Vibrations & AWSoM

I thought I should write up some notes about the infrasonic/low-frequency stuff I’ve been researching; my initial aim was to investigate ways of recording and transmitting/playing back very low frequency sounds - including infrasonics - on site at Two Tree Island, to feed into AWSoM.

There are two issues with the infrasound aspect of low frequency sound a) recording it and b) generating it. As Chris Watson warned us, the problem with most domestic/easily commercial available & affordable recording equipment (and I’ve found, also playback/speaker equipment) is that it is just not designed for anything beyond the 20hz-20khz range (approximately human hearing range).

Starting with recording infrasound – as most mics don’t go outside the 20-20, I’ve obtained three 3 Mark Products 4.5 HZ horizontal geophones (thank you e-bay) for about £15.

‘The sensor is a L28LBHH with a 385 ohm impedance. There is at least 2' of leads attached to one end. This geophone can be used to build your own earthquake monitor. It can also be used to sense ground vibration to build an intruder alarm or vehicle detection alarm which senses ground vibration. Excellent to use as an educational demo on ground vibration/earthquakes using a computer with an oscillosope card. You can feed the signal into a computer sound card.’

‘Geophones, with periods around 1 Hz, are short period devices designed to record local or large regional events. The higher frequency sensors (4.5 Hz and above) are mainly used for oil and geological studies. Geophones with periods around 4.5 Hz will record local events, but they are not designed for this purpose. Short period sensors will not record distant (teleseismic) events very well, if at all.’

So these geophones would pick up an earthquake in Leigh, but not necessarily one in Kent. They can be connected to a computer to show a visual wave signal or a sound signal. I have no idea how to connect them up though. Best information I’ve found so far is the ‘Table Top Earthquake’ stuff (http://pubs.usgs.gov/of/1998/ofr-98-0767/), which states that you need to connect the geophone to an ‘Analog to digital converter (AD) and display software’. The file is dated 1999 so a lot of the equipment references are out of date. Dataq Instruments are still in business (www.dataq.com ) and still sell analogue to digital converters, but not the model number stated. Might be something better somewhere else anyway that’s more Linux compatible.

There is a lot of other ‘build-your-own’ stuff around seismometers, but these seem to be pretty much motion sensors which can be bought for the 1-wire weather system anyway, so I’ve ignored that.

So – it might be possible to record data using the geophones, but we’d need to test the principle in the shed before moving out to the site and seeing if there is any ‘sound’ there.

There are other possibilities for collecting sounds using the NXT & MD recording set-up Stuart and I have put together (which allows me to feel sounds through the NXT exciters as they are being recorded). Problem with mics is the 20-20 issue again (plus MD compatibility). The Sony digital mics seem to mostly have a 100-15000hz range with only the ECMS957 going beyond that at 50-18,000hz. I haven’t been able to find a lower frequency mic suitable for working with the MD.

In terms of playing back any data recorded, we are back to the old 20-20 problem. Other people have investigated this too so I know just how difficult this is.

‘The secret of making good infrasound is a simple matter of scaling up; double the length of a music pipe and you halve its resonant frequency, making it sing an octave lower…Our infrasonic pipe…is based on a plastic sewage pipe 7m long. We use a loudspeaker…to vibrate a column of air around 4.7m long with a resonant frequency of 18hz…our loudspeaker needs to move a huge volume of air…a domestic speaker does not have enough ‘stroke’…so we are using a servo driven speaker instead. This uses a motor to drive the speaker back and forth’. (Soundless Music: Sarah Angliss)

For AWSoM this creates issues of power supply to drive the sound and portability (don’t fancy moving 7m pipes around, even if they are plastic). Additionally, although you won’t be able to hear it, there could be problems with resonant frequencies in other objects, including AWSoM itself ‘..the pipe began to resonate strip lights, furniture and other loose odds and ends’.

I have been playing about with Audacity and my Logitec NXT speakers (battery powered, uses NXT Surface Sound technology). Logitec don’t give information on the frequency range it is supposed to generate, but I could detect frequencies – or something - as low as 10hz – this could potentially be a side effect of something else in the speaker vibrating at a sympathetic (but higher) frequency, so would be good if someone could listen to it (if you can’t hear anything but I can feel it, it could be infrasound, but extreme bass needs more power to cross the threshold of human hearing, so it could be about ‘volume’). That said, the best vibrations are coming a bit higher up, 20-100hz. Some sounds give a better feel at higher frequencies than this – it depends on the nature of the sound - but these ones go right up your arm. I don’t know what we would be able to get with the Sonic Impact pads (stick-on NXT exciters designed to turn flat surfaces into NXT speakers) and the T-amp (battery powered portable amp with a dynamic range of 102dB). We might be able to build a panel which will generate more lower frequency sounds, but the exciters and amp are not that powerful, might be an issue in the open air:

‘With flat-panel speakers, the lowest frequency is dictated by the size of the panel. The bigger the panel the deeper the sound it makes…The lighter the material the louder the resulting sound, whereas the softer the material the softer the outcome of bass.’

NXT speaker manufacturers are using materials such as polyurethane rigid foam, which is not the most environmentally friendly stuff you can get, but it is also manufactured for foam-core insulation panels, so if we find this works, it could be reused somewhere to keep someone’s heating bills down! I think it needs to be pretty thin too.

So this is where I’m up to:

- I might be able to record data using the geophones, but need some help in connecting them to a computer/working out how to collect the data
Any data collected using the geophones may only be available as a visual waveform, we may not be able to convert that into an infrasonic sound (might be fun to have some visual data for inaudible sounds anyway)
- I can go out and make some field recordings using the md/sony mic/ NXT set up we’ve developed but we won’t get frequencies under 100Hz with that. ‘Low frequency’ is subjective anyway, if you’re hearing you may find the extreme lows unpleasant, even if I like ‘em!
- I may be able to manipulate sounds to take them down to infrasonic range, but there is still the issue of playback (visual data for inaudible sounds) Or I could see what can be done in Pure Data to adjust frequencies?
- There are some interesting possibilities with NXT speakers but I need to understand more about the final set-up/sound generation to see where any low frequency/infrasonic sound playback could fit in.