where recording in 96Khz, the anti-alias filter parameters where higher, so you wouldn’t have the problem.
If you increase the Bit Depth, such as to 32 or 48 you are increasing the resolution of the mix, as mentioned before it determines how many times within each sample the signal is read, so the wave forms transient will be smoother, and have more clarity.
However if it converted to a lower bit depths, all that extra data has to be simplified down to fit the new format, this could result in quantisation errors occurring where two “bits” of data get combined creating an un accurate representation of the sound.
You would also have to use dither to fill in the gaps created by the lower bit depth, creating shaped noise to even out the waveform.
Since the bit depth determines the dynamic range (amount of bits times by 6 to get the overall maximum dB) dither is used to fill in the gaps before conversion, with more having to be used depending on the difference in Bits.
Sine Sweep that induces aliasing: Practical Demonstration: (M)
I decided to try inducing aliasing in different bit depth and sample rate, starting at 44.1Khz and 16 bit, sweeping a sine wave from 20hz-20Khz, seeing at what frequencies aliasing might occur at, due to Pro Tools internal playback engine.
I started by setting up a session in Pro Tools 10, with a 44.1Khz Sample Rate and 16 Bit Depth, and created an auxiliary input with a signal generator plug in and a mono audio channel.
In order for me to record the frequency sweep, I had set the output of the signal generator to bus 1, and then set the input of the audio channel to the same, so when I record enable, it will send the signal into the audio track.
I also need to automate the sine wave sweep, which involves going to the signal generator automation menu, and adding the frequency control to the automation list, getting it to sweep from the minimum (20Hz.) to maximum (20Khz.) for a duration of 10 seconds.
When I played it back out through Pro Tools, with a Blue Cat Frequency Analysis on the audio channel, it was saying it was fine, even though visibly the waveform wasn’t completely smooth.
We then played the signal through an external device, In this case our lecturers iPad, and found that an aliased frequency had been produced at 215Hz, when the sweep got to 258Hz. We can tell this from the fact that the aliased frequency is slightly lower dB to the main frequency, and when this frequency is produced, the integers of the harmonic create a distorted waveform.
I also did the same test but at a 48Khz Sample Rate and 24 Bit Depth, and the rogue frequency wasn’t present, creating a more accurate image.
When a Signal on the mixing desk in Studio 1 is at 0dBVU it is not at 0dBFS in Pro Tools, discuss the implications of the difference: (D)
The difference starts with the fact that in order for Pro Tools to receive any signal from the desk, which is an analogue piece of equipment, it has to be converted into a digital signal via a analogue to digital converter.
The analogue to digital converter reads the incoming signal and “samples” the signal at a number of pre-determined intervals per second, which is determined by the sample rate, reading the frequency response. Within each of these samples the amplitude reading of the signal is determined by the bit depth, which sets the threshold of the dynamic range of the signal, and converts it into “bits”.
This is called quantisation, where Pro-Tools then has two sets of data, one for frequency (time) and one for amplitude (volume.), which is then correlated to create a series of points which become the digital signal.
However an Analogue’s input signal is measured in dBVU (decibels) which can exceed 0dB and go into positive dB, however a digital system cannot exceed 0dB and any signal which goes above unity will clip, and distort, causing false frequencies to appear on the track.
In order to avoid this happening, Pro-Tools sets the incoming signal from the desk to be lower than 0dBFS, so it has enough headroom for the signal to come in, and be read without it distorting, such as -18dBFS=0dBVU. There isn’t a universal standard for the ratio between FS and VU, though the BBC use -18dBFU as there calibration point.
The Nyquist Theorem is used to reduce errors in the recording of digital music. How? (D)
The Nyquist Theorem dictates that the minimum sampling rate in a digital audio system needs to be at least twice the range of human hearing, which is from 20-20,000Hz.
This is because If you where sampling the highest frequency being, 20,000Hz, at a sample rate of the same value, you would be sampling it only once per cycle, (frequency is calculated determined by how many times a signal repeats per second.)
This would result in a continuous line, which would be quantised as such when converted from analogue to digital, resulting in a quantisation error and distortion of the signal.
However if you decided to sample at twice this rate, at 40,000Hz, you would be reading the signal twice every cycle, so instead of the series of points being picked up at the same place each cycle, it would now have two alternating values, so when the signal was recreated, it would resemble a triangle wave instead, having twice as many harmonic values.
Also at both sample rates you would still have to take into account aliasing, which would be caused by when the signal is being quantised certain higher frequencies would be rounded down and anything that went higher than that being rounded down by the anti-aliasing filter. As mentioned earlier with microphones with a high frequency response, by increasing the sample rate the parameters of the anti aliasing filter would be moved, no longer affecting the higher frequencies.
So by sampling at 44.1KHz, you are sampling higher frequencies a decent amount of times to maintain clarity, plus giving yourself enough extra space to take into account aliasing.
Show and discuss the types of cabling found in studio 2 with a definition of the function of the function of each. (P)
Balanced ¼ Inch Jack:
¼ Inch connection, it is able to send two channels of audio through its Tip, Ring, Sleeve design (TRS) Tip and Ring carry the left and right channels while the sleeve acts as a common earth. Because the Desk has an option for a mix A and a mix B, plus 4 Stereo Buses which channels on the desks can be routed to.
Un-Balanced ¼ Inch Jack:
Acts as the cable, which transfers digitised data into the Mac, which then uses a DAW to read the data.
A balanced connection, using three pins, with both male (sends signal) and female (receives signal) ends used for microphone level inputs from the microphones in the live room (via the wall box) or the control room, into the desk.
Un-Balanced RCA Cable:
An RCA cable, or a Phono cable, is an unbalanced mono connection, which consists of two separate connectors being the left and right channel. In Studio 2 this goes from the output of the CD player into a “tape in” on the desk, allowing CDs to be played.
Explain how the audio interface is central to the working of studio 2:(M)
The Interface fulfils many different purposes with the studio, the first of which it acts as a AD/DA converter, as it takes the signal routed through the desk, which have been gain staged and possible EQ’d and converts the Analogue signal into a digital one via quantisation.
Once it has been quantised it then sends the information out through the firewire connection to the Mac, which would have no way of receiving the analogue signal.
It controls the amount of sources you can record at any one time, which can be expanded by connecting another audio interface with an ADAT cable, or accept more inputs from an SPDIF.
It also can be used a master clock, if no other master clock is present, to make sure multiple units quantisation rates are the same, which then it sends to the Mac via the Fire wire.
How digital connections ADAT, AES/EBU, and SPDIF are used in the studio: (M)
ADAT (Alesis Digital Audio Tape) allows you to add an additional 8 Mic level inputs into your audio system with the addition of one cable. This can be used to connect two audio interfaces together, with one acting as the master and the other as a slave, meaning that both their clocks will follow the masters clock source.
AES/EBU and SPDIF are very similar as they both send data in stereo and both transmit audio data in the same form. However the S/PDIF connection is unbalanced while the AES-EBU is balanced. However, S/PDIF carries extra data such as DAT and CD track IDs while the AES-EBU does not. AES-EBU does not recognise this data, which can be both an advantage and a disadvantage. This can be used to send an additional 8 Mic level inputs into a Audio interface at 44.1 and 48KHz, however if you where to record at a higher sample rate, such as 96Khz, you would have only be able to use 4 of them.
Function of Clocking in Digital Studio Systems:(M)
The main purpose of clocking in a Digital studio is to make sure that everything is running in sync with each other such as having multiple audio interfaces running alongside one another, to making sure that your DAW is reading the incoming data correctly so that it knows when each of its 44.1K samples per second are in the right place.
In order for this to happen, everything needs to be sync’d to the same source, which you’re DAW usually takes its master clock from the first audio interface it is connected to, with any additional clocks being slaved to the first one, avoid conflicts and quantisation errors such as Jitter.
Other errors caused due to not having an in sync clock can result visually in your meters randomly spiking, or audibly in the form of clicks and bouts of static.
The basic digital clock type is word clock, which runs at whatever the sample rate is with the DAW and the interface working together, with an however if you have to set your DAW to sync with an external clock. Pro Tools TDM (Time Divison Multiplexing) mix environment can handle streaming 256 digital audio streams, but this could not be done with the word clock, as it would need to be running 256 times faster, hence the requirement for a Superclock which does exactly that.
The Pros and Cons of Digital Verses Analogue Mixing:(D)
Up to only 20 years ago, Analogue mixing was the only way records where mixed, while now that the cost of recording equipment has become more and more affordable, and thus a lot of people who buy this equipment can mix on their laptop computer.
One of the benefits of digital, firstly your not limited to mixing one console that you would be on analogue, as because the data has been quantised and your signals are represented as a series of bits, if the computer you wish to mix on has a DAW installed you don’t have to be tied down. However the downside of this is that because the data has been quantised, it can be subject to timing errors, such as jitter, or aliasing, causing the information to become distorted. On an Analogue desk the incoming signal would be printed on magnetic tape, where it wouldn’t need to be converted into digital format, therefore not be subject to the same problems.
Not only would you be able to take your mix around with you if it was in a digital format, but you can save all panning and fader positioning so you can continue where you left off, while on an Analogue system, you would have to redo all of this every time a song was mixed, as well as being to add automation without having to do it manually when recording it to tape. Also a benefit of digital mixing is the use of plug-ins, software equivalents of hardware units, such as EQs or Compressors, where you could potentially have one or more on each track, while in analogue you would be limited to the amount of units you physically had with you.
Editing Audio is a lot simpler in digital, as it can be duplicated, stretched, and manipulated to the extreme, however this means that due to the number of possibilities, you can spend ages tinkering the mix, which might take away some of “feeling” of the mix, though Analogue it is a lot harder to duplicate tape, as it deteriorates with each successive replication, resulting in you wanting to get the correct take the first time.
Some of the pit falls of Digital Mixing are shared with digital devices in general, such as a file needs to be backed up manually while an Analogue mix is already backed up since its on tape.
A digital mix has the potential to become corrupted, while an Analogue cannot.
Digital has many different file formats you can export into, so compatibility maybe an issue, as both an advantage and disadvantage, with the later meaning you might not be able to use the audio because it was saved in the wrong format.
A digital system is also susceptible to crashing, while an analogue your desk will keep working as long as the power doesn’t go out. Also a digital system will have a certain shelf life before its needs to upgrade to the next model, while on an analogue desk you can use any type of tape recorded at anytime and it will still be compatible with the desk.