Speech Acoustics 5 - vowel formants

[@iamgerer]

Thank you very much , a great job explaining each and everything clearly

[@betulk.7073]

Thanks a lot . U saved my day, I have an exam tomorrow !

[@borafalaringlesbrasil]

Wow!!!! That's SO COOL!!! Thanks for this class! I'm really interested in the visual representation of the differences between standard pronunciation and accent. I want to help "nerdy" ESL students visualize that although there's variation in spoken English as a "lingua franca", such variation has limits and that's why pronunciation is so important. (Do you know any references on that? I'd really appreciate any tips. I'm a newbie in this field).

[@Ahstval]

Amazing video. Clear, concise, to the point, very helpful. Thank you.

[@voicelab9884]

thank you! this was helpful for teaching. Can you make one on the singer's formant?

[@PhilSkill96]

Pretty good content! Keep up the good work!

[@marceloandradey]

Great! Thanks for the information. How can we tell if a non-native speaker is approaching a native speaker pronunciation? Is there an accepted range of divergence in terms of the formant values?

[@GraveyarDisciple]

Amazing I just discover your channel pretty amazing. Thank you neat info.

[@raulgarciacardoza9799]

Amazing content. I have a question though. I've been strugling for a while trying to flip the vowel chart unsuccessfully. How can I do this? Thanks!

[@raulgarciacardoza9799]

Solved it, nevermind!

[@NarFlux]

I want copies of your charts arranging the frequencies...

[@meryp.2357]

Hello. I would like to ask you, how can I measure the schwa formant? Is there any frequency from which I know that the F1/F2 belong to the schwa sound? Thank you

[@EGMusic12]

are r-coloured vowels basically changing the 3rd formant?

[@listenlab_umn]

yes. /r/-colored sounds tend to have a lowered F3 (or you can think of it as F3 being close to F2)

[@EGMusic12]

@@listenlab_umn ok thank you! So you could technically make a 3 dimensional IPA vowel chart with r coloured vowels lol

[@user-rf4zy9uf1r]

감사합니다 교수님 이거로 제 미쿠쨩 발음을 쫌더 또렷하게 할 수 있겠어요😃

[@CamillaBookTok]

It is true that : The first formant of vowels shows that the higher the tongue, the higher the first formant frequency?

[@listenlab_umn]

it's just the opposite. If the tongue is higher, the first formant will be lower

[@iamgerer]

Doesnt this vowel mapping on graph affected by individuals voice

[@joepike1972]

f = c/L. So yes, frequency is inversely related to vocal tract length. Bigger people have deeper voices.

[@desireerubio2334]

How do you know which vowel it is ?

[@joepike1972]

He didn't show you on the video but Praat allows you to click on the bottom and you progress line will go across as the corresponding audio is played. Also you did get a feel for which numbers feel right for which vowels.

[@craigberryman]

Brilliant! Thank you.

[@abdulgafoor7076]

Sir how vocal folds vibrate at multiple frequencies at the same time Fundamental vibration and series of harmonics simultaneously how is it possible sir

[@joepike1972]

Impedance. Essentially the differences of mass in connection result in phase shifts of the frequency. Resonance occurs for waves resulting in ambient pressure at the mouth. You can run your own experiments using string and yarn, different masses of string. Tie them together to replicate differences of front of mouth more open vs back of vocal tract more expanded. Fant Gunnar is the one that tied the analogy of the human vocal tract to transmission line theory. Air pressure acts as voltage, volume velocity acts as current. It is the unequal distribution of pressure and air flows caused by differing cross-sectional areas that result in deviations from the standard f(n)=(2n-1)c/(4L) that you get from a standard open-closed tube that represent the 1:3:5 ratios of neutral vowels. The equations: F1 = c/L(atan(sqrt(A2/A1))/pi) F2 = c/L(1-atan(sqrt(A2/A1))/pi) F3 = c/L(1+atan(sqrt(A2/A1))/pi) can be used to simulate the variations of multiple frequencies by varying A2 towards the lips oral cavity, vs A1 towards the glottis pharyngeal cavity's cross-sectional areas. The model is based on tube segments of equal length to promote tractability.

[@adamzhang2874]

LING 311

[@shadialomari9067]

What's the name of the program you are using?

[@listenlab_umn]

this is just powerpoint

[@deniseadams4121]

Which. Program?

[@joepike1972]

Praat

[@codeparity]

how many formants in total can we have?

[@joepike1972]

Infinite, you wont get instruments fine enough to recover them though. Generally you set your setting to 5 formants in a 5500 Hz range, or you might want to change that depending on the speaker and vowel, your main concern is clean data. Generally you will see only up to F3 reported in most data and that will be what you will tend to focus on as well. The higher formants tend not to tell as much about the vowel and more about anatomical aspects. Not that you will ever know specifically which, since your really can't recover a vocal tract based on formants. Many vocal tracts map to the same formants making the information unrecoverable.

[@yousifaldailami4241]

Am i the only one who came here for sound design?

[@NarFlux]

I am here trying to deside how color code vowels by their frequency...which isn't quite answering my question here.

[@iamgerer]

Yes ig

[@kwaku.a.b.anningemmanuel5206]

please how did plot on excel?

[@4lpha0ne]

It's a scatter plot (or X Y plot)

[@user-ut1go9dn7j]

Please, please remove the script because it hides most of the vsual screen.

[@listenlab_umn]

you do not need to make this comment on all of our videos. captioning is an option on your browser, and is not part of the video :)

[@AliHamza-t8t7w]

osm