Internet Archive: Beehive Sounds

Mysterious sounds generated by a hive of honeybees. They resembles a purring cat The day sounds (lmys1.wav) differ from those produced at night (lmys2.wav). These are very rare sounds. I've only heard them a few time in 40 years of beekeeping.

Internet Archive: Inside a rooftop beehive, Rue de Flandre

Inside a rooftop beehive, Rue de Flandre, Brussels

Stack Exchange: How to make very long recordings (of bees)

An idea on gathering audio data: have you considered performing Fast Fourier Transforms on the data before sending?

A Fourier Transform is basically an analysis of sound data (or anything involving waves) that determines the spectrum of frequencies dominating the sample. It's basically a conversion of data from the domain of Amplitude vs. Time into the domain of Amplitude Vs. Frequency. This could be handy for the purpose of this project that is to identify certain frequencies of bees humming associated with certain behaviors.

Here's a good write-up giving a basic overview using the Arduino to perform FFT: http://www.arduinoos.com/2010/10/fast-fourier-transform-fft/

Using FFT you could simply gather samples for a specified amount of time on a recurring interval and transfer the analysis of the audio samples rather than the full recorded audio. This would be a significant trade-off on data to be transferred and battery life. The computation is somewhat complex, but if an Arduino can do it the Spark Core should be able to handle it. The question is if an embedded device can get the desired sample resolution that would be needed, which would require a little bit more research.

I don't have significant experience implementing FFT yet, but it seems like a good idea for this project.

Edit: from a quick google search on FFT and bee hives: http://www.beehacker.com/wp/?page_id=103

Recording hive sounds in a very simple way, I have fitted a simple lapel mic to a Tascam DR-05 digital audio recorder. The mic is placed within the roof space of the hive, on top of the frames within a layer of hessian material that forms a condensation catchment. The hive is a modified 20 frame Langstroth hive and there is a reasonably sized hive resident inside. All of the frames are active with either brood or food. The recorder is continuously powered from the USB data port. I will post some pictures once I'm allowed to do so in this forum but it's quite a simple setup and very easy to implement.

Further to the post above

Geolocation data : Melbourne Australia. https://www.google.com.au/maps/place/Catherine+Ave,+Mt+Waverley+VIC+3149/@-25.7789565,141.1982772,5z/data=!4m2!3m1!1s0x6ad66abcb0caf5a9:0xb595280b78aba0ab

Bee species : Apis mellifera

Hive type : Modified 20 frame Langstroth hive. Sitting above the hive space is a layer of hessian enclosed wood shavings which helps with condensation. The lapel mic was placed above the central brood comb within a hessian layer.

The lapel mike was sourced from a local electronics store for $20

.

The Tascam recorder is fitted with an 32GB memory card and continuously records a stereo, 44.1kHz MP3 file in 512MB blocks (around 9H of data). The recorder and a mains power supply are mounted into a weatherproof enclosure around 2.5m from the hive.

A link to the audio files is here https://mega.co.nz/#F!ZBNxhRTR!OCjEdBzodkbMWmW0chFcyg

My ultimate goal is to record swarming activity. We are in the middle of this period now although local swarming appears to be have been delayed. Hopefully this hive will come through with a swarm in the next few weeks. I am not that experienced with swarm activity other than it is a very active hive and I fully expect it to do so. The last two years have been quite difficult for the bees with local drought conditions.

This is still pretty new so any comments are welcome!

That is a really cool set-up. We have a 20 frame Langstroth that I was going to do something similar to - this is a great way to maintain easy access to the recorder.

This recorded hive swarmed on the weekend. The audio is quite interesting with "piping" calls evident. The noise on the pickup increased incredibly at swarm time. The internal hive temperature and humidity were also logged and show a marked peak at the time of swarming. Images for audio (not to time scale) and an environmental data plot are shown below. Swarm time was around 12:30pm in the afternoon on 26th Oct. A data link will be posted soon.

Congrats! Great to see that you captured that audio of the swarm. We are copying your set-up now, looks like that is definitely the way to go.

Hello All,

I am currently working on an installation art piece combining video, audio and interactive sensors embedded in podiums. The project emphasizes human collaboration with bees (Apis mellifera in particular). OSBH is beyond perfect for this project. I am developing an interface with Arduino/XBee and Max/MSP to pull in live data from human interaction in the space and was hoping to pull in live data recordings from hives. I see many audio and data recordings posted above and I was wondering if anyone could point me in the direction to find a live data feed. If not, I was hoping to gain permission to use the audio snippets above in my collaboration.

Feel free to scope out works in progress and development on my blog, http://www.zoedoubleday.com/?page_id=425

Thank you so much for your time and consideration!

I'm fine with this, and am quite sure @AaronM and @jonathan will be as well.

@Jagungal - does this sound alright with you? Could you share a link where @ZoeDoubleday could download some of the audio? I imagine soundcloud or somewhere we can easily share around might be a good option, since there seems to be quite a demand for your audio!

Hi @ZoeDoubleday. This looks like a nice project and am certainly happy with supporting in any way that we can. We will look at getting you a live feed for this as soon as a suitable one pops up online. Cheers @AaronM @trizcs @Jagungal

I was thinking of doing something like that actually - Soundcloud looks pretty good. I will try to put up one of the smaller (400MB) files and share it so that artists can sample off it.

If I get time, I will upload one of the larger files too, this will give @ZoeDoubleday a couple of hours of audio that she can play on a loop. Not quite a live feed, but still kinda fun.

@Jagungal Hey Cormac - if you hand pick the most interesting audio clip and send it to us, we'll upload it to SoundCloud and invite the whole community in our newsletter to check it out and make comments.

No problem @AaronM, one of my colleagues is a musician and has an audio editing suite, so she is going to analyse the datasets on our server and see if she can find any interesting sections that can be clipped out into a smaller audio file for sharing.

Once that is done I will upload to Soundcloud or similar and post the link here.

She has a big dataset to play with, but she also is a bit freaked out by bugs - we will see how much bee singing she can stomach!

The spirit of collaboration is strong within you @Jagungal! You're an asset to the project.

Hey @ghellard could you send @Jagungal your swarm event recording?

Hope your colleague can handle the buzz!

@Jagungal @ghellard

Hi Guys - you can download Greg's swarm audio here:
https://mega.co.nz/#!yQ5FjaxK!b3D5Hf9JiV_D_0VPGvdsvN7AlGgYHKN4S_IUrRUTFoI

Thanks

Thanks Aaron. I was just about to send a link. If you need more information just post and I can answer any questions.

Thank you so much all! I am grateful for the support and effort! I will make sure to include all proper attribution in all documentation of the work.

OK, here is a sample of the in-hive audio as a .wav file with some of the background noise scrubbed out. Looking at the data this appears to be around the time that the colony realised that the queen was missing, there are a lot more short spikes in the audio data after this:

Missing queen sample.wav (1.3 MB)

The one interesting thing with the audio is how relatively uniform it is - this is a good thing for detecting swarming, as this seems to be a very different to most other hive events.