Free YouTube Transcribe

Video transcript

Ghost AI - Module 2 3

Drew Trainor · 2,288 words · 11 min read

Want to search this transcript, jump the video from any line, or download it as TXT, SRT, or VTT?

Open in the transcript tool

Full transcript

0:01Okay guys, now we are moving on to your

0:03second ghost which is going to be the

0:05mini do style. Now this is the one that

0:07I love because it's super simple to

0:09make. We don't have to worry about

0:11having the lip syncing happening and at

0:14the same time these work even better in

0:17my opinion. Um so it can be either in

0:19first person narration or it can be in

0:21second person narration. But let me

0:23explain what we have to do.

0:26Okay. So the structural difference is

0:28this. In the UGC, uh the character talks

0:32in the mini do the character acts

0:34silently and then there's a narration in

0:36the background.

0:39So the narration script is pretty much

0:41the same idea, right? So hook, problem,

0:44solution, call to action. It's always

0:46those four steps. Same eight frames,

0:50different B-roll. Okay? So this works

0:53really well. If you put them into

0:54different scenes, you can spend a lot of

0:56time kind of going, "All right, in in

0:58scene one, you know, she's in the

1:01kitchen grabbing a coffee. In scene two,

1:03she sits down at her computer. In scene

1:06three, uh, she jumps up for joy." You

1:09know, it's it's it's a lot of fun to

1:11have to have ideas for the different

1:14scenes because it doesn't really matter

1:16to make their mouth match, right? So, in

1:18the background, that narration is

1:20happening and they work really, really

1:22well. So voice over voice over and

1:24assembly. So first off we've got to go

1:28get the narration done. All right. So we

1:30have to go

1:32down here. Just going to find this

1:39prompt.

1:48Okay. So it's right here. So 2.2 to

1:51narration prompt. So, what we're going

1:52to do is we're going to grab this

1:58and it's split up a little bit like the

2:01last one. So, we've got to grab it in

2:02two parts.

2:04We're going to come back into claw

2:06because we want to keep that memory of

2:09all that other information.

2:12And so, in this case, remember, I can go

2:15ahead and

2:25take this stuff,

2:30use the mission brief

2:35above. So, I don't need to fill that in

2:37manually. And I'm going to come back and

2:39grab the rest of the prompt.

2:49I'm going to paste that in and I'm going

2:51to get the narration.

3:03All right. So now we have the narration.

3:06So first year ebook state blah blah

3:08blah. Okay. Great. Perfect. So now we

3:11have to go back and we're going to do

3:12the same process basically as last time

3:14to get the images. So the image prompt

3:17is here.

3:19I'm going to copy in the mission brief.

3:21I'm going to copy in the narration

3:25as I create my images. So we're going to

3:27use

3:28Meta AI again for this

3:31because it does a really good job of

3:33just giving us the eight images that we

3:35need.

3:37Before

3:42I grab the mission brief and and

3:43narration, I got to get the rest of this

3:45prompt here.

4:00Okay, so that's all in there. So, now I

4:02got to go grab the

4:04mission brief, which is going to be way

4:06at the top somewhere.

4:10It's good to have these things kind of

4:11handy,

4:14which I did not. So,

4:19learn from my mistakes. Grab the mission

4:22brief.

4:24Put it here.

4:27And then we got to grab the narration.

4:31which is down at the bottom now.

4:43Put that here and we get Meta AI to

4:46create our

4:48eight images.

4:51All right. So, while that's happening,

4:53I'm going to explain to you a little bit

4:55more about Vibes.ai

4:57AI because this is the tool that that's

5:00really really good for this because it's

5:01unlimited. Make as many videos as you

5:03want and as an example it does a really

5:07really good job. So what you're going to

5:09do is you're going to create a new

5:10project while the images are coming.

5:14Give the project the title.

5:16So this is mini do

5:19video.

5:25And what we want to do is we want to

5:28create each scene at a time. So we're

5:32going to go start end frame and we're

5:36going to have like we had last time 1 A

5:381 B. So I would put in 1 A and I would

5:41put in 1 B and I would describe it and

5:43that is basically the B-roll of the

5:46narration. Right now the other thing

5:49that we're going to be doing is using a

5:51free tool for text to speech. When I

5:54discovered this tool, I was pretty

5:56excited because a lot of the text to

5:58speech stuff out there is, you know,

5:59like 11 Labs, for example, is is pretty

6:01expensive. So, this is 100%

6:04[clears throat] free. It's called

6:06speechma.com.

6:08And if you come here, you're going to

6:10have to watch an ad first, and that's

6:11how it's free. Um, there's tons of ads.

6:14You're gonna have to deal with that. And

6:16then you can create as many text to

6:18speech prompts as you want and up to

6:202,000 characters. So you can go, you got

6:24to put this in first. So I have to put

6:26in that number 48

6:301 27.

6:33And what I would do is I would just grab

6:36my script.

6:39So, I'm going to say just give me the

6:42script

6:43by itself

6:46so I can copy

6:49and paste it

6:51into a

6:54text to speech

6:58app.

7:04Okay,

7:08now I come back over to Speechma. I'm

7:11going to put that in and I'm going to

7:12choose my voice.

7:16So, let's try Anna.

7:19Oh, that's a little girl, so we can't

7:21use that.

7:25And it's actually good quality, too.

7:29Okay, I'm going to use Ava

7:32and I'm going to go ahead and generate

7:35audio.

7:37Oh, I've got to do the number again. It

7:39timed out. So, we got to go 8

7:424 33 9. Oops.

7:478 4 339.

7:50And I'm going to generate the audio.

7:55All right, let's see how Meta AI is

7:57doing with the images. So, it got them

7:59done, which is perfect. So, I'm going to

8:03have to download these again and I'm

8:04going to have to rename them just like I

8:07did last time. And so, we'll come back

8:09for that. And then we're going to have

8:11our narration.

8:19There we go. So, the narration is

8:22perfect. So, in this case, it's second

8:23person. Um, you can also have it first

8:26person where Dana is talking as herself,

8:28but we're going to use this one as the

8:30example. So, I'll come back in a bit and

8:33I'll have the images ready to go and

8:34then we'll go ahead and create the

8:36video.

8:39All right. So, I've got the images

8:41downloaded now. They're not the best

8:43quality. Um, there's obviously going to

8:45be some some messing up with with Meta

8:47AI. It's just a free video create or

8:49free free image creator. So, what I'm

8:51going to do is I'm going to show you how

8:52I use Google Flow to get the images that

8:54are a lot better. This one works okay

8:56for the UGC one. Um, but when it starts

8:59to have objects in it, it's not the

9:01best. So, I don't want to, you know,

9:03tell you guys to to make it with this

9:05one because it just doesn't have great

9:07quality. So, what I did is I went to

9:09Google Flow and because I'm not using

9:10Google Flow to use to make the video, I

9:13can use some credits to make some better

9:15images. So, I used the exact same pro

9:17the exact process I did here in the

9:19demo. And I went over to Google Flow and

9:22and did it and it came up with these

9:25eight images which is really really

9:27good. So it starts from here 1 A 1 B 2 A

9:312 B etc. So that's what we need for our

9:35mini do. And if we go over to speech

9:38mode we have our speech ready, our

9:40narration ready. So, we would download

9:42this

9:45and then we would go back over to here

9:48and I'm going to give you

9:51two different tools that you can use to

9:53do this to make the video.

9:56One of them is called Snapjai and I'll

9:59just explain this here quickly.

10:04I think I already have an account with

10:06this Gmail.

10:11All right. So, with Snapjen,

10:15there is one video generator on here

10:17where you get unlimited

10:20amount of of uh generations. So, if you

10:23go into studio, what you're going to see

10:26here is you can pick the video

10:28generation provider.

10:30And if you go to Google Vio one,

10:36it says generation will cost zero

10:39credits per video.

10:41So it's 100% free.

10:45And then you can also use this one right

10:47here is also free. This one actually

10:49gets pretty decent quality. Says

10:51generation will cost zero credits. So

10:54it's unlimited. You have 10 credits when

10:56you sign up, but you're not going to use

10:57any of these credits. So this is a great

11:01little tool to be able to do this with

11:02and it has the option to do first frame

11:04and last or first image last image. So

11:07we would do 1 A and 1B and then we would

11:11give it the prompt which is essentially

11:13the B-roll. The other one is going to be

11:17vibes. It's exact same idea. So you

11:20start with a frame here. I'm going to do

11:22this one right now um for the demo. So

11:25I'm going to go and upload

11:43Just got to find that one a

11:51which is right

11:53here. It was created by Google Flow. and

11:57I'm going to upload that.

12:02The key here is we just want the video

12:06to be around 20 seconds to match the

12:08narration. And then we can also add go

12:11to Cap Cut, any video editor that you

12:13use. And we can add subtitles and we can

12:17add a little call to action at the end.

12:20That is not hard uh to do. So, we're

12:24going to pick this one. We're going to

12:26add it to video. And then we need the

12:28end frame. So that's going to be 1B.

12:34Upload that one.

12:45Okay. Upload.

12:49And we would do the same process with

12:51Snapchat.

12:541 A 1 B. And it also does a great

12:57result. So, and unlimited. So, we don't

12:59have we're never going to get a

13:02situation where we're out of credits.

13:05Okay. So, now what I need to do is I

13:08need to get back to the narration

13:11and I need to give it the scene.

13:14So, the visual

13:17right here. I'm going to come back,

13:21put that in,

13:23and I'm going to create the video.

13:28And it actually with Vibes AI, it also

13:30gives you four options to choose from,

13:33which is great as well because then you

13:35can choose the one you like the best.

13:48And I have my narration downloaded. So

13:51what I would do is, for example,

13:55I'm not going to go through the whole

13:57process for this because

13:59it's simply just piecing all the videos

14:02together. Vibes AI does also have a

14:07video editor editor built in. So it's

14:09got actually got a timeline right here.

14:12So you can add media into your timeline.

14:14So, for example, what I would do is I've

14:18got these generations made.

14:23So, for example, this one. And then you

14:25can choose. So, I like

14:30I mean, that's not bad at all. Well,

14:32that one's not great because she morphs

14:35into a different person, but that one's

14:37actually really good here. So, I like

14:39that one. So, what I would do is I would

14:41go to timeline and I would add media.

14:46Which one? I think it was this one here.

14:48Yeah. So, it just kind of clips to the

14:51other

14:53part of the scene, which is nice. So,

15:00there you go. And then what I would do

15:01is I would simply go back

15:05to generate again and I would do the

15:07next scene. Okay? So, I use 2 A and 2 B

15:10and I would add it to the timeline. So,

15:12it's going to be 20 seconds because each

15:14one is five seconds. And then you have

15:18your narration.

15:20So, if I go back here and I want to add

15:23audio,

15:25um, I can now see this here

15:30has not been working. Otherwise, I would

15:32honestly just do it all here. Um, I can

15:34try again right now. It was not working

15:36for me before. So, let me just see if I

15:40can get

15:42this to work.

15:46If this works, it's a lot easier.

15:49Now, see, it's TT TTS failed. So, the

15:53narration uh option is not working. So,

15:56unfortunately, we uh cannot use that.

15:58But we can do we can add text if we

16:01wanted. But what I would do is I would

16:02just put those four clips together. I

16:06would download them. Literally all I

16:08have to do now is do the next one,

16:10right? So I go back to

16:14go back to my project

16:17and I would do the next one. I would do

16:19start end frame 2 A 2B and then I would

16:24pick the video I like the best that it

16:26generated. I would come back

16:29to the timeline and would add that one

16:31next one to the timeline. Okay, so it's

16:35that easy. You're going to have your 20

16:37second clip and then you all you have to

16:39do is go to your video editor, put in

16:42your

16:44narration

16:46with the video,

16:48and then you have your mini do style

16:51video. All right, we're going to move on

16:53to the next one uh right after this.

16:56Talk to you in the next video.

More from Drew Trainor

Recently added transcripts

Browse the whole transcript library

This transcript was generated from the captions YouTube publishes for this video. Get the transcript of any YouTube video atfreeyoutubetranscribe.com, free, unlimited, no sign-up.