Audio editting - Sermon recording
I've been recording Sermons at our local christian fellowship for over a year now (since August 2002). I thought it was about time I shared a little of the methodology used and gave a quick review of some software options open to you. This information is relevant to all types of live audio recordings, not just Sermons.
Sony Minidisc recorder
We have a Soundcraft Spirit M8 mixing desk at the fellowship building which offers 8 channels of inputs. Currently we just use 2 standing mics, 1 12-string acoustic guitar with pickups, 1 electronic keyboard and a radio mic for the speaker.
I attach a Sony 707 NetMD walkman (european model purchased 2002 for c.£200, picture here) to the direct out of the radio mic channel, and depress the direct out button on that channel. The direct out is then unaffected by changes to the mix (movement of the sliders) so I can adjust the volume without affecting the recording. I then go ahead and record the speaker. Sometimes I use the "mono sum" output on the board if I want to capture several channels at once; this board also has returns which you can mix the different channels on - only two of these are pre-fade and we use those for the stage monitors.
uploading to PC
The minidisc has a very annoying feature in that you can't digitally upload live recordings - this is a DRM feature. What this means is that you need to connect a miniplug-miniplug from the minidisc headphone socket to the microphone socket of your PC. Put the minidisc on charge while you do this step and you won't forget later - it might not charge up whilst being used, however
Programs to record from minidisc to PC
You'll need some software to record the sound from the minidisc player as it passes through the soundcard and out of the speakers. Some applications are:
- Goldwave - This is Windows only shareware ($40). It's very good (or at least v.4 was and has a long trial period. I've not bought this yet but may do if my Linux trials don't go well. It's fully featured and reasonably easy for basic editting.
- Sweep 0.9 - Open Source. Just started with this one. Seems to handle all the basics well and allows use of LADSPA plugins. Has a few dependencies so check first before installing. Allows writing of speex encoded files. Also all the usual formats including Vorbis.
- ReZound - Open Source. Not yet tested.
- Audacity - Open Source, available for MSWindows, Mac and *nix. Not yet tested.
- Protux - Open Source. Not yet tested.
- Nero Wave Editor 2 - This comes bundled with Nero Burning ROM. I've used the previous version of this extensively, it had very little processing ability but is fast for trimming the endpoints of recordings before I go on to mess with them further in, eg, Goldwave.
Commercial programs include Sound Forge from Sonic Foundry at over $300 and CoolEdit from Syntrillium at over $200. You don't need these for simple voice edits, maybe for commercial level sound production, maybe.
Record
Setting volume levels
You need to mess with a mixer now to ensure that the volume levels input to your computer don't peak outside the allowed range - if they do they get clipped and the sound is dreadful.
In MS windows you'll use the volume level tool or some version of it - this appears as a small speaker icon in your tasktray, double click to open and mess with the options until the microphone control is shown. Now press play on the minidisc player and adjust and volume settings so that the sound stays mainly within the green display (just like adjusting the gain on your mixing board), you'll probably want to use the advanced mic boost feature. OK?
With Linux use a mixer tool like kmix, rexima or alsa-mixer. The alsa-mixer comes with alsa utils and allows the mic boost feature to be turned on (use 'M' to toggle mute for the relevant section of the display). Make sure the volumes are set at appropriate levels by playing the minidisc and watching that the level doesn't hit red too much.
Now record
Easy, press play on the minidisc and then hit the record button on your chosen program. Fiddle with the settings a bit and play back a short section to ensure it's ok. If the sound is all shsshsh-y in the loud bits use the mixer/minidisc volume control to reduce the peak values.
... and edit
I use my editting program now to clear up the start and end of the recording. Usually my computer can't cope with the large file size so the very first thing I do is split the recording into tracks [Goldwave can do this at cue points, Sweep - I cut and past-as-new] of about 10 minutes in length. The tracks are easy to work with as the computer reacts faster and also when the CD is produced they create handy little chunks of listening, also Audio CD players don't seek well and can't easily restart mid-track.
Burning
File formats
A typical sermon recording is well over 100MB in wav format. Don't be tempted to store as MP3 (patented), OGG (open source), speex or whatever at this stage - these are compressed formats that throw away part of the sound and so degrade the quality. Keep the .wav format files - CD's generally use 16bit signed 44.1KHz stereo wav format but your burner may be able to convert to the correct format if it's not already in this format. Note .cda is not an audio format but is a marker used by a computer to find an audio track.
CD writing
Burn the files onto a CDR disc (eg use Nero, or k3b or xcdroast or whatever) and check them in your audio CD player.
Tips
- Don't get rid of original discs or editted wav files (of important recordings) until someone has listened to the CD through. You may have introduced errors that will need rectifying with the original copy.
- I usually burn the files as an image (.iso or .nrg) then burn CDs using the image - I can then clear the minidisc, convert the wav's to mp3/ogg/speex files for archiving and get rid of the wav's off my hard-drive.
- Having done the above, every now and again I'll write a bunch of archived files to a CDR.
- Always keep a master copy of sermon CDs so I can make a copy at a later date if necessary.
- Keep a record of all the recordings, dates, speaker, topic, relevant subject matter (such as which scripture it's based on).
- Name files with meaningful names - I use initials of the speaker followed by date and then either 'full' for the full length or 'tr001' for the track numbers, eg WG-30July03-full.wav or MT-23July03-TR002.wav.
- The largest files I've used have been c.350MB - in order to work with these on linux I used "xwmconfig" to switch my window manager to twm which uses less memory than KDE. I killed ("killall -kill appName") all apps I wasn't using and used "top" to check for daemons that I knew I could safely kill (check the man page, eg "man inetd") ... this seemed to improve the system performance giving me chance to split the files.
If there's anything you want to query in this information, you could try contacting me (especially if your part of the Church) and I'll see if I can help.
