Continuing on the previous post of the perfect performance…

Until the time comes when all narrations are done with no need for correction, we should try to optimize the process. After all, the less time spent editing, the more recording we can do in a given time, or eventually the more money we can make. I like the sound of that, since Voice over Work hopefully can become a profitable exercise.

There is a lot of chatter out in the WWW of the nirvana of editing called “Punch and Roll,” or as some call it, “Roll and Punch.” The term comes from audio engineers of yore (I was one of those once) who used tape for recording. The artists would make a first pass of the performance, and if it was perfect, they would probably call it a day and go get some grog. But if there was a mistake, or something on which to improve, they realized it would be easier to re-record just the small bad portion rather than the whole performance.

The engineer would roll the tape of the original performance, and the artist would listen to get the timing, tone, and timbre matched to what happened the first time around. Just when the bad part rolled up, the engineer would punch the record button and the new performance would be recorded. Many times, I myself have gotten the punch in right (this on a Tascam 8-track recorder utilizing standard cassette-tape cartridges) and missed an important part – the punch out – and thusly created the need for another round of punch and roll. Sigh.

However, Audacity doesn’t really do punch and roll. I have seen posts on plug-ins or hacks that are supposed to add the functionality, but they didn’t sound promising or worth the effort to implement at this point. I do just fine with old school editing methods. (There is an incongruous juxtaposition there where old school refers to something old in the modern era. I’m using old DAW methods, which are much newer school than old tape methods)

Excluding PnR, there are two ways to manage your edits. I have tried both of these for various projects, and found that I like the first one – mainly because of my physical setup, not necessarily because it is more productive or efficient.

Record it all then clean it up.

When you make the occasional error, you don’t worry about it – just recollect and re-read the section over, as many times as it takes to get the take you think it correct. You may need to take a few deep breaths, a sip of water, clear your throat, whatever. Audacity is running, so when you are ready to go, so is the computer. Eventually, you will make it through the piece or you will run out of gas.

At this point, you stop the DAW, SAVE YOUR WORK, and start back from the beginning of the recording listening for the errors (or looking for them) and deleting the mistakes (and saving your work) until you have only the good stuff left! Then, save your work.

This process can be made much quicker by doing one simple thing – use some kind of noise (or lack thereof) that will be easy to see in the wave form or the recording software. I’ve tried clapping, snapping fingers, hitting the desk, tapping the mic (as a sound guy, I should say you should never tap on a mic), or the perennial favorite dog clicker – something that will record as a high level peak that you can scan for on the monitor. Or, long enough silence that it will show up easily. Once you have these visual cues, finding and fixing the vocal errors gets a lot easier.

Edit as you go.

This could be considered the poor man’s PnR. If you record in the same space as your DAW, when you make an error just stop the recording and position the record head to a break just before the error, hit record and start narrating immediately. The difficulty here is that you need to start recording quickly to avoid a pause you will need to edit out later (see process above), and it can be difficult to get back into the script effectively.

The hugest upside with this method is that when you are done, you have nearly finished audio. Of course, you will need to normalize it (as we already learned) and maybe some other post-production (which we will learn soon), but this is a fairly polished narration ready for the final touches.

It is rather gratifying to finish recording and not have to edit the file in another session. But in reality, you are probably still going to do a proof listen to find mouth noise, breaths, or other noises that need to be addressed, so you aren’t avoiding the above editing process completely.

What methods do you use? Has anyone accomplished PnR effectively in Audacity? Does Audacity have a preroll for recording? If so, I haven’t been able to find it. Let us know!

Categories: Voice over Work