Friday, September 24, 2010

Week of September 20, 2010

Here are the reading notes we've had this week:

Chapter 10: Software Mixers

Tracks
-Audio
-Aux
-MIDI
-Instrument

Mixer Strips
-Input Selection
-Output Selection
-Insert Slots
-Send Slots

Solos

Control Grouping

Audio Grouping

Sends and Effects

Naming Buses

Internal Architecture

-Integer Notation
~The highest amplitude a 16 bit sample can handle is 65,535. Anything above this results in clipping

-Floating-Point Notation
~16-bit sample can theoretically handle any amplitude

-How they work together
~Pro Tools allows two hot signals to be summed without clipping. When bouncing in Pro Tools, the audio is converted from float into integer. If you bounce onto a 16-bit file, you lose 54dB of range

Dither
-To avoid producing repeating decimals, processors round off this data. Since the data is now incorrect and rounded off in the same way every time, distortion is produced. Dithering randomizes the rounding off so that a "low level of random noise" is created.
-Most audio sequencers ship with dither capabilities
-To avoid producing repeating decimals, processors round off this data. Since the data is now incorrect and rounded off inn the same way every time, distortion is produced.
-Most audio sequencers ship with dither capabilities

Normalization and the Master Fader
-Normalization
~Brings all signal level on a track up to the highest peak, without clipping, but rounding errors can occur, resulting in distortion, especially with 16-bit files. Use with CAUTION.
-Master Fader
~Scales mix output to the desired range of values
~Sometimes clipping will occur, even when no channels are overshooting the clipping threshold

Playback Buffer and Plugin Delay Compensation
-Playback Buffer
~Determines latency of input signals. Lower buffer size results in less latency, which is better for recording
~The mixdown should utilize a higher buffer size, because the system needs to read the information faster than it is played back

-Plugin Delay Compensation
~Plugins that run on DSP expansion, like a UAD card
~Plugin delay occurs when processing involves algorithms requiring more samples than available by each playback buffer


Chapter 11: Phase

What is Phase?
-Relationship between two or more waveforms, measured in degrees
-We only consider phase in relation to similar waveforms
-Identical waveforms are usually signs of duplication
~ex: Duplicated snare, one dry and one reverb

-Waveforms of the same event are two microphones capturing the same musical event (or recording)
~ex: A kick mic and overheads, both with kick in it

-3 Types of Phase Relationships between Similar Waveforms
~In phase or phase-coherent: waveforms start at exactly the same time
~Out of phase or phase-shifted: waveforms start at different times
~Phase inverted: both waveforms start at the same time, but amplitude is inverted

-Problems arise when similar phase shifted or phase inverted waveforms are summed
~Comb Filtering: If phase off less than 35ms, frequencies attenuated, tonal alteration and timbre change
~If waves are phase-inverted, level attenuation. If phase inverted and equal in amplitude, cancel each other out completely

-Phase in Recorded Material
~Comb filtering caused by a mic a few feet from guitar amp, picking up reflected frequencies as well as the direct sound. Not much a mixing engineer can do to fix. Caused by having two or more tracks of the same take of the same instrument can be treated by the mixing engineer:
(A) top/bottom front/back tracks: Microphones that are placed on opposite sides of an instrument are likely to pick up opposite sound pressures. Fix it by inverting the phase of one of the microphones.
(B) Close-mic and overheads: Close-miced kick or snare might interact with overhead microphones to cause phase shifting or inversion. Fix it by taking the OH as a reference and make sure the kit is phase coherent
(C) Mic and Direct: The signal from a bass guitar that is recorded DI will travel much faster that a signal that goes from guitar to an amplifier to a microphone to your console. Fix it by zooming in and nudging the track

Phase Problems During Mixdown:
-Delay caused by plug-ins
-Delay caused by digital to analog conversion when using outboard gear
-Short delays may cause comb filtering
-Equalizers cause delay in a specific range of frequencies

Tricks:
-Two mixing tricks are based on a stereo setup with both identical mono signals being sent to a different extreme, and one of the signals is either delayed or phase inverted

-Haas Trick
~Helmut Haas discovered that the direction of the sound is determined solely by the initial sound providing that (1) successive sound arrive within 1-35ms from the initial sound and (2) successive sounds are less than 10dB louder than the initial sound
%Takes the original signal panned to one extreme and the other, phase-inverted signal is sent to the other extreme with a delay of 1-35ms
%One way involves panning a mono track hard to one channel, duplicating it, and panning the duplicate hard to the opposite channel and nudging the duplicate by a few milliseconds
%Second way involves loading a stereo delay on a mono track, setting one channel to have no delay and the otehr to have a short delay between 1-35ms
~Used to:
%Fatten sounds on instruments panned to the extremes making them sound more powerful
%As a panning alternative
%To create more realistic panning, since the human ear can use teh amplitude, time, and frequency differences to locate sound
~Haas Trick controls amount of delay, level

Out of Speakers Trick
-Like Haas Trick, but instead of delaying the wet signal, just invert the phase. Results in the sound coming from all around you rather than directly at you.


Chapter 12: Faders

Sliding Potentiometer
-Simplest basis for an analog fader
-The amplitude of teh analog signal is represented in voltage
-Contains a resistive track with a conductive wiper slides as the fader moves
~Different positions along the track provide different amounts of resistance
~Different degrees of level attenuation

-Can not boost the audio signal passing through it (unless a fixed-gain amplifier is placed after it)
-Audio signal enters and leaves

VCA Fader
-Combination of a voltage controlled amplifier and a fader
-VCA is an active amplifier that audio signal passes through
~Amount of boost or attenuation is determined by DC voltage

-Fader only controls the amount of voltage sent to the amplifier
~No audio signal flows through the actual fader

-Allows a number of DC sources to be summed to a VCA
~Shortens the signal path

Digital Fader
-Determines a coefficient value by which samples are multiplied
~Doubling a coefficient of 2 results in a boost of around 6dB
~0.5 results in around 6dB attenuation

Scales
-Typical measurement is in the scale unit dB
~Strong relationship to how the human ear perceives loudness

-Scale is generally based on steps of around 10dB or 6dB
~6dB is approx. doubling the voltage (or sample value) or cutting it in half
~10dB is doubling or halving the perceived loudness

-The 0dB point is called unity gain
~Where the signal is neither boosted nor attenuated

-Most faders offer extra-gain
~Generally around 6, 10, 12dB boosts
~Only used if signal is still weak while at unity

-Area between -20dB and 0dB is the most crucial area

Level Planning
-Faders are made to go up and down
-When mixing the levels start by coming up
~Generally ending up at around the same positions

-Problem
~A natural reaction to not being able to hear a track is to bring the fader up
%Bringing a snare up in the mix might begin masking vocals, so you bring up fader on vocals, then bass masked, etc.
~Eventually, end up back where you started

-Solutions
~Having a set plan for levels before bringing up faders so the extra-gains settings are left alone
~Setting the loudest track first and bringing up the rest of the tracks around it

Extremes - Inward Experiment
-Take the fader all the way down
-Bring it up gradually until the level seems reasonable
-Mark the fader position
-Take the fader all the way up (or to a point where the instrument is too loud)
-Bring it down gradually until the level seems reasonable
-Mark the fader positions
-You should now have two marks that set the limits of a level window. Now instrument level within this window based on the importance of the instrument



Chapter 13: Panning

How Stereo Works
-Alan Dower Blumlein
~Reasearcher and engineer at EMI
~December 14, 1931, applied for patent called "Improvements in and relating to Sound-transmission, Sound-recording, and Sound-reproduction System"
~Was looking for a 'binaural sound', we call it 'stereo' today
~Ironically, first stereo recording published in 1958 (16 years after Blumlein's death and 6 years after EMI's patent rights had expired
-Stereo Quick Facts
~We hear stereo based on three criteria: (EX: trumpet on your right)
%amplitude (sound be louder in R ear than L)
%time/phase (sound will reach L ear later than R)
% frequency (less high freq in L than R)
~Sound from a central source in nature reaches our ears at the same time, with the same volume and frequencies. But, with two speakers, no center speaker, so phantom center
~Best stereo perception when triangle acheived

Pan Controls
-Pan Pot (Panoramic Potentiometer)
~First studio with a stereo system was Abbey Road, London
~Splits a mono signal L and R, and attenuates the side you're not favouring
-Pan Clock
~Hours generally span from 7:00 (L) to 17:00 (R)
-Panning Laws
~A console usually has only one panning law, but some inline consoles have one for channel path and one for monitor path
%two main principles:
^if two speakers emit the same signal at the same level, listener in the center will perceive a 3dB boost of what each speaker produces.
^when two channels summed in mono, half of each level is sent to each speaker
~0dB Pan Law: doesn't drop the levels of centrally panned signals. The instrument level will drop as we pan from the center outward, with 3dB increase of perceived loudness when centered
~-3dB Pan Law: when panned center, there is a 3dB dip (generally best option when stereo mixing)
~-6dB Pan Law: used for mono-critical applications. Provides uniform level in mono, but a 3dB dip when in stereo
~-4.5dB Pan Law: compromise between -3 and -6dB laws. 1.5dB center dip when in stereo, 1.5dB center boost in mono
~-2.5dB Pan Law: gives a 0.5dB boost when panning to the sides so instruments aren't louder when panning.
-Balance Pot
~Input is stereo, unlike pan pot. 2 input channels go through separate gain stages before reaching stereo output. Pot position determines how much attenuation applied on each channel.
~never cross-feeds the input signal from one channel to the output of the other

Mono Tracks
-Problem with dry mono track is it provides no spatial perception
-Dry mono tracks always sound out of place, so add reverb or some other spatial effect to blend it
-Some mono tracks include room or artificial reverb that doesn't sit well with a stereo reverb of the whole mix

Stereo Pairs
-Coincident Pair (XY) technique provides the best mono-compatibility given that the diaphragms of the two mics are so close in proximity, and there's no need to worry about phase complications
-Spaced Pair (AB) involves two mics a few feet apart, is certain to have phase issues, and is not mono-compatible
-Near-coincident pair is two mics angled AND spaced, with less drastic phase problems

Multiple mono tracks
-Multiple mics on the same instrument
-Mirrored panning widens and creates less focus on the instrument in the stereo image
-Same panning gives a more relative stereo image, and is easier to locate

Combinations
-Like mirrored panning but less extreme

Panning Techniques
-Look at the track sheet and get a basic idea of a tentative pan plan
-Panning strategies differ with every mix
-Small tweaks in the near final stages can greatly improve mix
-Panning instruments in the same place causes masking. Panning different directions and mirroring can avoid masking
-When panning, think of a sound stage or try to visualize an actual performance
-Center and extremes in the panning field tend to be the busiest areas wehre masking is more likely to occur
-Level and frequency balance are the main concern when panning
-Be aware of the rhythmic structure of the tracks and keep them balanced
-A close-to-perfect stereo mix is basically a good mono mix, although there is still room for imbalances
-Stereo effects (reverb/delays) can be panned towards th dry track to put the desired effect in clearer focus
-Mono effects benefit more from panning the effect farther from the dry track and enhance the stereo image

Beyond Pan Pots
-Autopanners: pans cyclically between the left and right sides
~Rate: Defined by Hz, cycles/second
~Depth: How far the signal will be panned. Higher setting = more apparent effect
~Waveform: Defines shape of panning modulation, how smooth/rigid the panning will sound
~Center: This setting defines the position of modulation


And here are the notes I took on Monday and Wednesday:

Echo/Delay/Effects Processing Automation
-Create dedicated Aux track
-Create send to that Aux track
-Go to waveform drop-down, choose your automation

Serial Compression
-Try two compressors, one at 2:1, the other at 5:1. Be fairly light on both of them

-TDM is th dedicated computer for Pro Tools
-RTAS is using the Mac's hard drive, so slight latency

-OOPS (Out-of-Phase Stereo)
~Taking it out of phase to hear what's hidden (what we did with Hey Jude)


And, for the rest of Track 2, Our production schedule is something like:
Tuesday, September 21:
-Finish EQ, gates, compressors in Pro Tools
-Experiment with automation for Aux sends

Thursday, September 23:
-Mono mix through the board! Must be finished by end of class!

Tuesday, September 28:
-Stereo mix through the board!

No comments:

Post a Comment