Adobe audition tutorial effects (noise reduction & compression & equalization)

Table of Contents

Noise reduction & restoration

De noise:

simple just move the slider to 0-100% to reduce amount of sound.

Enable output Noise only: listen only the noise level useful to cut the noise.

Gain:

Then apply gain to get actual value from reduced lower sound.

Pros: simple just like noise reduction technique in other softwares like audacity.

Adaptive noise reduction: (more parameters)

standard noise reduction only for post processing, not for live processing, so adaptive noise reduction must for live podacasts or streaming games.

Note: it’s realtime processing so need little bit extra cpu power otherwise delay due to heavy processing.

Reduce Noise By: 6-30db better, 20db by default,

Just it substract the noise. (you can monitor it while adjusting)

Noiseness: it’s indicates amount of noise in original audio.(so we can reduce accordingly)

Fine tune noise floor: manually adjusting it above or below the automatically calculated noise floor.

Signal threshold: manually adjusting desried audio level above or below the automatically calculated threshold.

Special decay rate: how quickly drops to 60db

Too short bubbles high reverb

Baseband preservation:

Saving frequency of vocals helps to boost audio quality. (Need to know about equalization)

FFT SIZE: length of frequency band. Or how many individual frequency bands are analyzed.

Processing focus:

All frequency focus

High frequency focus

High low frequency focus

Mid frequency focus

Low frequency focus

THIS PAGE

Techniques for restoring audio
Noise Reduction effect (Waveform Editor only)
Sound Remover effect
Adaptive Noise Reduction effect
Automatic Click Remover effect
Automatic Phase Correction effect
Click/Pop Eliminator effect
DeHummer effect
DeReverb effect
DeNoise effect
Hiss Reduction effect (Waveform Editor only)

Fliter and Equalizer

FFT Filter Effect:

Fast Fourier Transform an algorithm quickly analyzes frequency and amplitude of sound wave.

High pass filter(HP): Passes the high frequency and cuts the lower frequencies.

Ex: for vocal anything below 100hz cuts.

Low pass filter(LP): passes the lower frequency and cuts the higher frequencies.

Ex: for vocals above 10000hz it’s cuts

Narrow band pass filter/ telephony : 350-4000hz here only the vocals of human to avoid background noise. Telephone uses this spectrum.

Notch filter: to remove precise frequency low 60hz, monitor it just up& down , find a noise in a specific range, then use notch filter and drag drop to remove entire sound.

FFT SIZE:

Tradeoff between frequency and time accuracy.

Higher steep

Lower reduced transient artifacts

Ideal 1024 & 8192

Window

Parametric Equalizer in Adobe audition

Q/width: controls width of the effected frequency,

Low q value effects high range frequencies,

High q value effects low range frequencies (ex 100hz)

1-10 ideal for general equalization.

Equalization curve & 12db /6db octave

Q value which is a ratio of width to center frequency

Width = frequency width.

Amplitude & compression AU

Amplitude: (amplify or gain)

Dynamic processing:(cpu intensive)

dynamic= based on upper &lower it adjusts.

Noise gate effect: stops the noise below noise threshold (ex threshold,-30db then it only passes above threshold db like -29 )

Expander: just like compressor, but it compress the lower noise ex: threshold -30 db compress ratio 1:2

Then 30db becomes 15db if ratio 1:10 30-3db= 27db.

compressor: compress the above limit threshold .

Ex: voice over ranges between -20 to -10db.

Then compressor threshold limit is -10

attack time: wait before compressing after reaching threshold

release time: how long it will keep compress

Above -10 ton-1db

Ratio = 1:2 , before total db level to -1db. Now it’s -5db.

Because it’s compressed remaining 10db into 1:2 ratio means half then 10 becomes 5.

Gain: after compressing the noise level reduces we can apply some gain to -6,-3 or -1 db.

Normalisation: boosting lower level audio signals upto peak level audio signal.(make sure to completely reduce background noise)

Ex: audio track has -15 db , and -10db – 6-db -3db.

When we apply normalisation . All audio singals gained to upto -3db it’s a peak in audio track.

Before applying normalisation make sure to cut excessive wave lengh. So the audio track is innsame wavelength.

Limiter: cutting loud noise

Presets: -1db more than -1db or 0 distortion or in bearble sound.

-3 broadcast specific

-6 mostly use for voiceovers

-10db some people use it for quite voice but not recommended.

Equalization: plays vital role to remove unnecessary noise based on frequency.

attack & release time

attack time:

Determines how many milliseconds it takes for the output signal to reach the specified level

(to avoid unnecessary processing at short sudden changes )

Release time:

Determines how many milliseconds the current output level is maintained

To see good attack and release times for different types of audio content, choose various options from the Presets menu.

Fast attack & release time for drums, music instruments.

Lower attack & release time:

You can know by just doing post processing audio. To see how much time the audio at peak level.(it’s a art)

Hold time:

Dynamic processing

Level meter & gain reduction meter:

It’s need a bit practice to understand.

This works by compressing lower & higher audio signals. To maintain a level.

(Noise reduction done by expander, loudness reduced by compressor, also added bandwidth/frequency specific processing)

Dynamic effects:

4parts

Autogate: passing the volume at certain threshold

Compressor: reducing peak voice with certain ratio

Expander: reducing lower threshold with certain ratio

Limiter: totally reducing noise level for a specific threshold.

Hard limiter effect:

Maximum amplitude:

-1,-3,-6

Input boost: before cutting the audio, we pre amplify the audio to avoid audio clipping.

Look ahead time/ attack time: time to wait

Determines how quickly starts

Release time: time to release back to normal.

Determines how quickly compression stops after reaching normal.

Multiband compressor effect

just like parametric Equalizer it’s also a good & advanced effect to process audio frequently.

Main: compressing at band or frequency specific.

3 bands

Low: 100hz

Mid: 2k

High: 10k

5bands (not a scientific values)

Low /bass: upto 100hz

Low mid: 100-500

Mid 500- 2000

High mid 2k to 10k

High/ treble: 10k to 20k

(select bands by either manually entering frequencies or presets and edit later)

Vocal frequencies at 80hz to 10k

telephone frequency: 300hz- 8k even 400hz-4k.

Voice Fundamentals falls beween 80-300hz

voice Harmonics at 400-4k**

Band specific controls in multiband compressor

Solo: only selected band.

bypass: ignore the band

Threshold slider

Input level meter:

Gain reduction meter

Threshold: -10db

gain: 4db (boost or cut +4db or -4db)

Ratio:1.10

attack: 10ms

release: 100ms

speech volume leveler effect

compression effect that optimizes dialogue, evening out levels and removing background noise.

Target volume level (db):

Vocals: -20 to –3db

Leveling Amount:

Low settings: amplify speech slightly without boosting the noise floor.

High settings: amplifys entire signal as it drops to close the noise floor

Boost low signals: shorter & low volume passages (for most speech content, skip this for smooth audio)

Target dynamic range:

Minimizes background noise while amplifying and leveling speech content.

specificy target dynamic rangage

ex: -45db to -60db

higher: upper -85db greater background noise but lower amplitude & leveling.

lower: (down -30db) higher amplitude & leveling but background noise also boosts.

monitor & adjust as per noise floor or your room.

Advanced settings:

compressor : maintains strong signal by gain if processed signal falls .

Noisegate: reduces noise

Adobe audition ideal effects for live streaming & podcast

Obs or streamlabs obs with Adobe audition

Sound mapping adobe audition to obs

virtual audio cable or (voice meter banana ,fl studio)

sound card to audition >> virtual cable inout device>>obs.

adobe audition effects:

Noisegate if through vst plugins (optional).

Parametric Equalizer effect:

enable high pass filter: to remove background noises

low pass filters: other than non vocals.

Use presets: like vocal enhancer or cusomize and save as preset.

Tip: vocal frequency 100hz – 10k even lower 400-4000hz

Test& repeat.

Denoise or adaptive noise reduction:

Denoise simple: just reducing amount of noise by 0-100%

still not reduced , windows>>sound settings>> input device level (dercrease) input device boost(reduce)

cons: overall weak audio signal.

if mixer& interface: reduce gain.

monitor & test.

Adaptive noise.

automatically adjusts between ranges dynamicai, unlike static denoise.