What did I say?

A common problem we face when dealing with interviews is getting the stuff transcribed. Sure there are transcribers out there who can do the job, but many of them work in a world of audio cassettes and no timecode. We need timecode and we need to avoid time-consuming real-time dubbing of source material.

I’ve discovered a tool that should revolutionize our process of transcribing material…

Quite simply, the Start-Stop Universal Transcription System is the missing link for video and film transcribers. It does what any transcription software should: works with your favorite word-processor, can handle multiple, digital audio file formats, works with a 3-pedal foot controller, and so on.

What is important to grasp though is the particular dilemma we face in the post-production environment when it comes to transcription. Our world revolves around timecode. If you don’t know, timecode is the numeric address of any given frame on a professional-grade videotape. It is The Way all this digital editing works. Typically written as hours:minutes:seconds:frames (01:03:14:10), timecode is how we editors find bites or shots on a tape. You log the tape and reference everything by that code & we get the right shots in the right order to make your show.

Until recently most transcriptionist/typists have required analog audio cassettes to do their work. And in order to work with timecode, a transcriber needed to have a special gizmo that could translate audible timecode into a digital signal. Not a very complicated task, but time-consuming and a little dangerous. Usually, you’d have to playback all your camera original tapes on a machine that could output timecode as an analog, audible signal and record the audio on a cassette tape – one channel of audio and the other of timecode. Unfortunately, this usually means you need to send your tapes out to a post facility to have them make these cassettes. And they charge a goodly sum for this work. You’re in pretty good shape if you shoot on a standard, broadcast format like BetaCam SP or Digital BetaCam, but you have to get a pretty specialized VTR to hear audible timecode from a miniDV, DVCAM, or DVCPRO tape. The cheapest miniDV deck that can give you audible timecode costs over $5,000! When you’re shaving dollars off the production budget, wouldn’t it be nice to take advantage of new technologies and workflows that could save even more dollars?

Along comes an application like Start-Stop. Not only does it eliminate the need for audio cassettes, it can work with digital files – the same things you’re using on a non-linear, digital editing system. Granted you still have to capture the camera originals in real-time into some sort of system that can create those files, but as I’m about to explain, it allows you to double up on some effort to minimize steps (& time) later on.

The other key part (at least in our world) to Start-Stop are its time stamping capabilities. It seems pretty basic at first, but when you look at what it can do for you, it becomes a lot more exciting. The feature allows you to time stamp the document you’re working on as you transcribe. Again, no big deal if you’re used to the magic, audible timecode gizmo from earlier. But it eliminates the need to go through the dubbing-to-audio-cassette process – saving time, money, and wear & tear on your sources. You can enter an offset which can be equal to the timecode number of the first frame of the clip you’re transcribing and every time you tap the timestamp pedal – POOF! – you get the matching timecode to where you are on the tape. Granted, Start-Stop annoyingly requires an offset to be entered as a Total Seconds Count, it’s easy to figure out.

All right, enough with the sales pitch! Here’s how simple it is:

1. Digitize all material in whatever kind of chunks you want (5 minute blocks, 1 question & answer, etc.)…just get it ALL into the system. This is a big potential time-saver. Lo-rez is fine and you’ll then be all loaded for the editing to start, or at least you’ll have numbers & clips that can be culled and recaptured when you need to bring them in. You’re doing work you’ll have to do anyway – just in a different order. Think different!

2. From the edit system, export all clips as 22khz, 16-bit, mono, WAV files. This is what works, so stick with it. You could lower the quality of the exported clip, but you’ll likely be increasing the chance a word will be rendered {inaudible} or transcribed incorrectly.

3. Burn to a CD-R (This is a data disc, NOT an audio CD). While Start-Stop works with audio CD’s, you can get more info on a given disc using slightly compressed WAV files than a full-quality audio CD. 72-80 minutes on an audio CD versus 3-4 hours on a CD-R with above formatted WAV files. You can also use any media that can transfer large files (10-100MBs) or you can use an FTP or Website to transfer the files. Bringing film-making into, at least, the 20th-century if not the 21st.

4. For TC to be accurate, we must know the starting TC or total seconds count for the first frame of each clip. For the mathematically inclined: hours times 3600 plus minutes times 60 plus seconds. Most non-linear editing systems will allow export of a tab- or comma-delimited file. Export a bin with (at least) the Clip Name, Starting TC, & Reel Name. This file can be saved to the CD along with audio files. We found it most helpful to go in and append the total seconds number to the file name. Here is a Spreadsheet that allows you to enter the hours, minutes, and seconds and compute the number of seconds for you.

There you have it. Transcription for film and video enters the information age. Enjoy

{Special thanks goes to Patti Pancoe for sharing this tool with me. If you need a transcriptionist/typist who already has Start-Stop, contact her}

Leave a Reply