Sound Synthesis & Sequencing. Mattia G. Bergomi. April 5, Contents 1. List of Figures 2. List of Tables 2

Transcript

1 Sound Synthesis & Sequencing Mattia G. Bergomi April 5, 2013 Contents Contents 1 List of Figures 2 List of Tables 2 I Sound Synthesis: Some Theoretical Hints and SoundForge 3 1 Introduction Main Digital Synthesis Techniques Taxonomy of Digital Synthesis Techniques Nyquist-Shannon Sampling Theorem, Aliasing and Anti-aliasing Complex Signal Sampling Anti-aliasing

2 3 SoundForge - Notes by Antonello d Aguanno Waveforms Some Advices on EQ Exercises and Conclusions II Reaper - Sequencing & Audio Editing 30 4 Introduction Comparison Some practice (1) Getting started Settings - Creating Your First Project Bibliography 39 List of Figures 2.1 Sine-wave at frequency 1Hz Sine-wave at frequency 1Hz - 2Hz sampling frequency Sine-wave at frequency 1Hz - 3Hz sampling frequency Sine-wave at frequency 1Hz - 1.5Hz sampling frequency complex signal sampling Waveforms: sinusoidal, square, triangular, sawtooth Reaper main screen Track controls Audio settings Midi settings VST settings Media settings Project settings List of Tables 1.1 A taxonomy of digital synthesis techniques

3 Part I Sound Synthesis: Some Theoretical Hints and SoundForge 3

4 Chapter 1 Introduction This part has been largely borrowed from [1] and [2]. With the aid of suitable output equipment, the numbers which a modern digital computer generates can be directly converted to sound waves. The process is completely general, and any perceivable sound can be so produced. The greatest potentiality of a computer is that there are no theoretical limitations to its performance as a source of musical sounds, in contrast to the performance of ordinary instruments. In some sense one could think to work with an infinite keyboard and to generate either infinite ascending or descending scale. Exercise 1. Try to understand how an infinite sound can be synthesized. [Hint: human ears can perceive a finite range of frequencies...] The two fundamental problems in sound synthesis are Problem 1. the vast amount of data needed to specify a pressure function (hence the necessity of a very fast program). Problem 2. the need for a simple, powerful language in which to describe a complex sequence of sounds. The first problem has been solved to a large extent by the march of technology. Digital processor performance has increased at the rate of more than 40% per year for the past fifteen years, and the trend shows no sign of weakening. At present, multiple voices of many synthesis techniques can be sustained in real time on a personal computer. Problem 2 remains unsolved, and cannot, in principle, ever be completely solved. Since it takes millions of samples to make a sound, nobody has the time to type in every sample of sound for a musical piece. Therefore, sound samples must be synthesized algorithmically, or derived from recordings of natural phenomena. In any case, a large number of samples must be specified or manipulated according a much smaller set of numbers. This implies a great sacrifice of generality. The fundamental difficulty of digital synthesis is finding the smallest collection of synthesis techniques that span the gamut of musically desirable sounds with minimum redundancy. It is helpful when a technique is intuitively predictable. Predictability is good, for example, when analogies exist with well-known musical instruments, familiar sounds from daily experience, or established forms of communication. 4

5 1.1. Main Digital Synthesis Techniques 1.1 Main Digital Synthesis Techniques This section is a summary of [18] Abstract Synthesis Algorithms: in the digital world, oscillators range from basic sine, triangle, square and other simple wave generators, to more complex and hybrid systems [10, 11]. Because oscillators in some form play a role in many aspects of synthesis, we may think of them as building blocks of synthesis. We classify them as abstract because they do not necessarily arise from one particular concrete paradigm. Abstract synthesis algorithms building on oscillators include amplitude and frequency modulation and waveshaping. Frequency modulation, or the modulation of one oscillator s frequency (or phase) by another s output, was discovered by Chowning to have interesting musical implications [14, 15]. More generally, waveshaping refers to the modification of an existing signal by a non-linear function [12, 13]. Other work on waveshaping includes [14] and [15]. The study of non-linearity and chaos has led to a range of abstract sound synthesis algorithms. Synthesis from Scratch: synthesis from scratch refers to the replication of realworld sounds using physical or perceptual models, without the raw material of existing audio samples. The model may represent a sound s physical source or environment, or the perceptual characteristics desired. A frequent advantage of such model-based synthesis is the option of high-level parametric control over the synthesized sound. Think about Pianoteq by Philippe Guillaume 1. Synthesis from Existing Sounds: a third category of synthesis is the creation of sound from existing sound. As this often involves some form of analysis of the existing sound, we can also broadly refer to it as synthesis-by-analysis. In particular, we look at concatenative techniques, in which existing samples are rearranged in the time-domain, and additive synthesis, which generally takes a more spectral approach. Concatenative Techniques - A well known concatenative technique is wavetable synthesis[19, 14], the periodic repetition of a set of time-domain samples, often to create pitched instrument sounds. Concatenative synthesis uses a large database of source sounds segmented into units, a target sound to be synthesized, and a unit selection algorithm that chooses the units best matching the target according to a set of unit descriptors. The selected units are transformed as needed and concatenated in the time-domain, possibly with a cross-fade. A range of techniques with varying levels of manual control and automation fall under this umbrella, including methods for speech and singing synthesis [20] and audio mosaicing [21, 22]. Another type of concatenative synthesis is granular synthesis [23], in which sound is created by concatenating usually short sound grains. Additive Synthesis - Additive synthesis generally involves spectral analysis, and addition of the signals synthesized based on this analysis Taxonomy of Digital Synthesis Techniques In the following table the main synthesis techniques are listed: 1 5

6 1.1. Main Digital Synthesis Techniques Processed Recording Spectral Model Physical Model Abstract Algorithm Concrète Wavetable F Ruiz Strings VCO,VCA,VCF Wavetable T Additive Karplus-Strong Ext. Some Music V Sampling Phase Vocoder Waveguide Original FM Vector PARSHL Modal Feedback FM Granular Sines+Noise (Serra) Cordis-Anima Waveshaping Prin. Comp. T Prin. Comp. F Mosaic Phase Distortion Wavelet T Chant Karplus-Strong VOSIM Risset FM Brass Chowning FM Voice Subtractive LPC Inverse FFT Xenakis Line Clusters Table 1.1: A taxonomy of digital synthesis techniques. Remark 1. More details are available in [2]. 6

7 Chapter 2 Nyquist-Shannon Sampling Theorem, Aliasing and Anti-aliasing T he main sources for this chapter are [3, 4]. The first attempt to recreate the graph of a time-varying function waveform using samples is due to Augustin Cauchy. What Cauchy could not do was to predict limitations of this kind of approach. In 1928 Harry Nyquist of Bell Telephone Laboratories ( Bell Labs ) presented to the American Institute of Electrical Engineers a paper entitled Certain topics in telegraph transmission theory. In this paper he laid the groundwork for how samples of a time-varying signal could be used to exactly recreate the original signal, if certain conditions are met. In 1949 Claude Shannon, also of Bell Labs, wrote a paper entitled Communication in the presence of noise where he presented the first formal proof of the general concept presented by Nyquist. Theorem 1. [Nyquist-Shannon] The sampling frequency should be at least twice the highest frequency contained in the signal: f s 2f c, where f s is the sampling frequency and f c is the highest frequency contained in the signal. One should have at least an intuitive knowledge of the reasons which led to this result. Consider for example a signal composed of a single sine-wave at a frequency of 1 Hz: amplitude time (sec) Figure 2.1: Sine-wave at frequency 1Hz Sampling with frequency 2Hz we have 7

8 amplitude amplitude amplitude time (sec) time (sec) time (sec) Figure 2.2: Sine-wave at frequency 1Hz - 2Hz sampling frequency It is easy to see that this sampling frequency allows us to capture each peak and trough of the curve. 2 A higher sampling ratio will give 1.5 us more details about the curve: amplitude amplitude amplitude time (sec) time (sec) time (sec) Figure 2.3: Sine-wave at frequency 1Hz - 3Hz sampling frequency Sampling at a lower frequency gives the following result: amplitude amplitude amplitude time (sec) time (sec) time (sec) Figure 2.4: Sine-wave at frequency 1Hz - 1.5Hz sampling frequency that won t be enough to capture all the peaks and trough of the curve. 1 2 Exercise 2. If f s < 2f c our sampling 0.5 is not only poorer in terms of information, but definitely wrong. The amplitude person receiving these samples, without any previous knowledge of the original signal, may well be mislead into thinking that the signal has quite a different form. Using the previous pictures, try to understand why the signal could be plotted in a wrong shape if the sampling ratio is too low. amplitude amplitude time (sec) time (sec) time (sec) Can you give an intuitive definition of the term aliasing? 8

9 2.1. Complex Signal Sampling 2.1 Complex Signal Sampling Now we are ready to think about the sampling of a complex signal composed of many frequency components. By Fourier s theorem, we know that any continuous signal may be decomposed in terms of a sum of sines and cosines at different frequencies. For example, the following waveform is given by 5X sin(nx) n=1 Figure 2.5: complex signal sampling n According to the theorem 1 we have to use a frequency f s = 10Hz that is the 2 max ˆfi s.t.f c = P o ˆfi. Thus, when a signal contains not just one but many different frequencies added together, the minimal sampling rate needed to avoid aliasing is just twice whatever the highest frequency is, irrespective of how many other frequency components there are. 2.2 Anti-aliasing The simple way to avoid aliasing of course is to always have enough samples to capture the spatial or temporal variations in a signal. But there are many situations in which the sample rate may be limited for example, film is limited to 24 frames/sec, or you may be designing a web page where the images need to be a minimum number of pixels so they download quickly. So, if the sampling rate f s is fixed, then there is only one other parameter we can play with, and that is the highest frequency contained in the signal, f c. This can be done by lowpass filtering, or in layman s terms, smoothing. The easiest way to smooth a signal is to blur it. In film this is easily done by adjusting the lens so the image is a bit out of focus. The more out of focus it is, the lower the spatial-frequency content of the image becomes. And as the spatial frequency content is lowered, the temporal frequency content is also lowered. As none of us notice mysterious harmonies and frequencies appearing in our digital audio, it s safe to assume aliasing isn t too much of an issue for us to worry about. In some ways this is true, certainly when it comes to digital recording. Any digital recorder will have an Anti-Aliasing Filter. This is pretty much self-explanatory. CD audio has a sampling rate of 44.1 KHz which means that up to KHz of audio is safe to be sampled. In this case it s likely that a filter would be present to start rolling off all audio from 20 KHz onward, cutting everything off by 22 KHz. This way no unwanted audio should make it to the converter. At this point you can get into a debate about inaudible sounds and whether we hear them in different ways, which is where 96 KHz or higher sampling rates could become useful as theoretically they would allow audio to be sampled higher. 96 KHz would however ensure that even up to 20 KHz would have at least 4+ samples per cycle, which would technically increase the quality, (but remember it does increase your CPU workload, and increase your file sizes, as a result). 9

10 10

11 Chapter 3 SoundForge - Notes by Antonello d Aguanno Laboratorio di Informatica Musicale Autore : Antonello D Aguanno Sound Forge Lezione 1 Controlli generali Per selezionare una parte di segnale in Sound Forge basta ciccare con il tasto sinistro del mouse sulla parte di segnale che si vuole evidenziare e quindi spostare il mouse fino a raggiungere la selezione desiderata. Sound Forge mette a disposizione 2 tipi di zoom accessibili tramite la rotella centrale del mouse. Scorrendo con la rotella in avanti si ha uno zoom temporale del segnale (asse x) tenendo premuto il tasto Ctrl e facendo scorrere la rotella centrale si ha uno zoom sull ampiezza del segnale (asse y). Se si seleziona un area del segnale lo zoom di Sound Forge pone l area selezionata al centro dello schermo e continua ad ampliare lo zoom solo nell area selezionata. Come Registrare da una fonte esterna in Sound Forge Vogliamo ora mostrare come fare per acquisire correttamente un segnale audio tramite un microfono esterno: Passo 1 : New 11 Figura 1 Comandi generali di Sound Forge

12 Passo 2: Una volta aperta la finestra riportata in figura 2: Monitor Tasto M Figura 2 Finestra di registrazione di Sound Forge 1. Dal Tasto New Selezionare i Recording attributes impostando la frequenza di campionamento il numero di bit e il numero di canali (mono/stereo). Il passo 1 può essere saltato nel caso in cui si stia registrando su un file già esistente (occorre però comunque verificare che i parametri del file e della registrazione sia uguali) 2. In caso il Flag Monitor non sia selezionato selezionarlo e impostare i volumi del mixer audio di Windows in modo tale che provando il microfono si eviti il clipping e si ottenga un buon livello (indicativamente al limite della zona gialla). 3. Selezionare il flag DC ad just in modo da poter eliminare il DC offset sbloccando il tasto Calibrate 4. Per eliminare il DC offset premere il pulsante Calibrate prima di ogni registrazione. Il DC offset tende a variare nel tempo! 5. Premere il tasto Prepare (o Arm nella versione 7.0) e attendere che il sistema sia pronto, quindi far partire la registrazione. La preparazione serve semplicemente a rendere più reattivo il sistema permettendo di avere inizi registrazione allineati con il momento di pressione del tasto. 6. Chiudere finestra di dialogo ( Close ) 12

13 3.1. Waveforms Il tasto Reset azzera il monitor eliminando eventuali segnalazioni di Clip (scritta in bianco su sfondo rosso che compare in alto sopra il monitor). La combo box Mode: seleziona come gestire eventuali prese multiple. Il tasto M inserisce ogni volta che viene premuto un nuovo marker di regione senza fermare la registrazione in corso. Le regioni in Sound Forge sono dei marker grafici utilizzati per evidenziare particolari pezzi di segnale di particolare interesse. Non hanno alcun valore dal punto di vista del segnale non andando ad applicare nessuna modifica al file. Per selezionare tutta l area di una regione basta fare doppio click sinistro per evidenziarla tutta. Per inserire un nuovo marker posizione il cursore di Sound Forge nel punto in cui si vuole inserire il nuovo marker e quindi andare nel menù: Special Insert Marker Una volta inserito un nuovo marker lo si può spostare semplicemente trascinandolo ed editare cliccando con il tasto destro sul marker stesso. Delete : Elimina il marker Edit: Permette di modificare il suo nome, di spostarlo in un altro punto e di modificare altre opzioni. Come creare una forma d onda in Sound Forge Per inserire una forma d onda in un file occorre anzitutto creare il file audio tramite il comando New di figura 1. In questo modo si crea un nuovo file vuoto. Il file creato manterrà tutte le impostazioni relative a: frequenza di campionamento, numero di bit e numero di canali impostate precedentemente. Per creare un nuovo file vuoto con parametri diversi da quelli precedenti occorre aprire il menù: File New Nella finestra che compare si potranno inserire i parametri richiesti. Per inserire una forma d onda nel file audio si dovrà andare nel menu: Tools Synthesis Simple Si aprirà la finestra riportata in figura Waveforms In this section I want to analyze some kind of waveforms you can generate. It is important to have at least and intuitive knowledge about how those weves are obtained. What you have to learn is that synthesis is creation. To create you need both ideas and rationality. Mathematical language can give you a really clear and quick way to understand, learn and fix in your mind, how music works in its mechanical part. 13

14 3.1. Waveforms Figure 3.1: Waveforms: sinusoidal, square, triangular, sawtooth Sine waves: The sine wave is a smooth periodic C 1 (period) function. One can write it down as a function of time: x(t) =A sin(!t + ') where! =2 f is the angular frequency. Thanks to the contribution of Joseph Fourier, we know that every kind of periodic waveform can be approximated using a serie of sinewaves. A tone with a sinusoidal waveform is called a pure tone. Sinewaves are generally unpleasant and it is dificult to localize them (think about sirens, I mean alarms). The sensation they give is described as filling the room. Sawtooth waves:these waveforms are an example on non-sinusoidal waveform. Their analytical form can be expressed using Fourier series as follows x saw = 2 1X k+1 sin(k!t) ( 1) k k=1 A sawtooth wave s sound is harsh and clear and its spectrum contains both even and odd harmonics of the fundamental frequency. Because it contains all the integer harmonics, it is one of the best waveforms to use for subtractive synthesis of musical sounds, particularly bowed string instruments like violins and cellos, since the slip-stick behavior of the bow drives the strings with a sawtooth-like motion[24]. Square waves: Here is another example of non-sinusoidal waveform. In this case the transition between each minimum to the correponding maximum is istantaneous theoretically. It is not possible to realize exactly this ideal waveform. This kind of waves sound hollow and are generally used to generate wind instruments. However the wave generated by a distortion on a guitar resemble a square wave. Just for the curious, the Fourier expansion is x square (t) = 4 1X sin((2k 1)!t) 2k 1 which can be written as follows 4 13 sin(!t)+ sin(3!t)+15 sin(5!t)+ k=1 14

15 3.1. Waveforms In an ideal square wave only odd harmonics should appear. T riangle waves: Like the square wave, a triangle wave is a piecewise continuous periodic real function. Only odd harmonics are generated. The triangle wave has a strong fundemental, much weaker and rapidly decreasing odd harmonics (much more so than the square wave), and no even harmonics. A pure triangle wave has a reedy, thin sound 1. Figura 3 Finestra del menù di sintesi semplice Lo slider Amplitude imposta l ampiezza della forma d onda creata. L ampiezza è rappresentata sia in percentuale che in db, con il riferimento posto all ampiezza massima del segnale. Di conseguenza un onda con ampiezza massima avrà un valore pari a 0 db mentre una ampiezza pari a 0 sarà segnalata con un ampiezza pari a inf. Il menù Waveform Shape imposta il tipo di onda che sarà sintetizzata (onda quadra, sinusoide, onda triangolare ) Lenght imposta la durata temporale dell onda mentre Frequency ne imposta la frequenza in Hz. Il menù Insert new waveform at: imposta il punto di inizio della nuova onda all interno del file (inizio del file, posizione del cursore, fine del file). Il tasto Preview permette di ascoltare la forma d onda generata. Save As salva le impostazioni inserite nei preset della sintesi semplice. I preset sono richiamabili direttamente dal menù Preset posto in alto, nella finestra in figura

16 3.1. Waveforms Come salvare un file audio Sound Forge mette a disposizione vari tipi di file audio in cui salvare. Saranno considerati solo 2 formati: WAV e Raw Per salvare un file in Sound Forge si deve andare nel menu File di figura 1 e selezionare l opzione Save As Salvataggio file WAV Per salvare un file nel formato WAV occorre anzitutto selezionare il formato: Wave (Microsoft) dal menu di scelta dei tipi posto nella finestra relativa all opzione Save As. Dopo aver selezionato il tipo di file verificare che le specifiche relative a: frequenza di campionamento, numero di bit e numero di canali siano corrette. Per modificare queste opzioni occorre premere il bottone Custom e impostare i parametri corretti. È da notare che il formato Wave è un formato scatola che può contenere vari tipi di stream audio (PCM, MP3 ) occorre quindi impostare il menù format presente sempre nella finestra aperta dal bottone Custom. Salvataggio file Raw Per salvare un file nel formato Raw occorre anzitutto selezionare il formato: Raw File dal menu di scelta dei tipi posto nella finestra relativa all opzione Save As. Dopo aver inserito il nome premendo il bottone Salva comparirà la finestra di dialogo riportata in figura 4. Da questa finestra si dovranno impostare tutti i parametri relativi al file: Frequenza di campionamento (Sample Rate), il tipo di campioni che saranno utilizzati (Sample Type), il Format che indica se il formato con cui saranno rappresentati i campioni sarà con segno o meno e la disposizione dei Byte se di tipo Big o Little Endian. Remark: RAW Audio format or just RAW Audio is a computer format for storing uncompressed audio in raw form. Comparable to WAV or AIFF in size, RAW Audio file does not include any header information (sampling rate, bit depth, endian, or number of channels)

17 3.1. Waveforms Figure 4 Finestra relativa al salvataggio di File Raw 17

18 3.1. Waveforms Sound Forge Lezione 2 L elenco puntato in basso evidenzia gli strumenti trattati a lezione. Sono divisi seguendo la separazione adottata da Sound Forge relativa a Process e Effects. Process o EQ o Fade o Insert Silence o Mute o Normalize o Reverse o Time Stretch o Volume Effects o Noise Gate o Pitch o Reverb PROCESS EQ Un equalizzatore è un banco di filtri che ha principalmente lo scopo di compensare la diversa risposta frequenziale offerta dai vari dispositivi audio. Sound Forge mette a disposizione 3 diversi tipi di equalizzatore, ne analizzeremo 2: Graphic L equalizzatore Graphic è un esempio classico di equalizzatore che tende a replicare via software i classici equalizzatori presenti nei comuni impianti stereo domestici. In sound Forge sono presenti 3 diversi EQ grafici. La prima modalità presentata è detta Envelope. In questa modalità l utente posiziona direttamente sulla Frequency Graph (vedi figura 5) un controllo, i quadrati in figura 5, considerando sull asse orizzontale la frequenza selezionata mentre sull asse verticale il gain (positivo o negativo) impostato. È da notare che i db del gain hanno come riferimento il segnale di origine stesso di conseguenza porre un controllo (quadrato) posizionato a 0 db non altera il segnale. Ovviamente la risposta frequenziale del banco di filtri è selezionato dalla linea spezzata che congiunge tutti i controllo inseriti. L Output gain seleziona un eventuale gain generale applicato dopo il processo di equalizzazione a tutto il segnale. Il tasto Preview permette di ascoltare il segnale filtrato prima della sua applicazione definitiva. La check box Bypass se selezionata mentre si ascolta la preview permette di saltare lo stadio di equalizzazione e di ascoltare quindi il segnale originale. 18

19 3.1. Waveforms Selezionando questi controlli si passa alle altre modalità del EQ grafico. Figura 5 Finestra del Graphic EQ in modalità Envelope Le altre modalità dell equalizzatore grafico di Sound Forge mettono a disposizione 10 o 20 bande pre-costruite che permettono di gestire direttamente il segnale. Come nel caso precedente è presente uno slider di gain generale applicabile dopo il processo di equalizzazione. Parametric L equalizzatore parametrico di Sound Forge a differenza dell equalizzatore grafico analizzato in precedenza non mette a disposizione un banco di filtri bensì un solo filtro interamente programmabile. La sua finestra di controllo è riportata in figura 6. Figura 6 L'Equalizzatore parametrico di Sound Forge 19

20 3.1. Waveforms Some Advices on EQ Here follows some suggestions about standard equalization. There are no general rules for this subject, the most useful thing you can do is to practice and play with EQ. One should try to cut about five times as often as you boost. We all know that the human ear hears about 20 to 20kHz. Well the first octave, (20 to 40 Hz) is generally made up of stuff like thunder and trains. So do not boost it. Be VERY careful about boosting anything around 200 Hz. This makes your whole mix sound muddy. A little can fatten things up, but hopefully you won t have to use this. If it sounds good, it is perfect. This is way a computer can not mix a track with an algorithm. As a rule, the following facts, are true: Kick Drum - Boost at 80 Hz, cut around Hz, (removes "cardboard" sound) Snare Drum - Fatness at 240 Hz, crispness at 5 khz. Hats/Cymbals - Gong sound at 200 Hz, shimmer/air between 8-12 khz. Toms - Boom between 60 and 200 Hz, scoop out 100 Hz to 4 khz, attack at 5 khz. Bass Boost between 80 and 150 Hz, to make a passive bass sound active, scoop around 350 Hz, Boost 800 Hz for clarity or punch, and the slappin poppin thing is around 2.5 k. Electric Guitar - Rolloff at 80 Hz, cut at 800 Hz to remove "cheap" sound, and edge is around 2.5 khz. Acoustic Guitar - Bottom at 120 Hz, body at 240 Hz, clarity at 2.5 to 5 khz. Piano/Organ - Just like acoustic guitar except piano might need a little boost around 10 k. Vocals - Scoop at 240 Hz if muddy, boost at 4 k for presence, and anywhere above 8 khz to brighten (sibilance is between 8 and 10 k). Useful frequency ranges: 60Hz-200Hz: Catches the boom "boom" of tom tom s * Cut to decrease bass "boom" and increase overtones to help bring out bass lines 80Hz: Good place to boost kick drums slightly to give it some low end area of the spectrum to cut thru the mix, also advisable to cut other instruments around that area. * Rolling off the electric guitar in this area is sometimes advisable as well. 80Hz-200Hz: A decent range to boost the bass instruments. Shuffle your boost around in that area and work to not interfere with other instruments in theis range. * Boost can help add warmth and fullness to guitars, vocals and horns. 100Hz: "Oomph" of the kick drum can be centered here 100Hz-300Hz: Cut gently to remove "muddiness" (center at 180Hz). 100Hz-4kHz: A subtle and gradual scoop out of toms can emphasize the more highs and lows to still get a good tom sound and provide room in the mix for the other instruments. 20

21 3.1. Waveforms Human ear cannot detect stereo below 100Hz 120Hz and lower: These frequencies are generally responsible for warmth in a recording. Too much and the recording will sound muddy. 120Hz-125Hz: Generally the cut off point for the subwoofer signal in dedicated subwoofer setups like 5.1, 6.1, 7.1 and greater surround sound systems. This provides the rumble of the low frequencies often associated with explosions in the theatre and other dynamic action content. Also the low end of music such as kick drums and bass guitar. This are also is the bottom of an acoustic guitar and piano, a slight boost or cut here can add good sound quality, which to do kind of depends on the other instruments involved in the piece. 120Hz-600Hz: These frequencies give depth to a recording, giving vocals and other instruments a strong sense of presence without being clinical. On the other hand, these frequencies are where you re most likely to experience problems with vocal resonance. Too much in this area can be particularly fatiguing. 200Hz: A finicky area. Be careful bosting in this area as it can guickly muddy up the sound. * This is a good area to get the "gong" out of cymbals. 200Hz-1kHz: Cutting in this region can make the sound more punchy, and enhance clarity and prevent mid-range muddle. 240Hz: Fatten up a snare in this area, varying slightly from snare to snare depending on head type, shell thickness and other factors. This are also is a good range to boost acoustic guitars slightly for more fullness to the sound. Scoop vocal here if muddy. 250Hx-1kHz: Cut to remove "nasalness". 350Hz-400Hz Common area to cut the kick to remove some of the "cardboard" sound of drums Also a decent area to try to scoop the bass guitar a little bit. 600Hz-3kHz: These frequencies also give presence but of a generally harder nature. High output in this region is fairly common in rock music as it gives it a hard edge that suites the genre. 800Hz Boost the bass guitar a little bit at this level to add punch to it s sound. Cut the electric guitar in this range to remove the "cheap" sound 1kHz: Can reduce "digital coldness" by cutting gently (~3dB). Could give a softer/warmer sound. 1kHz-3kHz: Can reduce "harshness". 2kHz-4kHz: In this area you can emphasize the "smack" of the kick s beater. 2.5kHz 21