Naked Science Forum
General Science => General Science => Topic started by: neilep on 07/10/2012 16:09:46
-
Dearest Batologists,
Meet Rover the Bat-dog !
[ Invalid Attachment ]
Genuine Non doctored bona fide
picture of a real dog-bat just moments ago !
He's on the dog-and-bone !
Can Rover the Bat use his echo detection super powers to determine if a moth is flittering around "on the other side" ?* (in a haunted ghouly echoy voice with vincent-priceish laughing at the end)
This is important to know because wifey is scared of moths and she wants to get her own bat-dog to determine if it's safe to conduct telephone verbosity in the knowledge that no moths are going to come near the receiver "on the other side "*
whajafink ?
ta
Neil
-
A typical telephone only carries sound frequencies up to about 4kHz, VOIP (https://en.wikipedia.org/wiki/VOIP) is better at about 8kHz ...
traditional, voiceband or narrowband telephone calls limit audio frequencies to the range of 300 Hz to 3.4 kHz. Wideband audio relaxes the bandwidth limitation and transmits in the audio frequency range of 50 Hz to 7 kHz or higher.
https://en.wikipedia.org/wiki/Wideband_audio
Telephone bandwidth is not adequate to transmit the sound of bats ...
Bats emit calls from about 12 kHz to 160 kHz
https://en.wikipedia.org/wiki/Bat_detector
-
A typical telephone only carries sound frequencies up to about 4kHz, VOIP (https://en.wikipedia.org/wiki/VOIP) is better at about 8kHz ...
traditional, voiceband or narrowband telephone calls limit audio frequencies to the range of 300 Hz to 3.4 kHz. Wideband audio relaxes the bandwidth limitation and transmits in the audio frequency range of 50 Hz to 7 kHz or higher.
https://en.wikipedia.org/wiki/Wideband_audio
Telephone bandwidth is not adequate to transmit the sound of bats ...
Bats emit calls from about 12 kHz to 160 kHz
https://en.wikipedia.org/wiki/Bat_detector
Poo !!
Thanks RD. Actually that's good news !.... [:D]..I did not like the idea of bat-dog answering my phone all the time !....
Does that mean though, if the bandwidth allowed.......... that moths could then be detected ?
-
Often, when on skype, you can hear your voice picked up on the other end's microphone and sent back to you, due to transmission delays. Probably no reason why suitable software couldn't detect echoes of your voice (or a suitable tone) sent back to you so you could see an echolocated model of the area around the other end. Just another thing to worry about, voyeurs pinging your answering machine. (I know, doesn't work until you pick up, so never answer the phone.)
-
The bat would need stereo too.
-
Echo location is possible using audible sound ... https://en.wikipedia.org/wiki/Human_echolocation
However the resolution is going to be much poorer than with ultrasound, it's proportional to the wavelength of the sound
The bat would need stereo too.
Stereo hearing not necessary if the task is just to judge how close something is: just to tell if something has crossed a beam.
Like an ultrasonic "tape measure" (one orifice) ...
[ Invalid Attachment ]
http://www.amazon.co.uk/Ultrasonic-Measure-Distance-Meter-Pointer/dp/B004BEX24M
-
I'd imagine that a bat trying to locate a fly for dinner is concerned with the location in space rather than just the distance.
That's tricky with just one "ear". You can do it slowly if the ear is readily steerable but I don't think that would be effective.
(Though since bats beat us to the invention of sonar and chirping, they might conceivably use synthetic aperture tricks too.)
-
You could also use passive detection - a computer with a microphone could pick up the sounds of a clock, or a fluttering butterfly.
With two microphones slightly separated, you could even get an idea of the direction and distance of a sound source.
You have even more flexibility if you have stereo access to a pair of speakers and a pair of microphones.
A couple of challenges:
- If the microphones and speakers are all in a horizontal line, you will not have any vertical discrimination.
- Most VoIP systems chop the sound up into chunks of 5ms or 10ms, and replays the "average" of the sound over this period. 10ms packetisation would limit the distance accuracy to about 3 meters.
- By selecting a "G.711 codec", you could theoretically control the sound with a resolution down to 0.125ms, giving a distance accuracy down to about 4cm.
- Most computers have a CD-quality sound port with a sampling rate of 44kHz, able to generate sounds over 20kHz. This will give much more precise distance estimation.
- While not competing with bats, 20kHz is above the frequencies that most adults can hear - but it might annoy the dog. It will use about 1 Mbps of data to transmit and receive these signals, which is more bandwidth than some people have available.
- Sound levels drop rapidly with distance - an inverse square law for passive monitoring, and inverse 4th law for echolocation.
- Most VoIP systems encode only the 1 or 2 loudest sounds, on the assumption that these will drown out any quieter sounds, so transmitting anything else is a waste of bandwidth. This means that something distant will be drowned out by echoes or sounds from closer objects.
- G.711 can handle a dynamic range of about 4000:1 from the loudest to the softest sound. Most computers have a 16-bit sound system, which can handle a ratio of about 60000:1, which gives much greater range.
- VoIP systems include echo cancelers, which are designed to delete precisely the type of signal that you are looking for. You can disable echo cancellation by including the right signals when you set up a call.
- If you are willing to wait longer, you can generate a pseudo-random sound (it sounds like hiss), and that will slowly build up a picture of the surrounding area. However, it won't help with a fast-moving target like a moth.
- Some people put their computer near the microphone, and the fan noise will drown out all quiet sounds.
Or, your wife could use the camera on the computer and see if there were any moths at the other end, and decide if these were the sort of people she wanted to talk to...
-
Another challenge if you are operating over the internet: Internet traffic can cause the sounds to be randomly delayed by 20ms or more. This could introduce random distance errors of 6 meters.
Internet protocols for carrying sound and video partially overcome these random delays by the sender attaching a timestamp to each packet of sound. The receiver can then use a local timer to delay all of the packets by the right amount so that the sounds play out with only infrequent changes in delay.
These large, frequent, random delays do not happen over the Plain Old Telephone Service (POTS), which in other countries is called the Public Switched Telephone Network (PSTN).