This article is specifically about capturing classical or other acoustic music with multiple microphones in a format that is suitable for playback on a 5.1 surround speaker system. 5.1 surround is desirable with classical music in particular to give the listener a sense of being in the midst of the music. If I can get 5.1 recording to work well, my use case is to record choral church music, which I have been recording in stereo for many years.
Note: I am specifically talking about recording in 5.1 in one take with multiple microphones. I’m not talking about upmixing stereo, or creating a 5.1 sound stage artificially by panning a multi-track recording. But those are probably fun things to try too! 😀
In a previous blog post about stereo recording techniques, I gave a description of the church building. It’s a bit odd because the choir sit at the front, while the organ is at the back. 5.1 will probably sound a bit wacky, but should be interesting. The church is a reverberant space and if I can capture some of the ambience and presence of the building, I will be very pleased.
Microphones
The first question is how to lay out the microphones. 5.1 miking is far less well established than stereo miking. DPA Microphones have quite a helpful page about surround microphone techniques with some examples. I looked at the microphone arrays they suggest but with the equipment I currently have, I wasn’t able build any of them. Many of the designs demand the use of figure-of-8 microphones (which I don’t have) or lots of omnidirectional microphones (of which I have just two). The Fukada Tree looks most promising but needs a few more microphones than I have.
After consulting several sources including this guide about general surround-sound miking and this document specifically about microphone array design, I decided to build a Williams array as I had enough of the right type of microphones, and it isn’t physically too big.
It looks quite simple as a diagram. In real life it’s not quite as pretty:
This Williams array is built from the following microphones:
- L and R: Behringer B-5 small diaphragm condensers with cardioid capsule
- Centre: Behringer B-1 large diaphragm condenser (cardioid)
- Ls and Rs: Behringer C-2 small diaphragm condensers (cardioid)
The B-1 isn’t an ideal choice for this application – a small diaphragm would have been better. However I thought the large diaphragm would pick up more bass and could also provide the feed for the LFE subwoofer channel.
As you can see, this microphone array is quite large and intrusive, and not really suitable for use in a church service, so to prove the concept I decided to record choir practice instead. The music is interrupted here and there, but is definitely sufficient to prove whether or not the recording technique works! I placed the array on the D-shaped platform near the front of the church.
Mixing
The mix is straightforward, since each microphone is mapped directly to one channel. 5.1 actually means 5 surround channels and 1 LFE subwoofer channel – 6 channels in total. As you can see, my array contains only 5 microphones so the subwoofer (LFE) channel will be created by pinching the feed from the centre channel and filtering out the treble with a low-pass filter.
Mixing was quite laborious for me – my audio workstation only has stereo monitors and I had to render the project to a single 5.1-encoded file and copy it onto another PC which is attached to my 5.1 home theatre amp to listen to it. Rinse and repeat! If I’m going to make a habit of 5.1 mixing I’ll need to either get 5.1 speaker for my audio workstation, or start doing my 5.1 mixing using the home theatre PC downstairs.
After tinkering with the mix a bit, I decided that the subwoofer channel sounded best when I summed all of the other channels and sent them through the low-pass filter – not just the centre channel on its own.
Mastering
The only mastering step I did in the traditional sense was to add light compression to boost the volume in the quiet passages. Choirs have enormous dynamic range!
I’m using an old version of Adobe Audition (3.0, from 2007) which has fairly poor support for surround encoding. You work in stereo while doing the mixing and at the end It’s basically a case of “assign each project channel to a speaker channel and press the button, and it spits out a multi-channel WAV file”. I understand more recent versions of Audition support 5.1 projects better.
Formats
Distributing 5.1 audio is a tricky issue. Most commercial 5.1 audio releases are sold either on a DVD-Audio disc or Blu-Ray audio disc (in both cases, these are not the same as video discs). I don’t have a Blu-Ray writer so I decided to try and make a DVD-A disc. It’s surprisingly hard to do – the specialist software is hard to find and expensive. I managed to find a free trial of one package and burn a disc which looked sane to me. However, my Sony Blu-Ray player refused to read the disc, so who knows. There is an open source project called DVD-Audio Tools, but I wasn’t able to compile it.
In the end I gave up. It seemed a bit of a waste of time as none of my friends have 5.1 audio systems, and I’m content just playing the files from my home theatre PC! I decided to convert the 6-channel WAV file to a 5.1 FLAC file, which is widely accepted (but not universal).
If you have a 5.1 system, you can download a sample FLAC file from this session (you might need the codec). This recording is Kyrie Eleison by Louis Vierne. It should play on stereo systems too, but obviously you won’t get the surround effect.
For anyone who wants to hear the recording but doesn’t have the ability to play FLAC in surround, here’s the stereo MP3 version – available for streaming or download.
Overall I’m happy with my first attempt at a 5.1 recording. There’s definitely room for improvement on the microphone side of things, but the biggest obstacles I face are with processing the recording and distributing it in a format that other people can play. The fact that a lot of people don’t have a 5.1 system at home or the interest in using it to listen to music means that this is a bit of a gimmick, too.
The Williams array I used is too obtrusive to use in a live service but could be used for “studio” recording sessions. For now, I will continue to record church music in stereo and distribute it as a good old audio CD, plus MP3s for streaming and download for those who want to listen on their iPod.
Jonathan Gazeley 