Our Ravenskil project has been ticking along for some years now and has reached the stage where the majority of the voice recording is complete and we’re starting to pull the episodes together. There are eight of these in total, each coming in at thirty minutes in length – so you can imagine the vast amount of raw audio we’ve recorded in the form of takes and re-takes (and out-takes) which needs sorting through and editing into a final product.

Gill Dean and Leighton Williams – two of our voice actors

Due to hardware and (lack of) studio constraints, voice actors were recorded individually. We’ll take a look at the H4N and associated hardware setup in a later blog post, but suffice to say that we ended up with this sort of thing:-

(Tom Butterworth ran a close second for the role of Emily…)

The next stage was to transfer it to a PC and sort through all the different takes of an actor’s lines for an episode and select the best one. In some cases where the line has been recorded with a number of different inflections, both/all might be taken so that the one that fits best can be used.

Normally plastered across two monitors…

Each line was exported as a single audio clip which could then be inserted into the appropriate character track on the master recording. We then inserted some ambient noise and sound effects for each scene to form a rough draft of the complete episode, so that the original clip now sounds like this:-

Although it sounds much better this is still not the finished product – for example in this clip Emily’s voice still has too much “room reverb” due to the recording being made in a room with a hard floor. Once the entire episode has been put together to this standard it will be edited down to the correct run time and then passed to one of our friendly neighbourhood sound engineers to tweak and fiddle with until we have the finished product. So until then it’s back to work!