Selecting and hearing some
voice overtones traces.
To see
pictures in their original size – just click them.
Sample 1. Create new
synthetic wave file by all traces available in original sound.
To estimate
the quality of the developed system “Signal Workshop” we propose you to take
part in next small experiment.
On the next
top picture you see PASFilm and can listen original well-known speech signal:
“Four-zero or higher” (it was taken from the CSL).
On the
bottom picture you see PASFilm and can listen new signal, synthesized by all
traces available and selected on original top PASFilm.
To
listen original sound (top PASFilm) click here.
To
listen new synthesized signal by all available traces (bottom
PASFilm) click here.
Visually
comparing the PASFilm of the original signal (on the top) and the PASFilm of
whole-synthesized signal (on the bottom) you can see some differences.
Especially on the bottom of the PASFilm on the 0 Hz frequency. But it’s very
hard to hear these differences by ears. Try to do it on Hi-Fi speaker system.
Sample 2. The residual of
subtracting original and synthesized signals from the sample 1.
On the next
bottom picture you see PASFiln of the residual of subtracting original and
synthesized signals from the sample 1. If you listen it, you can hear voiceless
whisper speech signal of the same talker.
To
listen original sound (top PASFilm) click here.
Sample 3. Compare waveform
of original and synthesized signals from the sample 1.
On the
right part of next picture you can see two waveforms of original signal (top)
and synthesized signal (bottom). It’s hard to found any differences. (On the
left part of the picture you can see PASFilm of original sound. All the markers
on power level, PASFilm and oscillogramms are at the same time point).
On the
right part of next picture you can see two waveforms of original signal (top)
and synthesized signal (bottom). It’s hard to found any differences too. (On
the left part of the picture you can see PASFilm of synthesized sound. All the
markers on power level, PASFilm and oscillogramms are at the same time point).
On the
right part of next picture you can see two waveforms of original signal (top)
and residual of subtracting original and synthesized signals from the sample 1
(bottom). You can see how pure voice signal with perfect pitch becomes
whisper-like.
All the
markers on power level, PASFilm and oscillogramms are at the same time point).
Sample 4. Selecting and
hearing only first voice overtone.
In this
experiment expert select only one – fist voice overtone – and synthesized
correspondent wave file.
On the next
top picture you can see PASFilm of original sound with yellow dots (selected
sound traces) and red dots (unselected sound traces). On the bottom of the same
picture you can see PASFilm of the synthesized wave file with the trace of only
one first voice overtone.
To
listen original sound (top PASFilm) click here.
To
listen only one first voice overtone of the same speech click here.
Sample 5. Selecting and
hearing only first and second voice overtones.
In this
experiment expert select two – fist and second voice overtones – and
synthesized correspondent wave file.
On the next
top picture you can see PASFilm of original sound with yellow dots (selected
sound traces) and red dots (unselected sound traces). On the bottom of the same
picture you can see PASFilm of the synthesized wave file with the traces of
first and second voice overtones.
To
listen original sound (top PASFilm) click here.
To
listen first two voice overtones of the same speech click here.
Sample 6. Selecting and
hearing first three voice overtones.
In this
experiment expert select three – fist, second and third voice overtones – and
synthesized correspondent wave file.
On the next
top picture you can see PASFilm of original sound with yellow dots (selected
sound traces) and red dots (unselected sound traces). On the bottom of the same
picture you can see PASFilm of the synthesized wave file with the traces of
first three voice overtones.
To
listen original sound (top PASFilm) click here.
To
listen first three voice overtones of the same speech click here.
Sample 6. Selecting and
hearing only second voice overtone.
In this
experiment expert select only second overtone and synthesized correspondent
wave file.
On the next
top picture you can see PASFilm of original sound with yellow dots (selected sound
traces) and red dots (unselected sound traces). On the bottom of the same
picture you can see PASFilm of the synthesized wave file with the traces of
second voice overtone.
In this
samples it is interesting to hear how male voice becomes like a female’s.
To
listen original sound (top PASFilm) click here.
To
listen only one second voice overtone of the same speech click here.
Sample 7. Selecting and
hearing second and fourth voice overtones.
In this
experiment expert select second and fours overtones and synthesized
correspondent wave file.
On the next
top picture you can see PASFilm of original sound with yellow dots (selected
sound traces) and red dots (unselected sound traces). On the bottom of the same
picture you can see PASFilm of the synthesized wave file with the traces of
second and fourth voice overtones.
The female
voice property becomes more brightly.
To
listen original sound (top PASFilm) click here.
To
listen second and fourth voice overtones of the same speech click here.
