ScalaColliderUGens-core/de.sciss.synth.ugen/Pitch

Pitch

object Pitch extends ProductReader[Pitch]

An autocorrelation based pitch following UGen. It is more accurate than ZeroCrossing , but more also more CPU costly. For most purposes the default settings can be used and only in needs to be supplied.

The UGen has two outputs: The first output is the frequency estimate in Hertz, the second output is a toggle hasFreq , which tells whether a pitch was found (1) or not (0). If the clarify argument is used, hasFreq has more fine grained information.

The pitch follower executes periodically at the rate specified by execFreq in cps. First it detects whether the input peak to peak amplitude is above the ampThresh . If it is not then no pitch estimation is performed, the hasFreq output is set to zero and the freq output is held at its previous value. Otherwise, the autocorrelation is calculated, and the first peak after the peak around the lag of zero that is above peakThresh times the amplitude of the peak at lag zero is reported.

===Examples===

// pitch-follower resynthesizing with saw tooth
play {
 // be careful and use headphones!
 val in      = Mix(PhysicalIn.ar(0, 2))
 val amp     = Amplitude.kr(in, 0.05, 0.05)
 val p       = Pitch.kr(in, ampThresh = 0.02, median = 7)
 val saw     = Mix(VarSaw.ar(p.freq * Seq(0.5, 1, 2), 0, LFNoise1.kr(0.3,0.1,0.1)) * amp)
 Mix.fold(saw, 6) { res =>
   AllpassN.ar(res, 0.040, Rand(0, 0.040), Rand(0, 0.040), 2)
 }
}

Companion: class

trait Product

trait Mirror

trait ProductReader[Pitch]

class Object

trait Matchable

class Any

Pitch

Type members

Inherited types

The names of the product elements

Inherited from: Mirror

The name of the type

Inherited from: Mirror

Value members

Concrete methods

Value Params

ampThresh: The minimum amplitude threshold above which the pitch follower operates. An input signal below this threshold is not analyzed. ''(init-time only)''
binsPerOct: A value which guides the search for the peak frequency in the first coarse step. Its setting does not affect the final pitch resolution; setting it larger will cause the coarse search to take longer, and setting it smaller will cause the fine search to take longer. ''(init-time only)''
clarity: If the clarity argument is greater than zero (it is zero by default) then the hasFreq output is given additional detail. Rather than simply being 1 when a pitch is detected, it is a "clarity" measure in the range between zero and one. (Technically, it's the height of the autocorrelation peak normalised by the height of the zero-lag peak.) It therefore gives a kind of measure of "purity" of the pitched signal. ''(init-time only)''
downSample: An integer factor by which the input signal is down sampled to reduce CPU overhead. This will also reduce the pitch resolution. The default value of 1 means that the input signal is not down sampled. ''(init-time only)''
execFreq: The frequency at which the pitch is estimated. This will be automatically clipped to be between minFreq and maxFreq . ''(init-time only)''
in: The signal to be analyzed.
initFreq: The initial value of the freq output, until the first valid pitch is found. ''(init-time only)''
maxFreq: The maximum frequency in Hertz to be considered for reporting. ''(init-time only)''
median: This specifies the length of a median filter applied to the frequency output estimation. With the default value of 1 the filter is defeated. Median filtering can help eliminating single spikes and jitter. This will however add latency to the output. ''(init-time only)''
minFreq: The minimum frequency in Hertz to be considered for reporting. ''(init-time only)''
peakThresh: This is a threshold used to find the first peak in the autocorrelation signal which gives the reported frequency. It is a factor of the energy of the signal (autocorrelation coefficient at zero). Set this value higher (e.g. to 1 ) to eliminate false frequencies corresponding to overtones. ''(init-time only)''