AUdIoCoUrSeS
Joined: 31 Oct 2002
Posts: 2014
|
| Week 4 |
 |
|
Here we go.
1. There are 2 forms of compression commonly used in audio; Dynamic compression and Data/File compression. Describe the different forms and a use of each.
2. Explain the term Psychoacoustic modelling
3. Detail the typical controls, and their uses, found on a dynamic compression unit/software application.
4. Explain the difference between lossless and lossy coding.
5. Calculate the maximum theoretical audio frequency and the signal to quantisation noise ratio in dB for the following systems:
a) Compact disc digital audio system.
b) A system utilising a sampling rate of 26KHz and 12 bit linear quantisation.
6. Explain the principles of error concealment.
7. Explain how interleaving works.
8. Shifting a sample word one step to the left or one step to the right will change the gain in dB of the sample by how much? And Why?
9. What is a Session file?
10. Discuss the Archiving of digital recordings.
11. Comment on some features, functions and parameters of disk recording systems .
_________________
It's all in the ears. - Learn the concepts not the software.
Audio Courses is a way into the music business for you
|
Sun Mar 25, 2007 3:41 pm |
|
|
resol69
Joined: 31 Dec 2002
Posts: 69
|
| from nancy |
|
|
1. There are 2 forms of compression commonly used in audio; Dynamic compression and Data/File compression. Describe the different forms and a use of each.
Dynamic compression
changes the dynamic range of a signal which allows for an increase in volume without distortion.
In audio recording, a compressor reduces dynamic range by using a variable-gain amplifier to reduce the gain of the signal. In analog systems, a voltage controlled amplifier has its gain reduced as the power of the input signal increases.
A limiter is a compressor with a higher ratio, and generally a faster attack time. There is no absolute consensus on what ratio constitutes limiting as compared with compression, but most engineers would consider anything with a ratio greater than 10:1 as limiting. Compression and limiting are no different in process, just in degree and in the perceived effect. Engineers sometimes refer to soft and hard limiting which are differences of degree. The "harder" a limiter, the lower its threshold and the higher its ratio.
Compression and limiting can be used prevent digital clipping.
In data/file compression
the original and the compressed file are identical.
In this class of data compression algorithms that allows the exact original data to be reconstructed from the compressed data. This can be contrasted to lossy data compression, which does not allow the exact original data to be reconstructed from the compressed data.
A ZIP file is an example of date file compression.
2. Explain the term Psychoacoustic modelling
Psychoacoustics
is the study of subjective human perception of sounds. The human ear can nominally hear sounds in the range 20 Hz to 20,000 Hz (20 kHz).
The psychoacoustic model
provides for high quality lossy signal compression by describing which parts of a given digital audio signal can be removed (or aggressively compressed) safely - that is, without significant losses in the (consciously) perceived quality of the sound.
3. Detail the typical controls, and their uses, found on a dynamic compression unit/software application.
Compressors usually have controls to set how fast the compressor responds to changes in input level, known as
attack,
and how quickly the compressor returns to no gain reduction once the input level falls below the threshold, known as
release.
Another control on a compressor is
hard/soft knee
. This controls whether the bend in the response curve is a sharp angle or has a rounded edge. A soft knee reduces the audible change from uncompressed to compressed, especially for higher ratios where the changeover is more noticeable.
Because the compressor is reducing the gain (or level) of the signal, the ability to add a fixed amount of
make-up gain
at the output is provided so that an optimum level can be used.
4. Explain the difference between lossless and lossy coding.
A lossy compression
method is one where compressing data and then decompressing it retrieves data that may well be different from the original, but is close enough to be useful in some way
There are two basic lossy compression schemes:
1. In lossy transform codecs, samples of picture or sound are taken, chopped into small segments, transformed into a new basis space, and quantized. The resulting quantized values are then entropy coded.
2. In lossy predictive codecs, previous and/or subsequent decoded data is used to predict the current sound sample or image frame. The error between the predicted data and the real data, together with any extra information needed to reproduce the prediction, is then quantized and coded.
In some systems the two techniques are combined, with transform codecs being used to compress the error signals generated by the predictive stage.
Most lossy compression formats suffer from generation loss: repeatedly compressing and decompressing the file will cause it to progressively lose quality. This is in contrast with lossless data compression.
Lossless data compression
is a class of data compression algorithms that allows the exact original data to be reconstructed from the compressed data. This can be contrasted to lossy data compression, which does not allow the exact original data to be reconstructed from the compressed data.
Lossless compression is used when it is important that the original and the decompressed data be identical, or when no assumption can be made on whether certain deviation is uncritical. Lossless compression is preferred for text and data files, such as bank records, text articles, etc.
5. Calculate the maximum theoretical audio frequency and the signal to quantisation noise ratio in dB for the following systems:
SNR=6 (#bits) + 1.8dB
(every bit adds 6dB of noise? Not sure why we’re adding the 1.8dB at the end)
The Nyquist theorem states that anything below ½ the sampling frequency will be distorted or lost, so the maximum theoretical audio frequency would be ½ the sampling frequency.
SO:
a) Compact disc digital audio system.
Music on a CD is stored at 16bit accuracy with a 44.1KHz sampling rate.
SNR=6 (16) + 1.8dB = 97.8dB
Max frequency = ½ (44.1Khz) = 22Khz
b) A system utilising a sampling rate of 26KHz and 12 bit linear quantisation.
SNR=6(12) +1.8dB= 73.8dB
Max audio frequency = ½ (26Khz) = 13Khz
6. Explain the principles of error concealment.
A digital to analog decoder incorporates three levels of protection against damaged data:
Error correction; an error is detected in the data and completely corrected by using the additional error-correction data specifically put there for the purpose.
Error concealment; an error is detected but it is too severe to be corrected. Missing data becomes a “guess” based on the surrounding data and the result hopefully will be inaudible. However, if you ever get chance to see a CD player that has correction and concealment indicator lights, you will notice that an awful lot of concealment goes on just to play an average disc. How well concealment is done is one of the factors that make different digital systems sound different.
Muting; in this case the error is so bad that the system cant’ even make a guess at it and shuts down momentarily rather than output what could be an exceedingly loud glitch.
7. Explain how interleaving works.
Interleaving
is a form of error correction during playback. The process of scattering data around the medium so if a large section is damaged, it results in many small manageable data losses that can be recovered using error correction. Interleaved data is reconstructed using algorithms the player understands. Since the data is scattered, if a big chunk is damaged or unreadable, the player can determine what the missing bits should be by using the algorithms.
Interleaving in video:
helps ensure that video and audio files are in sync. Although video clips have simultaneous audio and video streams that play at the same time, the files that contain them are a single stream. Interleaving fakes having two streams by slicing the audio and video streams into chunks and mixing them together in chunks by time. A player reading an interleaved file receives a little bit of audio, then a bit of video, and then more audio, etc., buffering them in memory for a short time before playing the two together.
8. Shifting a sample word one step to the left or one step to the right will change the gain in dB of the sample by how much? And Why?
Word length is determined by the number of bits used. Each bit adds 6dB. Do I’m guessing that shifting a sample work one step to the left would lower it by 6dB and shifting it to the right would raise it dB???
9. What is a Session file?
A session file is where all your work or “mix” on a digital audio file is stored. Information stored includes volume, gain, results of plug-ins and outboard equipment like EQ and compression.
10. Discuss the Archiving of digital recordings.
Archiving is safely storing your digital recordings. Digital recordings can be archived on portable disk drives, CDs, DATS.
11. Comment on some features, functions and parameters of disk recording systems.
One major advantage of recording audio to a hard disk is that it allows for non-linear editing. Audio data can be accessed randomly and therefore can be edited non-destructively, that is, the original material is not changed in any way. In addition, hard disk recorders offer some disadvantages, including the limited capacity and relatively high cost of replacement drives, as well as a reduced ruggedness of hard disk recorders as compared to tape-based systems.
Hard disk recorders are often combined with a digital mixing console and are an inherent part of a digital audio workstation. In this form complex tasks can be automated, freeing the audio engineer from 'performing' a mix.
The major constraints on any hard disk recording system are the disk size, transfer rate, and processor speed. Some systems use "lossy" digital audio compression to minimize the first two factors.
Size – This is expressed in Gigabytes, and drives of up to 250Gb are on the market as of the time of writing.
RPM – this is the speed of the drive in terms of how quick it spins. The most common speeds are 5,400rpm & 7,200rpm, although the new SATA format is producing drives with speeds of 10,000rpm. The quicker the drive spins the quicker information can be read from the drive, although higher rpms may also result in higher noise levels (see below)
Transfer Rate – The numbers in these descriptions represent the (theoretical) maximum speed of the drive in MB/s. So an ATA133 drive can theoretically shift data at 133MB/s. However, there are 2 limiting factors: the motherboard has to support the relevant ATA ‘speed’. If your motherboard only supports ATA66 for instance, any ATA100 or 133 drives will not exceed a speed of 66MB/s. Second, these figures only really represent the ‘burst rate’ which is the maximum data transfer rate, not the sustained rate. Most of the time the sustained rate will be around 75%-80% of the maximum speed, but it can be even lower depending upon the setup of your PC.
Seek Time – This is the amount of time, in ms, that it takes the drive to find the required data. Shorter values are preferable.
Cache – The cache, sometimes called a buffer, is a ‘storage area’ whereby data is held until it is needed by the PC. Data can be read very quickly from the buffer. A drive with a large cache will ensure that any short-term slowdown in data transfer does not result in a loss of audio as there will already be data stored in the buffer.
Interface – By far the most common (at least at the time of writing) is the IDE interface. This is supported on all but the oldest motherboards and is a mature protocol. As explained above, these refer to the maximum possible speed of the drive in MB/s. However, your motherboard also needs to support the same ATA value to get the best performance. The second interface available at present is SATA. This is a serial rather than parallel interface and can offer speeds (theoretically) much higher than standard ATA/IDE. It also uses much smaller, and easier to handle, cables. |
|
Sun Apr 01, 2007 4:45 pm |
|
|
|
|
Rbro
Joined: 07 Sep 2005
Posts: 12
|
|
|
|
1. There are 2 forms of compression commonly used in audio; Dynamic compression and Data/File compression. Describe the different forms and a use of each.
Dynamic Compression: A dynamic compressor is a tool which allows the user to change the dynamics (volumes) of a piece of audio. A dynamic compressor affects predefined loud volumes levels by lowering the output volume when the sound reaches a set threshold. In order to keep the rest of the piece unchanged a ‘make up gain’ control boosts other, none affected parts and this leads to low volumes also being effected (The also raises the floor noise and is why noise gates are generally used before compressor). The main contols of a dynamic compressor are:
Threshold – Volume where the compressor kick in
Ratio - The amount the audio will be compressed
e.g. 4:1 ratio= 4db input/1db output
Dynamic compressors are also used as effects. The two main effects they produce are called ‘breathing’ and ‘pumping’. An example of this can be heard on @Call on Me’ by Eric Prydz.
Data/File Compression: A data/file compressor is used to allow you to change a large file into a smaller file (and back again) using algorithms. This type of compressor was invented to allow easier transfer and storage of files. The two main audio algorithms are ‘lossy’ and ‘ lossless’ (see below). The algorithms look for ways to make the long lines of code smaller e.g. if the code repeats it will write (2) at the end instead of writing it out twice. It is obviously far more complex than this but works along similar principles. Depending on how the data is compressed will define what the quality is like when uncompressed.
2. Detail the typical controls, and their uses, found on a dynamic compression unit/software application.
Threshold: The volume at which the compressor will kick in
Ratio: The amount the sound it to be compressed e.g. 4:1
Attack: The amount of time it takes the sound to reach full volume
Decay: Time taken from attack - sustain
Sustain: The held volume after the decay (this parameter refers to volume not time)
Release: The amount of time for the sound to go from sustain to silence.
3. Explain the difference between lossless and lossy coding.
Lossless: File compression that allows the user to retrieve most of the original audio.
Lossy: File compression that compresses more than lossless but uncompressed the quality is not as good as uncompressed as lossless and the original cannot be retrieved.
4. What is a Session file?
A session file is a new file that is created each time you open the sequencer and select new project. The session file is kept in the session directory and contains all the information about the seesion e.g. tracks, automations etc. If you import audio during the session the session file only references the import and doesn’t store it in the file. This can cause problems when you backup work. |
|
Sun Apr 01, 2007 5:38 pm |
|
|
resol69
Joined: 31 Dec 2002
Posts: 69
|
are we adding the additonal dB to account for the dymanic range of the signal? |
|
Sat Apr 07, 2007 4:43 pm |
|
|
 
|
|
All times are GMT. The time now is Sat May 17, 2008 5:22 am
|
|
|
|
| |
