TeachingHearingFilter
March 5, 2006 4:23 AM Subscribe
I'm going to be teaching about sound and hearing to some 12/13 year-olds next week and I'd like to show them how it's impossible to talk if your speech is relayed to your ears (by headphones) half a second or so late. How can I implement this in software?
I need a software-based solution because unfortunately hardware isn't a possibility. I've got microphones/headphones so what I need is some sort of (preferably free) software that will allow me to take input from the mike, delay it for a pre-determined amount of time, and then play it back.
Bonus points if you can tell me what the name for the effect mentioned in the question above is called - I think they use it in insurance to check if someone really is profoundly deaf or just faking it.
I need a software-based solution because unfortunately hardware isn't a possibility. I've got microphones/headphones so what I need is some sort of (preferably free) software that will allow me to take input from the mike, delay it for a pre-determined amount of time, and then play it back.
Bonus points if you can tell me what the name for the effect mentioned in the question above is called - I think they use it in insurance to check if someone really is profoundly deaf or just faking it.
this happens with communication over long distances - it's quite common to hear yourself, delayed slightly. while it's annoying, it most certainly does not making talking impossible.
a quick google search turns up this example which says "for some people". and i'd question whether it's impossible even for those people, or whether it's a more general problem due to confusion, embarassment and general giggles while at the exploratorium.
"impossible" is a very strong condition; i doubt very much that the brain is that rigid (or, perhaps more correctly, the process of speech processing is that simple).
and you can check whether someone is faking deafness without fancy electronic equipment - burst a paper bag behind them.
posted by andrew cooke at 5:28 AM on March 5, 2006
a quick google search turns up this example which says "for some people". and i'd question whether it's impossible even for those people, or whether it's a more general problem due to confusion, embarassment and general giggles while at the exploratorium.
"impossible" is a very strong condition; i doubt very much that the brain is that rigid (or, perhaps more correctly, the process of speech processing is that simple).
and you can check whether someone is faking deafness without fancy electronic equipment - burst a paper bag behind them.
posted by andrew cooke at 5:28 AM on March 5, 2006
It defiantly doesn't making talking impossible. When using Skype, for some reason I hear my voice echoed back to me, and I can't seem to fix it. It was annoying at first, but I could still talk, and now I just ignore it. Where did you hear it from? If you're looking for other cool sound and hearing things, I've always liked beat frequencies.
posted by Orange Goblin at 5:35 AM on March 5, 2006
posted by Orange Goblin at 5:35 AM on March 5, 2006
I want to say the term you're looking for is propigation delay, for example, the 0.25 second delay bouncing a signal off of a geosynchronous communications satellite. The examples I'm finding on google all seem to refer to that in terms of digital signals though.
posted by Fat Guy at 6:16 AM on March 5, 2006
posted by Fat Guy at 6:16 AM on March 5, 2006
For free delay effects try this page on Hitsquad.com for WinTel and this page for Mac. Some are slightly crippled try-beofre you buy versions, so you may have to experiment a bit.
posted by ZenMasterThis at 7:01 AM on March 5, 2006
posted by ZenMasterThis at 7:01 AM on March 5, 2006
Response by poster: Quite a few people are suggesting this effect doesn't work. All I have to say in my defence is that it worked on me and a room full of highly-sceptical physics/biology undergraduates a couple of years ago.
I imagine that this effect doesn't come in to play when talking on the 'phone/Skype because you are generally only using one ear, and the volume of the 'echo' is very quiet when compared to the volume of your voice and the surroundings. I'd be interested to hear results from anyone who can do this with a set of decent headphones at reasonably high volume.
Orange Goblin: I'm doing beats with 440/442Hz tones generated in Audacity; it works fantastically.
posted by alby at 7:38 AM on March 5, 2006
I imagine that this effect doesn't come in to play when talking on the 'phone/Skype because you are generally only using one ear, and the volume of the 'echo' is very quiet when compared to the volume of your voice and the surroundings. I'd be interested to hear results from anyone who can do this with a set of decent headphones at reasonably high volume.
Orange Goblin: I'm doing beats with 440/442Hz tones generated in Audacity; it works fantastically.
posted by alby at 7:38 AM on March 5, 2006
You can mark me down as another one who questions your premise. In fact, this very technique is used to assist people who stutter. So searching for stuttering+delay+software should give you a bunch of different software options to play around with. The effect is generally called Delayed Auditory Feedback.
posted by Galvatron at 8:15 AM on March 5, 2006
posted by Galvatron at 8:15 AM on March 5, 2006
This does work, like alby said.
I tried it recently in a setup at the Ontario Science Centre. They used two phones that you held to each ear (so it might indeed be because it has to be through both ears, but actually I think you talked in one receiver and heard through the other) and you were given a paragraph to read out loud. I often talk long distance with the annoying echo on the phone and that doesn't bother me at all, but this thing totally screwed me up.
However, another reason why it might have been so hard to do (and different than on the phone) was that I kept hearing what I just read - those words were still within eye-sight, while I was trying to say what was written half a sentence later. You don't read your phone conversations from a script.
posted by easternblot at 8:26 AM on March 5, 2006
I tried it recently in a setup at the Ontario Science Centre. They used two phones that you held to each ear (so it might indeed be because it has to be through both ears, but actually I think you talked in one receiver and heard through the other) and you were given a paragraph to read out loud. I often talk long distance with the annoying echo on the phone and that doesn't bother me at all, but this thing totally screwed me up.
However, another reason why it might have been so hard to do (and different than on the phone) was that I kept hearing what I just read - those words were still within eye-sight, while I was trying to say what was written half a sentence later. You don't read your phone conversations from a script.
posted by easternblot at 8:26 AM on March 5, 2006
You don't need software or any new hardware—just bring two cell phones, use one to call the other, and have them hold both phones up to their ears.
posted by crabintheocean at 8:49 AM on March 5, 2006
posted by crabintheocean at 8:49 AM on March 5, 2006
This VST effect should be able to delay the sound for a specified time.
Then you need a 'VST host' - software which can run the effect and deal with the audio in and out. I don't know which is best, but try googling 'free vst host'
posted by lunkfish at 9:22 AM on March 5, 2006
Then you need a 'VST host' - software which can run the effect and deal with the audio in and out. I don't know which is best, but try googling 'free vst host'
posted by lunkfish at 9:22 AM on March 5, 2006
DeMuDi linux can do this for you, really easily. I'm not positive, but I think the Live disc version (no HDD install) may work for you as well. All you need to do is run JackRack, add a delay, connect your microphone to Jackrack with QJackctl and JackRack to your output soundcard. It's really, really, easy, no linux-y knowledge required really, unless you decide to install, but that's pretty tame as well.
posted by phrontist at 11:02 AM on March 5, 2006
posted by phrontist at 11:02 AM on March 5, 2006
my sound blaster live does this with mike input and monitoring without any special software at all ... i cannot monitor what i'm singing because of this and have to turn monitoring off
in short, cheap or non-professional sound cards (or onboard audio) may do this for you automatically ... stick a mike in, put the thing on monitor and listen on headphones and see if your hardware does this
it might
(and to think that i thought this was a bug ...)
posted by pyramid termite at 12:22 PM on March 5, 2006
in short, cheap or non-professional sound cards (or onboard audio) may do this for you automatically ... stick a mike in, put the thing on monitor and listen on headphones and see if your hardware does this
it might
(and to think that i thought this was a bug ...)
posted by pyramid termite at 12:22 PM on March 5, 2006
crabintheocean wrote...
You don't need software or any new hardware—just bring two cell phones, use one to call the other, and have them hold both phones up to their ears.
I second this suggestion. In fact, all you need is one cellphone and a landline.
I experienced the effect quite frequently while writing VOIP software.
posted by tkolar at 2:36 PM on March 5, 2006
You don't need software or any new hardware—just bring two cell phones, use one to call the other, and have them hold both phones up to their ears.
I second this suggestion. In fact, all you need is one cellphone and a landline.
I experienced the effect quite frequently while writing VOIP software.
posted by tkolar at 2:36 PM on March 5, 2006
Half a second is a very long time in this context. I'm skeptical about 'impossible' too, but it isn't like any long distance phone call you will ever have.
I'm not sure you are interested in more experiments...
I just saw a lecture with an interesting demonstration involving Stairway to Heaven played backwards. The idea being that you can't really hear anything much when it is played backwards until somebody tells you what you are supposed to hear. Once you know what you are supposed to hear you start to think those words are actually in the sound. You can watch it at the perimeter institute's public lecture archive, it's currently first in this list, titled View: The Big Bang (Sorry, no direct link). The demonstration starts at around 20 minutes in.
Also, the effects of quantization and dither are pretty easy to demonstrate and quite interesting. Low resolution quantized audio sounds distorted, but if you add a sufficient level of white noise you hear it as a 'perfect' audio signal with a lot of noise - it doesn't sound distorted at all. This might be a bit advanced for 12-13 year olds, but it is worth consideration maybe...
posted by Chuckles at 5:46 PM on March 5, 2006
I'm not sure you are interested in more experiments...
I just saw a lecture with an interesting demonstration involving Stairway to Heaven played backwards. The idea being that you can't really hear anything much when it is played backwards until somebody tells you what you are supposed to hear. Once you know what you are supposed to hear you start to think those words are actually in the sound. You can watch it at the perimeter institute's public lecture archive, it's currently first in this list, titled View: The Big Bang (Sorry, no direct link). The demonstration starts at around 20 minutes in.
Also, the effects of quantization and dither are pretty easy to demonstrate and quite interesting. Low resolution quantized audio sounds distorted, but if you add a sufficient level of white noise you hear it as a 'perfect' audio signal with a lot of noise - it doesn't sound distorted at all. This might be a bit advanced for 12-13 year olds, but it is worth consideration maybe...
posted by Chuckles at 5:46 PM on March 5, 2006
I concur with those who say that the delayed-feedback effect can be overcome. But it is very difficult initially, so your 12/13 year olds will be flummoxed. You have to work to overcome your natural instincts to not listen to your own voice, which is difficult, expecially if your voice is as beautiful as mine. :-)
posted by breath at 8:26 PM on March 5, 2006
posted by breath at 8:26 PM on March 5, 2006
The effect pyramid termite talks about is called soundcard latency. Most serious soundcards let you adjust the latency, but if you set it too low, the sound sometimes drops out when your machine can't keep up. Some soundcards also have a direct monitoring option that routes the input directly to the output, without going through the PC, to avoid latency.
My laptop soundcard, using the default Windows drivers, has a latency of about 300ms. I can get down to 8ms using an M-Audio USB soundcard with ASIO drivers designed for music recording.
I find that with more than about 15ms latency, I can't play the guitar in time (not that I can play that well in time with less latency). I'd expect to be able to stand a much higher latency for speech, but half a second sounds like a lot.
If you use a delay effect like some people suggest, you'll want to set the output to "100% wet", which means that the effect only outputs the delayed signal, without passing through the original ("dry") signal. You also want closed headphones so that people don't hear their own voice directly.
posted by fuzz at 1:05 AM on March 6, 2006
My laptop soundcard, using the default Windows drivers, has a latency of about 300ms. I can get down to 8ms using an M-Audio USB soundcard with ASIO drivers designed for music recording.
I find that with more than about 15ms latency, I can't play the guitar in time (not that I can play that well in time with less latency). I'd expect to be able to stand a much higher latency for speech, but half a second sounds like a lot.
If you use a delay effect like some people suggest, you'll want to set the output to "100% wet", which means that the effect only outputs the delayed signal, without passing through the original ("dry") signal. You also want closed headphones so that people don't hear their own voice directly.
posted by fuzz at 1:05 AM on March 6, 2006
Response by poster: Thanks to everyone who's made suggestions. I'll be looking at the DeMuDi suggestion, I already run Mandriva so linuxy business isn't a problem. I'll also check out the other suggestions - I think the key issue is going to be finding software that will play (with delay) whilst also recording but I'll post a follow-up here when I know more.
Thanks again everybody for your help/suggestions.
posted by alby at 3:10 AM on March 7, 2006
Thanks again everybody for your help/suggestions.
posted by alby at 3:10 AM on March 7, 2006
This thread is closed to new comments.
As far as hardware not being an option - there's a number of units which employ 1/4" jacks. Assuming you're using non-XLR mics, connections shouldn't be an issue.
posted by Smart Dalek at 4:50 AM on March 5, 2006