If BMD is looking for new acquisitions :)

I was watching this video and went what the hell

This is crazy thing to be able to edit like this

Anyone here love Descript??
When i login to tiktok and upload some of my work use resolve, most of they ask me why i don't use Capcut, best video editor ever in the world is capcut.
Every app has a unique thing.

Did she use the app to edit her YouTube explainer video... yeah, nah!

This would make a great plugin. I can see myself using it on some content I work on. Looks kinda of an apple-type product probably they will buy it and cannibalize it.

Ricardo Marty

The thing that always strikes me about a huge majority of YT videos, is that they are always but always sat at a desk, to present. This was something that disappeared from TV over 50 years ago! Nowadays with AF etc. surely they could be a little more imaginative in their presenting? Oh and the pre-title sequence, with the false bonhomie, 'Hey whatsup guys?', gets more than a little tedious. Is there a single template for these things?

The jump cuts make the video unwatchable.

More people need to get that.

The jump cuts make the video unwatchable.

More people need to get that.

You meant "make the video unwatchable for you, I assume?

I didnt have a problem with it.

Maybe you need to get that

Interesting technologies - the VO replacement is wild. Avid has a script-sync technology that helps you find the sections of video as they relate to a known script and so you can grab sections of video to use based on the script. The technology shown in this thread really takes that to the next level from what I'm seeing. The idea that you create your video as iJustine showed is not as compelling for me, but the idea that you could extract hours of footage (such as documentary interview footage) based on keywords and whole sentences, etc. could be very useful (creating clips that could then be edited into a timeline). Seems like the sort of extra toolkit that has its place.

Consider that the latest iZotope RX10 has a new feature that will automatically extract the text from audio and let you move things around and paste based on words, etc. - sort of a very rudimentary version of what we see in iJustine's presentation.

Interesting.

One of the cringiest stuff (besides AI art) that I've seen lately. So instead of doing a mouse click equivalent you type in a slash in some script? What time does that save? Especially when you still have to go and fix the stuff the Ai didn't "get right" afterwards and still fix the "gaps", basically saving nothing timewise.

And since editing is a visual flow kind of environment how can this be useful for anything except horrible hyper jump cuts. Even the "vo" part sounded robotic and cringy.

My 2 cents. Yikes.

I don't think editors need to worry about this. As good as the AI is getting, they are still far off from being able to puke out an edit that elicits an emotional response. This is great for boring YT vids and frankly I would maybe even use it for some point and shoot corporate gigs. It's kind of like Myspace and you can probably easily add flashing orbs and cheesy music. There is so much more to editing than a forward slash and an exposure slider can offer.

The jump cuts make the video unwatchable.

99% of YouTube videos are made with jump cuts.

Jump cuts are being exaggerated these days to show how .... (um, what is the current slang for "hip" or "cool"?) ... you are.

Jim ain't hip or cool. Well, I don't seem to be so either!



Sent from my SM-S908U using Tapatalk

The jump cuts make the video unwatchable.

99% of YouTube videos are made with jump cuts.

Yeah, awful and painful to watch, if I even watch them at all few seconds thereafter.

From a narrative or documentary editing perspective, it would be nice in DVR to bring in the script and have clips sort and line up with the script in the CUT page, then the editor can just go through the source tape and work on editing.

The jump cuts make the video unwatchable.

More people need to get that.

The footage is at least HD You can place b roll or do cu am mcu. very easy for the right content but its rent-ware so im not looking any further.


Ricardo marty

ok, i'm really old school, hate jump cuts, unless really necessary, but if your talent can't read a script for more than a few words before screwing up, you need better talent

however, i do have many hours of interview footage that i would dearly like transcribed. i have been using 'way with words' for years, but would like to know if there's any software that i simply load (upload) a video to and get back a transcription. i would prefer free, but some cost would be acceptable?

suggestions welcome,

stay well

ok, i'm really old school, hate jump cuts, unless really necessary, but if your talent can't read a script for more than a few words before screwing up, you need better talent

however, i do have many hours of interview footage that i would dearly like transcribed. i have been using 'way with words' for years, but would like to know if there's any software that i simply load (upload) a video to and get back a transcription. i would prefer free, but some cost would be acceptable?

suggestions welcome,

stay well

Have a look here:
viewtopic.php?f=21&t=168403

I was watching this video and went what the hell... This is crazy thing to be able to edit like this.

I would refuse to buy it purely on the basis that the person in the ad has a shrill, harsh, and strident voice and hurts my ears after :10 seconds.

I would refuse to buy it purely on the basis that the person in the ad has a shrill, harsh, and strident voice and hurts my ears after :10 seconds.

Me too. It hurt my eyes too. I don't know why this unique "text for cuting video" is great for them. For them, it is amazing. For me, it is just a very little thing of cutting.

Jump cut = Director screwed up during shoot!

I would refuse to buy it purely on the basis that the person in the ad has a shrill, harsh, and strident voice and hurts my ears after :10 seconds.

Me too. It hurt my eyes too. I don't know why this unique "text for cuting video" is great for them. For them, it is amazing. For me, it is just a very little thing of cutting.

Didn't mind watching it for a while as compared to all the other YouTubers (mostly males), for what it is worth she is somewhat of an eye candy.

SAS. The biggest reason I made the switch from Adobe to DaVinci. (Not the reason I stayed though.)

And the pricing on this SAS is absolutely RIDICULOUS!

I might have bought it as a plugin for some sort of social media special use case for $150 max, but that subscription price. Oh hell no.

however, i do have many hours of interview footage that i would dearly like transcribed. i have been using 'way with words' for years, but would like to know if there's any software that i simply load (upload) a video to and get back a transcription. i would prefer free, but some cost would be acceptable?

You can upload the footage to YouTube and then download the autogenerated subtitles.

roger - hadn't thought of that, thanks.

but, having watched a few auto generated subs on yt, i'm not so sure i'd be able to keep a straight face

@robert - thanks for that link, most interesting. after a quick perusal, it sounds lust right for my needs, however, the talk about python and installation is rather off-putting since, other than basic basic, and programming keystrokes, such things remain a complete mystery ;-(

however, it's certainly worth looking into. thanks again...

however, i do have many hours of interview footage that i would dearly like transcribed. i have been using 'way with words' for years, but would like to know if there's any software that i simply load (upload) a video to and get back a transcription. i would prefer free, but some cost would be acceptable?

You can upload the footage to YouTube and then download the autogenerated subtitles.

Leslie, I use this company for multi-language subtitles and captions, and video transcription. It’s not free. You can have humans transcribed for a much higher rate or use the AI automated transcriber for $0.25usd per minute. I believe the first 45 minutes is free.

https://www.rev.com/services/auto-audio-transcription

They say their auto transcriber is 90%+ accurate and if you’re interview audio and dialogues are clean and clear, I’ve gotten perfect transcripts from it.

@ ellroy - they're much like the company i've used in the past:[url] https://waywithwords.net/
[/url]

with the material i have to transcribe, i'm not that fussy about accuracy, it's just for building blocks, and since there's a great deal of it, i think i'll opt for any free option i can take with reasonable results. neither am i time constrained - though i can see light at the end of the tunnel, it's been, and continuing to be, a very long one

@robert - thanks for that link, most interesting. after a quick perusal, it sounds lust right for my needs, however, the talk about python and installation is rather off-putting since, other than basic basic, and programming keystrokes, such things remain a complete mystery ;-(

however, it's certainly worth looking into. thanks again...

So I did install and try to run it myself - after a bit digging around because it failed first to do anything I found out how to get it up and running with no fuzz and wrote you a how to:

Steps to get StoryToolkit up and running with Resolve:

Press WIN + S
Type in

Code: Select all

cmd

and select open as administrator

This opens a command line

Go to
https://community.chocolatey.org/course ... olatey#cmd

and click there the blue copy icon where it says "Install with cmd.exe" or get it from here:

Code: Select all

@"%SystemRoot%\System32\WindowsPowerShell\v1.0\powershell.exe" -NoProfile -InputFormat None -ExecutionPolicy Bypass -Command "iex ((New-Object System.Net.WebClient).DownloadString('https://community.chocolatey.org/install.ps1'))" && SET "PATH=%PATH%;%ALLUSERSPROFILE%\chocolatey\bin"

go back to the command line window and right click to paste the command
Press ENTER and when asked press Y for yes

After the package manger Choco has been installed, close the command line and open it again as administrator

Copy and Paste this command

Code: Select all

choco install ffmpeg

Press ENTER

Wait till installation has been finished. Now ffmpeg has been installed onto your computer.

Download StoryToolkitAI.0.17.5.WIN_archived.exe
https://github.com/octimot/StoryToolkit ... ag/v0.17.5

Maybe put it into a not too deep folder structure.
Double click to extract the archive.


Install Phyton just with the offered default settings:
https://www.python.org/ftp/python/3.9.1 ... -amd64.exe

Now start Resolve and load a timeline.
Go into your StoryToolkit folder and double click StoryToolkitAI.0.17.5.WIN.exe

This will open a window with several buttons.
Click Transcribe Timeline to start with the process

wow! robert, thanks so much for your walk through...

must say i'll approach it with some degree of trepidation, but approach it i will

will probably be next week, all things going according to plan.

again, thanks!!!!

well, everything seems to have installed correctly, but i don't know where to find the actual storytoolkit?
if i open resolve it's not listed in any of the menus (?), and if i click the exe in the storytoolkit folder (installed in c/programs/story...) all i get is a command window with blinking cursor - no buttons?

i have my fingers crossed that i'm just being dumb?

Leslie, have you also installed Phyton?

And to use the tool make sure Resolve in the settings has scripting set to local.
Then start Resolve first and Storytoolkit afterwards.
The tool won’t be listed in Resolve. It should open a window on its own.

hi robert, thanks for your patience...

yep, followed your instructions to a T. incl. phyton - however, i hadn't done anything in resolve settings since i didn't realise there was anything to change

not at my rig at the moment, will try tomorrow.

btw. so clicking story... .exe would have produced just the blank cmd window i was getting?

again, thanks

ok, back at my pc - checked local scripting. loaded interview to time line, started story... but still getting blank cmd window:

cmd.jpg (140.72 KiB) Viewed 2644 times

btw. i have to use explorer to go to story... there's no menu item for it? i put it here -

C:\Program Files (x86)\StoryToolkitAI.0.17.5.WIN

what am i missing?

thanks

ok, was getting so frustrated i simply started from scratch again, only this time:

a. i installed python directly from the windows store (yuck!) it was a later version 10.?

b. installed storytool... onto my d: drive (rather than c:) and created a shortcut to the exe on my desktop.

c. double-checked that ffmpeg was correctly installed.

d.turned the whole thing off and went and had a coffee.

e. came back to an active cmd screen - ran a 45 min interview with two people and not only got a word-perfect transcription, but bloody srt for subtitles to boot!

i am truly astounded - bloody amazing.

have a very happy whatever season you subscribe to, because i'm going to

stay well

Hi Leslie,

I'm stuck where you were, I think, and don't understand how you solved it with steps d and e.

I installed everything, shut down my computer. I start Resolve, then start StoryToolkit AI, I get a cmd window showing running storytookit, ffmpeg found, using cuda tor Torch/Whisper, then after about a minute the cmd window goes away and there is nothing happening in Davinci. I don't see any new button panel. I've installed latest version of Python.

Also, sheesh, what a lot of hoops to run through to get to this point! Thanks, Robert, for the walk-through, even though it's still not working for me. Am I supposed to walk away and drink coffee to get it to work?

-Stephen

If you have a Nvidia GPU, you just need to download the windows version, and have ffmpeg installed and visible to the app. He shows away of doing that, but if you already have ffmpeg setup as an environmental variable you may not need to
https://github.com/octimot/StoryToolkitAI/releases (Only Nvidia GPU's on Windows)
https://www.thewindowsclub.com/how-to-i ... windows-10

Stephen, I had to wait for another coffee while the CMD window connected to download a language model or something. did the window disappear or simply close ? mine remained open while downloading. took about 20 minutes.

not at home at the moment, but will follow up once I get back.

If you have a Nvidia GPU, you just need to download the windows version,

Why does a transcription program need a specific GPU - that makes no sense to me?
Is this more AMDpartheid.

If you have a Nvidia GPU, you just need to download the windows version,

Why does a transcription program need a specific GPU - that makes no sense to me?
Is this more AMDpartheid.

Because NVIDIA has the far better developer environment for this kind of application.
Everything AI related can make use of that for accelerated computing.

It's up to AMD to provide an alternative here.

And you can always use your CPU, it's just slower.

Back to the jump cut issue and watchability, if you look closely at the statistics for complete viewing of most pieces the completion rates are down in the low single numbers, and the average watch times are 2-3 minutes.

ok, back in the studio...

here's what i get following starting resolve > loading timeline > starting story...

Screenshot 2022-12-16 133901.jpg (96.26 KiB) Viewed 2313 times

are you sure you've got scripting set to local?

Anyone here love Descript??
When i login to tiktok and upload some of my work use resolve, most of they ask me why i don't use Capcut, best video editor ever in the world is capcut.
Every app has a unique thing.

how to install capcut?

So I finally got it running.
Rendering the first transcription as we speak.

But I have some questions.

1. When I tried to save my preferences, all I got was a "Error: invalid API token"
Any ideas what that is?

2. When I transcribe a timeline, I put all other audio tracks than dialog on mute - cause why make it more difficult with music and FX - right? Looks like all tracks got turned on again for the wave rendering.
Is this the way it should work?

3. I got a German Timeline, but need English subtitles. Do I press Transcribe Timeline first and than Translate to English, or can I just press "Translate Timeline to English" in the first place?