BIRDSONGS
Syllable Tutorial
Define Objects¶
Import package and define matplotlib backend
In [1]:
# the following line enable interact with figures,
# you can make zoom and save images from a poup matplotlib window
# to select points: qt ipympl tk. To just visualize: notebook
%matplotlib ipympl
import birdsongs as bs
from birdsongs.utils import *
playsound is relying on another python subprocess. Please use `pip install pygobject` if you want playsound to run more efficiently.
Define directories path and show audio files
In [2]:
# root_path = "path_to_repository\\'
# audios_path = "audios_path\\'
# bird_name = "Zonotrichia capensis"
#audios_path = "C:\\Users\\sebas\\Documents\\GitHub\\audios\\Dissertation-xeno\\"#rufous-collared-sparrow\\humbolt\\"
paths = bs.Paths()#audios_path=audios_path) # root_path, audios_path, bird_name
ploter = bs.Ploter(save=True) # to save figures save=True
paths.ShowFiles()
The folder has 4 songs: 0 humman.wav 1 XC513182 - Rufous-collared Sparrow - Zonotrich... 2 XC11293 - Rufous-collared Sparrow - Zonotrichi... 3 XC104508 - Ocellated Tapaculo - Acropternis or... dtype: object
Define, visualize and hear birdsong. The same funcion allow to us select the time of interest enabling SelectTime_on option
In [9]:
birdsong = bs.BirdSong(paths, no_file=1, NN=256, flim=(1e3,20e3), tlim=(0.,5),
Nt=5000, umbral=0.05, split_method="amplitud")
ploter.Plot(birdsong, FF_on=False, SelectTime_on=True)
AudioPlay(birdsong)
/home/saguileran/miniconda3/lib/python3.8/site-packages/scikit_maad-1.3.12-py3.8.egg/maad/sound/input_output.py:113: WavFileWarning: Chunk (non-data) not understood, skipping it. fs, s = wavfile.read(filename)
The son has 6 syllables
Out[9]:
Select time inverval of interest
In [11]:
#time_interval = Positions(ploter.klicker)[0]
time_interval = np.array([1.7250, 1.9457])
print("Time interval selected: initial time {:.4f} s, final time {:.4f} s".format(time_interval[0], time_interval[1]))
Time interval selected: initial time 1.7250 s, final time 1.9457 s
Define syllable by the previous time interval selected
In [12]:
syllable = bs.Syllable(birdsong, tlim=time_interval, umbral_FF=1.0, NN=birdsong.NN, Nt=30, ide="syllable")
syllable.no_syllable = 1
ploter.Plot(syllable, FF_on=True)
AudioPlay(syllable)
Out[12]:
In [13]:
plt.figure(); plt.plot(syllable.FF, 'o'); plt.show();
3D plot (in process)
In [ ]:
# ploter.Plot3d(birdsong)
In [ ]:
# ploter.Plot(birdsong, FF_on=False, syllable=syllable)
Define syllable with the default syllable extractor
In [ ]:
# no_syllable = int(input("Enter the syllable number (1 to {0}): ".format(bird.no_syllables)))
# syllable = birdsong.Syllable(no_syllable=3)
# ploter.Plot(syllable)
# AudioPlay(syllable)
In [ ]:
# ploter.Plot(birdsong, FF_on=False, syllable_on=True, chunck_on=False)
Optimal Synthetic Syllable¶
Define optimization method and its parameters
In [14]:
brute = {'method':'brute', 'Ns':11} # , "workers":-1
optimizer = bs.Optimizer(syllable, method_kwargs=brute)
#?lmfit.minimize
Find optimal $\gamma$ for all the syllables computed by the syllables extractor. You can also just find the optimal for the single syllable
In [15]:
# optimal_gamma = optimizer.OptimalGamma(syllable)
optimal_gamma = 37000 # optimizer.AllGammas(birdsong)
syllable.p["gm"].set(value=optimal_gamma)
In [19]:
syllable.p
Out[19]:
| name | value | initial value | min | max | vary |
|---|---|---|---|---|---|
| a0 | 0.01000000 | 0.11 | 0.01000000 | 0.25000000 | False |
| a1 | 0.00000000 | 0.05 | -2.00000000 | 2.00000000 | False |
| a2 | 0.00000000 | 0.0 | 0.00000000 | 2.00000000 | False |
| b0 | -0.25000000 | -0.1 | -1.00000000 | 0.50000000 | False |
| b1 | 1.00000000 | 1 | 0.00000000 | 2.00000000 | False |
| b2 | 2.00000000 | 0.0 | 0.00000000 | 2.00000000 | False |
| gm | 37000.0000 | 40000.0 | 10000.0000 | 100000.000 | False |
Optimal $\alpha$ and $\beta$
In [ ]:
# [a0, a1, a2, b0, b1, b2]
# syllable.Set([0.01, 0, 0, -0.25, 1, 2])
# syllable.p["b2"].set(value=2)
In [ ]:
#optimizer.OptimalParams(syllable, Ns=5)
Display optimal parameters and generate a synthetic syllable
In [20]:
Display(syllable.p)
synth_syllable = syllable.Solve(syllable.p)
| name | value | initial value | min | max | vary |
|---|---|---|---|---|---|
| a0 | 0.01000000 | 0.11 | 0.01000000 | 0.25000000 | False |
| a1 | 0.00000000 | 0.05 | -2.00000000 | 2.00000000 | False |
| a2 | 0.00000000 | 0.0 | 0.00000000 | 2.00000000 | False |
| b0 | -0.25000000 | -0.1 | -1.00000000 | 0.50000000 | False |
| b1 | 1.00000000 | 1 | 0.00000000 | 2.00000000 | False |
| b2 | 2.00000000 | 0.0 | 0.00000000 | 2.00000000 | False |
| gm | 37000.0000 | 40000.0 | 10000.0000 | 100000.000 | False |
/home/saguileran/miniconda3/lib/python3.8/site-packages/librosa/util/decorators.py:88: UserWarning: Empty filters detected in mel frequency basis. Some channels will produce empty responses. Try increasing your sampling rate (and fmax) or reducing n_mels. return f(*args, **kwargs)
Visualize and plot the synthetic syllable
In [21]:
ploter.PlotVs(synth_syllable)
ploter.PlotAlphaBeta(synth_syllable)
ploter.Result(syllable, synth_syllable)
ploter.Syllables(syllable, synth_syllable)
AudioPlay(synth_syllable)
Out[21]:
In [22]:
AudioPlay(syllable)
Out[22]:
Write synthetic and real syllables
In [23]:
syllable.WriteAudio()
synth_syllable.WriteAudio()
/home/saguileran/miniconda3/lib/python3.8/site-packages/scikit_maad-1.3.12-py3.8.egg/maad/sound/input_output.py:390: UserWarning: Values for bit depth should be 8, 16 or 32. Argument ignored.
warn('Values for bit depth should be 8, 16 or 32. Argument ignored.')