Is this some new innovation in editing software?
Hold on George Lucas, don't get too excited. Let me explain.
Let's say I have a rough cut of a surgery (or any documented event for that matter).
I have cut this to the best of my ability, without any sort of Vulcan mind meld to know exactly what the client wants. Now that would be cool.
Next step is to have a meeting with the client, to review the video together. Let's say that the client has such limited time that the only time available to even review the edit is face to face for 1 hour a few days after the edit was completed.
Step 1
View edit on laptop - point video camera at laptop, lav on client, roll tape.
Step 2
Do not stop tape.
Step 3
Review video - client narrates important sections. Some of the narration is perfect, some segments need to be done several times to get just right. The narration is ad libbed, so if the first try does not work, we pause the timeline playback and try it until it sounds right.
Step 3a
In addition to narrating important segments, we also have some stream of consciousness editing decisions. Here's an example:
Client: This part is good, take a few seconds of that.
Me: Ok.
Client: We don't need any of this, we can skip ahead to the posterior dissection.
Me (hand on mouse scrolling timeline): How about this part?
Client: Yeah, that looks good. Right THERE - that's perfect - show some of that, super the artery right...here. And then maybe I should say the following: "Here we have a, um, good view of the critical structures and the, ah, the, um, instrument of choice here is the waddayacallit-o-matic."
Me: Can you say that again.
Client: Same thing but smoother.
**Sometimes rather than saying "Can you say that again without the ums and ahs," all you have to do is ask them to repeat something and it stands a decent chance of coming out better.
Me: Great.
Client: Ok, now this segment, what should we do with this?
Client's colleague: Well, we may want to show it, but minimize it, and say something about XYZPDQ.
Client: Good point.
Etc.
So forth.
So on.
So at the end of the session I have about 1 hour of tape, with some usable narration, a bunch of editing directions, some with a lot of latitude (show some of this), all with a visual reference, since the camera is pointed at the screen. All the furious note taking in the world can not accurately recreate the fluidity and accuracy of the recorded edit session.
So there you have it, Voice Controlled Editing. Instead of the client speaking to a computer, he speaks to me, and using the visual reference recorded to tape, I can leave the meeting knowing I have all my editing notes, without ever lifting a pen!
So now comes the editing.
Step 1
Digitize the video
Step 2
Since the lav was on Ch. 2, I use the Fill Right effect to make the good audio full track.
Step 3
I place the unedited video from my session at the end of my 1st edit timeline.
Step 4
Save a copy of the timeline as Edit 2 (even though no editing has been done to edit 1, this keeps edit 1 pristine)
Step 5
Make my audio edits to the raw video. I leave spaces in the edited audio, since it does not all go back to back.
Step 6
Copy and paste each segment of narration to the approximate point in the edit 1 portion of the timeline. I target a hidden video track and open audio track so the pasted clip does not interfere with the actual video in Edit 1. Notice I left the edited audio at the end of the timeline for the time being. I like to do one thing at a time.
Step 7
Save a copy of the timeline, and delete the edited narration from the end of the timeline.
Step 8
Save a copy of the timeline, and work on accurately lining up the new narration with the appropriate cuts in the edit.
Step 9
You guessed it - save another copy of the timeline. Version control in a complex multi-step process is vital, in case I screw something up, I never have to go further back than one step.
Now I need to decide how much extra video footage I need to retain from edit 1, beyond that which is exactly timed to the narration. This is where "use some of this and a little of that" comes into play, and my job as an editor and producer is to interpret these comments and attempt to follow the intent of the client. These nuances of editing go beyond rote matching up of video and audio. You have to know what you are looking at and know what the client wants. One hopes so anyway.
Step 10
Save a copy of the timeline. Now you can ripple delete, being careful that you have clips on the same track where you ripple delete, otherwise you can have clips line up improperly. Sometimes I select and drag rather than ripple deleting, for greater accuracy when there are many tracks involved. Once you have cleaned up the timeline, you need to watch it in real time, and fix any edits.
Step 11

With the correctly cleaned timeline, you then need to watch it again for content, and make any edits necessary to tighten up or lengthen shots. This version is 30 seconds shorter than the previous.
Step 12
This whole project is in fact a reshoot, which is replacing a similar finished section from a previously finished video. So all that is left is to make a copy of the previous finished project and delete the old section. Copy the finished content from the new edit and paste it into the open space on the copy of the original project. But we're not done yet. Since the two projects (old version and new edit) have different track layouts, the new material pasted into the old timeline did not line up exactly as far as active and visible tracks. So one more pass to check that video and audio appearing on each track are actually supposed to be visible/audible - there are numerous fixes to make, but nothing too serious.
With the final cleaning done, I can add whatever transitions are needed, fix any audio levels to match the old with the new, and watch it one last time. Not too bad at all.
Render out a WMV file so the client can review and hopefully approve online, and we can put this one to bed. Maybe.
Hope you learned something.
Mike
Very cool idea
Hi Mike,
That's a very cool idea. May have to borrow that for my event stuff.
At first I thought you came up with a way to use a PC mic to say the commands to edit the tracks in premiere.
I could not help thinking it would be a good idea if the voice recognition technology was really good. I gave up on voice recognition at Dragon Speak.
However, I'd have to say from looking at your articles 'you are the king of organization'
Bully!! ole boy
thnx
Joe