Finally I have a good idea for the headband. It’s similar to the idea’s in the previous post but a little bit simpler. Instead of using multiple elements I want to use one single core that which is made of a bendable material, much like a normal headphone’s headband. It’s tension is used to clamp to the head. With a cable approximately a centimeter higher as the core on the top of the headband. When pulling the cable, thus shortening it, the headband will straighten. The tension of the core also pulls the headband back into its original position after releasing the cable.
Doing a little math I found out that, if the head is 14 centimeters wide and we widen the headband by 30 degrees on a distance of 1 centimeter, the cable only shortens half a centimeter. I expected this would take quite a lot of force to pull. To find out I made a prototype with polystyrene and wooden blocks as spacers.
As expected I didn’t need to pull far but it turns out it doesn’t take as much force as I’d thought, which is a good thing. So I think I’ve found a good method that will work for my purpose, the next thing to do is make a nice design.
I’m still trying to figure out a way to make the headband expand when the user wants to put it on. Instead of continuing with the springs idea, it isn’t nifty enough for my taste, I tried to come up with other ways to accomplish the same. One of them is shown in the images below, and is inspired by snakes and an elephants trunk. Who bend they’re body by expanding and contracting muscles in line with the spine. That way the body bends. What I’ve done is make six turnable elements. I connected rubber bands to one side of the elements to make it constantly bend. If you we’re to put a string through the other end of each element and pull it, the elements would turn the other way, thus straighten.
When I build this I noticed that it is crucial that all rubber bands are equal in strength, otherwise some elements bend more easily as others. Therefor I decided to try to control the movement of every element very precisely. What resulted is demonstrated in the flash application below.
This method of bending isn’t perfect either, when increasing the amount of elements (to reduce the height of the headband) the rotation of each element has to decrease, which results in tiny movement, which in turn needs tiny gears, which is weak and might break easier.
One of the user goals I defined before was that it should be possible to use the device with only one hand. Putting headphones on your head is a task that rarely succeeds with only one hand. To improve this I’ve made a few lego models of a way to widen the headphones so your head will fit in.
The first model is inspired by those toys with a grabbing hand at one end of a stick and a gripping mechanism at the other end.
Although this mechanism worked, it’s a huge difference between collapsed and expanded state, and there isn’t much control. If you want to expand it it goes like a spring and bangs your head at the other end. Also it takes quite some force to do it.
The next prototype I made is based on springs and three solid parts as a headband. Instead of the headband being bendable it now is not flexible and can only change its shape by pressing the springs.
As you can see in these images the headband gets wider, this enables the user to put on the headphones with one hand. The springs are being contracted by a wire that runs from one end to the other end of the headband. The mechanism used to pull the string is shown below.
By pushing the handles the anchor point to which the wire is attached moves down, thereby pulling the wire. This contracts the springs and thus making the headband wider. This model worked quite well but was way to heavy. So I made a new lighter version.
As you can see in the movie, when I press the handless the springs contract. The next step was to improve the handless so that a horizontal slide is possible instead of a weird contraption that also moves the headphone upwards while pushing the handless.
In this latest version only the pulling mechanism is showed. The small grey parts that moves up and down is the anchor point for the wire. Which in the final version will be a bike brake cable, this enables me to adjust the ear pads in height without interfering with the length of the wire. With my bottom hand I’m holding the part that is going to be connected directly to the headband. My other hand moves the handle the user will press. This is also the way I want to build it in the final headphone model. The brake parts (with the springs) is something I still need to work out. But in short, this is the plan and my process of coming up with a non-electronic solution to a user problem.
For the next headset we will build I’ve made a schematic plan. In the image below you can see the different electronic parts and they’re position. On the bottom side the laptop layout is presented. The blue lines and boxes are the audio, it’s color-coded, otherwise it would be total chaos…
To get the system to work in the right way, we have to know what the natural body movement is from an adult while walking. We tested in a gym on a treadmill. We measured the link between the speed someone is walking or running and the movement from that someone his head.
Sadly, this during this test we only used ourselves to test upon. But on the other hand it delivered good information. As you can see on the graph below, the movement between Martijn and Peter differ from each other.
Peter has a bit more movement during his test. We think this is because of his lack of physical condition. Martijn as a climber has a much better condition. Conclusion is that he is more in ease when doing little exercises.
Conclusion
In the next graph you see what we think is as a result from the test. While walking under 6km/u the natural body movement can be measured with the formula: 2x + 0,5. The tipping point is at a little bit behind 6km/u. and is than linear whit the formula: 0,5x + 10.
Future work
Sadly, this information itself is not useful for us, or for the system. To get this information meaningful, we also need to know what people recognize as ‘the natural loudness’. This is because only that we can link those two variables together.
Discussion
For discussion is open the length of someone. Peter is 10cm. longer than Martijn. The values from Peter could be higher because the length of the movement are longer.
Today, on my way home I tested the automatic volume adjustment program we’ve written. What it does is adjust the volume of the music to the environmental sounds. If the user is in a loud environment the music will also be louder, if the environment is quiet the music volume will be turned down. The volume is based on the average sound level of the past 10 seconds.
In our test I put my laptop in my bag, running a Flash application that reads the sound levels of the internal microphone (of my laptop) and plays music files with adjusted volume. I used headphones to listen to the music. I also used my wireless mouse to enable the skipping of songs.
I started walking in school to the exit. Then I got on my bike and rode to the center of Hilversum, there I walked in a few shops. The effect of the volume adjustment was barely noticeable. Only when a few cars drove by I could really hear the sound getting louder. But that’s because the volume adjustment has a small delay. The adjustment occurs 10 seconds later as the car drove by, so when the cars are gone and it is quiet again I heard the volume go louder.
Conclusion
It works quite well, but the average sound level should be based on a longer period than 10 seconds. This way a car driving by will have less influence on the volume. Also I used the internal microphone of my laptop, which was hidden away inside my bag. Because of this the sound levels measured were a bit low. It would be better to have the microphone directly in the outside world. In the headset would be the perfect place. A second test is required to get a better way of adjusting volume. Below is a graph of the volume I heard while biking and walking around. 0 is no volume and 1 is maximum volume, to avoid not hearing music at all I set a minimum of 0,15 volume.
