-
Art doesn’t pay
-
Transcripts depends on the project.
I am working on a massive doc, with 30 hours of interviews from 15 people, and counting. Transcripts allows me to review the content much more quickly. It also allows a search of words and phrases. Unfortunately FCPX is bad at getting to a specific word from a transcript. Yes, sometimes there are instances where a phrase on paper isn’t like as good in video, but with transcripts I can easily scan or search to see if the subject comes up again.
Also, since this doc is already a year in the making, transcript allow a quick refresh of the material, especially when it’s not the only project I’m working on. It also helps in keywording in FCPX much more quickly. SpeedScriber allows adding favorites so even as it’s being transcribed, I can start the organization prep. I can easily get from favorites to keywords in FCPX.
I don’t get transcripts on every job, but I appreciate it when I do get them.
And now that FCPX has a caption option, my guess is that we will get very quick transcript to caption handling (or text box) in the future, which will be very helpful and useful.
-
Ive worked on a few content style trailers with talking heads giving quips about the movie/show.
Transcripts became invaluable through the revision process. Clients will ask for stuff that doesn’t exist all the time and the transcripts allow you to at a glance see whats possible way faster than I could ever skim through the footage or review my “tags”. I need an “s” sound to turn something plural that was once singular I can quickly see all the words that may work.
In my normal day to day of spots and trailers, transcripts are null and void but we use dialogue breakdowns (every sentence of the movie/show subclipped and labeled with exactly what it says and by who) to sort of achieve a similiar approach to finding stuff and forming cheats,
-
[greg janza] “I have no need to tag if I can instantly search the entire transcript and with each search have an in-point set and edit it into my timeline.
“There we differ.
Look, I totally understand the benefits of transcripts. And for many types of (especially long form work) yes, they’re a HUGE boon. And I’m sure they’ll continue to be used and continue to drive lots of efficiency into many workflows.
AND it’s great that they are being automated and developing even more usefulness in the era of speech to text and more AI in our software.
But I just feel that text on a page, by itself, will always be a distance from the speech you’re trying to locate.
In processing interviews, not having the ability to range reject OR select something that “reads well” but doesn’t PLAY well – seems to me a step backwards from what I’m doing now. Scanning everything is NOT the process I want. I typically want to “reduce to the best” as my initial pass. THAT is what saves me the most time going forward.
Basically, for me, unless you’re encoding that type of interpretative data, you’re really just settling for a shotgun to shoot at a bulls eye.
But that’s just me. Each editor has their own needs and style, as do each different project, often.
And so it goes.
Creator of XinTwo – https://www.xintwo.com
The shortest path to FCP X mastery. -
[Bill Davis] “In processing interviews, not having the ability to range reject OR select something that “reads well” but doesn’t PLAY well – seems to me a step backwards from what I’m doing now. Scanning everything is NOT the process I want. I typically want to “reduce to the best” as my initial pass. THAT is what saves me the most time going forward”
I tend to work in a reductive fashion, too. However, when a client says “but, didn’t they say…” you have no answer without going back and reviewing the footage again. Maybe they never said that, or maybe you just rejected it and moved on and it’s now out of mind. Having a transcript – even if you use it only after the fact – helps in those situations.
– Oliver
Oliver Peters – oliverpeters.com
-
[Bill Davis] “Basically, for me, unless you’re encoding that type of interpretative data, you’re really just settling for a shotgun to shoot at a bulls eye.”
We’ll agree to disagree.
I guess if you primarily do sizzle reel type pieces the need for transcripts would be far less but for any story driven piece with interviews, transcripts are the blueprint from which all content is analyzed and selected. Even the two minute corporate promos I work on are made much easier by using transcripts because there’s a tendency these days to over interview folks and so often a corporate exec is allowed to pontificate for 30-45 mins in raw form by the producer. And if you have half a dozen interviews the time needed to look at those interviews in real time goes up exponentially.
It’s also possible that fewer producers are providing transcripts to editors and so maybe some folks have gotten used to making selects from just watching the content but I’d argue that it’s an inefficient use of edit time.
Windows 10 Pro | i7-5820k CPU | 64 gigs RAM | NvidiaGeForceGTX970 | Blackmagic Decklink 4k Mini Monitor |
Adobe CC 2018 |Renders/cache: Samsung SSD 950 Pro x2 in Raid 0 | Media: Samsung SSD 960 PRO PCIe NVMe M.2 2280 | Media: OWC Thunderbay 4 x 2 Raid 0 mirrored with FreeFileSync -
[Bill Davis] “In processing interviews, not having the ability to range reject OR select something that “reads well” but doesn’t PLAY well – seems to me a step backwards from what I’m doing now. Scanning everything is NOT the process I want. I typically want to “reduce to the best” as my initial pass. THAT is what saves me the most time going forward. “
It’s not an either/or scenario though. You can have transcripts in order to leverage what transcripts do best and you can have ranges, markers, keywords, etc., in order to leverage what ranges, markers, keywords, etc., do best.
[Bill Davis] “I think it’s time to look beyond the transcript. It’s pretty limited, when you think about it.”
I don’t think having a complete record of what people said contained in a small, extremely portable, shareable, keyword searchable, platform agnostic, universally understood, easily editable, printable format that allows any literate person to build edits, find pertinent information, give feedback, etc., from anywhere in the world can be described as “pretty limited”.
IMO transcripts are a high water mark for how flexible, transferable and editable all metadata should be.
[Bill Davis] “We’re all getting pretty universal transcripts soon enough if Google Speech to Text continues to develop. “
I think this gets more to the point of why some people don’t see the value in transcriptions because it’s an added layer of work and an added layer of cost. I imagine when we get the point of every NLE being able to generate accurate transcriptions upon media import and those transcriptions are automatically synced with the media (ala ScriptSync) that people will wonder how they ever got by with anything less.
Take it a step further and imagine connected apps/devices where someone could highlight sections of a transcript (or even do very blunt word processing) and those text selections would automatically show up as a stringout in the NLE. And you could even add keywords, markers, ranges, etc., to the transcript and those markups would automatically be added to the media in the NLE.
That would even further open up the editorial processes (though not necessarily editing itself) to non-editors and/or non-technical people. Or even make an editor’s life more pleasant. Imagine sitting down with an iPad and a cup of coffee on the couch and as you leisurely highlight and tag your transcripts. When you are done you walk over to your NLE (or your editor walks over to their NLE) and, based on your transcripts, all the relevant clips are tagged, sorted, strungout, etc.,.
Sure it’s not going to lead to a perfectly edited cuts but the goal isn’t to get a perfectly edited cuts out of it.
-
[Andrew Kimery] “Take it a step further and imagine connected apps/devices where someone could highlight sections of a transcript (or even do very blunt word processing) and those text selections would automatically show up as a stringout in the NLE. And you could even add keywords, markers, ranges, etc., to the transcript and those markups would automatically be added to the media in the NLE.
“Uh, you’re late to this idea.
Phil and Greg at Intelligent Assistance with —Builder — does pretty much EXACTLY this inside FCP X right now.
FWIW.
Creator of XinTwo – https://www.xintwo.com
The shortest path to FCP X mastery. -
[Bill Davis] “Uh, you’re late to this idea.
Phil and Greg at Intelligent Assistance with —Builder — does pretty much EXACTLY this inside FCP X right now. “
I’ve had the idea for years but I’m not a software developer so that and a couple of bucks will get me a cup of coffee…
Thank you though for beautifully illustrating my point about how powerful and useful transcripts can be.
-
[Andrew Kimery] “Thank you though for beautifully illustrating my point about how powerful and useful transcripts can be.”
In their niche, VERY powerful. However, their niche isn’t particularly large, IMO.
It goes back to the same old argument. Is the larger “professional editing” landscape more like the one you are imagining or the one I have to work in daily.
My friend Braden Storrs (who helms one of the larger FB groups X oriented groups at around 30,000 members – is currently running a poll about the primary use for X among their members.
Right now, Corporate/Business is well in the lead. Commercials and Ads are next, falling off about 30%. Documentary and Non-Fiction is in 4th place, with about HALF the practitioners of Corporate.
You can interpret that in LOTS of ways, but for me, primarily, it says that most people use X for business videos simply because that’s the client pool that invests the most broadly and consistently in acquiring content. It’s where more money is spent in more places.
Are transcripts useful in some of those areas? Of course. But for every interview driven piece where transcripts can be key – there are lots of scripted and interview free content where transcripts are meaningless.
People are basically following the money.
And I do think the Machine Learning and Algorythmic stuff like that which Phil and Greg are working on, will eventually get rolled into all the NLEs. So transcriptions will become just another software provided expectation.
And so it goes. ????
Creator of XinTwo – https://www.xintwo.com
The shortest path to FCP X mastery. -
Given the number of X seats in the world, even if all 30,000 of the Facebook group respond, it may say more about the demographics of that group than how X is really being used as it’s a tiny percentage of the millions of X installs. As for how useful transcripts are that surely is in the realm of personal preference, program style and the way you might have been trained or self trained.
Reply to this Discussion! Login or Sign Up