Editing tools

Recording

Interview editing steps

Create a transcript

  • Start with a recording that has both voices on at once
    • You get this by default with a Zoom recording
    • in Squadcast, they’re separate by default, but you can get both voices in one recording like this:
      • Click the down arrow to the right of your session (default is Reusable session)
      • Select the two you’re interested in (using checkboxes on left)
      • Click Mix Recordings
  • Upload the mixed recording to Temi (use the big button on the home page)
    • Create a transcript

Create first pass of section headings

  • First get a feel for the broad outline
    • SKIM the transcript
    • Quickly insert headings broadly describing each section - these will have descriptive titles rather than the T001, etc you’ll be using later.
    • Use heading 2 (so it doesn’t conflict with all the heading 3s you’re going to use later on)
    • Keyboard shortcut in Word to create heading 2 is Ctrl + Alt + 2
    • Think about which sections you want to cut
      • This way you won’t waste ages listening to those sections in detail - you can just ignore them from the start
      • You can use Ctrl + Alt + S in Word to strikethrough the sections you want to lose (In Google docs, it’s Alt + Shift + 5).
      • It’s also worth suffixing headings with y, n and m for Yes, No and Maybe to indicate whether sections will be kept or not.

Make a note of timestamps

  • Now start inserting timestamped labels in Audacity and Word
    • Have Audacity and the transcript open side by side
    • Do two passes:
      • On the first, just create an Audacity label + timestamp for each of your broad headings that you inserted before.
        • Go all the way to the end of the process - ie calculate total timings in the spreadsheet.
        • That way you can get a pretty good idea of what you want to keep / discard.
        • When you number the labels, leave gaps because you will be adding more granular labels later
        • For instance, T10, T20 etc (use your judgement re how big the gaps should be - leave bigger gaps than you think you’ll need)
      • On the second, add more granular timestamps, but only in the sections you’re keeping
    • Whenever you reach a point that may mark the start or end of a clip, give it a label (I just give them labels like T002, T003)
      • In Audacity:
        • Use Ctrl + mouse wheel to zoom in/out - but only one scroll click at a time! Otherwise it goes haywire.
        • click X to pause the recording there (means it will resume at the same point, which it wouldn’t if you used spacebar)
        • Ctrl + B to add a label
          • To remove a label:
          • click in the label text
          • Home
          • Shift + end
          • Delete
          • Delete again
        • Alt left/right to move between labels
      • In Word:
        • Insert the Audacity label text at the correct place in the transcription text (create a new line)
        • Add a space, and then the timestamp
          • Not in brackets
          • Include the hour even if the recording is not over an hour long (So 3m 21s would be 00:03:21)
          • Don’t include fractions of seconds
        • Add a space, and “y” for keep (= yes) or “m” if you’re not sure (= maybe), or “no” for don’t keep (= no) (followed by strikethrough)
        • It should look like this (the formatting matters for the spreadsheet later on): T020 00:03:21 y
        • Ctrl + Alt + 3 to highlight the label using heading 3 (makes it easier to find later)
          • !! Watch out though! Somehow all my text ended up being in Heading mode, even though it looked “normal” - and I have no idea how it happened. It seemed to happen after I was using Outline mode. I think maybe it happened because I started editing when I was still in Outline mode? Anyway, keep an eye out for this. The quickest way to check whether it’s happened is View => Navigation pane and check that all your label headings are appearing separately.
        • Ctrl + Alt + S to strikethrough passages you want cut from the recording
        • Ctrl + Shift + N to undo the heading style (goes back to “normal”)
  • Make sure you create one last label which marks the end of the recording.

Use timestamps to calculate rough episode length(s)

  • When you’re done, transfer timestamps to a spreadsheet to get an idea of length:
    • In Word:
      • Click View => Outline
      • Make sure that “all levels” is selected from the “Show level” dropdown
      • Use Ctrl + Alt + Shift + S to bring up styles dialog
      • Click the dropdown next to Heading 3 and choose “Select All XX Instance(s)”
      • Ctrl + C to copy them all into clipboard
    • In Excel or Google sheets:
      • Have headings Label, Timestamp, Include? and duration
      • Format the duration column as a time (select col, Format => Number => Time)
      • Paste clipboard
      • Data => Split text to columns => choose Space
      • Now in the fourth column (“duration”), use the formula =B3-B2 in the top non-heading cell, then drag down into all cells in this column (except the last, which should mark the end of the recording)
      • Now create min and max time totals (wherever you like), with the following formulas:
        • Min: =SUMIF(C2:C300, "=y", D2:D300)
        • Max: =SUMIF(C2:C300, "=m", D2:D300)+[cell containing min total]
        • (you’ll have to format those cells as times too - Format => Number => Time)

Audacity: Pull snippets / audiograms / teasers out into a new track

In Audacity…

  • Use File => Import to load your wav or mp3 file into Audacity.
  • Top tip before you start: Hold the Ctrl key down and use the scroll wheel on your mouse to zoom in on the waveform
    • But it’s slow to react, so don’t scroll quickly like you normally would
    • Instead, scroll ve-ery slowly - just one notch at a time - and wait for it to react after each notch
    • There are a few other useful Audacity tips here
  • Highlight the relevant section manually, or use start and end times:
    • Note start and end time
    • Place your cursor in the source track
    • Select “Start and end of selection” from the dropdown at the bottom
    • Enter start and end time
      • I do it like this:
      • Place cursor at the start of the snippet
      • Use Ctrl + B to add a bookmark at that point (labelled something like “B start”)
      • Place cursor at the end of the snippet
      • Use Ctrl + B to add another bookmark at that point (labelled something like “B end”)
      • Make a note of the time stamp (you can see it at the bottom, under “Start and end of selection”)
      • Use Alt + left to move back to the previous bookmark
      • Now the timestamp of the first bookmark will be in both boxes under “Start and end of selection”
      • Edit the numbers in the second box (the end of the selection) to match the timestamp you just made a note of
      • Now the selection will match the two bookmarks (it will be highlighted in white)
    • Press play (spacebar) to check it worked
  • Use Ctrl + D to copy the selection out into a new track
  • Do NOT use Effect => Truncate Silence to move the snippet back to the start of its new track
    • Instead, just zoom in and highlight the empty section before the snippet… and use Ctrl + X to delete it.
  • Press Solo on the new snippet track
  • File => Export => Export as mp3