All in the <head> – Ponderings and code by Drew McLellan –

Recording Conference Audio

Recording Conference Audio

When preparing for the first dConstruct conference back in 2005, organiser Andy Budd sent an email to a few friends enquiring as to what he’d need to do to record then audio of the presentations for later release as a podcast series. Having a fair idea what was involved (sound engineering was my career path before I got distracted by the Web), I began writing a fairly lengthy description in reply. At the end of the email, fearing that my notes would completely put him off the idea, I offered to come along to the conference and help out. And so I did. (That’s a photo of my 2005 setup above.)

Since then, I’ve been back to Brighton and recorded audio for eleven dConstructs, two (soon to be three) Ampersands, and three Responsive Days Out. It’s something I enjoy doing, and I know how valued recorded sessions are to those who are unable to attend a conference. What I’ve never done is written up the process to help others who are planning to do the same.

This post describes what I do to record conference audio. In another post, I’ll describe how to edit it and produce final MP3 files for distribution.

What you’ll need

I’m presuming the venue you’re in already has a public address system, and the presenters are using a microphone of some sort. If that’s not the case, you have a bit more work to do, and that’s probably out of scope for this post.

The first thing you’ll need to do is get an audio feed from the sound desk. This means arriving in plenty of time before the first presentation starts and chatting to the technician who’s operating sound for the event. Tell them you’ve been tasked with recording the audio for the event, and ask if you could please have a feed from the desk and some power. The feed needs to include all the mics, any laptops, but no house music.

That brings us to the first bits of equipment you need – a good quality, balanced audio cable to take the feed from the desk. Most modern, professional desks will have XLR outputs for their auxiliary outputs (‘sends’), although smaller venues or places with older desks (club venues especially) may well take 1/4 inch jacks. I carry one of each, and have used both.

The second thing you’ll need is a reliable multi-socket power strip with a good length of extension cable on it. If you don’t know the venue and where you’ll end up setting up, be prepared with a good length of cable. In case you need to run your cables across a walkway, bring some gaffer tape to safely tape them in place.

So you have power, you have a mono audio feed coming from the sound desk via either XLR or 1/4 inch jack, and it’s all taped down and safe. What do are you plugging this into?

Belt and braces

The obvious pressure of recording a live event is that if you have an equipment problem and fail to capture something, there’s no going back and doing it again. For this reason, I build a little redundancy into my setup. I don’t go completely over the top (it would be embarrassing and a shame to miss something, but it’s not life and death) but I do split the signal into two paths and make two separate recordings of each event. If one of those fails, I have the other.

I used to do this by recording to both MiniDisc and into a Mac. A couple of years ago I replaced the MiniDisc recorder with a digital SD card recorder – a Tascam DR-07mk2. There are a few options from Tascam at different feature levels and price points. If you’re going to do a lot of recording, consider one with XLR inputs. Mine has a minijack line input, which is fine, but not as hard-waring and not ‘balanced’. A company called Zoom also make similar recorders and a cheaper price point – Zoom are Dell to Tascam’s Apple, they do the same thing, but one leaves you in a better mood.

I split the signal by taking the audio feed into a small mixer. The one I have is a Soundcraft Notepad 102. I chose it because it’s small, light, and my existing large mixing desk is from Soundcraft, so I’m familiar with the layout. It’s discontinued now, but almost any small mixer would do. You need to be able to take an XLR and 1/4 inch input, set the input level, and independantly set two different levels to two different outputs. I use the main mix and an auxiliary out (pre-fade) for this.

I send the main mix out to the audio interface of my computer, and the aux send out to the SD recorder. I can independently control the output level of each, which gives good flexibility in finding a good recording level for each of my recording devices.

Recording

My primary recording device is a Mac running Adobe Audition. I used to use Apple Soundtrack Pro for this (now discontinued) but switched to Audition because I have access to it as part of my Creative Cloud subscription. Anything would do – Audacity is a good candidate if you don’t mind assaulting your eyes with its repugnant UI. Garage Band might be another option.

I use a Focusrite Saffire firewire interface with my Mac to provide audio inputs and outputs. Don’t use your computer’s built-in analogue audio input, it’s cheap and designed for a chat headset, primarily. The interface I mainly use is (again) discontinued. I also have a Focusrite Scarlet 2i2 USB interface, which is really great for the price. It’s smaller and lighter than the Saffire, so have used that when travelling light.

The key to making a good recording is to watch the levels. You want to get a good strong signal without overloading the inputs. If the incoming audio is too quiet, you’ll be unable to amplify it without also amplifying the background noise (hiss) and the recording will sound noisy. If the incoming audio is too loud, you’ll overload the inputs and the signal will ‘clip’ – all values will be at 100% and the dynamics of the sound are lost. Clipped audio is harsh and hard to listen too.

There will usually be a number of sets of levels you can watch – one on the mixer, and one on each of your recording devices (for me, that’s the Tascam SD recorder and Adobe Audition). Most meters have a traffic light system of green, yellow and red zones. As a general rule, you want to see plenty of green for normal levels, touching into yellow if something really loud happens, like a cough. If you see no lights, it’s too quiet and the signal level needs to be increased. If you’re hitting red, it’s too loud, and you need to back it off.

A lot of software will show you the waveform as it records. This is basically a graph of the signal. The x-axis is central, like a typical sine wave graph you used at school. As you record, you should see a strong signal painted onto the graph – you’re looking for big hills, but keep it away from the top – no mountains.

When no one is speaking, you should have a flat line like an ECG when someone dies. If it’s not flat, you may have too much background noise creeping in, so check.

Hit record, watch the levels, and off you go. I tend to stop recording during breaks, but I’ll often just leave it running in switchovers when there are two talks back to back. Those moments of quiet when the mics are all still live will be useful later.

Conference venues are dark, so take a small torch (flashlight) so you can see to make adjustments once the lights are down.

Listen!

My top tip is to listen to what you’re recording. Sounds obvious, but unless you listen with a reasonable set of headphones, you’ll not hear any problems while you’re still in the position to sort them out. There’s been times when a laptop on stage has been sending out a low buzz that could ruin a recording – the sort of thing that wouldn’t normally be easy to hear over the PA. It’s usually easy for the sound engineer to just mute any input causing noise, but they won’t if you don’t point it out. It’s possible to remove background noise in post-processing, but much, much easier if you can fix it live.

You need some reasonable headphones or earphones to listen with. They need to be comfortable to wear for an extended period, and need to be able to reproduce a good range of frequencies so you can hear hisses and rumbles. I use a pair of AKG K271 MKII headphones – they’re what I use to listen to music at my desk and I just take them along. They’re comfortable enough to wear all day without my ears getting tired. I’ve also used a good pair of Sennheiser in-ear earphones when travelling light. They did a good job, but my ears fatigued more quickly.

If your recording software has the ability to add markers (in Audition, I press the M key) it can be useful to add a marker at any point in the recording where something of note happens. For example, if the speaker has a coughing fit, or there’s a few moments of mic trouble, or anything like that. You can then quickly find those spots later when editing.

At the end

After the presentations have finished, make sure you have saved the recording, and then power down your devices. Keep your torch to hand as you pack up to find anything you might have left on the ground, and be sure to thank the sound engineer for their help.

Hopefully you’ll have a good solid recording to form the basis for your subsequent edit, which I’ll cover next time.

Equipment list

Here’s a handy list of the equipment I carry. I keep most of it squirrelled together in a box, and just load it into my bag each time it’s needed.

  • Length of XLR male-female cable
  • Length of 1/4 inch jack-jack cable
  • Multi-socket power strip
  • Gaffer tape
  • Interconnecting cables for your recording devices
  • Spare batteries for anything that uses them
  • Computer
  • USB or FireWire interface
  • Optional secondary recording device
  • Headphones
  • Torch / flashlight

Next time we’ll talk able the process of editing and getting the final file ready to distribute.