Final exam is next week, the last day of class. Study the Final Exam Review sheet and Chapters 1 -8 of the book. Let me know if you have any questions.
If you didn't give a presentation in class, to still get some credit - email me your final paper.
At the start of next week's class, you will need to turn in your completed Final Exam Review.
Make sure you have 7 Sound Journal entries completed by this Sunday at midnight. Any additional entries will be counted towards extra credit.
Monday, December 8, 2008
Monday, November 24, 2008
Remaining Class Information
A couple of things to remember:
1. No class for Wednesday's section this week, November 26. No class for the Monday section next week, December 1.
2. Papers and presentations are due in 2 weeks! Be sure that you have handed in a proposal to me and that we've talked about your topic. Check your syllabus or the post below for more information about the papers and presentations.
3. Final exam is in 3 weeks! Make sure you have read chapters 1-8, and start reviewing now...
4. If you want to make up any sound journals or other assignments, be sure to let me know. You should have 6 sound journal entries at this point in time. A sound journal entry is not required for the Thanksgiving holiday weekend, but if you want to use it as a make-up entry, feel free.
As always let me know if you have any questions.
1. No class for Wednesday's section this week, November 26. No class for the Monday section next week, December 1.
2. Papers and presentations are due in 2 weeks! Be sure that you have handed in a proposal to me and that we've talked about your topic. Check your syllabus or the post below for more information about the papers and presentations.
3. Final exam is in 3 weeks! Make sure you have read chapters 1-8, and start reviewing now...
4. If you want to make up any sound journals or other assignments, be sure to let me know. You should have 6 sound journal entries at this point in time. A sound journal entry is not required for the Thanksgiving holiday weekend, but if you want to use it as a make-up entry, feel free.
As always let me know if you have any questions.
Loudspeakers & Monitoring
The quality of any sound is based on what you hear from the loudspeaker interacting with the room acoustics. But choosing the "best" loudspeaker depends on many factors, both for the loudspeaker and also the location it will be in. Key requirements though are that it must be able to handle a range of frequencies from 30-16kHz and reach 120db-SPL.
A loudspeaker is a transducer that converts electric energy to acoustic energy - essentially the opposite of a microphone. Like a mic, speakers come with different elements. The moving coil is the most common and is used for a wide range of monitors because it is sturdier and handles a high SPL. Ribbon speakers are used for high frequencies because they have smooth high frequency response. Electrostatic (capacitor) is rarely used because they are expensive and have a low dB rating.
There are two ways of powering speakers. Passive power uses external amplifiers, while active power uses a built in amplifier to power the speaker internally. Speakers can't linearly output all frequencies - a large speaker that generates low frequency sound waves can't efficiently create high frequencies and vice versa. So multiple speaker drivers are needed to create the whole frequency range. Using a crossover network to divide the frequency spectrum into high/low sections, a large driver creates the low frequencies (woofer for the bass) and a small driver creates the high frequencies (tweeter for the treble). The point where the divide occurs is called the crossover frequency. One division is a 2-way system loudspeaker, with the crossover at 1500-2000 Hz. Two crossover frequencies is a 3-way system loudspeaker, with divides at 400-500 Hz, and 3500-4000 Hz.
A passive crossover network has an external amplifier before the crossover frequency divide, which may cause distortion. An active crossover network has the crossover divide before the two internal amplifiers power it, also called Biamped. This is less expensive, causes less distortion, requires less power, and gives a better transient response.
A speakers specications re based on tests in an anechoic chamber (room with no reflections), so there will be different performance in every room because of the different acoustics. Some things to look for include the following.
Wide frequency response - at least 40-20kHz needed for professional audio. Studios use 3 sets of speakers: low quality with midrange response only like in car radios, average quality ones with more high/low, and good high powered wide speakers.
Linearity - where the input loudness is reproduced at the same level iwht no more than +/- 3 db variance.
Amp power - sufficient power to create loud sound levels without distortion. 30 Watt tweeters and 100 Watt woofers.
Dynamic range - pro audio needs 40dB-SPL for soft sounds and 120dB for loud for an 80 dB range. Most consumer speakers only have 50dB to 105dB capabilities for a 55 dB range.
Sensitivity - tells you the overall efficiency of the speaker. Should be 93dB or higher.
Polar Response - how a speaker focuses sound at the monitoring position. You want to monitor with the fewest number of reflections - no affect of the room. Use bass traps for the low frequencies because the longer wavelengths can't be controlled by room dimensions.
Arrival Times - reproduced sound must reach the listener within 1 ms of one another.
Polarity - if the driver cone motions are opposite (compression versus rarefaction) then they are out of phase and must be rewired.
Distortion is the appearance of a signal in the reproduced sound that was not in the original sound. It can be created anywhere in the sound chain. There are several types of distortion: Intermodulation (IM) - 2 or more frequencies occur at the same time and create combination tones and dissonances unrelated to the orginal sounds. Rating of 0.5% IM or lower is best.
Harmonic - introduced harmonic not in the original signals, occurs when input and output of sound system are nonliner.
Transient - inability of audio component to respond quikly to a rapidly changing signal.
Loudness - overload - signal recorded or played back at level greater than the system can handle.
Monitor placement also affects the sound quality, dispersal and arrival time. Flush mounting the speakers keeps the walls and protusions from affecting the sound. When monitoring stereo, there are 2 dimensions, so 2 symmetric speakers placed in an equilateral triangle with the listener is best for monitoring. Far field monitors are large high quality speakers flush with the wall, while near field are on or near the console closer to the listener for just direct sound. Surround sound expands the depth by placing the listener in the middle of the aural image instead of in front. 5.1 surround has 6 discrete channels - 5 full range and one low frequency subwoofer.
To calibrate the speaker system you need an obective measure of the correlation between the monitor sound and the room sound. A spectrum analyzer will display the frequency response over time or in real time. An SPL meter will give approximate levels in real time (RTA). Fix the problems at the source by modifying the monitor position, amplifier, acoustics instead of attempting to use signal processing/EQ to fix it.
Reference levels for film are 85dB, TV is 79dB, and music is 79-82dB.
Headphones give a wide, flat uncolored frequency response, that is consistant across different studios and allow you to hear sutle changes. However you don't get the same sound quality as from monitors, and it is an unaturally wide aural image.
The 60%/60 min rule states that you should not listen to your headphones for more than 1 hour per day at 60% of the maximum volume.
Make your own loudspeakers
A loudspeaker is a transducer that converts electric energy to acoustic energy - essentially the opposite of a microphone. Like a mic, speakers come with different elements. The moving coil is the most common and is used for a wide range of monitors because it is sturdier and handles a high SPL. Ribbon speakers are used for high frequencies because they have smooth high frequency response. Electrostatic (capacitor) is rarely used because they are expensive and have a low dB rating.
There are two ways of powering speakers. Passive power uses external amplifiers, while active power uses a built in amplifier to power the speaker internally. Speakers can't linearly output all frequencies - a large speaker that generates low frequency sound waves can't efficiently create high frequencies and vice versa. So multiple speaker drivers are needed to create the whole frequency range. Using a crossover network to divide the frequency spectrum into high/low sections, a large driver creates the low frequencies (woofer for the bass) and a small driver creates the high frequencies (tweeter for the treble). The point where the divide occurs is called the crossover frequency. One division is a 2-way system loudspeaker, with the crossover at 1500-2000 Hz. Two crossover frequencies is a 3-way system loudspeaker, with divides at 400-500 Hz, and 3500-4000 Hz.
A passive crossover network has an external amplifier before the crossover frequency divide, which may cause distortion. An active crossover network has the crossover divide before the two internal amplifiers power it, also called Biamped. This is less expensive, causes less distortion, requires less power, and gives a better transient response.
A speakers specications re based on tests in an anechoic chamber (room with no reflections), so there will be different performance in every room because of the different acoustics. Some things to look for include the following.
Wide frequency response - at least 40-20kHz needed for professional audio. Studios use 3 sets of speakers: low quality with midrange response only like in car radios, average quality ones with more high/low, and good high powered wide speakers.
Linearity - where the input loudness is reproduced at the same level iwht no more than +/- 3 db variance.
Amp power - sufficient power to create loud sound levels without distortion. 30 Watt tweeters and 100 Watt woofers.
Dynamic range - pro audio needs 40dB-SPL for soft sounds and 120dB for loud for an 80 dB range. Most consumer speakers only have 50dB to 105dB capabilities for a 55 dB range.
Sensitivity - tells you the overall efficiency of the speaker. Should be 93dB or higher.
Polar Response - how a speaker focuses sound at the monitoring position. You want to monitor with the fewest number of reflections - no affect of the room. Use bass traps for the low frequencies because the longer wavelengths can't be controlled by room dimensions.
Arrival Times - reproduced sound must reach the listener within 1 ms of one another.
Polarity - if the driver cone motions are opposite (compression versus rarefaction) then they are out of phase and must be rewired.
Distortion is the appearance of a signal in the reproduced sound that was not in the original sound. It can be created anywhere in the sound chain. There are several types of distortion: Intermodulation (IM) - 2 or more frequencies occur at the same time and create combination tones and dissonances unrelated to the orginal sounds. Rating of 0.5% IM or lower is best.
Harmonic - introduced harmonic not in the original signals, occurs when input and output of sound system are nonliner.
Transient - inability of audio component to respond quikly to a rapidly changing signal.
Loudness - overload - signal recorded or played back at level greater than the system can handle.
Monitor placement also affects the sound quality, dispersal and arrival time. Flush mounting the speakers keeps the walls and protusions from affecting the sound. When monitoring stereo, there are 2 dimensions, so 2 symmetric speakers placed in an equilateral triangle with the listener is best for monitoring. Far field monitors are large high quality speakers flush with the wall, while near field are on or near the console closer to the listener for just direct sound. Surround sound expands the depth by placing the listener in the middle of the aural image instead of in front. 5.1 surround has 6 discrete channels - 5 full range and one low frequency subwoofer.
To calibrate the speaker system you need an obective measure of the correlation between the monitor sound and the room sound. A spectrum analyzer will display the frequency response over time or in real time. An SPL meter will give approximate levels in real time (RTA). Fix the problems at the source by modifying the monitor position, amplifier, acoustics instead of attempting to use signal processing/EQ to fix it.
Reference levels for film are 85dB, TV is 79dB, and music is 79-82dB.
Headphones give a wide, flat uncolored frequency response, that is consistant across different studios and allow you to hear sutle changes. However you don't get the same sound quality as from monitors, and it is an unaturally wide aural image.
The 60%/60 min rule states that you should not listen to your headphones for more than 1 hour per day at 60% of the maximum volume.
Make your own loudspeakers
Sunday, November 23, 2008
Signal Processors
Signal processors are devices or software used to alter a characteristic of sound. A plug-in is an add-on software tool that provides a DAW with more signal processing alternatives than what is built in.
Spectrum processors affect the frequency aspect of the signal.
Equalizers (or EQ) alters the frequency response by increasing or decreasing the level of a signal at a specific portion of the spectrum. The increase or decrease is down around a center frequency, which is the most affected value while those around it have gradually lesser changes. The range of frequencies affected is called the bandwidth. Shelving affects all frequencies above or below the selected frequency equally. There are fixed frequency EQs with a certain number of fixed center frequency options, generally knobs for high, upper middle, lower middle and low frequencies. Graphic EQ uses sliders instead of knobs. Parametric EQ allows for continuously variable frequencies and bandwidth size, which gives more flexibility and precision.
Filters attenuate bands of frequencies, usually at a preset and with a steep drop. High pass (low cut) cuts frequencies below the preset point. Low pass (high cut) cuts frequencies above a preset point. Band pass filters have both a high and low cutoff point and allows the frequencies between them through. A Notch filter cuts out an extremely narrow band, such as the hum at 60 Hz.
Psychoacoustic processors dd clarity and definition with EQ and harmonics.
Time processors affect the time relationships in a signal.
Reveration is created by random multiple blended repetitions of a sound. Dry sound is without reverb, we sound is with added reverb.
Digital reverb is the most commonly used, where the original signal is delayed and attenuated multiple times and then added to the original signal. Predelay is the time between the direct sound and the early reflections, and tells you the room size.
Convolution reverb is a sample based process that multiplies the spectrums of 2 audio files. The first is the acoustic signature of a space, called the Impulse Response. It is acquired by recording the room response after a quick impulse is played in the room. The second file is the source signal.When the two files are multiplied, it applies the room characteristics to the original signal, overlaying any room or location.
Plate reverb is a from a mechanical-electronic device with a thin steel plate that vibrates and is miked and sent through a console.
Acoustic chambers are dedicated rooms that create realistic reverb, but are very expensive to build.
Choose a reverb that sounds natural, with bright highs and clear lows. Listen to vocals to check for clearless, and sharp transients like a drumbeat for density.
Delay is the time interval between a sound and its repetition. Digital delay routes audio through an electronic buffer and holds it for a specific amount of time. The delay time is how long the sound is held. Feedback is how much of the delayed signal is returned. Higher feedback increases the number of repetitions and longer decay. No feedback means just one repetition.
Uses of delay include doubling, chorus, slapback echo, and prereverb delay.
Flanging is the original signal combined with a 0-20ms time delayed replica. Phase cancellations occur and create comb-filter efects with peaks and dips in the frequency response that lead to a holly, swishy sound. In phase (positive flanging) accents the even harmonics for a metallic sound, while out of phase (negative flanging) accents the odd harmonics for a warm sound.
Phasing uses a phase shift instead of a time shift, giving more irregular peaks and dips for a wavering vibrato and less effect on the pitch.
Morphing is a continuous seamless transition from one signal to another. It is not crossfading - the signal takes on actual characteristics of the other sound. Examples are available here.
Amplitude processors affect the dynamic range of the signal.
Compressors output level increases at a slower rate than the input level increases, and therefore restricts the dynamic range for peak signal limitations. The compression ratio establishes the proportion of change between the input and output levels. This can range from 1.1:1 to 20:1, which means that every change in 20db for the input only gives 1db in output change. The compression threshold is the level where the ratio takes effect. When it is reached, compression begins, which reduces the gain according to the amount the signal exceeds the threshold level and the ratio set. Knee is the moment the compressor starts gain reduction; hard knee is abrupt while soft knee is smoother. Broadband compressors act on the dynamic range of the input signal across the entire frequency spectrum, while split band compressors affect the input signal independently by splitting the audio into multiple bands.
A Limiter is a compressor where the output level stays the same at a preset point, no matter what the input level is. Basically gives a ceiling on the loudness of a sound, and reduces the high frequency response.
De-essers are a fast acting compressor that attenuate high frequencies to remove hissy consonants in s, z, ch, and sh vocals.
Expanders increase the dynamic range, essentially the opposite of compressors. They are triggered when the signal falls below a set threshold at a set ratio. A noise gate is used to reduce or eliminate low-level noise from amplifiers, ambience, rumble, noisy tracks, etc.
Pitch Shifters use time compression and expansion to change a flat or sharp pitch to be in tune. With time compression, the signal runs faster and raises the pitch. Time expansion runs it slower and lowers the pitch.
Noise processors reduce noise using DSP (Digital Signal Processing), removing clicks, cracks, humming, etc. Multieffects processors combine a number of functions into one unit.
Spectrum processors affect the frequency aspect of the signal.
Equalizers (or EQ) alters the frequency response by increasing or decreasing the level of a signal at a specific portion of the spectrum. The increase or decrease is down around a center frequency, which is the most affected value while those around it have gradually lesser changes. The range of frequencies affected is called the bandwidth. Shelving affects all frequencies above or below the selected frequency equally. There are fixed frequency EQs with a certain number of fixed center frequency options, generally knobs for high, upper middle, lower middle and low frequencies. Graphic EQ uses sliders instead of knobs. Parametric EQ allows for continuously variable frequencies and bandwidth size, which gives more flexibility and precision.
Filters attenuate bands of frequencies, usually at a preset and with a steep drop. High pass (low cut) cuts frequencies below the preset point. Low pass (high cut) cuts frequencies above a preset point. Band pass filters have both a high and low cutoff point and allows the frequencies between them through. A Notch filter cuts out an extremely narrow band, such as the hum at 60 Hz.
Psychoacoustic processors dd clarity and definition with EQ and harmonics.
Time processors affect the time relationships in a signal.
Reveration is created by random multiple blended repetitions of a sound. Dry sound is without reverb, we sound is with added reverb.
Digital reverb is the most commonly used, where the original signal is delayed and attenuated multiple times and then added to the original signal. Predelay is the time between the direct sound and the early reflections, and tells you the room size.
Convolution reverb is a sample based process that multiplies the spectrums of 2 audio files. The first is the acoustic signature of a space, called the Impulse Response. It is acquired by recording the room response after a quick impulse is played in the room. The second file is the source signal.When the two files are multiplied, it applies the room characteristics to the original signal, overlaying any room or location.
Plate reverb is a from a mechanical-electronic device with a thin steel plate that vibrates and is miked and sent through a console.
Acoustic chambers are dedicated rooms that create realistic reverb, but are very expensive to build.
Choose a reverb that sounds natural, with bright highs and clear lows. Listen to vocals to check for clearless, and sharp transients like a drumbeat for density.
Delay is the time interval between a sound and its repetition. Digital delay routes audio through an electronic buffer and holds it for a specific amount of time. The delay time is how long the sound is held. Feedback is how much of the delayed signal is returned. Higher feedback increases the number of repetitions and longer decay. No feedback means just one repetition.
Uses of delay include doubling, chorus, slapback echo, and prereverb delay.
Flanging is the original signal combined with a 0-20ms time delayed replica. Phase cancellations occur and create comb-filter efects with peaks and dips in the frequency response that lead to a holly, swishy sound. In phase (positive flanging) accents the even harmonics for a metallic sound, while out of phase (negative flanging) accents the odd harmonics for a warm sound.
Phasing uses a phase shift instead of a time shift, giving more irregular peaks and dips for a wavering vibrato and less effect on the pitch.
Morphing is a continuous seamless transition from one signal to another. It is not crossfading - the signal takes on actual characteristics of the other sound. Examples are available here.
Amplitude processors affect the dynamic range of the signal.
Compressors output level increases at a slower rate than the input level increases, and therefore restricts the dynamic range for peak signal limitations. The compression ratio establishes the proportion of change between the input and output levels. This can range from 1.1:1 to 20:1, which means that every change in 20db for the input only gives 1db in output change. The compression threshold is the level where the ratio takes effect. When it is reached, compression begins, which reduces the gain according to the amount the signal exceeds the threshold level and the ratio set. Knee is the moment the compressor starts gain reduction; hard knee is abrupt while soft knee is smoother. Broadband compressors act on the dynamic range of the input signal across the entire frequency spectrum, while split band compressors affect the input signal independently by splitting the audio into multiple bands.
A Limiter is a compressor where the output level stays the same at a preset point, no matter what the input level is. Basically gives a ceiling on the loudness of a sound, and reduces the high frequency response.
De-essers are a fast acting compressor that attenuate high frequencies to remove hissy consonants in s, z, ch, and sh vocals.
Expanders increase the dynamic range, essentially the opposite of compressors. They are triggered when the signal falls below a set threshold at a set ratio. A noise gate is used to reduce or eliminate low-level noise from amplifiers, ambience, rumble, noisy tracks, etc.
Pitch Shifters use time compression and expansion to change a flat or sharp pitch to be in tune. With time compression, the signal runs faster and raises the pitch. Time expansion runs it slower and lowers the pitch.
Noise processors reduce noise using DSP (Digital Signal Processing), removing clicks, cracks, humming, etc. Multieffects processors combine a number of functions into one unit.
Synchronization & Transfers
Synchronization is the ability to lock 2 or more digital signals or devices so that they operate at exactly the same rate. This is used to synch DAWs, consoles, synthesizers, outboard effects, and other digital equipment.
SMPTE is the time code (TC) developed for film based on the frame rate. HR:MIN:SEC:FRAMES is the format used. With 30 frames per second used, every 1/30 second of audio has a unique indentifying number called the TC address. MIDI TC translates SMPTE to MIDI to control MIDI devices like a keyboard.
A word clock is a signal generated in a digital audio system to control the sampling frequency. You must synchronize the word clocks in individual devices so that when tranferring the data signal is not degraded. A master clock sends a synchronization signal to the slave devices to synchronize all of the devices to the same sampling frequency.
Jitter is a variation in time from sample to sample that causes changes in the shape of the audio waveform and is caused by a degradation in word clock signals among digital devices. It results in lower detail and harsher sound. Using a master clock generator with low jitter and balanced, well-sheilded digital cables will minimize the effect.
An audio driver is a low-level program that allows the transfer of audio signals from/to audio interface. DAE (Digital Audio Extraction) is a multichannel driver for Pro Tools, Logic and Digital Performer. Latency is the period of time it takes for data to get through the audio driver and interface to an output. Low latency is desirable.
To synchronize sound and film, a clapslate can be used. The clapstick stops the TC and shows the exact time for the audio.
Transfer, or dubbing, audio from one device to another. Analog to analog results in lower quality and worse SNR (signal to noise ratio). Analog to digital transfers must take into account that there is not headroom in digital like there is in analog. Digital to digital dubbing creates an exact replica.
SMPTE is the time code (TC) developed for film based on the frame rate. HR:MIN:SEC:FRAMES is the format used. With 30 frames per second used, every 1/30 second of audio has a unique indentifying number called the TC address. MIDI TC translates SMPTE to MIDI to control MIDI devices like a keyboard.
A word clock is a signal generated in a digital audio system to control the sampling frequency. You must synchronize the word clocks in individual devices so that when tranferring the data signal is not degraded. A master clock sends a synchronization signal to the slave devices to synchronize all of the devices to the same sampling frequency.
Jitter is a variation in time from sample to sample that causes changes in the shape of the audio waveform and is caused by a degradation in word clock signals among digital devices. It results in lower detail and harsher sound. Using a master clock generator with low jitter and balanced, well-sheilded digital cables will minimize the effect.
An audio driver is a low-level program that allows the transfer of audio signals from/to audio interface. DAE (Digital Audio Extraction) is a multichannel driver for Pro Tools, Logic and Digital Performer. Latency is the period of time it takes for data to get through the audio driver and interface to an output. Low latency is desirable.
To synchronize sound and film, a clapslate can be used. The clapstick stops the TC and shows the exact time for the audio.
Transfer, or dubbing, audio from one device to another. Analog to analog results in lower quality and worse SNR (signal to noise ratio). Analog to digital transfers must take into account that there is not headroom in digital like there is in analog. Digital to digital dubbing creates an exact replica.
Monday, November 10, 2008
Project Proposal
As you know, there is a paper and presentation due the second-to-last week of classes. [Monday, December 8 or Wednesday, December 10 depending on which section you are in.] In order to get you started early on this, a proposal of your topic is due next week. Please write a couple paragraphs or an outline of what you would like to research for the paper and presentation. Please turn in your proposal in class, not on your Sound Journal. To remind you, here are the guidelines for the project:
Topic can be any of the following:
- Historical uses or developments of audio
- An event, device, or person involved in the history of audio
- An alternative use of sound
- An area of audio that you are interested in pursuing
The final paper will be 2 pages in length, double-spaced with standard font and font sizes.
Presentations must be 5-10 minutes long and discuss your topic and findings.
Please let me know if you have any questions!
Helpful links:
AIA Library - search the online databases and the library resources
Google Scholar - has more validated sources than standard Google
Topic can be any of the following:
- Historical uses or developments of audio
- An event, device, or person involved in the history of audio
- An alternative use of sound
- An area of audio that you are interested in pursuing
The final paper will be 2 pages in length, double-spaced with standard font and font sizes.
Presentations must be 5-10 minutes long and discuss your topic and findings.
Please let me know if you have any questions!
Helpful links:
AIA Library - search the online databases and the library resources
Google Scholar - has more validated sources than standard Google
Thursday, November 6, 2008
Digital Recording
In analog recording, the waveform of a signal being processed resembles the waveform of the actual sound itself. The analog signal is a continuous waveform, made up of an infinite number of points. In order to translate the analog signal into discrete digital levels, there are a 2 main parameters to determine- the sampling rate and the bit rate.
Sampling rate is the frequency that a digital sample is taken of the analog waveform. This is measured in Hertz, and is defined as the number of samples per second. An example of a common sampling rate is 44.1 KHz, which means that 44,100 samples are taken per second.
The Nyquist Frequency is defined as the minimum sampling rate that can be used without losing data about the signal. It is 2 times the highest frequency. For example if the highest frequency in a song is 20kHz, you must double this to get the minimum sampling rate of 40kHz. Higher sampling rates (or "oversampling") have a wider/flatter frequency response, leading to more clarity and detail in the sound, with low noise and low distortion.
Sampling is the time component, while quantization is the level component (like amplitude for the amplitude signal). Quantization is the process that converts voltages of each sample into a discrete quantity and assigns bit values.
Bits are binary digits, using a base 2 system with either 1 or a 0 value. Decimal is a base 10 system that you are used to - using ten values, 0 -9. In binary, a voltage on represents a 1 while voltage off represents a 0. A digital word is a combination of 1s and 0s that create a distinct value. 10 is a 2 bit word - each n-bit word makes 2^n discrete levels.
1 bit = 2 levels = 1, 0
2 bit = 4 levels = 00, 11, 10, 01
3 bit = 8 levels = 000, 001, 011, 111, 100, 101, 010, 110
etc.
The longer the word length the more quantizing levels that are available and the more dynamic range allowed, or resolution. The higher resolution allows for a more accurate representation of an anolog signal.
Some additional information on binary is available here.
Digital audio is expressed with both items: 16 bit resolution and 44.1 kHz sampling rate.
An example of an early digital system is the DAT recorder, or Digital Audiotape. However, the audio industry is generally going more towards a tapeless system, using flash/hard disk recorders instead.
Compression is a method to reduce the size of an audio file, by representing information with the fewest number of bits possible and removing redundancy. MP3s are an example of a codec algorithm that reduces the size from a WAV or AIFF audio file.
Interfacing with digital systems uses different protocols. The most common is the AES/EBU which is a pro audio connection interface standard that is used in XLR cables. Firewire and USB are low cost, flexible, compatible standards used as well.
ISDN (integrated services digital network) is bases on a public telephone all-digital network so that audio professionals can record from across the country without audio quality losses.
Sampling rate is the frequency that a digital sample is taken of the analog waveform. This is measured in Hertz, and is defined as the number of samples per second. An example of a common sampling rate is 44.1 KHz, which means that 44,100 samples are taken per second.
The Nyquist Frequency is defined as the minimum sampling rate that can be used without losing data about the signal. It is 2 times the highest frequency. For example if the highest frequency in a song is 20kHz, you must double this to get the minimum sampling rate of 40kHz. Higher sampling rates (or "oversampling") have a wider/flatter frequency response, leading to more clarity and detail in the sound, with low noise and low distortion.
Sampling is the time component, while quantization is the level component (like amplitude for the amplitude signal). Quantization is the process that converts voltages of each sample into a discrete quantity and assigns bit values.
Bits are binary digits, using a base 2 system with either 1 or a 0 value. Decimal is a base 10 system that you are used to - using ten values, 0 -9. In binary, a voltage on represents a 1 while voltage off represents a 0. A digital word is a combination of 1s and 0s that create a distinct value. 10 is a 2 bit word - each n-bit word makes 2^n discrete levels.
1 bit = 2 levels = 1, 0
2 bit = 4 levels = 00, 11, 10, 01
3 bit = 8 levels = 000, 001, 011, 111, 100, 101, 010, 110
etc.
The longer the word length the more quantizing levels that are available and the more dynamic range allowed, or resolution. The higher resolution allows for a more accurate representation of an anolog signal.
Some additional information on binary is available here.
Digital audio is expressed with both items: 16 bit resolution and 44.1 kHz sampling rate.
An example of an early digital system is the DAT recorder, or Digital Audiotape. However, the audio industry is generally going more towards a tapeless system, using flash/hard disk recorders instead.
Compression is a method to reduce the size of an audio file, by representing information with the fewest number of bits possible and removing redundancy. MP3s are an example of a codec algorithm that reduces the size from a WAV or AIFF audio file.
Interfacing with digital systems uses different protocols. The most common is the AES/EBU which is a pro audio connection interface standard that is used in XLR cables. Firewire and USB are low cost, flexible, compatible standards used as well.
ISDN (integrated services digital network) is bases on a public telephone all-digital network so that audio professionals can record from across the country without audio quality losses.
Mid-Term Exam
The mid-term exam will be next week (November 10 @ 6pm or November 12 @ 8am). Please study Chapters 1, 2, 4, and 5 from the required textbook. You are still responsible for a Sound Journal entry by Sunday at midnight this (and every) week! Start thinking about what you might want your final paper and presentation to be on, as proposals will be due in 2 weeks. Let me know if you have any questions, as always.
Monday, October 27, 2008
Week 4 Homework assignment
Write a couple good paragraphs on the SSL Duality console. Research it at the SSL website here - read the documentation, watch the videos, etc. Look also in Mix magazine or other online sources - do a Google search for "SSL Duality" and you should find plenty. Compare what the Duality has to what is discussed in Chapter 5 reading about consoles. Is it an analog or digital console? Does it have VU meters? Is there anything else about the console that interests you? etc.
Turn this in next week during class.
Turn this in next week during class.
Sunday, October 26, 2008
Audio Consoles
Consoles (also known as boards or mixers) take input signals and amplify, balance, process, combine and route them for broadcasting or recording. Some consoles also store operational data.
There are a few broad categories of consoles: analog or digital, on-air broadcast or production/post-production, and software based virtual consoles.
Analog consoles use physical electric circuits and wiring to control that audio signals flow throughout the board.
The patch bay is used for flexibility in the signal flow. Jacks are wired to console components to allow for re-routing of signals both within the console as well as with other equipment outside of the board, such as an external Equalizer or Compressor unit. Patch cords are the cables used to plug into the jacks to direct signal flow.
Digital consoles convert the analog signal to digital at the input. Routing is done on board as well as through patching cords.
Virtual consoles use a simulated console on a computer screen with specialized software.
On-air broadcast consoles handle audio sources that are immediately available to an audience, along with pre-packaged audio for playback. It has more control for rapid transitions between the DJ, music, advertisements, guests, call-ins, and others.
Inputs bring the signals from the microphone, CD player, recorder, phone calls, etc. into the console. Faders control the loudness of each channel.
The master fader handles the combined signals feeding to it from the input channels and then outputs the combined final signal.
Features found on most consoles include the input/output channel strips, input selector (high versus low level signal), phantom power (48 V), mic pre amp (boosts weak mic signal to a voltage of useable level to minimize noise introduced by console), trim/gain (changes input sensitivites- boosts low level or prevents overload in high levels), pad (lowers power of signal when trim isn't enough), channel assignment and routing, pan pot (panoramic potentiometer - shifts proportion of sound to any point between L and R), EQ and filtering (alters signal's frequency response by raising or lowering level of a range of freencies or attenuating freq at certain levels), signal procession, FX sends to external signal processing, and Solo/Mute buttons.
There are also VU (Volume Unit) meters that indicate the loudness level based on a perceived volume, reading the average of the combined waveforms. Includes the percent of modulation in relation to the maximum of the system's ability as well as the dB level. Peak Meters are electronic instead using LEDs but have the same purpose as the VU.
Automation is used to store mixer settings for recall, in order to compare mix settings of a recording to one another. Write - creates automated mix where data is monitored and stored. Read - plays back automation data. Update - makes changes to automation.
Balanced cables are used because they resist electrical interference and therefore result in lower noise.
Digital consoles do not have an physical connections or circuits, all functions go through the central processor. This removes duplication of many channels, and increases the number of operations that can be done in a smaller area.
Digital Control Surfaces are similar to consoles but there are no audio signals in them, only control circuits that send digital instructions to other units for signal processing. These are used with Digital Audio Workstations (DAWs) like Pro Tools.
Helpful Links:
Cable information (XLR, TRS, etc)
There are a few broad categories of consoles: analog or digital, on-air broadcast or production/post-production, and software based virtual consoles.
Analog consoles use physical electric circuits and wiring to control that audio signals flow throughout the board.
The patch bay is used for flexibility in the signal flow. Jacks are wired to console components to allow for re-routing of signals both within the console as well as with other equipment outside of the board, such as an external Equalizer or Compressor unit. Patch cords are the cables used to plug into the jacks to direct signal flow.
Digital consoles convert the analog signal to digital at the input. Routing is done on board as well as through patching cords.
Virtual consoles use a simulated console on a computer screen with specialized software.
On-air broadcast consoles handle audio sources that are immediately available to an audience, along with pre-packaged audio for playback. It has more control for rapid transitions between the DJ, music, advertisements, guests, call-ins, and others.
Inputs bring the signals from the microphone, CD player, recorder, phone calls, etc. into the console. Faders control the loudness of each channel.
The master fader handles the combined signals feeding to it from the input channels and then outputs the combined final signal.
Features found on most consoles include the input/output channel strips, input selector (high versus low level signal), phantom power (48 V), mic pre amp (boosts weak mic signal to a voltage of useable level to minimize noise introduced by console), trim/gain (changes input sensitivites- boosts low level or prevents overload in high levels), pad (lowers power of signal when trim isn't enough), channel assignment and routing, pan pot (panoramic potentiometer - shifts proportion of sound to any point between L and R), EQ and filtering (alters signal's frequency response by raising or lowering level of a range of freencies or attenuating freq at certain levels), signal procession, FX sends to external signal processing, and Solo/Mute buttons.
There are also VU (Volume Unit) meters that indicate the loudness level based on a perceived volume, reading the average of the combined waveforms. Includes the percent of modulation in relation to the maximum of the system's ability as well as the dB level. Peak Meters are electronic instead using LEDs but have the same purpose as the VU.
Automation is used to store mixer settings for recall, in order to compare mix settings of a recording to one another. Write - creates automated mix where data is monitored and stored. Read - plays back automation data. Update - makes changes to automation.
Balanced cables are used because they resist electrical interference and therefore result in lower noise.
Digital consoles do not have an physical connections or circuits, all functions go through the central processor. This removes duplication of many channels, and increases the number of operations that can be done in a smaller area.
Digital Control Surfaces are similar to consoles but there are no audio signals in them, only control circuits that send digital instructions to other units for signal processing. These are used with Digital Audio Workstations (DAWs) like Pro Tools.
Helpful Links:
Cable information (XLR, TRS, etc)
Thursday, October 23, 2008
Music Industry information
Music Industry journal - interesting information on the latest developments in the music industry.
Do you have any other good music industry links? Please let me know!
Do you have any other good music industry links? Please let me know!
Wednesday, October 22, 2008
Tuesday, October 21, 2008
Microphones
A microphone is a transducer, which converts acoustic energy into electric energy. The element used in a mic names it.
Moving Coil (Dynamic) mics are generally more rugged, with less self-noise, and higher SPL without distortion, as well as less expensive. However the transducing element has more mass and therefore has a slower response to transients (quick attacks).
Ribbon mics are good for voice and music recording, but are more expensive and delicate. The transient response is better than the Dynamic.
Capacitor mics have the best transient response, but are also most epensive. They are good for distance miking and have low noise.
Polar Response Pattern is the direction a mic will receive sound. It is NOT related to the transducer in the mic.
Omnidirectional - all around the mic
Bidirectional - front and back
Unidirectional - front only
Cardioid - heart shaped
For all patterns, higher frequencies result in a more directional pattern, while lower frequencies have a less directional pattern.
Mic Sound Responses depend on:
- Magnetic Induction mics have a fixed magnet and a movable diaphragm.
- Moving Coil, or Dynamic, mics have a small lightweight coil in the magnetic field that moves in response to the sound waves hitting the diaphragm and creates a voltage proportional to the sound pressure level.
- Ribbon mics have a metal ribbon instead of a coil. The ribbon vibrates from the SPL and creates a voltage. Older ribbon mics used a vertical piece of metal, while newer ones use a horizontal ribbon which allows for the capsule to be smaller and lighter.
- Active Ribbon use an amplifier system that requires Phantom Power - remotely powering along an audio cable. This allows for a higher output and sound sensitivity.
- Variable Capacitance mics transduce energy using voltage variation instead of magnetic
- Capacitor mics have 2 parallel plates separated by a space . The front plate is a thin metalized plastic and the fixed back plate form a capacitor that holds charge. As sound hits the diaphragm, the capacitance changes and is translated into a relative level.
- Tube mics use a tube circuit to sound more present and livelier than the standard mic.
- Lavaliere mics use an electret diaphragm hat permanently holds a charge.
Moving Coil (Dynamic) mics are generally more rugged, with less self-noise, and higher SPL without distortion, as well as less expensive. However the transducing element has more mass and therefore has a slower response to transients (quick attacks).
Ribbon mics are good for voice and music recording, but are more expensive and delicate. The transient response is better than the Dynamic.
Capacitor mics have the best transient response, but are also most epensive. They are good for distance miking and have low noise.
Polar Response Pattern is the direction a mic will receive sound. It is NOT related to the transducer in the mic.
Omnidirectional - all around the mic
Bidirectional - front and back
Unidirectional - front only
Cardioid - heart shaped
For all patterns, higher frequencies result in a more directional pattern, while lower frequencies have a less directional pattern.
Mic Sound Responses depend on:
- Frequency response - the range of frequencies that a mic will reproduce at an equal level with a margin of 3 dB. Essentially the "coloration" of the mic.
- Overload - how well the mic handles the distortion if sound levels are too high. Ribbon mics are very sensitive to this and possibly damaged if overloaded.
- Max SPL - the level where output starts to distort - harmonics are introduced by the audio system that are not in the original signal. 120 dB = good, 135 dB = very good, 150 dB = excellent
- Sensitvity - measure the volltage the mic produces at SPL that indicates loudness. A capacitor has the loudest response, awhile ribbon has the quietest and requires more gain.
- Self Noise - the SPL that creates the same voltage as inherant electrical noise. 40 dB = fair, 30 dB = good, 20 dB or less = excellent
- Signal to Noise Ratio (SNR) - the difference between signal and noise levels in dB. [SPL/self noise] The larger the SNR, the less noise in the signal. 64dB = fair, 74 dB= good, 84 dB = excellent
- Proximity effect - bass frequencies increase when close to microphone. Pressure is raised at lower frequences, so Bass Roll Off is used to attenuate the bass frequencies and negate the effect.
- Hum - stray Alternating Current at 50-60 Hz occurs especially in dynamic mics. Use balanced cables to reduce this. Humbucking cancels induced hum with a 50/60Hz component.
Acoustics
Acoustics is the science of the sound wave's behavior. It is based upon the interaction of the sound waves in spaces with surfaces, where the sound is reflected, absorbed, or diffracted. Psychoacoustics is the study of how humans hear and perceive sound.
Direct sound is the first wave that reaches a listener's ears, directly from the sound source.
Early Reflections are the sound waves that bounce off the walls or other objects in the room and reach the listener withing 30 ms after the Direct sound. These Early reflections combine with the Direct sound to create the entire Initial sound.
Reverberant sound is comprised of the rest of the reflections as the amplitude and time between them decreases.
Early reflections add loudness and fullness to the direct sound and help define the subjective size of the room. They also hold the most energy of the sound.
Reverb Time (RT) is defined as the time it takes a sound to decrease 60 dB.
A longer RT gives the perception of a larger and hard-surface room, as well as being farther from the sound source. An RT of 1 sec or more is considered "live", while an RT of 1/2 sec or less is considered "dry".
An Echo is a distinct repetition of the Direct sound, and therefore occurs 35ms or more after the Direct sound, so that it will not be combined with the direct sound like early reverberations are.
The Haas, or precedence effect, occurs when two identical sound waves of the same intensity come from two separate sources, but one is delayed by 30-40ms. The sound emitted by the closest location is heard first. This creates the impression that the sound comes only from that closer location. The Haas effect is what allows us to perceive sounds as coming from between two stereo speakers.
Acoustic aspects:
Warmth - achieved with lower frequencies and increased RT
Intimacy - early reflections around 15-20 ms give a sense of closeness to the source
Clarity - the ability to hear sonic details, such as lyrics and attack of notes, when there is more direct sound than reverberant
Auditory Source Width (ASW) - the perception of the sounds width versus the actual, which is affected by the room, the microphone used, and the panning
Reverb Times used for:
Speech = 0.3-0.5 sec
Pop/Rock = 0.5 - 0.6 sec
Orchestral = 1.5 - 2 sec
Studio design is based on 4 factors: Sound Isolation, Room Dimensions, Room Shape, and Acoustics. All four have the ultimate goal of lowering noise - which is essentially any unwanted sound.
Noise Criteria (NC) is a rating system used to identify background noise levels. Similar to the Equal Loudness Curves, there is a graph that shows the dB needed for every frequency level and different NC values.
Isolation - Keep noise out and the sound source from going out of the studio.
Dimensions - The dimensions of a room add coloration to the recording, as certain frequencies are reinforced through resonance. Resonance occurs when a material has the same natural frequency as the sound wave and results in vibrating together in phase so that the amplitudes increase. Remove resonance in a room by creating Height X Width X Length ratios that are not multiples of one another (ie 10 X 20 X 30).
Shape - Parallel walls reinforce sound waves, since the angle the sound wave hits the wall equals the angle of reflection and bounces right back to the opposite side. To break up these parallel reflections, design the room with non-right angle corners or rounded walls.
Acoustics - Different surfaces interact with sound waves differently - they may absorb, reflect, diffract, or diffuse the sound. Absorption soaks up the sound, which results in a lifeless, dry sound. Reflection results in a complete bouncing off of the sound from the material, giving reverberation. Diffraction bends the sound around the surface, while diffusion reduces the amplitude of a wave and spreads it in time.
Link to more information on sound diffusion
Link to more information on sound diffraction
The Sound Absorption Coefficient goes from 1.0 for complete absorption to 0.0 for complete reflection. Different materials are assigned a value in this range based on how they react to sound.
Listen to interesting binaural microphone effects here
Direct sound is the first wave that reaches a listener's ears, directly from the sound source.
Early Reflections are the sound waves that bounce off the walls or other objects in the room and reach the listener withing 30 ms after the Direct sound. These Early reflections combine with the Direct sound to create the entire Initial sound.
Reverberant sound is comprised of the rest of the reflections as the amplitude and time between them decreases.
Early reflections add loudness and fullness to the direct sound and help define the subjective size of the room. They also hold the most energy of the sound.
Reverb Time (RT) is defined as the time it takes a sound to decrease 60 dB.
A longer RT gives the perception of a larger and hard-surface room, as well as being farther from the sound source. An RT of 1 sec or more is considered "live", while an RT of 1/2 sec or less is considered "dry".
An Echo is a distinct repetition of the Direct sound, and therefore occurs 35ms or more after the Direct sound, so that it will not be combined with the direct sound like early reverberations are.
The Haas, or precedence effect, occurs when two identical sound waves of the same intensity come from two separate sources, but one is delayed by 30-40ms. The sound emitted by the closest location is heard first. This creates the impression that the sound comes only from that closer location. The Haas effect is what allows us to perceive sounds as coming from between two stereo speakers.
Acoustic aspects:
Warmth - achieved with lower frequencies and increased RT
Intimacy - early reflections around 15-20 ms give a sense of closeness to the source
Clarity - the ability to hear sonic details, such as lyrics and attack of notes, when there is more direct sound than reverberant
Auditory Source Width (ASW) - the perception of the sounds width versus the actual, which is affected by the room, the microphone used, and the panning
Reverb Times used for:
Speech = 0.3-0.5 sec
Pop/Rock = 0.5 - 0.6 sec
Orchestral = 1.5 - 2 sec
Studio design is based on 4 factors: Sound Isolation, Room Dimensions, Room Shape, and Acoustics. All four have the ultimate goal of lowering noise - which is essentially any unwanted sound.
Noise Criteria (NC) is a rating system used to identify background noise levels. Similar to the Equal Loudness Curves, there is a graph that shows the dB needed for every frequency level and different NC values.
Isolation - Keep noise out and the sound source from going out of the studio.
Dimensions - The dimensions of a room add coloration to the recording, as certain frequencies are reinforced through resonance. Resonance occurs when a material has the same natural frequency as the sound wave and results in vibrating together in phase so that the amplitudes increase. Remove resonance in a room by creating Height X Width X Length ratios that are not multiples of one another (ie 10 X 20 X 30).
Shape - Parallel walls reinforce sound waves, since the angle the sound wave hits the wall equals the angle of reflection and bounces right back to the opposite side. To break up these parallel reflections, design the room with non-right angle corners or rounded walls.
Acoustics - Different surfaces interact with sound waves differently - they may absorb, reflect, diffract, or diffuse the sound. Absorption soaks up the sound, which results in a lifeless, dry sound. Reflection results in a complete bouncing off of the sound from the material, giving reverberation. Diffraction bends the sound around the surface, while diffusion reduces the amplitude of a wave and spreads it in time.
Link to more information on sound diffusion
Link to more information on sound diffraction
The Sound Absorption Coefficient goes from 1.0 for complete absorption to 0.0 for complete reflection. Different materials are assigned a value in this range based on how they react to sound.
Listen to interesting binaural microphone effects here
Monday, October 13, 2008
What is Sound
What is sound? Sound is the vibrations of molecules in the air as the energy is passed in waves of compression and rarefactions. It can be pictured similar to when a stone is dropped in a lake, and the radiation of the waves from that point.
The five main components of sound are: Frequency, Amplitude, Velocity, Wavelength, and Phase.
Frequency is the number of cycles a wave completes in one second, and is measured in Hertz (Hz). For example, if 500 cycles per second are completed, the frequency is 500Hz.
Pitch is how humans psychologically perceive frequency.
Amplitude is the wave's maximum height (crest) and minimum (trough). This is perceived as loudness. Loudness is measured in decibels, which is on a logarithmic scale in powers of 10. In order to double loudness, an increase of 6 dB-SPL is needed.
Velocity is the speed of a sound wave. Generally this is 1130 ft/sec at 70 degrees F at sea level through air. Thicker materials slow down the molecules - 4800 ft/sec in water, 11700 ft/sec in wood. Also, as air warms up, the speed of sound increases, and the inverse is also true.
The wavelength is exactly what is sounds like - the distance a sound wave travels to complete one cycle. Longer wavelengths correlate to lower frequencies, while higher frequencies have shorter wavelengths.
Sound movement examples: http://www.fearofphysics.com/Sound/dist.html
Decibel vs. loudness charts: http://www.gcaudio.com/resources/howtos/loudness.html
Reflection examples: http://www.kettering.edu/~drussell/Demos/reflect/reflect.html
Questions:
1. What is timbre? How does it define different instruments?
2. How does the inner ear convert sound waves to electrical pulses for the brain to understand?
The five main components of sound are: Frequency, Amplitude, Velocity, Wavelength, and Phase.
Frequency is the number of cycles a wave completes in one second, and is measured in Hertz (Hz). For example, if 500 cycles per second are completed, the frequency is 500Hz.
Pitch is how humans psychologically perceive frequency.
Amplitude is the wave's maximum height (crest) and minimum (trough). This is perceived as loudness. Loudness is measured in decibels, which is on a logarithmic scale in powers of 10. In order to double loudness, an increase of 6 dB-SPL is needed.
Velocity is the speed of a sound wave. Generally this is 1130 ft/sec at 70 degrees F at sea level through air. Thicker materials slow down the molecules - 4800 ft/sec in water, 11700 ft/sec in wood. Also, as air warms up, the speed of sound increases, and the inverse is also true.
The wavelength is exactly what is sounds like - the distance a sound wave travels to complete one cycle. Longer wavelengths correlate to lower frequencies, while higher frequencies have shorter wavelengths.
Sound movement examples: http://www.fearofphysics.com/Sound/dist.html
Decibel vs. loudness charts: http://www.gcaudio.com/resources/howtos/loudness.html
Reflection examples: http://www.kettering.edu/~drussell/Demos/reflect/reflect.html
Questions:
1. What is timbre? How does it define different instruments?
2. How does the inner ear convert sound waves to electrical pulses for the brain to understand?
Subscribe to:
Posts (Atom)