0:00:08.964,0:00:12.948
[APPLAUSE]

0:00:12.948,0:00:13.910
HELGA VELROYEN: Hi.

0:00:13.910,0:00:15.145
Can you hear me?

0:00:15.145,0:00:15.930
Ah, nice.

0:00:15.930,0:00:18.370
That's what it's all about.

0:00:18.370,0:00:19.920
Welcome to my talk.

0:00:19.920,0:00:22.160
I will talk about hearing[br]aids and what the

0:00:22.160,0:00:24.010
state-of-the-art is.

0:00:24.010,0:00:26.420
There will be a little hacking,[br]but not my own.

0:00:30.910,0:00:33.890
I know that there are some[br]people interested in my talk

0:00:33.890,0:00:36.570
that can not hear very[br]well or not at all.

0:00:36.570,0:00:41.380
So I will publish slides that[br]have very detailed speaker

0:00:41.380,0:00:43.060
notes so that you can[br]read it afterwards,

0:00:43.060,0:00:45.390
if you missed anything.

0:00:45.390,0:00:47.430
I hope there will be a[br]recording available.

0:00:47.430,0:00:51.145
And if it's possible to add[br]subtitles, I will do those.

0:00:53.970,0:00:54.330
OK.

0:00:54.330,0:00:56.000
First, a few words about me.

0:00:56.000,0:00:57.275
I'm a software engineer.

0:00:57.275,0:00:58.490
I am based in Munich.

0:00:58.490,0:01:03.480
Some people might know me from[br]my time in Cologne, as well.

0:01:03.480,0:01:05.840
I'm more a software geek[br]than a hardware hacker.

0:01:05.840,0:01:07.910
So all this is also new to me.

0:01:07.910,0:01:13.280
From university, I have a[br]background in data mining and

0:01:13.280,0:01:14.690
signal processing.

0:01:14.690,0:01:17.040
I worked in the medical[br]industry for a while.

0:01:17.040,0:01:20.860
But that had nothing to[br]do with hearing aids.

0:01:20.860,0:01:22.710
Also, my current job[br]has nothing to do

0:01:22.710,0:01:23.250
with hearing aids.

0:01:23.250,0:01:28.220
So this is really just[br]my personal project.

0:01:28.220,0:01:32.400
I'm hearing impaired for[br]about 3 and 1/2 years.

0:01:32.400,0:01:36.510
So this is when I started[br]to dig into the topic.

0:01:36.510,0:01:40.295
And well, yeah, that's just[br]what I will talk about.

0:01:44.830,0:01:47.910
Since I haven't seen many talks[br]about audiology here at

0:01:47.910,0:01:49.940
the Hacking Congress, I will[br]start with a short

0:01:49.940,0:01:53.220
introduction and the process[br]of how you get

0:01:53.220,0:01:55.030
hearing aids, actually.

0:01:55.030,0:01:58.190
Then what are the current[br]hearing aid models, and what

0:01:58.190,0:01:59.830
can they actually do.

0:01:59.830,0:02:01.900
Some words about the peripheral[br]hardware.

0:02:01.900,0:02:04.610
There's quite a lot[br]of it outside.

0:02:04.610,0:02:07.650
And there is some hacking.

0:02:07.650,0:02:10.360
And another point is[br]self-tuning, that are people

0:02:10.360,0:02:13.160
that tune their own hearing[br]aids, although they are not

0:02:13.160,0:02:14.410
audiologists.

0:02:18.180,0:02:20.170
So this is an audiogram.

0:02:20.170,0:02:24.040
This is a result of a hearing[br]test that you usually do at

0:02:24.040,0:02:27.020
ENT doctors.

0:02:27.020,0:02:32.250
The x-axis shows the frequency[br]in kilohertz, and the y-axis,

0:02:32.250,0:02:35.520
the loudness, level of volume.

0:02:35.520,0:02:38.180
The silence is at the[br]top, the really loud

0:02:38.180,0:02:39.660
sounds at the bottom.

0:02:39.660,0:02:43.020
And the green line you see[br]here is that result of a

0:02:43.020,0:02:45.230
normal-hearing person.

0:02:45.230,0:02:49.460
And this is obtained by the[br]audiologist. Or the doctor

0:02:49.460,0:02:51.530
plays sounds in the different[br]frequencies.

0:02:51.530,0:02:54.810
It starts at a very[br]low volume.

0:02:54.810,0:02:58.610
And as soon as you hear it, you[br]hit a buzzer or say, yes.

0:02:58.610,0:03:02.790
And then they create[br]this curve.

0:03:02.790,0:03:06.592
So the blue curve is a[br]typical curve of a

0:03:06.592,0:03:08.410
hearing impaired person.

0:03:08.410,0:03:12.330
So what you can see here-- that,[br]at the low frequencies,

0:03:12.330,0:03:13.700
their hearing is quite well.

0:03:13.700,0:03:15.610
That's very typical.

0:03:15.610,0:03:20.120
Their hearing starts to get[br]worse in the high frequencies.

0:03:20.120,0:03:24.660
And bear in mind that the[br]decibel scale is actually

0:03:24.660,0:03:25.780
logarithmic.

0:03:25.780,0:03:31.015
So if you have a hearing loss of[br]60 decibels, it's 1 million

0:03:31.015,0:03:33.420
worse than 10 decibels.

0:03:33.420,0:03:36.790
So it's not linear.

0:03:36.790,0:03:41.270
Another thing that gets measured[br]at an audiogram is

0:03:41.270,0:03:43.900
actually the maximum that[br]you can hear or that

0:03:43.900,0:03:45.530
you can stand hearing.

0:03:45.530,0:03:49.130
So the audiologist raises the[br]volume more and more, and you

0:03:49.130,0:03:51.600
have to say yes until[br]it hurts.

0:03:51.600,0:03:53.780
So what you see here[br]is the red curve.

0:03:53.780,0:03:56.290
It's the level of discomfort.

0:03:56.290,0:04:00.370
And also the typical thing is[br]that it raises exactly at the

0:04:00.370,0:04:02.870
areas where the hearing[br]goes bad.

0:04:02.870,0:04:06.110
This is a very complicated thing[br]for tuning aids because

0:04:06.110,0:04:09.610
they can not just amplify[br]everything.

0:04:09.610,0:04:14.830
Because you would hurt people[br]very soon, as soon as you get

0:04:14.830,0:04:16.080
below the red line.

0:04:18.269,0:04:22.019
Another thing is the area where[br]speech takes place.

0:04:22.019,0:04:23.560
This is called speech banana.

0:04:23.560,0:04:26.570
Actually, that's the technical[br]term for it.

0:04:26.570,0:04:30.710
It will rise, of course,[br]according to the language and

0:04:30.710,0:04:31.790
the speaker.

0:04:31.790,0:04:33.970
For example, female speakers[br]have a little slightly

0:04:33.970,0:04:37.740
different curve than[br]male speakers.

0:04:37.740,0:04:44.030
And this is the area where[br]hearing aids target, which is

0:04:44.030,0:04:47.650
they are used to make you[br]understand speech again.

0:04:47.650,0:04:49.830
So they focus on this[br]area, actually.

0:04:49.830,0:04:52.290
And you can see the[br]blue line here.

0:04:52.290,0:04:54.960
So half of the banana[br]is actually cut.

0:04:54.960,0:04:57.800
And this is the high[br]frequencies, which, in speech,

0:04:57.800,0:05:02.360
are the consonants like[br]S and F, for example.

0:05:02.360,0:05:07.010
The vowels are usually[br]understood quite well.

0:05:07.010,0:05:11.090
To give you an impression how[br]I hear, I made a sample.

0:05:18.040,0:05:20.730
So this is a song called, "Sad[br]Robot" from Pornophonique.

0:05:20.730,0:05:24.520
It's a nice band which[br]makes music with a

0:05:24.520,0:05:26.480
Game Boy and a guitar.

0:05:26.480,0:05:27.155
And the original--

0:05:27.155,0:05:27.430
[SAD ROBOT, BY PORNOPHONIQUE]

0:05:27.430,0:05:29.690
HELGA VELROYEN: --has[br]really nice high

0:05:29.690,0:05:30.910
frequencies at the beginning.

0:05:30.910,0:05:34.110
This is why I use[br]it for testing.

0:05:34.110,0:05:35.730
And later, there's also[br]some singing.

0:05:40.130,0:05:42.570
So this is the original.

0:05:42.570,0:05:48.230
And now I'm going to play my[br]version, so with less high

0:05:48.230,0:05:50.990
frequencies and a tinnitus[br]as well.

0:05:50.990,0:05:53.710
There's actually a web site[br]where you can download

0:05:53.710,0:05:55.245
tinnitus sounds.

0:05:55.245,0:05:58.095
[LAUGHTER]

0:05:58.095,0:06:00.920
HELGA VELROYEN: And it says that[br]you should actually turn

0:06:00.920,0:06:04.140
down the volume when you start[br]broadcasting that because it

0:06:04.140,0:06:07.750
can hurt the audio equipment.

0:06:07.750,0:06:10.290
So I hope I don't destroy[br]anything.

0:06:10.290,0:06:15.080
So I will turn it down first,[br]then slightly increase it.

0:06:15.080,0:06:26.915
[SAD ROBOT, BY PORNOPHONIQUE]

0:06:26.915,0:06:28.060
HELGA VELROYEN: So this[br]is not a feedback

0:06:28.060,0:06:30.120
loop, it's the tinnitus.

0:06:30.120,0:06:32.980
And without the high[br]frequencies,

0:06:32.980,0:06:36.860
this is how it is.

0:06:42.400,0:06:49.490
So it's really hard to actually[br]hear the high

0:06:49.490,0:06:50.740
frequencies.

0:06:53.090,0:06:57.440
Yeah, so this is just an[br]impression of what hearing

0:06:57.440,0:07:00.860
aids have to work with.

0:07:00.860,0:07:04.090
Getting hearing aids,[br]I like to compare

0:07:04.090,0:07:05.400
it to getting glasses.

0:07:05.400,0:07:08.380
So one day you wake up and[br]everything is really blurry.

0:07:08.380,0:07:11.800
And you decide well, this[br]is very not so good.

0:07:11.800,0:07:12.990
I should go to a doctor.

0:07:12.990,0:07:14.590
And then you go to a doctor.

0:07:14.590,0:07:15.860
You make some tests.

0:07:15.860,0:07:18.130
Then he sends you to an[br]optometrist. He makes some

0:07:18.130,0:07:19.610
more tests.

0:07:19.610,0:07:22.060
Then you choose a model[br]for your glasses.

0:07:22.060,0:07:25.550
And the optometrist orders[br]the glasses and puts

0:07:25.550,0:07:26.300
them into the frame.

0:07:26.300,0:07:29.150
And then you're happy,[br]seeing nerds.

0:07:29.150,0:07:30.540
And then you can see ponies.

0:07:30.540,0:07:32.460
[LAUGHTER]

0:07:32.460,0:07:34.210
HELGA VELROYEN: Getting hearing[br]aids, unfortunately,

0:07:34.210,0:07:35.040
is not that easy.

0:07:35.040,0:07:39.530
So one day you realize you can[br]not hear very well any more.

0:07:39.530,0:07:40.400
You go to a doctor.

0:07:40.400,0:07:41.360
You make some tests.

0:07:41.360,0:07:43.690
He sends you to an audiologist.[br]He makes some

0:07:43.690,0:07:44.870
more tests.

0:07:44.870,0:07:49.010
You choose from the shop[br]of the audiologist

0:07:49.010,0:07:52.240
some hearing aids.

0:07:52.240,0:07:57.480
And then the audiologist has[br]to adjust the hearing aids

0:07:57.480,0:07:59.532
according to your audiogram.

0:07:59.532,0:08:01.970
And then it doesn't stop.

0:08:01.970,0:08:04.010
So then, actually,[br]the work starts.

0:08:04.010,0:08:07.340
You have to go through all the[br]difficult hearing situations

0:08:07.340,0:08:10.480
to test if it works with[br]that tuned hearing aid.

0:08:10.480,0:08:11.570
So you drive a car.

0:08:11.570,0:08:14.440
You listen to music or other[br]people in the car.

0:08:14.440,0:08:18.610
You try to have someone whisper[br]something in your ear.

0:08:18.610,0:08:22.440
You listen to the TV. Or you go[br]to a party where a lot of

0:08:22.440,0:08:25.030
people are talking to each[br]other, and you have to make

0:08:25.030,0:08:28.610
out the person that's talking[br]directly to you.

0:08:28.610,0:08:31.165
Or you listen to a talk like[br]that where this speaker is

0:08:31.165,0:08:33.350
actually quite far away from[br]your hearing aids.

0:08:33.350,0:08:36.340
And if you have done all that,[br]you go back to the

0:08:36.340,0:08:39.640
audiologist. And you have to[br]tell him why it doesn't work

0:08:39.640,0:08:42.610
or in what situations it[br]doesn't really work.

0:08:42.610,0:08:45.140
And then he does some changes[br]in the parameters.

0:08:45.140,0:08:47.070
And then you have to[br]do that again.

0:08:47.070,0:08:49.420
And sometimes you switch[br]to a different

0:08:49.420,0:08:50.300
hearing aid, as well.

0:08:50.300,0:08:54.060
So this whole process, these[br]iterations, it takes weeks or

0:08:54.060,0:08:57.580
months until you have something[br]that is actually

0:08:57.580,0:08:59.650
fitting to your ears.

0:08:59.650,0:09:04.800
And after that, you're[br]sort of happy.

0:09:04.800,0:09:07.230
Actually, I haven't met a person[br]that has hearing aids

0:09:07.230,0:09:11.170
that actually compensate for the[br]hearing loss completely.

0:09:11.170,0:09:14.600
So whenever you are done with[br]that, you're usually just

0:09:14.600,0:09:16.990
stopping because you don't[br]want to spend any

0:09:16.990,0:09:18.200
more time on that.

0:09:18.200,0:09:22.950
And it works well enough.

0:09:22.950,0:09:26.470
Compared to glasses, this is[br]actually a lot more effort and

0:09:26.470,0:09:28.260
a lot more frustrating.

0:09:28.260,0:09:32.050
Another thing is that hearing[br]aids are really expensive.

0:09:32.050,0:09:36.990
A good hearing aid starts at[br]like, 1,500 up to 3,000.

0:09:36.990,0:09:41.530
And I only have the numbers for[br]the German health system.

0:09:41.530,0:09:45.060
The normal German insurance[br]pays 500.

0:09:45.060,0:09:47.305
So there's a lot of money you[br]have to pay for yourself.

0:09:58.610,0:10:01.490
Hearing aid models and[br]their features.

0:10:01.490,0:10:04.920
There are roughly three types,[br]in-ear units that go

0:10:04.920,0:10:08.590
completely into the ear canal.

0:10:08.590,0:10:11.110
A more common one is behind-ear[br]hearing aids.

0:10:11.110,0:10:15.980
They are for mediocre to[br]severe hearing losses.

0:10:15.980,0:10:18.440
The main part is[br]behind the ear.

0:10:18.440,0:10:22.360
And another special thing are[br]cochlear implants where parts

0:10:22.360,0:10:26.670
of it are implanted into the[br]head and some is attached from

0:10:26.670,0:10:28.580
the outside.

0:10:28.580,0:10:32.360
I will mostly talk about the[br]behind-ear hearing aids,

0:10:32.360,0:10:36.760
because that's what I have and[br]where is a lot of variety on

0:10:36.760,0:10:39.380
the market.

0:10:39.380,0:10:42.190
Hearing aids got pretty[br]invisible.

0:10:42.190,0:10:44.670
These are pictures of me wearing[br]my hearing aid and not

0:10:44.670,0:10:45.240
wearing it.

0:10:45.240,0:10:48.640
So except for this little wire[br]on the right side, you can not

0:10:48.640,0:10:51.120
really see it.

0:10:51.120,0:10:53.510
Most people that don't[br]have hearing aids

0:10:53.510,0:10:56.080
find this an advantage.

0:10:56.080,0:10:59.530
People who have a hearing aid,[br]actually, are not that sure

0:10:59.530,0:11:03.700
about it, because sometimes,[br]when you have to ask someone

0:11:03.700,0:11:07.240
to repeat a sentence, if they[br]know you're wearing a hearing

0:11:07.240,0:11:10.460
aid, they think, OK,[br]she didn't get it

0:11:10.460,0:11:11.510
because of the acoustics.

0:11:11.510,0:11:14.660
And if they don't see it, they[br]think, she didn't get it

0:11:14.660,0:11:16.370
because she was stupid.

0:11:16.370,0:11:19.550
So it's really not[br]that obvious.

0:11:19.550,0:11:23.240
And sometimes it just helps that[br]people see that you have

0:11:23.240,0:11:25.630
a handicap.

0:11:25.630,0:11:26.820
They also got pretty small.

0:11:26.820,0:11:29.920
This is an example of my hearing[br]aids with a $0.50

0:11:29.920,0:11:32.460
piece, so you have an impression[br]of the size.

0:11:32.460,0:11:35.610
And you can see that, actually,[br]have just by the

0:11:35.610,0:11:36.860
battery compartment.

0:11:40.930,0:11:44.062
Out of curiosity, I took[br]my hearing aids apart.

0:11:44.062,0:11:44.503
And--

0:11:44.503,0:11:45.826
[LAUGHTER]

0:11:45.826,0:11:46.880
HELGA VELROYEN: Of course.

0:11:46.880,0:11:47.380
[LAUGHS]

0:11:47.380,0:11:52.145
[APPLAUSE]

0:11:52.145,0:11:55.500
HELGA VELROYEN: You can see they[br]have shells which you can

0:11:55.500,0:11:58.620
take off, which is like[br]for mobile phones.

0:11:58.620,0:12:01.340
You can switch the color and[br]choose a different one.

0:12:01.340,0:12:03.480
And the body--

0:12:03.480,0:12:05.640
so I know this picture[br]is not really good.

0:12:05.640,0:12:08.050
There's still a lot of[br]plastic around it.

0:12:08.050,0:12:11.610
And the white part there is two[br]microphones, the signal

0:12:11.610,0:12:14.220
processor and antenna.

0:12:14.220,0:12:16.120
That's useful peripheral[br]hardware.

0:12:16.120,0:12:17.436
I will come to that.

0:12:17.436,0:12:19.970
And the speaker is actually[br]at the part that

0:12:19.970,0:12:21.640
goes into the ear.

0:12:21.640,0:12:24.900
And what you can also see here,[br]the part that goes into

0:12:24.900,0:12:29.490
the ear has also holes where[br]the natural sound can still

0:12:29.490,0:12:30.210
come to the ear.

0:12:30.210,0:12:35.660
So if you still hear low[br]frequencies, then you can

0:12:35.660,0:12:37.410
receive them naturally.

0:12:37.410,0:12:41.890
And the hearing aids only add[br]what you can not hear.

0:12:41.890,0:12:43.460
This is called open[br]hearing aid.

0:12:43.460,0:12:47.300
And it has the advantage that[br]you still hear natural sounds,

0:12:47.300,0:12:51.310
which is really nice if you're[br]really sort of an audiophile

0:12:51.310,0:12:53.850
person that likes to[br]listen to music.

0:12:53.850,0:12:57.590
And so the first thing you do[br]is actually choose that one.

0:13:00.550,0:13:02.820
Yes.

0:13:02.820,0:13:06.200
I think in the last[br]two centuries,

0:13:06.200,0:13:08.260
hearing aids got digital.

0:13:08.260,0:13:12.250
And with that, they come with a[br]lot of new features that you

0:13:12.250,0:13:15.040
couldn't do with analog[br]hearing aids.

0:13:15.040,0:13:16.800
And right now, they[br]are standard in

0:13:16.800,0:13:18.880
most first world countries.

0:13:18.880,0:13:24.950
And the most important thing is[br]that they can analyze the

0:13:24.950,0:13:29.450
situation and react to it[br]instantly, since they have

0:13:29.450,0:13:34.310
signal processing in it much[br]more sophisticated than analog

0:13:34.310,0:13:35.560
hearing aids did.

0:13:38.350,0:13:43.310
Coming back to the audiogram,[br]this is actually a feature

0:13:43.310,0:13:46.460
that can also be done by[br]analog hearing aids.

0:13:46.460,0:13:49.300
I don't know in what[br]extent, actually.

0:13:49.300,0:13:52.500
So if you have a look at this[br]audiogram and only consider

0:13:52.500,0:13:57.960
one frequency band, like for[br]four kilohertz, here the

0:13:57.960,0:14:02.090
hearing loss is between[br]60 and 90 decibels.

0:14:02.090,0:14:04.540
Over 90 decibels, it[br]gets too loud.

0:14:04.540,0:14:06.717
And the input of the environment[br]still has the

0:14:06.717,0:14:07.410
whole range.

0:14:07.410,0:14:12.230
So the hearing aid has to map[br]0 to 130 decibels to this

0:14:12.230,0:14:14.840
small area between 60 and 90.

0:14:14.840,0:14:16.360
It can not just amplify[br]everything,

0:14:16.360,0:14:18.080
because that will hurt.

0:14:18.080,0:14:19.450
And this is called[br]compression.

0:14:19.450,0:14:22.900
This should not be mistaken[br]with compression in audio

0:14:22.900,0:14:24.360
files like MP3 or something.

0:14:24.360,0:14:26.850
It's a little bit different.

0:14:26.850,0:14:31.740
And if you have a look at the[br]software that is used to

0:14:31.740,0:14:33.835
adjust hearing aids,[br]it looks like that.

0:14:40.420,0:14:44.270
This controls here at[br]the amplification.

0:14:44.270,0:14:48.830
And this here reduces[br]the maximum level.

0:14:48.830,0:14:53.470
So my hearing aids can, at[br]maximum, do 108 decibels.

0:14:53.470,0:14:56.310
And if you have minus 12, you[br]just subtract it from that.

0:14:59.590,0:15:04.740
This has the problem that, if you[br]compress it, the volume

0:15:04.740,0:15:08.440
gets increased and decreased[br]all the time.

0:15:08.440,0:15:12.610
And this can actually make it[br]harder to understand speech,

0:15:12.610,0:15:14.990
because the hearing aid is[br]adjusting all the time.

0:15:14.990,0:15:20.310
And so to avoid that, they don't[br]compress every time.

0:15:20.310,0:15:28.640
So the first area here is[br]linearly amplified.

0:15:28.640,0:15:32.390
And only after a certain level,[br]they start compressing.

0:15:32.390,0:15:34.850
And this is called knee point.

0:15:34.850,0:15:41.790
So they usually try to avoid[br]compressing the speech signal

0:15:41.790,0:15:45.350
and only something above that.

0:15:45.350,0:15:48.220
And hearing adjusting software,[br]it looks like that.

0:15:48.220,0:15:50.430
This is actually an example[br]only from Siemens.

0:15:50.430,0:15:52.590
Every branch has their[br]own adjusting

0:15:52.590,0:15:55.440
software or tuning software.

0:15:55.440,0:15:57.290
So this is just an example.

0:15:57.290,0:15:59.960
So here you have the knee point[br]in the first row and the

0:15:59.960,0:16:02.800
factor by what it[br]is compressed.

0:16:02.800,0:16:09.050
And the third row is a[br]factor regarding how

0:16:09.050,0:16:10.300
fast it should react.

0:16:12.580,0:16:15.890
So within one syllable, it's[br]adjusted to the right volume.

0:16:18.770,0:16:20.020
Ah, OK.

0:16:25.500,0:16:28.680
The problem is what do you do[br]when your hearing loss is so

0:16:28.680,0:16:32.230
bad that the red and the blue[br]curve actually meet each other

0:16:32.230,0:16:34.850
or the blue curve goes[br]all the way down?

0:16:34.850,0:16:38.150
And this is actually a problem[br]because then you can not do

0:16:38.150,0:16:41.150
compression any more, at least[br]not in the original sense.

0:16:45.320,0:16:50.050
So this is a simplified[br]audiogram.

0:16:50.050,0:16:52.220
This is the area that is dead.

0:16:52.220,0:16:54.840
So instead of compressing in[br]one frequency band, you

0:16:54.840,0:16:56.920
compress the frequencies.

0:16:56.920,0:17:01.120
So you reserve a part of[br]the still alive hearing

0:17:01.120,0:17:03.920
frequencies and map it there.

0:17:03.920,0:17:06.440
This is called frequency[br]compression.

0:17:06.440,0:17:10.660
And it works only if you have[br]closed hearing aids, meaning

0:17:10.660,0:17:13.589
those where you can not hear[br]natural sound any more, where

0:17:13.589,0:17:16.980
the ear canal is really blocked[br]by the hearing aid,

0:17:16.980,0:17:19.530
because, otherwise, it would[br]be really confusing.

0:17:19.530,0:17:23.079
And this is actually offered by[br]only one brand right now,

0:17:23.079,0:17:25.859
by a company called Phonak.

0:17:25.859,0:17:32.830
And it's actually quite hard[br]to get used to that.

0:17:32.830,0:17:36.800
The brain has to adjust to that[br]for a very long time.

0:17:36.800,0:17:37.470
What I heard.

0:17:37.470,0:17:40.370
I mean, I don't have this.

0:17:40.370,0:17:42.710
But still, it's interesting[br]that they try

0:17:42.710,0:17:46.020
to do it like that.

0:17:46.020,0:17:48.580
A very common problem with[br]hearing aids is feedback

0:17:48.580,0:17:51.130
loops, especially if you[br]have open hearing aids.

0:17:51.130,0:17:54.880
Then it can happen that the[br]hearing aid captures its own

0:17:54.880,0:17:57.680
sound and amplifies it a lot.

0:17:57.680,0:18:00.540
There is just this squeaking,[br]what you get when you get too

0:18:00.540,0:18:02.270
close to them.

0:18:02.270,0:18:04.140
And this is really annoying.

0:18:04.140,0:18:06.260
It happens every time[br]something gets

0:18:06.260,0:18:08.060
close to your ears.

0:18:08.060,0:18:10.540
It can simply be hair,[br]wearing open.

0:18:10.540,0:18:12.190
Or you put on a hat.

0:18:12.190,0:18:15.035
Or you hold a telephone handle[br]next to your ear.

0:18:15.035,0:18:17.200
Or you just want to lie[br]down on the sofa.

0:18:17.200,0:18:21.070
Or especially when you hug[br]someone, you start squeaking.

0:18:21.070,0:18:23.020
Like you give bionic feedback.

0:18:23.020,0:18:24.972
[LAUGHTER]

0:18:24.972,0:18:28.570
HELGA VELROYEN: Yeah, this is[br]really, really annoying.

0:18:28.570,0:18:31.140
And what the hearing aids[br]do, they try to

0:18:31.140,0:18:32.770
detect feedback loops.

0:18:32.770,0:18:36.030
So they look for clear[br]sinus signals.

0:18:36.030,0:18:40.930
And when they detect one, they[br]send an un-hearable flag.

0:18:40.930,0:18:42.280
So oh, I detected one.

0:18:42.280,0:18:46.950
And then the affected[br]frequencies get damped until

0:18:46.950,0:18:49.800
it doesn't squeak any more.

0:18:49.800,0:18:51.070
They can adapt in real time.

0:18:51.070,0:18:53.600
So it actually works[br]really fast,

0:18:53.600,0:18:55.000
but it's still hear-able.

0:18:55.000,0:19:00.130
So the problems with that is[br]that music contains clear

0:19:00.130,0:19:01.330
sinus signals.

0:19:01.330,0:19:04.870
And those get then damped,[br]which make your music

0:19:04.870,0:19:09.470
experience a lot worse.

0:19:09.470,0:19:13.060
And also the damp frequencies[br]can be in the speech banana.

0:19:13.060,0:19:16.240
And then that means whenever you[br]put on a hat, then you get

0:19:16.240,0:19:16.940
a feedback loop.

0:19:16.940,0:19:18.380
Then the frequencies[br]get damped.

0:19:18.380,0:19:22.510
And then you can not understand[br]anyone any more.

0:19:22.510,0:19:24.870
Analog hearing aids did not have[br]a measure against that.

0:19:24.870,0:19:28.210
So this is something that is[br]clearly new with the digital

0:19:28.210,0:19:29.740
hearing aids.

0:19:29.740,0:19:31.800
The screenshot here[br]is also taken

0:19:31.800,0:19:33.520
from the tuning software.

0:19:33.520,0:19:36.930
When you have tuned your hearing[br]aid, you can make a

0:19:36.930,0:19:40.060
feedback loop test, where you[br]place a lot of different

0:19:40.060,0:19:44.400
sounds and tries if it detects[br]the feedback loop.

0:19:44.400,0:19:49.360
And then it reduces the maximum[br]power of the output of

0:19:49.360,0:19:50.540
your hearing aid.

0:19:50.540,0:19:53.070
That means you spend hours[br]tuning your hearing aid, and

0:19:53.070,0:19:54.690
then everything gets[br]reduced by that.

0:19:57.740,0:20:00.820
A very common problem for people[br]with hearing impairment

0:20:00.820,0:20:03.550
is the cocktail party problem.

0:20:03.550,0:20:06.660
This is when you're in an[br]acoustic setting where a lot

0:20:06.660,0:20:09.540
of people are talking and you[br]have some background noise.

0:20:09.540,0:20:11.130
And then someone is[br]talking to you.

0:20:11.130,0:20:14.180
And you have really problems to[br]figure out the person that

0:20:14.180,0:20:15.870
is talking to you.

0:20:15.870,0:20:18.485
And there are several[br]factors in that.

0:20:18.485,0:20:22.450
And hearing aids react to[br]that in several ways.

0:20:22.450,0:20:25.580
So first of all, directional[br]hearing is impaired when you

0:20:25.580,0:20:27.340
have a hearing impairment.

0:20:27.340,0:20:31.345
The human ear uses two[br]ears and the brain

0:20:31.345,0:20:33.110
to locate the sound.

0:20:33.110,0:20:37.280
And we use the pinna, that is[br]actually the outer part of the

0:20:37.280,0:20:40.858
ear, ohrmuschel, in German.

0:20:40.858,0:20:45.090
If you have behind-ear hearing[br]aids, of course, most of the

0:20:45.090,0:20:47.370
microphones and everything[br]else is behind the ear.

0:20:47.370,0:20:49.640
So you can not use the pinna.

0:20:49.640,0:20:52.300
You have to simulate that[br]in a different way.

0:20:52.300,0:20:56.100
And this is why both hearing[br]aids have two microphones.

0:20:56.100,0:20:59.540
So you have four microphones[br]in total, when you have a

0:20:59.540,0:21:02.050
hearing impairment[br]in both ears.

0:21:02.050,0:21:05.860
And this way, they can detect[br]if the sound comes

0:21:05.860,0:21:07.640
from front or back.

0:21:07.640,0:21:12.830
And they talk to each other, so[br]they can also detect if the

0:21:12.830,0:21:14.550
source of the signal is[br]right or left to you.

0:21:18.170,0:21:23.420
Additionally, they try to[br]recognize the situation and

0:21:23.420,0:21:27.170
automatically focus on the[br]person that is talking to you

0:21:27.170,0:21:32.660
and also to reduce the[br]background noise in general.

0:21:32.660,0:21:35.240
The in-ear hearing aids I[br]showed before, they, of

0:21:35.240,0:21:37.600
course, still can use the[br]features of the pinna.

0:21:40.380,0:21:43.740
This is also a screenshot from[br]the tuning software.

0:21:43.740,0:21:46.390
You can actually test[br]your directional

0:21:46.390,0:21:47.740
hearing in real time.

0:21:47.740,0:21:49.230
So you can wear your[br]hearing aids

0:21:49.230,0:21:51.150
connected to the software.

0:21:51.150,0:21:56.430
And then you can do something[br]like that, and see if it's

0:21:56.430,0:21:59.070
recognized correctly.

0:21:59.070,0:22:01.790
It works, more or less,[br]in a silent room.

0:22:01.790,0:22:07.410
But yeah, not in a[br]cocktail party.

0:22:07.410,0:22:11.990
Yeah, generally, it's really[br]hard to extract foreground

0:22:11.990,0:22:16.780
from background noise, because[br]foreground noise has all the

0:22:16.780,0:22:19.140
high frequencies and background[br]noise doesn't.

0:22:19.140,0:22:23.190
And if you don't hear high[br]frequencies at all, everything

0:22:23.190,0:22:28.082
is one blob of sound.

0:22:28.082,0:22:33.800
Hearing aids help with that[br]because they mostly focus on

0:22:33.800,0:22:35.510
high frequencies.

0:22:35.510,0:22:38.420
And they have filters[br]to filter out

0:22:38.420,0:22:39.850
the background signal.

0:22:39.850,0:22:44.210
But actually, that doesn't[br]really help so much, because

0:22:44.210,0:22:47.920
those situations, recognitions,[br]they tend to

0:22:47.920,0:22:48.900
fail as well.

0:22:48.900,0:22:51.690
So sometimes it's likely that[br]a person that is talking to

0:22:51.690,0:22:55.774
you gets faded out because it[br]is considered as noise.

0:22:55.774,0:22:57.256
[LAUGHTER]

0:22:57.256,0:23:01.320
HELGA VELROYEN: But sometimes[br]they also work very well.

0:23:01.320,0:23:04.620
I mean, it's also coincidence.

0:23:04.620,0:23:07.070
Sometimes you're in a setting[br]which has exactly the

0:23:07.070,0:23:10.100
situation that the hearing[br]aid can work well with.

0:23:10.100,0:23:13.520
So sometimes you are here with[br]a hearing person, actually,

0:23:13.520,0:23:15.600
and then you hear[br]him quite well.

0:23:15.600,0:23:18.380
And then you start talking in[br]a normal voice, because you

0:23:18.380,0:23:19.970
can actually understand[br]everything.

0:23:19.970,0:23:21.820
And then that hearing[br]person is asking

0:23:21.820,0:23:23.410
you to repeat a sentence.

0:23:23.410,0:23:24.970
This is really weird[br]sometimes.

0:23:28.110,0:23:30.560
The tuning software for hearing[br]aids also has a real

0:23:30.560,0:23:34.280
time monitor where you can[br]see some parameters.

0:23:34.280,0:23:37.950
So you wear your hearing[br]aids and then, for

0:23:37.950,0:23:39.580
example, listen to music.

0:23:39.580,0:23:47.090
And then you can see here the[br]dark areas are where the

0:23:47.090,0:23:50.000
hearing aid actually[br]started to work.

0:23:50.000,0:23:51.860
Below that, it doesn't[br]need to amplify.

0:23:51.860,0:23:55.590
And the grey thing here[br]is the speech banana.

0:23:55.590,0:23:57.940
So you see that it's optimized[br]in a way that it starts

0:23:57.940,0:24:01.100
amplifying in the[br]speech banana.

0:24:01.100,0:24:05.340
It also has the situation[br]recognition

0:24:05.340,0:24:06.410
and different settings.

0:24:06.410,0:24:08.790
And one of those is[br]music, actually.

0:24:08.790,0:24:11.970
I tried it with different[br]types of music.

0:24:11.970,0:24:14.120
Actually, if you like heavy[br]metal, you will never see

0:24:14.120,0:24:14.960
music here.

0:24:14.960,0:24:16.170
It's usually noise.

0:24:16.170,0:24:21.856
[LAUGHTER]

0:24:21.856,0:24:23.260
HELGA VELROYEN: A problem[br]with hearing aids

0:24:23.260,0:24:25.010
is humidity, actually.

0:24:25.010,0:24:27.380
So most hearing aids[br]are not waterproof.

0:24:27.380,0:24:32.320
That doesn't sound so bad, but[br]actually, a lot of things are

0:24:32.320,0:24:33.060
related to that.

0:24:33.060,0:24:36.590
So that means no swimming with[br]friends, no pool parties, no

0:24:36.590,0:24:39.880
water sports where you'll[br]have to talk to someone.

0:24:39.880,0:24:41.740
Sweat is a problem, especially[br]for people

0:24:41.740,0:24:43.970
who do a lot of sports.

0:24:43.970,0:24:45.770
No audio books in the bathtub.

0:24:45.770,0:24:46.680
No heavy rain.

0:24:46.680,0:24:49.680
So if you go to an open air[br]concert and it starts raining,

0:24:49.680,0:24:52.610
you will really have to[br]take care of that.

0:24:52.610,0:24:57.880
Also, just like wet hair, when[br]you go out of the shower, you

0:24:57.880,0:25:01.340
have to wait until your hair is[br]dry until you can put your

0:25:01.340,0:25:03.490
hearing aids in again.

0:25:03.490,0:25:07.730
A very recent development is[br]that Phonak also offers

0:25:07.730,0:25:10.200
hearing aids that are waterproof[br]or water-resistant.

0:25:13.310,0:25:16.180
They claim that you can hold it[br]under water for 30 minutes,

0:25:16.180,0:25:21.020
and then there will be no[br]irreparable damage.

0:25:21.020,0:25:22.350
I'm not really sure[br]what that means.

0:25:22.350,0:25:24.250
So I guess you probably[br]have to dry it.

0:25:24.250,0:25:25.531
[LAUGHTER]

0:25:25.531,0:25:30.290
HELGA VELROYEN: Or if you have[br]to give it into repair for

0:25:30.290,0:25:31.760
weeks until you get it back.

0:25:31.760,0:25:35.250
Oh yeah, by the way, you don't[br]have a spare hearing aid.

0:25:35.250,0:25:38.260
So whenever your hearing aids[br]break, you have to go to an

0:25:38.260,0:25:41.560
audiologist. And you get spare[br]hearing aids, which is like,

0:25:41.560,0:25:43.940
when you drive a Porsche,[br]you get a tractor.

0:25:43.940,0:25:48.800
And also, audiologists have[br]opening hours for the elderly.

0:25:48.800,0:25:51.580
So if your hearing aids break[br]on Friday night, you have to

0:25:51.580,0:25:53.840
wait until Monday until you[br]actually can hear again.

0:25:56.790,0:26:00.780
Yeah, those were the important[br]features of hearing aids.

0:26:00.780,0:26:03.890
But there is some peripheral[br]hardware that I

0:26:03.890,0:26:06.060
would like to present.

0:26:06.060,0:26:10.800
So there are different[br]interfaces for which you can

0:26:10.800,0:26:12.950
use to plug into your[br]hearing aids.

0:26:12.950,0:26:15.850
So the oldest one is the direct[br]audio input, which is

0:26:15.850,0:26:17.970
just here, a cable.

0:26:17.970,0:26:20.950
So it looks like that.

0:26:20.950,0:26:23.830
And it's usually connected to[br]the hearing aid with some sort

0:26:23.830,0:26:26.170
of shoe, which looks[br]like that.

0:26:26.170,0:26:30.610
And it has the pros and cons of[br]cables, of course, if you

0:26:30.610,0:26:31.720
like being on the leash.

0:26:31.720,0:26:35.180
But you also have no[br]interference with other

0:26:35.180,0:26:37.330
wireless stuff.

0:26:37.330,0:26:40.470
It's usually used to plug[br]something else in, for

0:26:40.470,0:26:43.415
example, FM or Bluetooth[br]adaptors.

0:26:43.415,0:26:47.570
And yeah, this is pretty[br]old technology,

0:26:47.570,0:26:48.980
but it's still around.

0:26:48.980,0:26:52.310
Although, for really small ones,[br]the plug is actually too

0:26:52.310,0:26:55.530
big, so they leave[br]it out there.

0:26:55.530,0:27:00.860
Another very common technology[br]is FM systems. You can buy

0:27:00.860,0:27:02.370
those from several vendors.

0:27:02.370,0:27:05.480
You have receivers[br]and transmitters.

0:27:05.480,0:27:07.940
They have different setups like[br]for a meeting, that you

0:27:07.940,0:27:10.960
can put a microphone on the[br]meeting table and hear all the

0:27:10.960,0:27:15.060
participants talk, or for[br]lecture halls like that, that

0:27:15.060,0:27:18.120
you can connect the microphone[br]to the transmitter

0:27:18.120,0:27:19.720
and listen to it.

0:27:22.260,0:27:26.080
There are some standards,[br]but most receivers and

0:27:26.080,0:27:29.200
transmitters don't work with[br]others from other companies.

0:27:29.200,0:27:33.730
But at least, if you plug them[br]in with these direct audio

0:27:33.730,0:27:38.490
input, you can choose an FM[br]system from a different vendor

0:27:38.490,0:27:40.626
than from your hearing aids.

0:27:40.626,0:27:43.760
The sound quality is said[br]to be quite good.

0:27:43.760,0:27:45.140
I actually could never try it.

0:27:45.140,0:27:51.210
But I heard that, actually, in[br]schools, hearing impaired

0:27:51.210,0:27:55.290
students listen to music while[br]actually they should listen to

0:27:55.290,0:27:56.180
the teacher.

0:27:56.180,0:27:59.130
And teachers call the[br]audiologists and ask, could

0:27:59.130,0:28:00.390
you make this stop?

0:28:00.390,0:28:02.070
They don't listen to[br]what I am saying.

0:28:05.880,0:28:08.860
A really common problem for[br]hearing impairment is calling

0:28:08.860,0:28:11.700
on the phone.

0:28:11.700,0:28:15.230
The problem is, first of all,[br]most people do lip reading.

0:28:15.230,0:28:17.660
So they don't rely only[br]on the audio input.

0:28:17.660,0:28:20.240
We use the visuals as well.

0:28:20.240,0:28:22.510
And of course, you don't[br]have that on the

0:28:22.510,0:28:24.330
radio or the telephone.

0:28:24.330,0:28:27.000
And also, for technical reasons,[br]the frequency range

0:28:27.000,0:28:29.780
of the phone line is reduced.

0:28:29.780,0:28:34.710
In Germany, it's like 300[br]hertz to 3.4 kilohertz.

0:28:34.710,0:28:38.560
That's the blue box[br]in this audiogram.

0:28:38.560,0:28:41.360
Also the background noise that[br]you have in the room where you

0:28:41.360,0:28:44.910
are talking on the phone, it[br]does have the full range.

0:28:44.910,0:28:46.650
So you get, actually, background[br]noise in a better

0:28:46.650,0:28:50.250
quality than the signal from[br]your person that you're

0:28:50.250,0:28:51.650
talking to.

0:28:51.650,0:28:55.670
Then, often, the signal is[br]altered and unnatural.

0:28:55.670,0:28:57.510
Sometimes you have[br]bad reception.

0:28:57.510,0:29:01.490
And also, you hear it only in[br]one ear if you just use a

0:29:01.490,0:29:02.880
usual telephone.

0:29:02.880,0:29:05.460
And if you hold it to your ear,[br]you get a feedback loop.

0:29:05.460,0:29:07.400
So a lot of things make[br]it really annoying

0:29:07.400,0:29:08.200
to talk on the phone.

0:29:08.200,0:29:11.200
And there are some technical[br]solutions for that.

0:29:11.200,0:29:14.935
The most old one is the[br]telecoil or T-coil.

0:29:18.610,0:29:20.910
And the source is connected[br]to an induction loop.

0:29:20.910,0:29:24.710
And you take off the[br]electromagnetics.

0:29:24.710,0:29:26.850
And this is from the telecoil.

0:29:26.850,0:29:30.610
It's there in the picture,[br]a really small antenna.

0:29:30.610,0:29:35.240
And there are different setups[br]for the induction loop.

0:29:35.240,0:29:38.030
So there are adapters that[br]have the induction loop

0:29:38.030,0:29:40.420
actually used to hang[br]it around your neck.

0:29:40.420,0:29:42.660
Or there are induction[br]loops installed in

0:29:42.660,0:29:43.920
lecture halls like this.

0:29:43.920,0:29:47.900
I don't know if there[br]is one here.

0:29:47.900,0:29:51.140
It's widely used in Europe,[br]especially in Scandinavia.

0:29:51.140,0:29:54.340
They even have laws where every[br]public building has to

0:29:54.340,0:29:59.640
have one or these public[br]lecture halls.

0:29:59.640,0:30:01.760
You have some pros and[br]cons, of course.

0:30:01.760,0:30:03.580
You have interference.

0:30:03.580,0:30:07.580
When you move inside the[br]induction loop, the level of

0:30:07.580,0:30:10.030
volume changes.

0:30:10.030,0:30:12.730
So it's nothing where you should[br]dance or something.

0:30:12.730,0:30:14.520
But if you sit still[br]in a theater or

0:30:14.520,0:30:17.320
something, it works.

0:30:17.320,0:30:19.840
Installing an induction[br]loop in a lecture

0:30:19.840,0:30:21.440
hall is quite expensive.

0:30:21.440,0:30:23.990
But there are actually[br]DIY kits available.

0:30:23.990,0:30:26.210
It's quite common that people[br]build their own ones.

0:30:28.820,0:30:31.430
And telephones also have[br]an induction loop,

0:30:31.430,0:30:33.010
even very new ones.

0:30:33.010,0:30:36.770
So all telephones which are[br]called hearing aid compatible,

0:30:36.770,0:30:38.870
they have an induction loop[br]that can be used with a

0:30:38.870,0:30:41.525
telecoil, even the new[br]iPhone, for example.

0:30:44.920,0:30:46.170
And then there's Bluetooth.

0:30:48.370,0:30:51.820
There are, right now, no hearing[br]aids that can do

0:30:51.820,0:30:54.940
Bluetooth directly because,[br]mostly, of the batteries.

0:30:54.940,0:30:59.840
Usually, hearing aid batteries[br]last one week to 10 days.

0:30:59.840,0:31:02.260
But with Bluetooth, I think[br]they would only

0:31:02.260,0:31:03.950
last a couple of hours.

0:31:03.950,0:31:08.100
Although there are no[br]Bluetooth-enabled hearing aids

0:31:08.100,0:31:12.140
on the market right now,[br]I've heard that they

0:31:12.140,0:31:13.770
are working on that.

0:31:13.770,0:31:17.220
Siemens is located in Erlangen,[br]which is not that

0:31:17.220,0:31:18.230
far from Munich.

0:31:18.230,0:31:23.510
So I have heard about people[br]who would test prototypes.

0:31:23.510,0:31:28.615
Right now, you have to use a[br]Bluetooth adapter to use your

0:31:28.615,0:31:30.460
hearing aids, attach that.

0:31:30.460,0:31:34.380
And there's different[br]versions of it.

0:31:34.380,0:31:36.940
So this is an example[br]of Phonak.

0:31:36.940,0:31:39.290
They built a really[br]nice thing.

0:31:39.290,0:31:41.910
This piece, you hang it[br]around your neck.

0:31:41.910,0:31:45.040
And the ribbon is actually the[br]induction loop that is used

0:31:45.040,0:31:46.340
with a telecoil.

0:31:46.340,0:31:51.030
And to this gadget you can[br]connect several things.

0:31:51.030,0:31:53.720
It has direct audio input, which[br]you can use to plug in

0:31:53.720,0:31:55.090
an FM system.

0:31:55.090,0:31:57.760
It also has aux-in that[br]you can directly

0:31:57.760,0:31:59.576
plug in your MP3 player.

0:31:59.576,0:32:01.930
And it has Bluetooth.

0:32:01.930,0:32:04.310
It actually also has a warning[br]to combine this with

0:32:04.310,0:32:05.800
pacemakers.

0:32:05.800,0:32:10.600
And I have a friend who has a[br]pacemaker and this thing.

0:32:10.600,0:32:11.920
He just ignored the warning.

0:32:11.920,0:32:12.790
But I think this is--

0:32:12.790,0:32:13.470
[LAUGHTER]

0:32:13.470,0:32:16.720
HELGA VELROYEN: Luckily,[br]he's still alive.

0:32:16.720,0:32:18.810
And I think this is a good[br]example for what we are

0:32:18.810,0:32:20.050
heading in the future.

0:32:20.050,0:32:21.970
We will get more and[br]more cyborgs.

0:32:21.970,0:32:26.680
And I doubt that every hearing[br]aid vendor is trying their

0:32:26.680,0:32:29.850
adapters with every pacemaker[br]there is.

0:32:29.850,0:32:33.320
So we will get a lot[br]of compatibility

0:32:33.320,0:32:34.680
problems in the future.

0:32:34.680,0:32:38.580
And if this really works, it[br]might be an option to kill

0:32:38.580,0:32:41.140
people really, really[br]silently, remotely.

0:32:41.140,0:32:41.485
[LAUGHTER]

0:32:41.485,0:32:44.694
HELGA VELROYEN: I mean, you[br]have to think of that.

0:32:44.694,0:32:47.600
[LAUGHS]

0:32:47.600,0:32:50.710
Siemens' solution was, of[br]course, not to use any of the

0:32:50.710,0:32:52.150
existing standards.

0:32:52.150,0:32:53.574
Just build something new.

0:32:53.574,0:32:54.068
[LAUGHTER]

0:32:54.068,0:32:57.032
[APPLAUSE]

0:32:57.032,0:33:00.530
HELGA VELROYEN: So they built[br]something called Siemens Tek.

0:33:00.530,0:33:03.720
And it hooks Bluetooth to the[br]phone, or to whatever you're

0:33:03.720,0:33:09.290
connected to, and some wireless[br]NFC protocol to the

0:33:09.290,0:33:10.840
hearing aids.

0:33:10.840,0:33:12.880
So the hearing aids also have[br]a small antenna in it, which

0:33:12.880,0:33:15.760
looks kind of like a telecoil,[br]but it is not.

0:33:15.760,0:33:18.810
And it is not compatible to[br]anything the telecoil is

0:33:18.810,0:33:19.930
compatible to.

0:33:19.930,0:33:22.100
So you can only use[br]it with this.

0:33:22.100,0:33:25.230
And it has a signal around[br]3.3 megahertz.

0:33:25.230,0:33:26.340
You can see it on[br]that picture.

0:33:26.340,0:33:28.030
I tried that.

0:33:28.030,0:33:31.600
And it's compatible with[br]every Bluetooth

0:33:31.600,0:33:33.420
speaking device, in theory.

0:33:33.420,0:33:35.820
And practically, you have to[br]check it with everything.

0:33:35.820,0:33:39.050
So whenever I get a phone at[br]work or whatever, I have to

0:33:39.050,0:33:40.590
check if it will[br]actually work.

0:33:40.590,0:33:44.170
So of course, it works best with[br]Siemens mobile phones.

0:33:44.170,0:33:46.670
But you can imagine[br]how old they are.

0:33:46.670,0:33:53.040
And yeah, they also are supposed[br]to work with land

0:33:53.040,0:33:53.670
line phones.

0:33:53.670,0:33:59.310
But on their websites, they say,[br]I think, it was only to

0:33:59.310,0:34:02.850
Siemens land line phones,[br]of course.

0:34:02.850,0:34:07.910
And other than that, they don't[br]guarantee that it works.

0:34:07.910,0:34:12.139
It comes with an additional[br]transmitter that you can

0:34:12.139,0:34:15.219
connect to your source that[br]is more far away.

0:34:15.219,0:34:18.760
The Tek itself has a[br]range of one meter.

0:34:18.760,0:34:22.880
So if you want to have a flat[br]screen TV, I don't know, five

0:34:22.880,0:34:26.210
meters away from you, you could[br]use the transmitter.

0:34:26.210,0:34:30.659
This thing costs about 400 euros[br]for just turning one

0:34:30.659,0:34:33.260
wireless protocol[br]into the other.

0:34:33.260,0:34:35.230
And no insurance is going[br]to pay for that.

0:34:35.230,0:34:38.949
So you have to pay for[br]this on your own.

0:34:38.949,0:34:42.050
If you have a generous[br]employer, you might

0:34:42.050,0:34:44.989
get some for it.

0:34:44.989,0:34:48.120
They also released a[br]new version of it.

0:34:48.120,0:34:50.040
This is on the right[br]side here.

0:34:50.040,0:34:50.929
It's called miniTek.

0:34:50.929,0:34:54.280
And it has, actually, less[br]features than the old one

0:34:54.280,0:34:57.440
because they removed[br]the display.

0:34:57.440,0:34:59.170
And they still want 400[br]euros for that.

0:34:59.170,0:35:03.290
And you don't get a discount[br]if you bought the old one.

0:35:03.290,0:35:06.940
Well, but I guess that's[br]marketing.

0:35:06.940,0:35:08.210
This is the sender with[br]the transmitter.

0:35:08.210,0:35:11.070
So you connect it to the[br]computer, and you wear the Tek

0:35:11.070,0:35:13.190
around your head.

0:35:13.190,0:35:16.370
It also has some patents for[br]the different programs. So

0:35:16.370,0:35:19.220
hearing aids have different[br]programs that you can change

0:35:19.220,0:35:23.250
manually so that you have one[br]for listening to music, or one

0:35:23.250,0:35:29.880
for your living room, and one[br]for outside, or whatever.

0:35:29.880,0:35:31.785
Of course, I took that[br]apart as well.

0:35:31.785,0:35:34.530
[LAUGHTER]

0:35:34.530,0:35:37.790
HELGA VELROYEN: You have to[br]couple this Tek with the

0:35:37.790,0:35:39.840
tuning software of[br]hearing aids.

0:35:39.840,0:35:43.110
So there is some kind of[br]authentification via a

0:35:43.110,0:35:45.480
7-character serial number.

0:35:45.480,0:35:49.330
I doubt that there is actually[br]a lot of encryption in there,

0:35:49.330,0:35:51.370
because the latency[br]is crucial.

0:35:51.370,0:35:53.130
Bluetooth already[br]has a latency.

0:35:53.130,0:35:55.320
And you don't want to[br]add that much to it.

0:35:55.320,0:35:58.680
But I also tried to use a[br]different Tek with my hearing

0:35:58.680,0:36:00.190
aids, and it actually[br]doesn't work.

0:36:00.190,0:36:03.910
So some kind of authentication[br]must be in there.

0:36:03.910,0:36:06.350
But if you're too lazy[br]to hack that, you can

0:36:06.350,0:36:08.100
still hack the Bluetooth.

0:36:08.100,0:36:11.550
There are lots of talks[br]about that here.

0:36:11.550,0:36:14.460
And of course, the PIN is 000.

0:36:14.460,0:36:19.390
And something that is not[br]directly about the Tek, but

0:36:19.390,0:36:21.520
the hearing aids also[br]communicate with each other.

0:36:21.520,0:36:23.970
So if I switch the program--[br]there's actually a small

0:36:23.970,0:36:25.370
switch on my hearing aids--

0:36:25.370,0:36:28.800
if I switch between the programs[br]on one ear, it also

0:36:28.800,0:36:30.480
tells that to the other ear.

0:36:30.480,0:36:33.010
And that one doesn't use[br]authentification.

0:36:33.010,0:36:35.600
I have heard that, when people[br]have the same model of hearing

0:36:35.600,0:36:37.480
aids and they're close to each[br]other-- for example, this

0:36:37.480,0:36:40.040
happens when couples buy[br]the same hearing aids--

0:36:40.040,0:36:43.310
then you switch your program,[br]and your spouse also gets the

0:36:43.310,0:36:44.485
program switched.

0:36:44.485,0:36:45.315
[LAUGHTER]

0:36:45.315,0:36:48.080
HELGA VELROYEN: And this can[br]actually only be changed by

0:36:48.080,0:36:50.430
the audiologist by changing[br]the channel.

0:36:50.430,0:36:54.190
So they have, like for wireless,[br]different channels.

0:36:54.190,0:36:55.590
And you'd set it to[br]a different one.

0:36:55.590,0:36:57.640
That's the security[br]about that.

0:37:00.184,0:37:03.590
Yeah, hacking.

0:37:03.590,0:37:06.200
When I started to dig into[br]that topic, I was really

0:37:06.200,0:37:09.950
disappointed that there is[br]not very much hacking.

0:37:09.950,0:37:12.530
So there's one forum called[br]hearingaidhacks.

0:37:12.530,0:37:15.160
livejournal.com.

0:37:15.160,0:37:19.320
This is the biggest[br]one I found.

0:37:19.320,0:37:23.810
But if you go through the[br]entries, it's mostly people

0:37:23.810,0:37:27.940
asking for technical advice,[br]So, I bought this and this

0:37:27.940,0:37:28.450
hearing aid.

0:37:28.450,0:37:32.090
What peripheral hardware[br]can I use with it?

0:37:32.090,0:37:35.610
I guess the reason for[br]that is that the

0:37:35.610,0:37:36.980
devices are really expensive.

0:37:36.980,0:37:41.470
And the warranty and the[br]insurances are really not that

0:37:41.470,0:37:43.300
nice if you break your[br]own hearing aids.

0:37:43.300,0:37:46.230
And you still have this problem[br]that audiologists

0:37:46.230,0:37:47.710
don't have opening hours.

0:37:47.710,0:37:50.920
They can not be used by people[br]who still have a life.

0:37:50.920,0:37:54.520
So people are a little resistant[br]to actually hack the

0:37:54.520,0:37:56.970
hearing aids.

0:37:56.970,0:38:00.270
But there is a little hacking[br]on the peripheral hardware.

0:38:00.270,0:38:04.920
I will show two examples[br]for that.

0:38:04.920,0:38:07.660
There's one guy called[br]Gertlex who built his

0:38:07.660,0:38:10.650
own Bluetooth adapter.

0:38:10.650,0:38:14.570
And he posted this on Flickr,[br]quite detailed.

0:38:14.570,0:38:18.810
And what you can see here, he[br]used a Sony wireless Bluetooth

0:38:18.810,0:38:24.220
headset, and hacked it in a way[br]where you can connect the

0:38:24.220,0:38:28.230
direct audio input cables[br]from the hearing aids.

0:38:28.230,0:38:30.920
The picture here in the[br]upper-right corner shows the

0:38:30.920,0:38:32.810
setup when he tested it.

0:38:32.810,0:38:35.020
So he actually didn't[br]test it with his

0:38:35.020,0:38:36.020
original hearing aids.

0:38:36.020,0:38:38.030
He used an old one[br]that he had.

0:38:38.030,0:38:41.330
And he even used an old MP3[br]player because he was afraid

0:38:41.330,0:38:43.040
of frying that as well.

0:38:43.040,0:38:49.770
So this is the precaution that[br]you have to do when you start

0:38:49.770,0:38:51.700
frying your hearing aids.

0:38:51.700,0:38:55.090
You can not only fry your[br]hearing aids, you can also fry

0:38:55.090,0:38:56.150
your hearing even more.

0:38:56.150,0:38:58.800
So you should be careful.

0:38:58.800,0:39:03.200
And there's another guy who also[br]made a Bluetooth adapter.

0:39:03.200,0:39:11.026
And he also took the DYI cables[br]here, those shoes that

0:39:11.026,0:39:16.080
you use to connect them, and a[br]Bluetooth mono thing here.

0:39:16.080,0:39:21.080
And this is actually the result,[br]that you connect it

0:39:21.080,0:39:22.830
directly to the hearing aids.

0:39:22.830,0:39:26.770
He also provided some nice[br]diagrams for that.

0:39:26.770,0:39:30.500
The slightly bigger scene is[br]actually the self-tuning

0:39:30.500,0:39:36.570
scenes for hearing aids because,[br]as I said, it's kind

0:39:36.570,0:39:38.460
of frustrating to get[br]hearing aids.

0:39:38.460,0:39:41.510
You have to go to the[br]audiologist a lot of times.

0:39:41.510,0:39:43.600
And he asks you, yeah,[br]what's wrong?

0:39:43.600,0:39:45.870
And then you have to describe[br]the situation.

0:39:45.870,0:39:48.480
But you're sitting in this[br]silent cavern at the

0:39:48.480,0:39:54.290
audiologist, so the adjustment[br]is not really done in

0:39:54.290,0:39:56.070
realistic circumstances.

0:39:56.070,0:39:58.690
And a lot of people get[br]frustrated about that.

0:39:58.690,0:40:02.930
So they spend weeks tuning[br]their hearing aids at the

0:40:02.930,0:40:05.570
audiologist. And still they're[br]unhappy with it.

0:40:05.570,0:40:09.120
So they try to get the hardware[br]and software that is

0:40:09.120,0:40:10.770
necessary for it.

0:40:10.770,0:40:13.190
And those are actually only[br]sold to doctors and

0:40:13.190,0:40:14.480
acousticians or audiologists.

0:40:17.150,0:40:18.820
And they are not sold[br]on eBay because

0:40:18.820,0:40:19.820
it is medical equipment.

0:40:19.820,0:40:25.310
And that is not supposed[br]to be sold on eBay.

0:40:25.310,0:40:30.590
So you have to use[br]other channels.

0:40:30.590,0:40:33.810
There's a black market for it.

0:40:33.810,0:40:39.200
It's kind of hard to put a price[br]on that, but I've seen

0:40:39.200,0:40:42.130
offers for the hardware,[br]which is called Hipro.

0:40:42.130,0:40:45.160
And there are different[br]versions for

0:40:45.160,0:40:47.670
serial use via Bluetooth.

0:40:47.670,0:40:49.920
And it starts with a couple[br]of hundred euros.

0:40:49.920,0:40:52.160
So you can imagine the pain[br]that people have when they

0:40:52.160,0:40:54.860
already spent 5,000 euros[br]for two hearing aids.

0:40:54.860,0:40:59.430
And then they spend even more[br]money, because they want to

0:40:59.430,0:41:01.830
tune them themselves.

0:41:01.830,0:41:04.760
And there is this self-tuning[br]scene.

0:41:04.760,0:41:07.475
And people really[br]hack the system.

0:41:07.475,0:41:12.100
The effort for people who[br]actually installed a fake

0:41:12.100,0:41:16.250
business, so they register a[br]business for an audiologist to

0:41:16.250,0:41:19.110
buy this hardware and then[br]stop the business again.

0:41:19.110,0:41:22.960
So there are actually people[br]doing a lot of effort to get

0:41:22.960,0:41:25.300
this hardware and software.

0:41:25.300,0:41:27.440
But of course, then you have[br]no customer support.

0:41:27.440,0:41:30.120
And when you fry your ears or[br]your hearing aids, then it's

0:41:30.120,0:41:31.640
your own fault.

0:41:31.640,0:41:32.540
There's one exception.

0:41:32.540,0:41:35.920
In America, there's a hearing[br]aid manufacturer called

0:41:35.920,0:41:37.410
America Hears.

0:41:37.410,0:41:41.460
They sell quite low-budget[br]hearing aids, up to

0:41:41.460,0:41:45.030
$1,000, if I'm right.

0:41:45.030,0:41:47.340
Then you send in an audiogram[br]of yours.

0:41:47.340,0:41:51.680
And they tune it at their place[br]for the first time.

0:41:51.680,0:41:56.320
And then you can download the[br]software and tune it a little

0:41:56.320,0:41:58.930
bit more at home.

0:41:58.930,0:42:00.910
Unfortunately, I've never[br]seen that software.

0:42:00.910,0:42:02.530
You can only order[br]it in the US.

0:42:02.530,0:42:06.370
But I would be interested to[br]have a look at that as well,

0:42:06.370,0:42:08.590
if someone has channels[br]for that.

0:42:11.240,0:42:15.580
Of course, some of this hardware[br]ended up in my hands.

0:42:15.580,0:42:19.705
This is a serial Hipro.

0:42:19.705,0:42:23.290
This looks really fancy, like[br]a modem from the '80s.

0:42:23.290,0:42:26.570
It's connected via serial.

0:42:26.570,0:42:28.740
And you connect the hearing[br]aids to it.

0:42:28.740,0:42:31.460
I have some close-up[br]pictures for that.

0:42:31.460,0:42:35.730
So you take out the battery and[br]put on a small cable which

0:42:35.730,0:42:39.740
has a contact of the battery,[br]the size of the battery, a

0:42:39.740,0:42:40.770
very flat cable.

0:42:40.770,0:42:42.610
And this is connected[br]to a bigger cable.

0:42:42.610,0:42:44.030
And that is connected[br]to the Hipro.

0:42:46.740,0:42:50.960
The Hipro is the same for nearly[br]all hearing aid brands,

0:42:50.960,0:42:54.600
so you can use it for Siemens,[br]and Phonak, and whatever.

0:42:54.600,0:42:58.355
But these small, flat cables,[br]they are different for nearly

0:42:58.355,0:42:59.930
every hearing aid.

0:42:59.930,0:43:05.160
So if you try to buy this Hipro[br]on the black market, you

0:43:05.160,0:43:07.610
also have to buy those cables.

0:43:07.610,0:43:09.880
There's also a Bluetooth[br]version that also

0:43:09.880,0:43:12.310
ended up in my hands.

0:43:12.310,0:43:13.070
It looks like that.

0:43:13.070,0:43:16.850
And it has the advantage[br]that you are not

0:43:16.850,0:43:19.900
connected to it via cable.

0:43:19.900,0:43:22.380
So you don't feel[br]like on a leash.

0:43:22.380,0:43:27.140
And I haven't really[br]used that much.

0:43:27.140,0:43:29.890
But this way, you could actually[br]go outside and tune

0:43:29.890,0:43:33.710
it in the subway or at[br]your office, because

0:43:33.710,0:43:35.790
you just need a laptop.

0:43:35.790,0:43:38.450
And it works without[br]your powerful wire.

0:43:43.560,0:43:45.890
The tuning software, I showed[br]you some excerpts from it.

0:43:45.890,0:43:47.850
This is just another[br]screenshot.

0:43:47.850,0:43:52.790
On the right side here, you see[br]the different programs. So

0:43:52.790,0:43:56.160
the universal one, here is one[br]for music, and one for the

0:43:56.160,0:43:59.158
tech, if you work with that.

0:44:02.340,0:44:05.700
When I was playing with the[br]tuning software, I found

0:44:05.700,0:44:07.020
something very interesting.

0:44:07.020,0:44:09.170
That it's actually[br]spying on me.

0:44:09.170,0:44:13.340
So it locks some stuff[br]and, for example,

0:44:13.340,0:44:15.570
how much I wear it.

0:44:15.570,0:44:18.990
Although I find 14 hours[br]a day a little bit--

0:44:18.990,0:44:22.550
I think I'm awake, actually,[br]longer than 14 hours.

0:44:22.550,0:44:24.740
But they use this, actually,[br]because people come and

0:44:24.740,0:44:26.820
complain, yeah, this doesn't[br]really work much.

0:44:26.820,0:44:29.960
And then they see they only wear[br]it half an hour a day.

0:44:29.960,0:44:31.950
And so, of course, you can[br]not get used to it

0:44:31.950,0:44:33.470
and adapt to it.

0:44:33.470,0:44:35.900
And you can also see what[br]different programs are used.

0:44:35.900,0:44:37.960
So mostly I use the[br]universal program.

0:44:37.960,0:44:42.730
And sometimes I have another one[br]called universal that is

0:44:42.730,0:44:45.950
actually tuned to have[br]less feedback loops.

0:44:45.950,0:44:48.760
That's the one I use when I put[br]on a hat so that I still

0:44:48.760,0:44:53.080
hear a little bit, but don't[br]have too many feedback loops.

0:44:53.080,0:44:56.910
And it also tracks how often I[br]was in a noisy environment or

0:44:56.910,0:44:58.480
was listening to music.

0:44:58.480,0:45:01.745
And since I like to listen to[br]heavy metal, this is actually

0:45:01.745,0:45:03.750
not correct.

0:45:03.750,0:45:05.320
Yeah, I found it really[br]interesting

0:45:05.320,0:45:06.260
what you can see here.

0:45:06.260,0:45:12.290
And yeah, I hope that it doesn't[br]record anything what I

0:45:12.290,0:45:15.840
talk about or what[br]I listen to.

0:45:15.840,0:45:19.420
We have a little bit more time,[br]so I will talk about the

0:45:19.420,0:45:21.470
cochlear implants, as well.

0:45:21.470,0:45:25.840
I mentioned that those are the[br]ones that are half implanted

0:45:25.840,0:45:29.760
in the head and also have[br]an external device.

0:45:33.650,0:45:37.420
This shows this part[br]is implanted.

0:45:37.420,0:45:42.200
And we have a wire that is[br]drawn into the cochlea.

0:45:42.200,0:45:44.590
And it's connected[br]to the nerve.

0:45:44.590,0:45:47.180
So actually, the whole[br]ear is circumvented.

0:45:47.180,0:45:50.020
Only the wire goes directly[br]to the nerve and

0:45:50.020,0:45:51.390
then through the brain.

0:45:51.390,0:45:56.350
And cochlear implants are what[br]I find really fascinating.

0:45:56.350,0:45:58.450
They really make deaf[br]people hear.

0:45:58.450,0:46:02.190
It's only applied to people[br]who have a really severe

0:46:02.190,0:46:05.580
hearing loss, like less[br]than 20% or something.

0:46:05.580,0:46:08.970
So they hear only[br]less than 20%.

0:46:08.970,0:46:13.630
Of course, it's a surgery[br]to insert that.

0:46:13.630,0:46:16.530
It destroys any remaining[br]hearing, because you poke a

0:46:16.530,0:46:17.640
wire into the nerve.

0:46:17.640,0:46:21.470
So everything else[br]is gone then.

0:46:21.470,0:46:23.400
It can also affect[br]other nerves.

0:46:23.400,0:46:26.460
So I have a friend who[br]had this surgery.

0:46:26.460,0:46:31.650
And they touched some taste[br]nerves as well, so everything

0:46:31.650,0:46:33.120
tastes as metallic.

0:46:33.120,0:46:33.830
It was kind of weird.

0:46:33.830,0:46:37.260
But it actually went[br]away after a while.

0:46:37.260,0:46:38.510
The signal is really[br]different.

0:46:40.980,0:46:43.810
The brain has to adjust[br]to that very long.

0:46:43.810,0:46:46.070
There are actually hearing[br]courses after you get this

0:46:46.070,0:46:48.440
surgery and the device[br]is started.

0:46:48.440,0:46:52.340
You have to really get used[br]to that, because it's like

0:46:52.340,0:46:56.040
electrical signals directly[br][? intruded. ?]

0:46:56.040,0:46:59.750
And the technology of this[br]device, this is usually behind

0:46:59.750,0:47:02.440
the usual hearing aid[br]technology, because it has to

0:47:02.440,0:47:07.360
be well-tested before you put[br]something into your head.

0:47:07.360,0:47:09.850
And like for the other hearing[br]aids, there are not many

0:47:09.850,0:47:13.920
standards, no interoperability[br]between the brands.

0:47:13.920,0:47:19.220
So if you decide to take the[br]cochlear implant of one brand,

0:47:19.220,0:47:22.120
you can never switch[br]to another one.

0:47:22.120,0:47:25.790
So yeah, you have to[br]think that through.

0:47:25.790,0:47:32.700
I have an example[br]of how it sounds

0:47:32.700,0:47:35.640
with a cochlear implant.

0:47:35.640,0:47:37.840
So it sounds kind of spooky.

0:47:37.840,0:47:41.930
It starts with a normal sample[br]that everyone can hear it.

0:47:41.930,0:47:45.380
And then they have different[br]channels and reduce the number

0:47:45.380,0:47:45.810
of channels.

0:47:45.810,0:47:47.530
And then it gets less[br]and less hear-able.

0:47:47.530,0:47:47.840
[AUDIO PLAYBACK]

0:47:47.840,0:47:49.980
-A boy fell from the window.

0:47:49.980,0:47:51.668
A boy fell from the window.

0:47:51.668,0:47:54.090
A boy fell from the window.

0:47:54.090,0:47:55.795
A boy fell from the window.

0:47:55.795,0:47:58.552
A boy fell from the window.

0:47:58.552,0:48:00.544
A boy fell from the window.

0:48:00.544,0:48:01.042
[UNINTELLIGIBLE].

0:48:01.042,0:48:01.540
[END AUDIO PLAYBACK]

0:48:01.540,0:48:04.730
HELGA VELROYEN: Yeah, so you[br]can imagine that it takes a

0:48:04.730,0:48:06.855
while until you actually can[br]understand speech with that.

0:48:09.700,0:48:12.790
Yeah, I'm coming to an end.

0:48:12.790,0:48:18.010
So there are a lot of things[br]that I want from the industry.

0:48:18.010,0:48:20.220
First of all, better service.

0:48:20.220,0:48:22.710
And that goes into the[br]consideration of young

0:48:22.710,0:48:27.490
people's needs, because you can[br]feel it every time, that

0:48:27.490,0:48:29.510
everything is designed[br]for the elderly.

0:48:29.510,0:48:32.250
They don't consider that people[br]actually have to work,

0:48:32.250,0:48:35.480
and have a life, and actually[br]want to go out and talk to

0:48:35.480,0:48:41.100
people, and not only in a[br]silent room one-on-one.

0:48:41.100,0:48:44.820
And there are a lot of things[br]where you miss the support

0:48:44.820,0:48:49.340
when you have hearing aids and[br]are not 60 or something.

0:48:49.340,0:48:53.330
Generally, I'd like to have[br]better signal processing.

0:48:53.330,0:48:58.290
Of course, regarding the size,[br]they already did a really good

0:48:58.290,0:49:01.110
job, if you consider what[br]they do already.

0:49:01.110,0:49:03.980
But actually, the cocktail party[br]problem is not solved.

0:49:03.980,0:49:06.280
So a lot of people who have[br]hearing aids and then who also

0:49:06.280,0:49:09.030
have really good hearing aids,[br]they just avoid social

0:49:09.030,0:49:09.960
situations.

0:49:09.960,0:49:11.140
So they don't go out.

0:49:11.140,0:49:13.180
They don't go to the Congress.

0:49:13.180,0:49:15.160
They don't go to parties.

0:49:15.160,0:49:19.550
So whenever you would ask them[br]to go for dinner, they'd

0:49:19.550,0:49:22.760
really carefully choose the[br]restaurant, if it's a more or

0:49:22.760,0:49:27.040
less crowded one so you can[br]understand people, actually.

0:49:27.040,0:49:29.460
What I'm really missing[br]is some standards.

0:49:29.460,0:49:32.800
It would be even cool if it was[br]open standards, because,

0:49:32.800,0:49:35.915
this way, you feel really[br]trapped as a patient.

0:49:38.790,0:49:43.635
There's this saying, "If you[br]can't open it, you don't own

0:49:43.635,0:49:48.730
it." And I really miss that[br]when having hearing aids.

0:49:48.730,0:49:52.820
So I have a lot of ideas[br]how to improve that.

0:49:52.820,0:49:56.300
But I don't see most of it[br]coming in the next 500 years,

0:49:56.300,0:50:00.280
because the companies are not[br]very open source friendly.

0:50:00.280,0:50:02.550
So what would be really cool,[br]to have something like a

0:50:02.550,0:50:06.960
hearing aid app market that you[br]can download the newest

0:50:06.960,0:50:12.720
feature for a background noise[br]remover or something, and that

0:50:12.720,0:50:17.170
you can write you own filters[br]and share those, and

0:50:17.170,0:50:22.090
especially exchange those[br]between different brands.

0:50:22.090,0:50:24.840
And the funny thing is, when[br]I was browsing through the

0:50:24.840,0:50:28.550
websites of the vendors, the[br]marketing of some hearing aid

0:50:28.550,0:50:31.460
companies actually got[br]this idea already.

0:50:31.460,0:50:34.880
They just call their[br]features apps.

0:50:34.880,0:50:37.460
So this is just an enumeration[br]of what the

0:50:37.460,0:50:38.720
hearing aid can do.

0:50:38.720,0:50:40.458
But they just call it app.

0:50:40.458,0:50:42.330
[LAUGHTER]

0:50:42.330,0:50:44.690
HELGA VELROYEN: And it would[br]be really cool if there was

0:50:44.690,0:50:45.490
something like that.

0:50:45.490,0:50:49.010
You can exchange those.

0:50:49.010,0:50:53.270
What I would also like, hearing[br]aids do not use

0:50:53.270,0:50:54.780
reoccurring situations.

0:50:54.780,0:50:57.810
Most people have a rather[br]steady lifestyle.

0:50:57.810,0:50:59.570
You live in the same[br]apartment.

0:50:59.570,0:51:01.790
Most of the time, you work[br]in the same office.

0:51:01.790,0:51:05.540
You take the same subway[br]to work every day.

0:51:05.540,0:51:10.430
And hearing aids are only tuned[br]for a general situation.

0:51:10.430,0:51:15.620
But I think signal processing[br]works well the more you know

0:51:15.620,0:51:17.030
about the surrounding.

0:51:17.030,0:51:19.770
So it would be really cool if[br]you have one program for the

0:51:19.770,0:51:24.100
office and one program for at[br]home and one for the journey

0:51:24.100,0:51:25.460
to the office.

0:51:25.460,0:51:30.520
And we carry something around[br]that knows all this.

0:51:30.520,0:51:33.720
I mean, we have smartphones, and[br]they have a calendar, and

0:51:33.720,0:51:35.050
it shows where you are.

0:51:35.050,0:51:39.350
And it knows the people that[br]you're talking to and even if

0:51:39.350,0:51:40.990
you talk to them on the phone.

0:51:40.990,0:51:43.950
So there could be parameters[br]for each person

0:51:43.950,0:51:44.880
that you talk to.

0:51:44.880,0:51:47.930
And that could be saved on the[br]smartphone, if there wasn't

0:51:47.930,0:51:51.810
enough space in the memory in[br]the hearing aids themselves.

0:51:51.810,0:51:54.140
So all the information[br]is actually there.

0:51:54.140,0:51:56.040
But I don't see any[br]of the hearing aid

0:51:56.040,0:51:59.390
vendors adapting to that.

0:51:59.390,0:52:03.990
And what I also think, there are[br]people building 3-D models

0:52:03.990,0:52:07.140
of houses, and you could take[br]this information into

0:52:07.140,0:52:08.550
consideration as well.

0:52:08.550,0:52:12.580
So if I have never been to the[br]BCC, but someone has made a

0:52:12.580,0:52:16.440
3-D model, you could also get[br]the characteristics of the

0:52:16.440,0:52:17.540
acoustics here.

0:52:17.540,0:52:20.740
And then, before you go to the[br]Congress, you could download

0:52:20.740,0:52:23.690
the acoustic settings for[br]your hearing aids.

0:52:23.690,0:52:25.020
That would be really cool.

0:52:25.020,0:52:28.240
But, well, this is just ideas.

0:52:28.240,0:52:33.690
Yeah, and then also regarding[br]the hardware, it would be nice

0:52:33.690,0:52:36.210
if there were some[br]open standards.

0:52:36.210,0:52:38.770
It would be cool if you could[br]print your own hearing aids

0:52:38.770,0:52:42.010
with a 3-D printer, at least the[br]part that goes into your

0:52:42.010,0:52:45.110
ear so that it fits[br]really well.

0:52:45.110,0:52:47.290
And there are a lot[br]of possibilities.

0:52:47.290,0:52:49.820
But the market is really,[br]really slow.

0:52:49.820,0:52:52.880
And they still try[br]to figure out how

0:52:52.880,0:52:54.510
do we connect Bluetooth.

0:52:54.510,0:52:59.440
And for me, it's just[br]way too slow.

0:52:59.440,0:53:03.680
I'd like to see more[br]progress in that.

0:53:03.680,0:53:04.030
Yes.

0:53:04.030,0:53:07.490
Actually, with that, I would[br]like to conclude.

0:53:07.490,0:53:11.413
And I have to thank some people[br]who helped me with this

0:53:11.413,0:53:18.210
work, with the talk itself and[br]all the stuff I talked about.

0:53:18.210,0:53:22.490
And yeah, I think we have some[br]minutes for questions.

0:53:22.490,0:53:23.375
FEMALE SPEAKER: Yeah,[br]and that's--

0:53:23.375,0:53:36.125
[APPLAUSE]

0:53:36.125,0:53:37.500
FEMALE SPEAKER: Wonderful.

0:53:37.500,0:53:40.960
So before we come to the[br]questions, I need to say three

0:53:40.960,0:53:45.410
things, which means please stay[br]seated while the question

0:53:45.410,0:53:47.020
and answers are going on.

0:53:47.020,0:53:49.990
Then please pick up your trash[br]and your water bottles and

0:53:49.990,0:53:51.540
take them out with you.

0:53:51.540,0:53:55.215
And then please leave through[br]the front door here, while the

0:53:55.215,0:53:58.290
last door over there is the[br]entrance for the new people.

0:53:58.290,0:54:00.820
So now we can go over to the[br]question and answers.

0:54:00.820,0:54:03.880
We have a signal angel again[br]in the IRC sitting and

0:54:03.880,0:54:05.130
watching Twitter.

0:54:05.130,0:54:07.850
And we have an audio angel in[br]the back running around with a

0:54:07.850,0:54:08.550
microphone.

0:54:08.550,0:54:10.050
So please, questions now.

0:54:18.159,0:54:19.130
AUDIENCE: Hi.

0:54:19.130,0:54:22.680
I have a question about the[br]cochlear implant [? hat. ?]

0:54:22.680,0:54:28.330
I'm a neuropsychologist. And I[br]know that, with eyes, there's

0:54:28.330,0:54:33.450
an approach to re-engineer[br]the signal

0:54:33.450,0:54:35.290
processing of the retina.

0:54:35.290,0:54:40.950
And with this kind of knowledge,[br]you can make better

0:54:40.950,0:54:42.480
retina implant [? hats ?]

0:54:42.480,0:54:47.250
to enhance the possibilities[br]of seeing with

0:54:47.250,0:54:48.610
this kind of aid.

0:54:48.610,0:54:52.850
And is there a similar approach[br]for hearing aids and

0:54:52.850,0:54:54.240
the cochlear implant[br][? hat ?]?

0:54:54.240,0:54:55.560
HELGA VELROYEN: There[br]are different

0:54:55.560,0:54:57.340
types of cochlear implants.

0:54:57.340,0:55:02.500
And this is really not much[br]of my field of expertise.

0:55:02.500,0:55:07.590
I know there are those who still[br]use the membrane in the

0:55:07.590,0:55:13.930
ear or some directly, the ones[br]that I showed, that directly

0:55:13.930,0:55:16.530
insert the wire into[br]the nerve.

0:55:16.530,0:55:19.990
But I am not that familiar with[br]the field, so I don't

0:55:19.990,0:55:21.240
know what's coming there.

0:55:23.570,0:55:27.570
AUDIENCE: OK, there's a question[br]from the front row.

0:55:27.570,0:55:30.490
I wonder, is there any[br]connection of research in

0:55:30.490,0:55:33.600
companies like Siemens or[br]whoever builds these parts?

0:55:33.600,0:55:36.180
Is that, in any way, connected[br]with other

0:55:36.180,0:55:39.170
consumer good research?

0:55:39.170,0:55:40.920
I'm not hearing disabled.

0:55:40.920,0:55:45.660
But if somebody would come up[br]with a decent set of in-ear

0:55:45.660,0:55:49.160
headphones that would fit and[br]would be able to have some

0:55:49.160,0:55:52.410
noise cancellation also, I'd[br]be quite interested in

0:55:52.410,0:55:54.730
spending money on that.

0:55:54.730,0:55:58.315
But this seems to be totally[br]uncorrelated altogether.

0:55:58.315,0:55:59.630
HELGA VELROYEN: Yeah,[br]that's true.

0:55:59.630,0:56:04.900
I mean, it's pretty close[br]to in-ear headphones.

0:56:04.900,0:56:08.970
But it's still, from what I've[br]seen, a different market.

0:56:08.970,0:56:11.710
You still can buy new[br]hearing aids, if you

0:56:11.710,0:56:13.600
don't really use them.

0:56:13.600,0:56:15.630
No one can prevent[br]you from that.

0:56:15.630,0:56:20.710
But I don't see any trend in[br]merging that very much.

0:56:23.770,0:56:24.715
Is there any more questions?

0:56:24.715,0:56:26.060
AUDIENCE: Hello.

0:56:26.060,0:56:30.710
I would like to add two things[br]to your wish list. One thing,

0:56:30.710,0:56:34.220
I would like the hacker[br]community to find out what the

0:56:34.220,0:56:37.260
real differences in devices from[br]different price ranges

0:56:37.260,0:56:40.870
are, what is done in hardware[br]and what is

0:56:40.870,0:56:41.910
actually only firmware.

0:56:41.910,0:56:44.620
My audiologist reported devices[br]coming back from

0:56:44.620,0:56:48.320
repairs or programming to[br]report as more expensive

0:56:48.320,0:56:51.380
devices, for instance, where[br]it's just a firmware thing.

0:56:51.380,0:56:57.390
And the other thing is I see an[br]interesting hacking attack

0:56:57.390,0:57:01.550
angle for hacking in the[br]protocol that the devices use

0:57:01.550,0:57:05.330
to communicate with each other,[br]with the devices which

0:57:05.330,0:57:08.830
transfer Bluetooth into the body[br]area network that really

0:57:08.830,0:57:10.090
speaks to the devices.

0:57:10.090,0:57:13.170
So do you know any off-the-shelf[br]components that

0:57:13.170,0:57:17.142
can speak these body area[br]network protocols?

0:57:17.142,0:57:19.380
HELGA VELROYEN: For the Siemens[br]thing I showed, I

0:57:19.380,0:57:20.230
haven't found anything.

0:57:20.230,0:57:25.270
So this is just that gadget[br]that I have. But I think

0:57:25.270,0:57:28.540
people are experimenting with[br]the audio induction loops and

0:57:28.540,0:57:34.290
the FM systems. But yeah, it's[br]not that much that you can see

0:57:34.290,0:57:35.950
documentation of that.

0:57:35.950,0:57:38.660
So I'd like to see more here.

0:57:38.660,0:57:42.260
I made this talk because I would[br]tell you hackers what's

0:57:42.260,0:57:43.000
on the menu.

0:57:43.000,0:57:45.490
And it would be really nice[br]if there was more

0:57:45.490,0:57:46.620
activity in the scene.

0:57:46.620,0:57:49.980
And it's kind of hard to start[br]that if you're all alone.

0:57:49.980,0:57:55.890
So I hope I've risen some[br]interest. And if you have any

0:57:55.890,0:58:00.360
pointers for me or anything I[br]didn't mention here, I'm also

0:58:00.360,0:58:01.870
happy if you send me an email.

0:58:01.870,0:58:05.186
Or I will be around the[br]next couple of days.

0:58:05.186,0:58:07.430
FEMALE SPEAKER: Before we[br]finish, there are some

0:58:07.430,0:58:10.240
questions from the internet.

0:58:10.240,0:58:13.640
The first one is Gilligan who[br]asked, does the fact that you

0:58:13.640,0:58:16.780
have tinnitus make the process[br]of tuning a hearing aid for

0:58:16.780,0:58:17.780
you harder?

0:58:17.780,0:58:20.700
Does a hearing aid have a[br]negative or positive effect on

0:58:20.700,0:58:22.885
your tinnitus when[br]you wear it?

0:58:22.885,0:58:23.890
HELGA VELROYEN: Of my what?

0:58:23.890,0:58:25.690
Sorry.

0:58:25.690,0:58:26.780
A positive or negative--

0:58:26.780,0:58:28.815
FEMALE SPEAKER: Does the hearing[br]aid have a negative or

0:58:28.815,0:58:33.150
a positive effect on your[br]tinnitus when you wear it?

0:58:33.150,0:58:35.510
HELGA VELROYEN: Actually,[br]yeah.

0:58:35.510,0:58:39.230
There are hearing aids that[br]claim to [? counter-fy ?]

0:58:39.230,0:58:42.130
the tinnitus, that[br]they generate the

0:58:42.130,0:58:44.440
counter-frequencies[br]against that.

0:58:44.440,0:58:49.560
I also asked my acousticians or[br]my audiologists about that.

0:58:49.560,0:58:50.750
They actually don't work.

0:58:50.750,0:58:53.230
I mean, they usually are[br]not really offered.

0:58:53.230,0:58:56.020
And even if you ask for it, they[br]tell you, yeah, usually

0:58:56.020,0:58:58.770
they don't really make[br]a difference.

0:58:58.770,0:59:01.880
My choice was actually just to[br]make a normal hearing aid.

0:59:01.880,0:59:06.300
And because the frequencies that[br]you don't hear very well

0:59:06.300,0:59:11.590
are amplified, the tinnitus,[br]in relation, gets as loud.

0:59:11.590,0:59:15.510
And you learn to ignore it.

0:59:15.510,0:59:18.080
That's not really magic, but[br]that's the only thing that

0:59:18.080,0:59:19.756
seems to work.

0:59:19.756,0:59:24.870
FEMALE SPEAKER: OK, the next one[br]is from [? Schpegmada ?]?

0:59:24.870,0:59:28.100
Are there hearing aids with[br]documented RP for the tuning

0:59:28.100,0:59:29.790
official doc reverse[br]engineered?

0:59:29.790,0:59:32.450
Can you get the official[br]tuning software to tune

0:59:32.450,0:59:36.240
yourself that you write[br]through any people who

0:59:36.240,0:59:37.200
manufacture them?

0:59:37.200,0:59:39.830
HELGA VELROYEN: I haven't found[br]anything about that.

0:59:39.830,0:59:43.430
The only thing you can do is get[br]the software on the black

0:59:43.430,0:59:46.160
market and disassemble[br]it, if possible.

0:59:46.160,0:59:48.630
But there are no API[br]system thing.

0:59:48.630,0:59:51.705
That's really what I'm[br]missing as well.

0:59:51.705,0:59:55.770
FEMALE SPEAKER: OK, the last[br]one is from Lucy, I guess.

0:59:55.770,0:59:58.810
What frequency range can[br]you usually hear

0:59:58.810,1:00:00.560
using a cochlear implant?

1:00:00.560,1:00:03.440
How fine can you resolve[br]frequencies?

1:00:03.440,1:00:06.990
HELGA VELROYEN: They are about[br]the range that you have in the

1:00:06.990,1:00:07.580
speech banana.

1:00:07.580,1:00:09.820
But actually, I'm not[br]that familiar

1:00:09.820,1:00:10.890
with cochlear implants.

1:00:10.890,1:00:15.195
So I can not give the[br]details here.

1:00:15.195,1:00:16.910
AUDIENCE: OK, hi.

1:00:16.910,1:00:22.330
You showed us the programming[br]tool for changing the settings

1:00:22.330,1:00:23.890
on the hearing aid.

1:00:23.890,1:00:27.365
I've forgotten the name,[br]but the 1980's modem.

1:00:27.365,1:00:30.210
Have there been any attempts[br]to clone that hardware or

1:00:30.210,1:00:33.530
reverse-engineer the spec?

1:00:33.530,1:00:36.010
HELGA VELROYEN: I'm not sure[br]if I understood correctly.

1:00:36.010,1:00:39.280
You mean like how you get[br]the hardware for this?

1:00:39.280,1:00:41.920
AUDIENCE: Yeah, so the moment[br]you have to buy the

1:00:41.920,1:00:43.740
programming tool,[br]as I understand,

1:00:43.740,1:00:44.760
to update the software.

1:00:44.760,1:00:48.730
HELGA VELROYEN: Since it's only[br]delivered and sold to

1:00:48.730,1:00:52.860
audiologists and doctors, you[br]have to be friends with

1:00:52.860,1:00:54.670
audiologists or doctors.

1:00:54.670,1:00:57.250
And somehow they can[br]give you that.

1:00:57.250,1:00:59.190
But there's no official[br]channel.

1:00:59.190,1:01:03.560
So you have to find[br]the black market.

1:01:03.560,1:01:06.970
AUDIENCE: Yeah, sorry, my[br]question was has anyone tried

1:01:06.970,1:01:10.937
to create an open hardware[br]variant?

1:01:10.937,1:01:11.152
HELGA VELROYEN: No.

1:01:11.152,1:01:12.365
AUDIENCE: Has anyone copied?

1:01:12.365,1:01:13.760
HELGA VELROYEN: No,[br]I haven't found

1:01:13.760,1:01:14.780
anything like that either.

1:01:14.780,1:01:19.210
So that was actually also on my[br]wish list, that makes the

1:01:19.210,1:01:21.060
tuning hardware open[br]source as well.

1:01:21.060,1:01:22.290
That would be really nice.

1:01:22.290,1:01:25.880
But it's a serial device, so[br]you could actually do some

1:01:25.880,1:01:26.830
sniffing there.

1:01:26.830,1:01:29.410
But I haven't seen any[br]approaches there.

1:01:29.410,1:01:30.610
AUDIENCE: Yeah, because[br]it looks pretty

1:01:30.610,1:01:31.530
easy, I would say.

1:01:31.530,1:01:32.016
[LAUGHTER]

1:01:32.016,1:01:33.960
AUDIENCE: Yeah.

1:01:33.960,1:01:36.580
FEMALE SPEAKER: So we are[br]running out of time.

1:01:36.580,1:01:40.590
Is there any really, really[br]important question left?

1:01:40.590,1:01:45.180
I mean, because she's around,[br]so you can meet her in the

1:01:45.180,1:01:48.550
next days on the Congress[br]again and

1:01:48.550,1:01:49.610
ask her more questions.

1:01:49.610,1:01:51.480
And thank you very much[br]for the talk.

1:01:51.480,1:01:52.470
It was very interesting.

1:01:52.470,1:02:13.680
[APPLAUSE]