Posted by: recordingsofnature | March 18, 2015

Recording of early spring in the forest by the motorway & Comparison of microphone setups

A nature recording from Brøndbyskoven on the first sunny spring morning, March 8th 2015. The weather forecast had predicted temperatures approaching 10C, however in the morning hours around 9, it felt windy and not very warm.

IMG_3784This forest located in the suburb of Copenhagen and is almost surrounded by motorways. On this morning only very few natural sounds were audible; a few tweets and hissing leaves. The all dominating sound was man-made and came from the motorway, -that is even for an early Sunday morning.
This forest would be completely different place if  the traffic noise was absent. It is easy to imaging how a constant, 24-7 background noise disturbs and stresses birds and animals. It simply makes it difficult to communicate over distances. Likely the birds will try to sing louder (and more harshly) in order to break though the background noise.

It is my clear experience from audio recordings in nature, that traffic noise, followed by air traffic, are by far the dominating man-made sounds. In this context, noise from wind turbines are exceeded by magnitudes, yet still wind turbines at desolate locations are yet another noise source that adds up in the total picture. Anyway wind turbines should not be a problem near busy roads since the noise is simply downed by the cars.

360 degrees panorama from the recording site

360 degrees panorama from the recording site. Forward direction of the recording corresponds to the center of the image. Location: 55.647841, 12.436417

With this field recording I am testings and comparing 4 different recording setups. It is my plan with a new binaural setup, one day, to be able to add an extra layer realism into the nature recordings. As by now, it has not really succeeded yet.

1. Standard nature recording setup

(Listen in headphones for best result)

This first clip is captured using my standard recording setup, which uses a stereo pair of Earthworks qtc40, with a baffle and attached pressure equalizers. Here, stereo image is created by the microphone spacing and dummy head baffle, which provides extra channel separation and stereo time delays.

IMG_3731   IMG_3736

Due to the high quality Earthworks microphones, this setup has a very clean and linear frequency response ranging from a few hertz up to 30-40 kHz. The attached pressure equalizers gives a slight enhancement of the forward directionality and adds an about 4dB presence gain in the 5-8kHz range (also described here). This is also seen from the measurement of the frequency response of the setup, as seen below.

8_std qtc

Frequency response for my standard nature recording setup. Different colors corresponds to different directions.


2. Experimental binaural setup

With this binaural setup, the Earthworks microphones were now positioned just at the opening of the ear canal of a (semi-) realistic human ear models on a plate. More details of the setup is described here.

IMG_3739   IMG_3740   IMG_3743

In addition to the latter microphone setup, the Binaural technique is supposed to capture extra directional information related to how the unique shape of the ears color the sound depending on the direction. Ideally, this will provide a clear sensation of sound sources with respect to depth, front/back and up/down.

The binaural technique is rather delicate and the effect will vary from person to person. It will definitely only sound good with headphones. The measured frequency response of this setup (below) is also far from flat, featuring a significant 6-10dB gain in the 4-8 kHz range. In order for the recording to sound normal, this gain has been compensated by use of EQ to reach a more flat response (diffuse field EQ). Out from the frequency response it can actually be seen that the used ear geometry is not fully correct compared to a real human ear, as the response curves do not really match with the standard measurements of human ears and HRTF’s, such as found here and here. It looks like the ear is too small.

8_qtc eras.

3. Binaural setup including ear canals

In this setup, the binaural setup was extended with ear canals, and the Earthworks microphones were now attached at the end of artificial ear canals.This is an experiment to explore how the sound actually is at the end of the ear canal at the tympanic membrane. As expected this recording technique does not sound very good, but I think it tells something about the nature of the human ear and for which frequency bands the ear has a high sensitivity.

IMG_3765   IMG_3764   IMG_3767

As seen from the frequency measurements, the recorded signal is now completely dominated by awful resonances caused by the ear canal (in this case L=2.5 cm and Ø=7 mm). Despite these resonances, this is what the tympanic membrane and the brain hears, and what it perceive as a flat response.

8_earcanalThe audio sample for this setup is again equalized in an attempt to compensate for these resonances. The recording without equalizing can be heard below:



4. Binaural Tree Ears setup

The final setup (Tree Ears setup) uses the same ear models as the binaural setup (with no ear canals) but now using the Primo 172 microphone capsules instead of the Earthworks microphones. This setup is also described here.

IMG_3773   IMG_3771   IMG_3798

The frequency response here is almost the same as for setup no. 2. and the for the audio sample the same EQ has been applied to reach approximate flat response.




The sounds at the forest for this recording session turned out be very sparse, and apart from the consistent traffic noise, not the best for comparison of microphone setups. I think all recordings sound pretty similar. I feel there is a bit more directional information in the binaural setups, but various design parameters are currently not optimal and needs to be adjusted.

Next step is to optimize the ear design.  I am convinced that there is a potential for very interesting nature recordings with the binaural method.



  1. Just some feedback: Your original setup (1) sounds good, the binaurals (2) & (3) appear to be hollow, lacking in centre audio field with (3) being the worst of all (it was very uncomfortable listening to that one). The tree binaurals (4) sounded good.

    I wouldn’t worry too much about the affect of traffic noise on bird calls, they appear to have excellent, selective audio mechanisms which allows their calls to be transmitted and received by their own species within the most noisiest of environments – it’s similar to our use of narrow band filters in radio transceivers when operating with CW signals (morse), which can similarly be heard amongst the noisy background radio interference/traffic.

    Best of luck with your experiments.



  2. Thanks for the feedback!
    Actually it was not the intention that anyone should listen to binaural(3) for more than a few seconds! I highly agree this one sounds uncomfortable.
    I going to try to optimize the ear shape + dummy head. I think the combination with the earthworks microphones should have good potential.
    I think you are right about the birds, their hearing is most likely focused on the frequencies of the own voice, and that is quite far from the low frequencies of the motorway rumble.
    best regards, Kristian


    • Liking what you are doing with the binaural audio, I’m streaming binaural audio 24/7 – home made binaural mics with 3D printed ears for the time being, hope to have some silicone ears of my own ears, and perhaps another set from a lady I know. Have you made your ears from making casts of your own ears? One would hope that your ear shape is that of the majority that listen ;o)

      my stream is here:

      Main website (another raspberry pi)

      I’ve even a IP camera in the garden too, has good night vision, user admin, password left blank

      Keep up the good work!



      • Thanks for the links Alan! Very useful information.
        I have been at your site and listened to your stream and i must say that your realtime streaming is just brilliant. It is very interesting and relaxing to listen to. Unattended, realtime recordings like these are true field recordings.
        At this moment testing updated versions of my ‘own’ ears. They are modelled in Art of Illusion 3D software, using pictures of my own ears from different angles as reference. I am planning to use my tree ears for some long unattended recordings and maybe one day streaming…

        all the best

        Liked by 1 person

  3. […] June 21, 2015, around noon. It is captured close to previous recording location from March. Oh, now the forest is now just so fresh, dense, and green from top to […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: