This is a proof of concept. Due to security restrictions on mobile web browsers head tracking is not supported by now. Please volunteer, if you would like to contribute to a native iOS or Android app.
Who was the nightingale singing to? The wind? The passing cars? The trees?
This is my very first em64d Eigenmike recording in the fields, powered by battery. I cycled around looking for one of the few remaining nightingales still singing, and I found this one. The bird was kind enough to just sit there in one place and sing for the recording. During post-production, I separated the bird's song from the rest of the audio signal, so that I could apply different equalisers and microphone array rendering types.
It's quite challenging to find the right amount of noise that is still acceptable, because some sounds and voices are noisy by nature. If you suppress the wrong frequencies, it sounds artificial right away. In my case, I wanted to record the nightingale. However, if I 'cut out' the bird and remove everything else, it doesn't sound right. Attempting to eliminate electrical noise by blindly attenuating high frequencies makes the recording sound cheap quickly, too. I am new to HOA recordings, but from what I have learned, a good step in the process of dealing with noise is, to start with (re)identifying all aspects, that were present in the scene.
The elephant in the room here was simply the sound of leaves whispering in the wind. I didn't notice it wile being enchanted by the bird.
A demonstration of the 10 different encoding methods, called 'Designs' the Eigenmike software offers.
My aim for this short piece was, to show the inherent tradeoff between noise and spatial information. With each new step you get more spatiality, at the cost of noise.
This platform decodes up to third order Ambisonics, so what you hear is the first three orders of this test. In the original 6th order Ambisonics mix, the difference is more prominent.
Which 'Design' do you like most? Tip: use the numbers on your keypad to jump.
em64d Eigenmike A‑Format to B‑Format encoded 10 times
The recording was made with a Schoeps MK2H. The signal was rotated in third order Ambisonics with the IEM MultiEncoder plugin. Impulse response reverb was added with Ambi Verb HD (Jeanne d'arc Church). Reaper was used for processing and rendering.
Thanks to talent and kindness of Ms. Daniela Niedhammer, I was able to realise one of my first Ambisonics recordings.
In order to get my A-levels I had to write a skilled work. Supported by Hermann Voßeler, Ingo Kube, Albert Frasch and my father Walter Brachtel (†), I could set up a first order Ambisonics microphone using four Schoeps MK4 cardioid condenser microphones, arranged in a similar way (tetrahedron) as disclosed in the patent of Michael Gerzon. In comparison with higher order microphones, localization is blurry.
What holds up is a distinct quality of reverb and a wonderful interpretation of Ms. Niedhammer.
Imagine standing among some of the kindest and most talented aspiring singers of your hometown – who helped with the equipment, and knew by heart how to hold you after the concert – as the voice of a cello emerges from behind you on the left and enters into conversation with a viola voice so salving that my late mother’s long service of care and sacrifice could finally be filed under unsuccessful practice.
The concert recording has been on my hard drive for over two years now. I tried to finish the spatial mix countless times, and, admittedly, that was a challenge. I had to overcome both technical and personal issues.
Deutsche Bahn made me two hours late, and we had less than thirty minutes left to set up eight microphones on three different systems. I wasn’t able to set levels with the choir prior to the concert. Consequently, when I corrected the Zylia ZM-1 midway through, crackling almost ruined the signal. I apologise for that.
On a personal level, every time I listened to this beautiful performance, I vividly remembered that special evening, and I found myself emotionally affected. I lacked the inner distance required to work professionally as an audio engineer. One needs to listen carefully for just a few parameters. They are more technical than we might think: easy to name, hard to identify. Balance. Tonality. Spatiality.
Only when I turned the room hard on the yaw axis all the way to the left, did the headlocked binaural downmix hold together; only then did the piece stop asking myself to be repaired. It started giving testimony: the cello left behind me, where my late father had always preferred the last row of the cello section, the viola at the centre where my mother’s practice had found its proper filing cabinet. The choir became what they were all that evening: not just singers, but attentive listeners.
Did I find a balance between the instruments and voices that would support the conductor’s interpretation? Above all, did I treat the composer’s intent with due respect?
One cold, windy day in December, I travelled to my hometown to record my brother’s graduation concert. The audience was excited by the performance and so was I.
That evening I could test my new third-order Ambisonics microphone for the first time. I also made a binaural recording from my desk back in the far corner with the Sennheiser headset within my pinnae. This helped me as second reference during mixdown.
Everything went perfect. Before the concert I managed to crash my brother's laptop's Dante Audio, which he needed for live echo effects. Puh. Thanks to a local technician, we got the network fixed. The mix you are listening to is a combination of the Zylia ZM-1 and the Schoeps MK5 (set to cardioid) + MK2H on stage, with some additional higher order convolution reverb.
Thanks to Mike Senior, a driving force behind the Cambridge Music Technology Forum, one can experiment with an excellent performed and executed classical multitrack recording.
I used the IEM Plugin suite and a bit of Ambi Head HD reverb. Reaper was used as DAW.
You can find the multitrack files and details of the original recording here:
This Werck was made, when I built my first Ambisonics rig at home. A cube with eight no name active speakers, fed via a Behringer ADA-8000 ADDA converter and a PCI RME Interface. Good old days. :´-)
I had pulled an all-nighter coding this plattform; the sun was about to rise. From many failed attempts to get into the zone, I realised that you have to be ready at all times. My Zylia ZM-1 is always on a stand. I only had to grab the iPad, and take a few steps outside, to capture this excited roundtable discussion.
However, it took a few attempts to match the sound quality to my vivid memory of it. How much gain for each order? Am I allowed to EQ? What is the nature of the altered signal?
When I was at school, there was an urgency around everyone. I don’t do esoterics or mysticism, but I remember something - something that kept me struggling on when luck wasn’t on my side. Unspoken words between like-minded humans.
A classmate - a double bass player - was preparing a little piano piece, and I fell in love with it dans un instant. I learnt it by heart swiftly and played it this way ever since. Please do yourself a favour, and don't look up the score to compare it note by note...
This remains one of the few recordings, where I can stand myself, Ferdinand. It shall be to you.
Alexander Skrjabin: Prelude in B-flat major, Op. 11, No. 21