Subject:
|
Re: NXT Sound Sensor
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 5 May 2006 15:02:25 GMT
|
Viewed:
|
2532 times
|
| |
| |
In lugnet.robotics, Tim Byrne wrote:
> In addition to distinguishing decibel levels...
Which it does pretty well, maxing out around 100 dB.
> they mentioned it featured "tone recognition"... What have
> the privileged been able to accomplish with the sound
> sensor?
Not as much (yet!) as I'd like. The sound sensor has two modes familier to
folks who work with dB noise levels. One is a "flat" frequency response curve,
and one is a frequency response curve that more closely mimics how a human hears
sounds. The result is some difference in response between those two modes in
certain frequency ranges (the difference seems greatest below 800 Hz). There are
two big problems with this method of detecting "tone" information. First, the
sensor takes a significant amount of time to switch between modes. And second,
while I can get reasonably good data using pure tones (I used a signal generator
from work to get response curves), most sound sources are, well... noisy. For
instance, have an RCX play a continuously rising series of tones at what is a
suppossedly constant volume, and you'll easily hear a lot of volume variation:
the NXT has to deal with this as well, and resonances within the casework or
structure you build, etc. And human voices are even tougher to deal with.
My goal is to be able to sing a series of high/low tones, and have the NXT
recognize the sequence. So far I need to working on my singing voice (not to
mention become a deep base... and no, before somebody asks, I've *already* had
my voice change, I can't depend on that :).
Using the NXT as a tuning fork is right out, as far as I can see. The
electronics might be up for it, but there are too many other complicating
issues.
--
Brian Davis
|
|
Message has 1 Reply: | | Re: NXT Sound Sensor
|
| (...) Hmmm is there a "raw data" mode for the sound sensor? Seems to me, it may be possible (maybe not yet but one day) to write a program that does FFT analysis, as long as it can get the raw wave data from the sensor. I'm assuming the "tone mode" (...) (19 years ago, 5-May-06, to lugnet.robotics)
|
Message is in Reply To:
| | NXT Sound Sensor
|
| The LEGO Education blog had some information a while ago on the NXT sound sensor. In addition to distinguishing decibel levels, they mentioned it featured "tone recognition". Can someone in the know comment on this feature. Can the sound sensor (...) (19 years ago, 5-May-06, to lugnet.robotics)
|
6 Messages in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
Active threads in Robotics
|
|
|
|