Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[iprosoundweb]

Hy, I'm a language teacher and I'm trying to visualize pitch changes in audio tracks for spoken sentences, in order to show the phrasal stress to my students.

For the moment, this is what I do :

using sonic visualiser, I'm using the pYin pluging (monophonic pitch and note tracking) from Matthias Mauch with stems as plot type. I'm doing this on an audio file which speed I have reduced by 60%. It shows me the "bumps" in the pitch. For some sentences this works fine, like this one :

http://fr.tinypic.com/view.php?pic=121ww76&s=8

but many times, I'm getting very hatched images, like this one :

http://fr.tinypic.com/view.php?pic=aag2vc&s=8

So I'm looking for a way to visualize the pitch in a different way, without white spaces between certain letters, f.ex. like a presentation in bars, like an equalizer :

http://fr.tinypic.com/view.php?pic=8ys7if&s=8

It doesn't have to be real-time analysis, I just want a static picture.
Can this be obtained by any means?

Any program which does the job is fine with me.
Thanks for your help...

[Fletcher]

Sorry, I didn't listen to any of the clips as I don't click on links from people with whom I've never shared a beverage [virus paranoid... what can I say].  I'm not quite sure what you're trying to accomplish with the graphics -- I mean absolutely no disrespect [ignorance of what you're trying to do perhaps, but absolutely no disrespect] -- won't they get the differences in pitch a whole lot easier by hearing them [vs. seeing them]?

I remember when my daughter was learning Chinese there were "pitch" and "note emphasis" elements in each of the characters [kinda like sheet music] that illustrated what pitch / emphasis went where.  She had to learn what the character elements sounded like, and then "sing" them back when she spoke them in a sentence.  My point being [and again, I could be very well ignorant of what you're trying to do], her whole thing was completely sound oriented... the "visuals" were only applicable to when she was reading [or more like "trying to read"... but I digress].

I'm not sure if there is a way you can post those links as pictures... but if you can, that might be helpful in getting you a more intelligent response.

Peace

[iprosoundweb]

ok, obviously the approach is sound-oriented. Actually, what I do, is to play a clip of the pitch, so one can see where the "bumps" come in. Believe it or not, but phrasal stress is not obviously to catch and imitate, especially for Asians... So all means are ok : both hearing and seeing.

I'm linking three pics and can add a short clip if wished.
1: ok
2: distorted
3: desired

Thanks for any help.
Here we go :

1

hébergeur d image gratuit

2

Hébergeur d'images

3

hebergement d image

[Fletcher]

Again, let me apologize in advance for being obtuse -- this references pitch somehow?  The first two pictures look like "amplitude", and the 3rd one looks like a spectrum analysis.  None of the associated graphs have any kind of reference nomenclature so its kind of difficult for me to understand them.

That said -- and please understand that I would like to understand -- so far, I don't... BUT... I'm wondering if any of the "auto-tune" programs might have some kind of "pitch" representation.  I have never used one so I can't say for sure -- but it seems like that might be a way to go if they have any kind of "initial pitch" graphic representation.

Wish I could be of better assistance!!  FWIW, I'm heading to the NAMM show in Anaheim this weekend... I will try to give a look-see - if I run across anything that might be of some assistance to you [or at least what I think you're looking for] I will absolutely post what I find on the board here.

Again, apologies for my ignorance on this subject... I'm really more of an "analog" / "I know which microphones to choose and where to put them" kind of guy.

Peace

[boz6906]

Greetings!

I may have some insight into your subject.

My wife is a speech-lang pathologist, they sometimes use waveforms similar to your fig 1 to illustrate vocal pitches or what they call 'formants' which which can be shown a couple different ways.

Your fig 1 show amplitude on the y-axis and time on the x-axis, like an oscillascope.
A better display is the spectrum as in your figure 3, showing amplitude on the y-axis and frequency (pitch) on the x-axis.
The spectrum display shows more detail of the various pitches (and their amplitude) that make up the formant.

Most audio editors offer a spectrum display that you may find useful, I use Samplitude which has a full suite of display options including spectrum analysis.

You also might consider the LSVT voice training program, they have custom software that shows real-time how well a client is staying on pitch and amplitude.  My wife is certified in LSVT, she uses the software with Parkinsons, stroke, TBI, etc patients to remediate voice deficits.

http://www.lsvtglobal.com/patient-resources/what-is-lsvt-loud

[Fletcher]

A better display is the spectrum as in your figure 3, showing amplitude on the y-axis and frequency (pitch) on the x-axis.
The spectrum display shows more detail of the various pitches (and their amplitude) that make up the formant.

Again, I'm trying to learn here... and kind of understand what I'm seeing in "figure 3"... but without any kind of nomenclature its really not something I understand... pretty picture, but I have zero point of reference to tell me what I'm looking at. 

It looks like a ⅓ octave RTA display... which doesn't seem to me to be what the original poster is seeking to accomplish -- but I could very easily be wrong.

Peace