Workflow Question: Audio Documentaries

Question

Hello there,
I am trying to figure out the most efficient workflow for 2 types of audio pieces that will end up in a podcast:
1) The first type is one long interview (an hr) that will be extensively edited (e.g., sentences and words rearranged; "ums" deleted, sentences shortened and tightened ) and inserted into a podcast with some narration and sound fx. I'm unsure if I should be marking (so I know what's where) and editing the interview in the waveform or multitrack. In the waveform, marking is easy; I can break the entire clip up into a series of chunks. But then what? Export them all as individual files? OR Export the entire file with the markers into multitrack? The problem with the first solution is that I then have a tons of files that are out of order. The problem with the latter is that any subsequent edits I make in the waveform puts the clips out of sync in the multitrack. Another option is to edit it all in the multitrack, but it seems that there's less ability to use features like zero-crossing and the markers apply to all tracks not individual ones. Suggestions?
2) The second type of piece is similar to the first, except it involves 10 or more different interviews that are going to be edited into one piece.
Any tips? Suggestions? Advice? Workflow resources?
Any and all input is really appreciated.
Best,
David

Answer 1

Steve's speech editing 101:
Long rambling interviews are a huge problem from this POV - if you did the interview yourself, you generally learn pretty rapidly that actually structuring the interview itself whilst recording it is a pretty good idea! So is the concept of getting the interviewee to include the question in the answer - this generally makes editing a whole lot simpler.
De-umming and -erring in waveform view is fine - you won't need to redo any of this, whatever you do next. So you can do this, and save the resulting file as a first pass (always keep the original though - you never know...). At this point, as you've realised, you have options. The only thing I can sensibly do is tell you what I do when presented with this sort of thing; it makes sense to me, and you may just about retain your sanity. And yes, the whole thing revolves around markers. The first thing you do is to identify individual chunks of interview, and mark each end of them. What you do then is turn these into a marker range, and most importantly give this a name that means something. There's more than one way to go about this - if it's really complicated then give the ranges numbers, and have a separate reference list. If you think you can put adequate information in the marker, then do this - but bear in mind that when it's all scrunched up, these aren't always so easy to read. The basic idea here is that, if you can manage it, these are like paragraphs - a short collection of related sentences, preferably based around one idea only. You don't want to keep them too long though - I tend to break longer collected sections into separate sub-numbered sequences.
The most important thing as far as marking is concerned though is to have a structure for it, especially if you have multiple speakers in different interviews. If you keep to a unified scheme, you should easily be able to identify the speaker and the particular point being made easily. And that means that you should be able to assemble a final piece without going completely mad...
Anyway, what you do then is batch process these marker ranges into separate files, using the marker name for the file name. And now, you can import the whole lot into multitrack view, and start to assemble your interview. At this point it should be fairly straightforward; you can reassemble the clips in any order you want relatively easily, and with a bit of luck, you won't have to do too much editing during assembly - although that option is still open to you.
So to sum it up, the key to this is organisation and structure. At first it won't seem like you're doing any 'editing' at all - except that really you are - you're doing the most important bit. This is a conceptual thing; editing isn't about cutting out the bad bits, it's about assembling the bits you want. And when you think about it like this, hopefully the foregoing will make some sense. Hopefully it will also be obvious that this scheme works for either single person or multiple person interviews.