Re: Audio Jack Sampling Rate?
Hi Charles, 2008/9/21 Charles Pax [EMAIL PROTECTED]: I'm considering the plausibility of a little hardware project that would turn a Freerunner or any computer with an adequate sound card into a mimio [1] device. [...] Any input from the community? Assuming I've understood this even half-way correctly... It sounds like the accuracy of your triangulation will be limited by the small size of the FR; and when you add in the hardware and software path latency, I wonder if this will work accurately enough. (I'm assuming the marker would be a few metres away from the FR, is that right?) I think that points in a similar direction to Al's comment, i.e. moving some of the solution outside the FR, over USB. That would allow you to have more distance between the inputs, and so get more accuracy. Regards, Neil ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
Here's another approach: Why not use two freerunner on different locations which do have one ir input attached. They synced the clock with ntp and both of them save the time difference between mic and ir data. You know the exact location of both and you could sync the data over wlan/gprs and calculate the resulting position. Just an idea i have when reading this thread. :) Ciao, Rainer Neil Jerram wrote: Hi Charles, 2008/9/21 Charles Pax [EMAIL PROTECTED]: I'm considering the plausibility of a little hardware project that would turn a Freerunner or any computer with an adequate sound card into a mimio [1] device. [...] Any input from the community? Assuming I've understood this even half-way correctly... It sounds like the accuracy of your triangulation will be limited by the small size of the FR; and when you add in the hardware and software path latency, I wonder if this will work accurately enough. (I'm assuming the marker would be a few metres away from the FR, is that right?) I think that points in a similar direction to Al's comment, i.e. moving some of the solution outside the FR, over USB. That would allow you to have more distance between the inputs, and so get more accuracy. Regards, Neil ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
On Sat, Sep 20, 2008 at 8:42 PM, Al Johnson [EMAIL PROTECTED]wrote: You probably want to look at these two. The main limitation is that you've only got one input channel via the jack, but you might be able to get round that using the mic switch detection. I incorrectly assumed there is stereo input on the jack. I'll have to go back and rethink this whole thing. What is the mic detection switch? Can I record audio from the mic input and the built-in microphone at the same time? Since you need external bits for the IR anyway you may be better off making a USB device instead. The arduino and other similar devices make the USB part relatively easy. Put a narrow bandpass filter on your audio transducer and you make the audio signal a simple on/off - use multiple frequencies if you need to track different markers or indicate different states. I'm hoping to find some novel way of doing this with the most simple hardware setup possible. Maybe we can have each transducer connected to a bandpass filter that activates an oscillator. If the two oscillators operate at different frequencies on the same mic-in channel, we should be able to process this in software on the Freerunner to calculate when each oscillator was triggered. This would allow us to use only dumb and cheap hardware that should work on any computer with a mic-in. The IR detector(s) can also be connected to a oscillator operating at a third frequency. What do you mean by external bits for the IR? Data bits or just little pieces of hardware? I figure a photodiode can be connected between signal and ground. When suficient IR light hit it the diode should make a spike on the mic-in channel. -Charles Pax ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
2008/9/21 Charles Pax [EMAIL PROTECTED]: All the Freerunner would be needed for is processing the input of the mic-in. The freerunner microphone isn't used. Imagine a stick with two microphones glued 400 cm apart and connected to a audio cable. There is more to it, but that's pretty much what it would look like. OK, I see now, that makes sense. I had thought you were talking about two microphones in the FR itself. Best wishes, Neil ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
why not try this technique through the bluetooth ? http://www.youtube.com/watch?v=QgKCrGvShZs that is, using a wiimote regards NiL ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
On Sunday 21 September 2008, Charles Pax wrote: On Sat, Sep 20, 2008 at 8:42 PM, Al Johnson [EMAIL PROTECTED]wrote: You probably want to look at these two. The main limitation is that you've only got one input channel via the jack, but you might be able to get round that using the mic switch detection. I incorrectly assumed there is stereo input on the jack. I'll have to go back and rethink this whole thing. What is the mic detection switch? There's a switch on the headset that shorts the mic when pressed. I've not looked into how this appears on the freerunner. Can I record audio from the mic input and the built-in microphone at the same time? Yes. Route headset (mic1) to the left ADC and handset (mic2) to the right ADC, then set the record mode to stereo. Since you need external bits for the IR anyway you may be better off making a USB device instead. The arduino and other similar devices make the USB part relatively easy. Put a narrow bandpass filter on your audio transducer and you make the audio signal a simple on/off - use multiple frequencies if you need to track different markers or indicate different states. I'm hoping to find some novel way of doing this with the most simple hardware setup possible. Maybe we can have each transducer connected to a bandpass filter that activates an oscillator. If the two oscillators operate at different frequencies on the same mic-in channel, we should be able to process this in software on the Freerunner to calculate when each oscillator was triggered. This would allow us to use only dumb and cheap hardware that should work on any computer with a mic-in. The IR detector(s) can also be connected to a oscillator operating at a third frequency. Sounds more complicated than making a usb device to me :-) What do you mean by external bits for the IR? Data bits or just little pieces of hardware? I figure a photodiode can be connected between signal and ground. When suficient IR light hit it the diode should make a spike on the mic-in channel. Bits as in components. I've not tried a photodiode on a mic input, but it sounds like the sort of thing someone might have done for lirc hardware. -Charles Pax ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
On Sun, Sep 21, 2008 at 7:46 PM, Al Johnson [EMAIL PROTECTED]wrote: Can I record audio from the mic input and the built-in microphone at the same time? Yes. Route headset (mic1) to the left ADC and handset (mic2) to the right ADC, then set the record mode to stereo. This is good news. Since you need external bits for the IR anyway you may be better off making a USB device instead. The arduino and other similar devices make the USB part relatively easy. Put a narrow bandpass filter on your audio transducer and you make the audio signal a simple on/off - use multiple frequencies if you need to track different markers or indicate different states. I'm hoping to find some novel way of doing this with the most simple hardware setup possible. Maybe we can have each transducer connected to a bandpass filter that activates an oscillator. If the two oscillators operate at different frequencies on the same mic-in channel, we should be able to process this in software on the Freerunner to calculate when each oscillator was triggered. This would allow us to use only dumb and cheap hardware that should work on any computer with a mic-in. The IR detector(s) can also be connected to a oscillator operating at a third frequency. Sounds more complicated than making a usb device to me :-) It probably is, but would make for an inexpensive piece of hardware. However, now that you've told me we can record from both the handset and headset the only necessary external hardware should be some wire with a 2.4 mm plug, a transducer, and a photodiode. Now we're pretty much back to my original idea, but with less necessary hardware. Thanks. What do you mean by external bits for the IR? Data bits or just little pieces of hardware? I figure a photodiode can be connected between signal and ground. When suficient IR light hit it the diode should make a spike on the mic-in channel. Bits as in components. I've not tried a photodiode on a mic input, but it sounds like the sort of thing someone might have done for lirc hardware. One time I decoded the signal from an IR personal response system [1] by pointing the transmitter at a photodiode connected to the mic-in port of my laptop, so I know that would work. -Charles Pax [1] http://en.wikipedia.org/wiki/Audience_response ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
On Sun, Sep 21, 2008 at 4:01 PM, Nicolas Laurance [EMAIL PROTECTED]wrote: why not try this technique through the bluetooth ? http://www.youtube.com/watch?v=QgKCrGvShZs that is, using a wiimote I think I will investigate this option. I've thought about it before, but couldn't find any Linux software. Now, however, I see there it can be done in Linux and there's even a nice GUI [1]. On Johnny Lee's site [2] I've found a really cool physics simulation application [3] that I'm really excited to try with my students. Thanks for the info. -Charles Pax [1] http://www.stepd.ca/gtkwhiteboard/ [2] http://www.cs.cmu.edu/~johnny/projects/wii/ [3] http://www.phunland.com/wiki/Home ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Audio Jack Sampling Rate?
What is the highest sampling rate I can get from the audio jack? What is the frequency range? I'm considering the plausibility of a little hardware project that would turn a Freerunner or any computer with an adequate sound card into a mimio [1] device. The principle is an expansion on the way you calculate the distance (d) of a lightning strike from you.You see the light instantly, but hear the sound with a delay (delta t) because sound travels at a much slower speed (v_s) of approximately 340 m/s. You can use the following equation to find the distance from you to the lightning strike d = (delta t)(v_sound) With this information you can draw a circle of radius d with you at the center. Observing the same lightning from two known points would allow us to draw two circles whose intersection would indicate where the the lightning hit the ground. This gives two points, but knowing the general direction of the strike would give us a single point. This is basically what I want to do with the stereo input of the Freerunner. However, instead of observing lightning and audible sound, I want to observe infrared light and sound at a higher frequency than humans can hear. The mimio device works in this way. A dry erase marker is placed into a case what emits IR light and sound when the marker is pressed down. I basically want to connect both an appropriate microphone and an IR detector on each of the two input channels of the Freerunner. There would be one mic-IR pair at each of points A and B. Points A and B would be seperated by some distance h along the vertical axis. Below is a simple hardware schematic diagram. /--micA || | GND | I--IR_detA | | |GND \--(to left channel) /--(to right channel) |--micB || | GND | \--IR_detB | GND Software would then monitor the two channels and compare the input. The software would know that a set of square waves from both the right and left channel occuring at the same time indicate a flash of IR light. It would then begin measuring on each channel the time interval to the middle of the next set of waves. Sample input (left channel) __|-|_|-|_|-|_|-|_|-|___|-|_|-|_|-|__ |--from IR_detA-| |--micA-| |---delta tA---| |--from IR_detB-| |--micB-| |delta tB--| (right channel) __|-|_|-|_|-|_|-|_|-|___|-|_|-|_|-|__ The following two equations would give the position of the dry erase marker relative to point A. (I may have mixed up my positive and negative directions somewhere, but you get the picture.) d_Ax = (v_sound)sqrt(t_A^2 - t_B^2) d_Ay = h - (v_s^2)(t_B^2 - t_A^2)/(2h) My first step in testing my idea is to connect ir detectors and mics to a headphone cable, plug it into my computer, and see if Audacity displays something similar to the sample input above. Then I'll hack around with that until I can make a reliable piece of hardware. I'm more of a hardware guy than a programmer, so writing this software would be very challanging to me. Is there anyone out there willing to hack some code if I were to supply you with the appropriate hardware? Any input from the community? -Charles Edward Pax [1] http://www.mimio.com/products/interactive/index.php ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Audio Jack Sampling Rate?
The following two equations would give the position of the dry erase marker relative to point A. (I may have mixed up my positive and negative directions somewhere, but you get the picture.) d_Ax = (v_sound)sqrt(t_A^2 - t_B^2) d_Ay = h - (v_s^2)(t_B^2 - t_A^2)/(2h) I just came up with this next idea as I was turning on the shower water. The x and y position can then be translated into screen coordinates (after calibration). This can then be passed through ReMoko to some other device. Combine this with a projector and you have yourself a nice smart board. -Charles Pax P.S. I'm thinking about all this because the high school where I teach physics gave me a $800 mimio device to aid a visually impaired student of mine. The problem is that the mimio device only works with Windows and OS X and I don't want to be limited in my operating system choices. i also think it would be cool to beat the system and do the same thing with $40 off-the-self components. ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
