I have used Audacity to edit the podcasts on The-Nexus.tv since December 2011. When I built my new computer in October 2012, I was able to edit three times as much audio as before. My workflow is efficient as I can know how to make it in Audacity right now.
But REAPER is different. REAPER is a DAW, or a Digital Audio Workstation. It makes Audacity look like the primitive and antiquated application it is. It allows me to edit without changing the original audio file, basically using my beloved diff to track changes, it has an interface that doesn’t scream Windows 2000 and best of all, I think it makes my workflow faster in the end phase. (Oh, and I like paying for nice things.)
I edited At The Nexus #59: Nothing Happens Intentionally last night entirely without Audacity and completely with REAPER. It took me four hours to experiment and actually get show and its associated fringe produced, but as with all new products, it takes a while to learn and the investments come back quickly.
In Audacity, my workflow follows:
- Record the entire show in a single project (file), and label various points with these markers: fringe start, show start, show end, title, fringe end
- Upon ending; save project (file) and export FLAC of RAW audio
- Remove background noise (coil whine, fan noise, furnace noise, any low decibel noise)
- Normalize the audio (literally,
Effect > Normalize)
- Compressor the audio at default values (in the past, I’ve used Chris’ Magic Audio Compressor, but I like default things built into Audacity more often than not)
- Normalize (yes, again)
- Clip the show between the start show and end show markers into a new project (file)
- Perform a truncate silence (with a threshold of -40 for a normal show, or -35 for a special show [or really long show])
- Edit that: title, show music, leads, end notes, and so on
- Clip the pre-show fringe and the post-show fringe sections into a new project (file)
- Between the two clippings, and some noise to indicate show change
- Perform a truncate silence with a threshold of -35
- Export each new file as FLAC
- Export each new file as MP3 at 64kbps CBR
This workflow has it’s advantages. The Normalize-Compressor-Normalize scheme brings the volume down a bit, the compressor shifts the audio so quiet bits are louder and loud bits are softer, and the second normalize cleans the form up again. The noise removal before splitting the show and fringe allows both episodes to have a consistently removed noise level, likewise, the NCN scheme keeps the audio consistent across files too. The real processing step is truncating silence. On shows I think are really long, I am harsher on the silence, otherwise, I find -40 to be a reasonable threshold (oh, and it’s default). Saving this for last keeps the show markers in place, which as it turns out, is insanely important if you’re going to use them.
I know my workflow in Audacity so well that it takes me maybe twenty minutes to produce a show and it’s fringe, but then another twenty minutes to actually render (see export to FLAC/MP3) them. Audacity’s biggest short coming is that it offers no way to enqueue multiple types of export (two MP3s, two FLACs). I have to wait for each one individually and then start the next (if I remember, and if I’m even there) export.
REAPER works differently. You never change audio when processing it (well, we’ll get to the nuances another time). Most processing effects happen in realtime. For instance, the noise removal tool allows me to find some noise, listen to it, build a profile, and then that’s it — I don’t have to process the entire show again right there. Audio is built on demand and I hear it when I need to. Likewise, when I run normalize or the compressor (which is a travesty of its own for another day), I just have to enable and then it’s done — there is no pre-processing and the original audio is completely intact.
I made the mistakes of a beginner unfamiliar with his tools in episode 59. We’ll see if you can hear them. Next time, I’ll detail the steps I take in REAPER to produce the show and explain the finer points of setting settings without explanations.