Research Project: Analyzing bee flight recordings

Audacity Help Forum Windows
1

Hello all!
First post, asking for general ideas. I am doing a project involving analyzing recordings taken of bees flying out in the field.
This is a proof of concept project so my system is quite cumbersome (made as inexpensively as possible.)

I am recording using an array of lapel mics (between 1 and 5) plugged into a converter plugged into a TASCAM 1800 USB interface connected to a Mac laptop. When out in the field, the interface and laptop are powered by a 12V battery using an inverter (an inexpensive $60 one that has been causing a lot of interference since it is not regulating power evenly.) When in the lab the interface and laptop are both plugged into a power strip plugged into a wall outlet- I still get a small amount of interference when plugged into the wall this way. The laptop is always as far away from the interface as the USB cord will allow (about 5ft.)

I am currently working with another researcher with a lot of experience with acoustics and its related equipment and processing to figure out ways to prevent/remove the interference from the inverter. He is sending us a voltage regulator so that we can connect the USB interface directly to the 12v battery safely.
My questions for you guys:
-Anyone have any other ideas to prevent the interference I am getting from either the inverter or the wall outlet? On both battery and wall outlet power, we have already tried having a 25ft extension cord between the inverter and the USB interface with no big difference in the amount of interference.
-The noise profile is different for each track I have recorded, so I at this time I am going through each using the noise removal tool to isolate the “flight buzz” I need to analyze- any other ideas for how I could remove the interference noise WITHOUT altering the peak frequency the buzz that I am isolating?
-Analysis- any efficient ways to find the peak/average frequency (most important) and peak amplitude of a track besides the plot spectrum, or best way to set up the plot spectrum for it to be the most accurate? Those are the data points I need for my project.

Super long post, thank you for bearing with me! I’d honestly appreciate any ideas you have particularly for post processing.

Specs: Audacity 2.0.5, recording with Mac running OS X 10.9.3 and doing post processing/analysis with a Toshiba PC laptop running Windows 7 32-bit.

2

Before we head down the wrong road… What happens when you run the laptop off its internal batteries?

the interface and laptop are powered by a 12V battery using an inverter (an inexpensive $60 one that has been causing a lot of interference since it is not regulating power evenly.)

There is a good chance that the laptop runs off of 12VDC, and it looks like the TASCAM runs off 12V also. So you may not need the inverter.

The voltages should be labeled on the power adapters. You’ll need to get the correct connectors and make sure the polarity is correct. If you don’t have access to a multimeter, it would be a good idea to get one so you measure the voltage & polarity. (You can get a “hobby” multimeter for $20 USD or less.)

I am recording using an array of lapel mics (between 1 and 5)

Do you get the same interference without the mics plugged-in? Are these low-impedance balanced mics with XLR connectors to properly match the TASCAM inputs?

I would think that low level signals along with wind and other environmental noise would be the big problem. I’m sure low signal levels are a BIG contributor to your signal-to-noise ratio problems.

3

If you need low frequency accuracy, set the “Size” to 65536 (maximum) and the “Axis” to “Log frequency”.
For even greater low frequency resolution, resample the track to a low sample rate (Tracks > Resample).

Note that you can export from Plot Spectrum as numerical data that can then be analysed or plotted in other (more specialised) applications.
http://manual.audacityteam.org/o/man/plot_spectrum.html