my beef with transcriptions in fcp – warranted?

[Adam Rose esq.]

FWIW, can let the timeline play in vegas while transcribing, either in separate app, or within vegas itself.

🙁

can then export to excel etc.

[Chris Bové]

From the angle of a doc editor, the industry has reached the point where it really doesn’t matter what system (Avid, FCP, Premiere…) is used… except for the handling of scripts and transcripts.

Be it FCP or whatever, nothing right now is as good as Avid Media Composer’s internal script program and their relatively new addition of ScriptSync.

https://www.avid.com/scriptsync/scriptsync.asp
https://www.scottsimmons.tv/blog/2007/07/19/avid-scriptsync-demo/

Perhaps not for this production of yours, but consider renting an Avid for the next one – even just for the interview-assembly stage.

______
/-o-o-
`(=)`/…Pixel Monkey
`(___)

A picture says 1000 words. Editors give them meaning.

[Mike Cohen]

I believe Premiere has a time code filter you can apply to clips, so you see the clip’s timecode as it plays. Find yourself a good transcription service that can either transcribe timecoded video, or simply add a beep at one minute intervals and have the transcriptionist note the beep in the transcript – you can then add the approximate times to your transcript, import to excel or whatever.

I am starting such a project with about 8 hours of interviews and may try this method, and will no doubt blog about it.

As far as making notes within an editing app, if an Avid is not feasible, you might work out a system using text slates, or even screen shots of 1 minute chunks of the transcript imported and superimposed over the video segments, as if you are working with index cards or whatever convention you use on your paper edit.

Here’s another idea. Given the 1 minute chunks of video/transcript – you can easily break up video on a timeline into 1 minute clips, to match the transcript. Then rename the clips with the numbering system.

Then edit your transcript to rough out your desired video edit, maintaining the clip numbering.

Then rearrange the clips on your timeline to match the paper sequence. You will still have to manually view the edit and trim your edits to match the sequence in your script, but it may be a way to eyeball the edit without having to manually find the rough edit points.

Now if only this could be done Minority Report style with a touch screen – the way Lightworks was supposed to revolutionize the computer user interface back in the 80’s.

Let us know what works, and i too will report back with my process.

Mike

[Ross Neugeboren]

Well if it makes you feel any better,
I’m working on a transcription plugin for FCP. – Though I’m so tied down at the moment I can’t give a solid timeframe for its completion.

————————
Ross Neugeboren
GetTheShot

[Timothy J. allen]

I’ve been going through hours of interviews for a project right now. Since I like to work out of the office and on several different operating systems, I needed a way to transcribe using my standard office PC laptop – which has no editing capability (other than Quicktime Pro)

(It doesn’t solve your timecode notation issue, but since its related to transcribing, I’ll go ahead…)

I exported my interview audio out to itunes and put the tracks on my ipod. Then I found a quiet place and opened up Microsoft Word on my PC. I turned on the “speech to text” feature in Word and plugged in my USB headset microphone, put one earbud from the ipod in my ear, then just repeated the interview into the headset. (Since I’ve used a voice prompter before, repeating my questions and the subject’s answers doesn’t seem that odd to me.)

Anyway, since I ran through a few training sessions with Text to Speech, my transcripts have been more than 95% accurate on the first read. Using the standalone ipod for audio, I can pause and scrub through the interview without having to switch applications.

I’m not that slow of a typist, but using speech to text cut my transcription time in half.

I always play through a second time to correct any errors from where Word “hears” me wrong. Usually, it’s just a phrase or per paragraph or two. By the end of that second run, I pretty much have the interviews memorized and the transcript is just as accurate as if I had sent it off to be done professionally.(I must learn faster when I hear things and repeat them than when I type them, because when I was simply transcribing the interviews by typing them, the content didn’t really stick with me like it does when I enter the text with speech to text.) The accuracy probably is a result of me actually being the one that was asking the questions in the original interview session – since I know the lexicon, I can easily correct things that those not familiar with the conversation might not understand right away. Our interviews tend to be pretty technical and the transcription services I’ve used in the past sometimes have trouble with all those weird NASA acronyms 😉

Since part of the reason I’m transcribing is to get familiar with the content, I’ve been saving the $100 per transcript fee that I had been paying and doing it this way for the past couple of months. So far so good…

I still have to say that I’m missing the capabilities of my old Avid Adrenaline HD these days.

[Adam Rose esq.]

wow, nice tip, Tim

danke

🙂

[Michael Phillips]

Perhaps Transcriber may help?
https://www.videotoolshed.com/?page=products&pID=27

Also, with Avid ScriptSync as mentioned here, you don’t need to have timecode embedded into the DOC since the indexing is done phonetically 30x faster than real time.

Michael

Michael Phillips

[Celeste Geer]

Hi I find this approach interesting and am currently seeking a good workflow for my own very long form doco in which there are many 70 – 90 min IV’s. One thing i cant wrap my head around though is how do you get any kind of synch between the time ref on your ipod and the time ref on your IV footage in FCP?
Would be very grateful to hear your thoughts on this,

Celeste