Good computer tool for speaking practice

[geraldc]

I was listening to the radio a few days ago when I heard an article about a new cure for stuttering, and this interested me as I stutter a lot when speaking Chinese. http://news.bbc.co.uk/1/hi/health/5327316.stm

Essentially what they do is add a delay and pitch shift to a person's voice, and then play it back to them via a tiny ear piece. I then realised that the same techniques could be used to practise speaking Chinese. Rather than speak long passages and then record it and play it back to yourself, if you just use readily available music software, you could just add a few hundred milliseconds delay to your own speech, and then listen to yourself as you speak.

I've found the freeware Audiomulch works pretty well, and if you follow the tutorial you should be able to set yourself up pretty quickly. http://www.audiomulch.com/

I just added a delay to my voice, and then added a frequency shift (as that's what they did for the stutterers) so now I can monitor myself as I practise reading Chinese etc. I never liked the rigmarole of recording myself on tape and then playing it back to myself, as it always took too long (I'm very lazy) and rewinding to the relevant point was always a chore, but with this, it's just fire up a program on your pc and just listen to yourself talk as you talk.

[geraldc]

As it's a BBC article I've cut and pasted it below.

'I'm not stammering any more'

By Julian Sturdy

Inside Out

Heidi King

A new treatment for stammering developed in the US involves fitting patients with an earpiece echoing their own voice.

Twenty-five-year-old Heidi King is one of the first British patients to try it.

Heidi is the life and soul of any party. Intelligent and vivacious, she has a wide circle of friends.

She loves to socialise and is a keen tango dancer.

But when Heidi opens her mouth to speak, her voice instantly commands the attention of everyone in the room.

A few simple sentences can take severe stammerer Heidi an age to get out.

It's painful to listen to and exhausting for her.

She's no victim though and she doesn't want pity.

In fact the only time her smile fades is if you try to finish off her sentences.

Heidi began stammering around the age of three. Growing up in Billericay in Essex, she had all the therapies, none really worked.

It will change the way I communicate

Heidi's diary

But she hasn't let it hold her back - her outgoing personality is testament to that - so too is a first class honours degree in psychology from the University of Kent.

She has a good job too - working on an older carers project for Age Concern in Norwich.

Despite not seeing her severe stammer as an impediment, Heidi volunteered for the pioneering treatment in America. She is one of the first Britons to try it.

She says: "I suppose deep down I would love a cure or something which helps my stammer."

The treatment is not available on the NHS and costs over £5,000, and there are no guarantees it will even work.

The 'choral effect'

Her speech therapist Mary Kingston says it will not be a miracle cure. "I don't think we should look at it as a cure.

"It is much more like a pair of glasses. It will hopefully ease it. That is what we are both hoping for."

Heidi travelled to America where she was fitted with the electronic implant called SpeechEasy.

It looks like an advanced hearing aid.

Gareth Gates

Gareth Gates is a famous example of someone whose stammer stops when he sings

Like most stammerers, Heidi can sing in unison without stuttering. The implant mimics that "choral effect" by sending out an echo of Heidi's voice.

It tricks her brain into thinking she is talking along with someone else and unblocks the impediment.

Heidi's initial response was very promising. For an hour she talked fluently - practising reading pages of text; recounting the days of the week, the months of the year.

"I don't feel like Heidi because I am not stammering," she said.

"It is almost as if I am detached. That is a strange person speaking, it isn't me.

"I feel as if I am on a drug because it is making me so relaxed. I am listening to a little man in my ear. I am just not struggling as much. It is just so strange to speak without stammering."

Emotional adjustments

It's not a total cure, says speech pathologist John Haskell who treated her in New York.

"Heidi is starting to hear herself differently. She heard her voice with a slight delay and with a higher pitch speaking with her.

She will have to expect moments of stuttering or blocks, but she is going to learn to deal with it. People around her will have to expect that she is not going to be 100%."

There have been emotional adjustments to make too. Heidi admits she's scared of using the device.

"It will change the way I communicate. It is almost like I have lost my control."

"Now I am more self-conscious. Today it is almost like I am a different person. I know I am not.

"I am trying to listen to the little man in my ear. It is quite hard.

"I go through life and I don't ever have to think about my speech. Now I am having to concentrate.

"It has been a long journey. I am still on that journey. It seems too much now. Too scary to look at what I can do with it. It's still me, whether I stammer or not."

Inside Out: Stammering is on BBC1 East on Monday 11 September at 1930BST or nationwide for satellite viewers on D-Sat 951.

[carlo]

Interesting. When I was 12 I tried to practice English this way (with an analog delay effect). It's like listening to the echo of your own voice over a bad telephone connection. Unless your pronunciation is already very close to your target, though, you may actually be reinforcing bad habits.

Are you sure audiomulch is freeware btw? It says 90 days trial on the website.

[smalltownfart]

geraldc:

Brilliant hack! Applied research at its best. Thanks for the tip.

[geraldc]

I think after 60 days audiomulch doesn't allow you to save anymore, and if you're using it to listen to yourself speak using the delay, you don't actually need to record the results. I'm not really an that au fait with audiomulch, I just asked a friend who's into music production to recommend a piece of free software that would allow me to add delay to my voice and monitor it. I still have over 50 days of the trial to go.

I can hear when other people mispronounce Cantonese, etc, it's just half the time when I mispronounce I know I've got it wrong, I just don't know how I've got it wrong, which is why I find this kind of delay-monitoring useful.

I know nothing beats going out and practising with real people etc, it's just I'm a bit of a geek, and I like to apply computers to my learning now and again, it helps justify the amounts I spend on them

Turns out what I thought was pitch shift on audiomulch wasn't actually pitch shift, but it was something called frequency shift, so I'm yet to see if pitch shifting your voice when listening to it adds any affect...

[teachinator]

It's definitely free and it's at audacity.sourceforge.net. I am not very familiar with using it except for the simplest functions, but I've seen it demonstrated and it seems to be a very versatile sound recording/playing/editing tool.

I hadn't thought of using it to record and playback myself speaking Chinese, but that's a great idea. Like you, I'm lazy about doing it with a tape

[imron]

I've used Audacity for quite sometime to record/playback my own voice, and it works quite well. You can set it up so that it plays the sound back as you record it, but I'm not sure if you can add a delay to that playback. I find listening to myself even several/tens of seconds after the event is still quite useful.

[smalltownfart]

I had a brief look at audacity and I don't see how it can be used for this purpose.

The "immediate" feedback of hearing your real voice is what you want here in order to automatically self-correct.

It sucks about the audiomulch trial period, but maybe there is another free alternative out there...

[imron]

If you go to Edit->Preferences->Audio I/O and check the "Software Playthrough" option, it will play back as you're recording it. I can't find any options to apply filters to the input stream, so adding your own delay/pitch shift might not be possible.

[roddy]

Quick walkthrough for how to set Audiomulch up to give you the delay.

1) Right click in the left pane > Input / Output > SoundIn

2) Right Click in the left plane > Input Output > SoundOut (actually, this will probably already be there)

3) Right Click in the left plane> Effects > SDelay

4) Double click on the SDelay contraption (their word, not mine ). Adjust the timings if you want (probably best left for now though, can come back to it later). You will probably also want to drag the feedback setting down to zero (you want a repeat, not an echo . . .echo . . . echo . . .

5) Click on the bottom left tab on the SoundIn contraption and drag the line down to the top left tab on SDelay

6) Click on the bottom left tab on the SDelay contraption and drag the line down to the top left tab on SoundOut

7) Press Play on the toobar

I didn't bother with any pitch or frequency changes - I don't see why you'd need them at the moment - perhaps to help you distinguish the two streams in continuous speech?

I can see how this could be pretty useful. You can do much the same with Audacity but it requires stopping and starting. This is very much 'in-line' feedback. I think the value of this would be in having it running in the background while you are working on something else - every time you come across a new word you can read it out and listen to yourself saying it, without having to switch to Audacity, start recording, playback, try again, etc.

Seems to be perfectly feasible to run it in conjunction with the likes of Skype and MSN / Yahoo voice chat. Things could get confusing though, as you'd be speaking, listening to yourself, and the other person all at the same time.

It's also quite fun finding out what all your sneezes, coughs and farts sound like to other people Turn the speakers and mic on, give it a 10 second delay, leave the room, and wait for someone else to come in - should be amusing.

That said, the $89 price tag is pretty steep if this is the only thing you're using it for. The 90 day trial will give you plenty of use though.

Attaching a screenshot of what the workspace should look like for the above set up.

[johnmck]

For free sound delay software try: http://babbleback.sourceforge.net/wiki/

[johnmck]

If you use BabbleBack and then capture the mp3's you will find at:

http://www.china-on-site.com/language/phonetic/phonetic.htm

you have quite a nice tool for practising your pinyin.

[edelweis]

@geraldc

bumping this, did you feel like your speaking improved using this method? do you still use it?

[geraldc]

No I don't really use it anymore. i probably stopped using it shortly after starting this thread. I'm always more interested in finding new ways to "learn" rather than practising. I think my hobby is learning to learn, rather than learning itself. 

[edelweis]

Thanks geraldc. Well, having a hobby is nice I think we share one or two.