Sound Samples and Basic Analysis Hive with Queen vs. Queenless

Background

For an initial proof of concept about sound analysis and honey bee specific sound data I’ve chosen a phenomena beekeeper talking about for ages: It is a common knowledge that “normal” colonies with a queen sounds different than queenless hives.

I have checked this hundred times and did fail a dozen times while regular hive inspections. You have sometimes heavily buzzing hives, but all is ok and the queen is there and sometimes you have the opposite facts: All sounds normal but you know there is no queen in the hive. So the human ear seems not to be the best sensor for honey bee buzzing recognition.

Introduction

By counting on technology in this special case I did a small experiment - no built-in hardware in the hive, no Arduino, no I2S mic - just an ordinary mobile phone with Android and an audio recording app Sony’s Audio Recorder: https://play.google.com/store/apps/details?id=com.sonymobile.androidapp.audiorecorder&hl=de

I started recording with the phone in my hand then I placed it under the hive. The base was not solid but had a mesh, so sound recording was easy without disturbing the bees by opening the hive.

Analysis

At home I used the Sonic Visualiser http://www.sonicvisualiser.org/ to visualize the sound.
I added a so called “Peak Frequency Spectogram” and tweaked the settings a bit, in the end I got this:

Colony with queen


Download: colony-with-queen-gruber.mp3

Colony without queen


Download: colony-without-queen-gruber.mp3

Further Analysis

See also the thread with the automatic OSBH algorithm based sound analysis Analyze recordings from queen vs. queenless hives using "audiohealth"

Conclusion

I think you can see clearly differences in the pictures, have also an ear for the sound. Do you hear differences? ;-) Isn’t this interesting??!

For me a colony with queen sound like driving in a car on the Autobahn. A deep sound is there “in the back” while the second track is more confused.

3 Likes

you think of: raw PCM audio? I’d like to analyze that too…

(the mp3’s uploaded are available via the context menu of the in-built player: it has “copy audio location” and “save audio as…” - at least here.)

I missed that detail, thanks!
Update: Also added explicit “Download” links.

10 posts were merged into an existing topic: Analyze recordings from queen vs. queenless hives using “audiohealth”

4 posts were split to a new topic: Sound Visualization

I have split up some postings to the new topic Sound Visualization in case you want to see some statistics about the here discussed data have a look at this new thread:

You can find here some more sound samples from queenless and hive with queens, the file Hive1 and Hive2 is the same as in my first posting but a bit longer.

Metadata can be found in the file name.

From the visual inspection hive 4 and 6 are outstanding: Hive 4 looks like a queenless hive. the main frequency peaks around 250Hz is not so clear visible. There is a sound peak also but with a broader frequency area. Hive 6 has this band much clearer but has no queen. So it looks - from the spectrogram - like a normal hive with queen.

Hive 1, queenless


Download: hive1_no-queen_nucleus-10-days-old_with-queen-cell_2017-06-05_15-27.mp3

Hive 2, with queen


Download: hive2_queen_production-hive-small_2brood-boxes_no-super_old-queen_2017-06-05_15-29.mp3

Hive 3, with queen


Download: hive3_queen_production-hive-big_2brood-boxes_2super_open-and-covered-brood_2017-06-05_15-33.mp3

Hive 4, with queen


Download: hive4_queen_production-hive-middle_2brood-boxes_1super_open-and-covered-brood_2017-06-05_15-35.mp3

Hive 5, queenless


Download: hive5_no-queen_nucleus-10-days-old_with-queen-cell_2017-06-05_15-37.mp3

Hive 6, queenless


Download: hive6_no-queen_nucleus-3-days-old_2017-06-05_15-40.mp3

Hive 7, queenless


Download: hive7_no-queen_nucleus-3-days-old_2017-06-05_15-42.mp3

1 Like

A post was merged into an existing topic: Web-Service “Audioanalyse”

I have some new samples, some of them from the same hives as in the last postings but some weeks later and also some new data from new nuclei:


Download: hive1_queen-nucleus_2017-07-21_12-38-18.mp3


Download: hive2_queen-nucleus-some-weeks-old_2017-07-21_12-39-36.mp3


Download: hive3_queen_production-hive-small_2brood-boxes_no-super_old-queen_2017-07-21_12-40-43.mp3


Download: hive4_queen-nucleus_2017-07-21_12-42-21.mp3


Download: hive5_queen-nucleus_2017-07-21_12-43-48.mp3


Download: hive6_queen_production-hive-middle_2brood-boxes_no-super_2017-07-21_12-45-08.mp3


Download: hive7_queen_production-hive-big_2brood-boxes_no-super_2017-07-21_12-46-22.mp3


Download: hive8_queen-nucleus-some-weeks-old_2017-07-21_12-47-35.mp3


Download: hive9_queen-nucleus-some-weeks-old_2017-07-21_12-49-14.mp3


Download: hive10_queen-nucleus_2017-07-21_12-50-42.mp3


Download: hive11_queen-nucleus-some-weeks-old_2017-07-21_12-52-10.mp3

3 Likes

I can also confirm from experience that although a lot of queen-less hives do really sound different , others can (to the regular ear) sound OK … but the recordings you shared are differently interesting

1 Like

via LongHive

Hi @clemens. I’m a software engineering student and I’m interested in the subject also I have some questions for you. What was the type of bees that were recorded? Were the hives located in the countryside or the city?

1 Like

Hi @cljimene, nice to have you also on hiveeyes here! The bees are Apis mellifera carnica or a F1 or F2 generation from a Carnica hive, in the neighborhood are some Buckfast hives also so perhaps the F1 or F2 has not Carnica genes only. The hives are located in Berlin, so a city, a street is nearby, not too much traffic, but there are for sure some audio artefacts from the city in the sound data.

1 Like