alexis-michaud
commented
5 years ago Interesting situation. What may be happening here? Rapid larynx movement, with a transition from phonation mechanism m1 to phonation mechanism m2? Rapid decrease in the amplitude of the modulation in vocal fold contact area could lead to losing the signal. It is reminiscent of signal loss (or sudden reduction in signal amplitude) in rising tones in Vietnamese data (D1, sắc) by male speakers.
Looking at the audio signal, it is clear that this is not just an artefact of EGG recording, since the audio signal, too, has much reduced amplitude at that point.
The functionality has now been added to peakdet_inter. It should be self-explanatory.
disp('If some of the glottal cycles went undetected, or extra cycles were erroneously detected:')
disp('- enter 1 (one) to change the settings for automatic detection, or')
disp('- enter 2 to make changes manually (by visual detection of cycles not detected by the script).')The f0 value for each cycle is simply the f0 of the spurious cycle (cycle 7 in your example) multiplied by the number of cycles detected visually by the user (in your example: five cycles).
This functionality is to be used carefully, of course.
The added cycles can be detected automatically because they only contain 3 values: the first 3 columns (beginning of cycle, end of cycle, and f0): the rest of the columns contain zeros.
Please test & send feedback!
MinhChauNGUYEN
and here is the corresponding dEGG signal, with the same time marks:
Peakdetcan't detect the five glottal cycles in the middle of a rhyme where the intensity of voicing decreases abruptly, as in the above figure. The five small cycles are all contained inside the 7th detected cycle. This leads to a spurious f0 value, as if there were no succession of glottis openings and closures in the whole interval. The reason is that as can be seen on the DEGG signal, there are no precise peaks for these small cycles. The oscillations take place between two clear closing peaks. They have a 'hill pattern', like the one found at soft offset of voicing (see issue #1 ).How to deal with this? Here is a suggestion for a solution.
It is not a signal error (artefact): in view of the audio, it seems clear that something really happens in the larynx such as a change of phonation mechanism (from m1 to m2: 'chest voice' to 'head voice'), maybe also with some movement of the whole larynx. Although there are no precise peaks for both closing and opening, the electroglottograph still captures this phenomenon: four small oscillations, which probably correspond to five glottal cycles. Therefore, PeakDet should implement a solution to add cycles where necessary. The user will do it roughly by looking at the signal.
PS Here is a link to the signals for this example: https://www.dropbox.com/s/yctplk9u9yrzq33/Method_PeakDet_PetitPeaks_M14_1541_audio.wav?dl=0