Recording B - Second Conclusions

4/1/08
My second recordings had again produced a Speed of Sound measurement that must be very wrong, so I had to check every stage. The DAT recorder was set correctly to 44.1 kHz; the Soundscape was also set to 44.1 kHz, 16 bit, 25 frames per second, and it was locked onto incoming digital synch for the transfer. Vari-speed was not switched on! I very carefully made one more test recording in a long, concrete farmyard; measured the yard and re-calculated the SOS. This time it came out at 348 m/s (on a damp day at about 7 deg C). Now the problem must lie with my measuring. I'd thought first that the wheel of the odometer could be skidding, so I'd used it slowly and carefully. I'd also measure the circumference of the wheel - as far as I could tell from a tape rule, it was exactly one metre. After further tests I discovered this was not the case: I laid out a 15 m course, checking it with three tape-measures - and the odometer measured it as 15.4 m. Aha!

I checked this several times - the odometer was reading 97.402% of the true distance.

Even so, my measurements appear to be far from accurate. This could be due partly to running an odometer over uneven grass, but more likely it's because I'm not very good at measuring long distances across a field (never having done it before). On reflection, I believe this job is best done by an experienced archaeologist with surveying skills, rather than a musician with none...

So, using the on-line SOS calculator (which is only accurate down to 0 deg C) I settled on an arbitrary SOS of 330 m/s and recalculated the distance to my recorded reflections. These new figures have been added to the original results (in red) and if measured against the map, they confirm that it is indeed the stones which are producing the echoes.

This reinforces my conclusion that for calculations concerning the Speed of Sound, the most important tool is a thermometer for measuring the air temperature. Humidity makes little difference and can be estimated; air pressure makes none. However, as Victor Reijs has pointed out, it is good to have some reference measurements. Victor suggests recording an echo from a stone not too far away, maybe only 10 m (using a standard tape measure) so that the SOS can be compared to the javascript one. To this I would add - it's best to face sideways to the stone whilst doing this, so the returning echo is recorded onto only one channel of the stereo. This makes it much easier to see the echo on a graphics display and thus measure it. Steve Waller has also indicated that a control is need - an extra recording where there are no stones in the area. I will include this next time.

When I posted my first results, I included a page of the maths and this has now been appended: MATHS

As one can of worms closes though, another opens - the question of reflections from trees will be explored at a later date!