Microphone design Home Analysis software
Record audio and extract calls automatically (updated April 28, 2015)
Recording sound from the night sky directly to computer hard drive is the most efficient way to document avian night flight calling. Ten hours of recording (22050 kHz, 16 bit sampling parameters) translates to ~ 1.6 Gigabytes of hard drive space. Recording directly to a hard drive gives you a digital soundfile that can then be analyzed for bird calls ~50x faster than real time with a moderately powerful PC running Old Bird's automatic detection software. This software can also work in real-time and run through the night depositing short files of potential bird calls in a folder on the computer. The Old Bird call detection software only works on PCs and only on mono, 22050 sampling rate 16 bit resolution wav files. Macs can be used with PC emulation software (e.g., Parallels). Raven sound analysis software from the Cornell Bioacoustic Research Program records sound, has extensive sound analysis tools, and has detectors that work with both PCs, Macs, and Linux.
For recording audio on PCs, third party software is available for recording at specific times and automatically naming and saving sound files to a specific folder on your computer. For example, see the inexpensive Easy H-Q and I-Sound (both ~ $25 USD). An important consideration for automatic nightly recording is to set up the computer so that in event of a power outage the computer automatically returns to automatic recording status once power returns. To do this one must go into the PC's BIOS and set up automatic bootup after power outage in the Power Management section. One must also make sure that the option to have the recording software (e.g., Easy H-Q or I-Sound) automatically initiate after the computer restarts is selected. The software can be set to automatically name the nightly files with the date and time of recording. The Easy H-Q recorder allows one to add a specific text name in addition. This is useful if you have multiple stations and want to also label a recording with a site name. One trick to using both Easy Hi-Q and I-Sound is that once you have selected your settings, you need to quit the program and reboot your computer in order for any settings you specify to carry over if there is power outage. Note also that there is a 2 GB file size limit for wav files - this correspondes to about a 12 hour record time.
Most PCs come with a soundcard installed, but there are many different types. Most have a microphone input and a line-level input. If you are running directly from a microphone to the computer, you will want to use the microphone level input. In most cases, the home-built design offered on oldbird.org will need a preamp to boost the signal up to microphone level. It will be important to test the input recording level to make sure it is set at a reasonable level - not too loud or too weak. On most PCs the recording input level adjustment can be set by double clicking the sound level control icon in the lower right of the Start menu bar (little, often gray-colored, megaphone-looking icon). You double click this volume control then select the options pull down menu (upper left on the volume control window). Select properties, then select recording, then select OK (make sure the microphone volume control is selected), and adjust the microphone volume. Many newer computer soundcards have a mic boost option than can be implemented in the volume controls, typically in an advanced controls section (accessible by clicking on options). This increases the loudness of the signal. A caveat is that many newer PCs have soundcards that come off the shelf only supporting a 16 kHz sampling rate, which is not high enough to render many high-pitched night flight calls -- spectrograms will only show sound energy up to about 8 kHz. In some cases you can adjust the computer's soundcard's options to access recording at higher sample rates. Another option is to buy a Turtle Beach Amigo II USB sound card (~ $25 USD) and bypass your computer's sound input.
You will need some way to assess how loud your input signal is. Audacity is free sound recording software that allows you to see your recording input levels. If you are using the Easy HiQ or I-Sound Recorder, you can simultaneously adjust the microphone input level and see the input sound level vary in the recorder window. For Easy H-Q, ideally you want the typical night time background noise at your recording site (quiet night) to be a bit to the right of center in the signal strength indicator. For I-Sound, background noise should show about a quarter of the of way up the signal strength indicator. This allows enough room for bird calls that are louder than the background noise to be recorded without being distorted. Inevitably, the calls of some low flying birds will be too loud and end up being distorted. In most cases these calls can still be classified to species when spectrographically analyzed (e.g., with Old Bird's GlassOFire software). The goal in setting the input recording level is to record the bulk of the calls loud enough for spectrographic analysis but without distortion. You will want to experiment with your input recording levels to find a good setting.
In most cases, even using the mic boost option on newer sound cards, running the
signal from your microphone directly into the microphone input of your computer
will produce too weak a signal. You will likely need a preamp to boost the signal
from your microphone before you run the signal into your computer. The
USB Dual Pre (~$75 USD) is one that works well with the circuit noted on the
microphone construction page. This preamp allows you to turn off its 48V
phantom powering and let the 9V battery power the circuit. It also has a second
channel to facilitate a stereo monitoring system and can operate on battery
power for off-grid operation.
The USB Dual Pre (~$75 USD) is one that works well with the circuit noted on the microphone construction page. This preamp allows you to turn off its 48V phantom powering and let the 9V battery power the circuit. It also has a second channel to facilitate a stereo monitoring system and can operate on battery power for off-grid operation.
A recording option is to run the microphone signal directly to the computer while a detector program is operating. In this case, the computer will only record snippets of sound when specific bird calls are detected (species recorded depend on the detector used). This does not use much hard drive space. The problem with this strategy is that you only get what the detector picks up and there are no detectors that pick up all types of bird calls. This type of call detection can be set up to run automatically using a PC's task scheduler and the simple scripts below:
Start and Stop Detector
Click links to download zipped MS-DOS batch file
The Start detector is set up to automatically start the Tseep-r warbler and sparrow flight call detector if that program is located on the C: drive (i.e., C:/Tseep-r.exe). To schedule the start of a different detector, right click on the Start Detector icon, select Edit in the popup menu, and simply change the Tseep-r text to the text name of the detector you would like to start and save the file.
Once you have downloaded the Start and Stop detector batch files, create a folder named Batch on your C: drive. Then, access the Scheduled Tasks icon in your PC's Control Panel, select add scheduled task and follow directions for setting up automatic tasks to run the batch detectors.
Moderately powerful PCs are capable or running the Tseep and Thrush
detectors at the same time as well as an all-night recording program like Easy
Moderately powerful PCs are capable or running the Tseep and Thrush detectors at the same time as well as an all-night recording program like Easy HiQ.