🚀 Add to Chrome – It’s Free - YouTube Summarizer
Category: N/A
No summary available.
00:00
I'm about to show you how to create insanely realistic AI videos that look completely real using just simple text prompts inside Google's brand new VO3. And yes,
00:20
even if you've never made AI videos before. Hi, I'm Simon and I'm having way too much fun making this tutorial.
So, in the next 12 minutes, I'm going to walk you through everything you need to know to get started with VO3. From how to get free access to how to write those killer prompts that
00:36
going to get you the best and most consistent results every time. And then we're going to go through together and create three completely different types of videos from cinematic scenes with people talking, an eye-catching sales generating product promo, and a consistent character video, all with just a few lines of text.
Trust me, by the end of this tutorial,
00:55
you're not just going to know how to use V3, but you're going to know exactly why, in my humble opinion at least, V3 is one of the best AI video generators out there right now. And first things first, let me show you how to access V3 for free. Well, kind of free.
You'll need Google AI Pro,
01:13
which is around $20 a month. But here's the good news.
Google offers a free onemon trial, so you can test it out and you can cancel anytime. that unlocks full access to all of Google's most powerful AI tools, including V3. And once you have that, there are two main ways to access V3.
01:30
And honestly, Google just made these way more confusing and complex than it has to be. So, let me break it down.
First up, Gemini. Head to gemini.google.com and sign in.
This is probably the fastest and the easiest way to use V3. And yes, I'll drop all the links in the video
01:47
description below so you can access them easily. The other option is Google Flow via Google Labs. So if you just go to labs.google.com and flow is basically Google's prolevel film making tool with full scene control, shot planning, all the bells and whistles. But honestly,
02:05
it's probably overkill unless you're planning, you know, like a big project, uh, you know, with lots of multiple scenes and lots of moving parts. So if you would like me to do a tutorial on Google Flow in the future, let me know down in the comments.
But for now, let's stick with Gemini since you get the same stunning results without all the unnecessary complexity. All right,
02:24
so here's the quick tour. So Gemini's layout is super simple, and if you've ever used Chat GPT before, then you'll really feel right at home. So first, look for the video icon at the bottom of the prompt box, and let's select that.
That's what switches it over to V3 mode. On the left,
02:43
there's a sidebar. You can open and close it to check your previous charts or if you want to start a new chat.
And here's the prompt box. So this is where all the magic happens.
So you just type in what you want, hit enter, and Gemini will take it from there. And that's really all there is to it.
02:58
So right now, V3 gives you three 8second videos per day. So let's make each of those count.
So let's jump right in then and create our first video together. And we're going to kick things off with a full-on action scene that looked like it came right out of a Hollywood blockbuster,
03:15
complete with characters and speech, so that we can see what V3 can really do. Now, when it comes to entering your prompt, you can go as simple or as detailed as you like.
But here's the five-part prompt formula that I use to get the best and most consistent results. So, you've got
03:32
subject, so who or what is in the video. Action, what's actually happening in the scene.
Setting, so where the action is taking place. Style, so this is how it should look and feel visually. And then audio, so any sound effects, music or dialogue that you want included.
So using this
03:50
simple prompt formula, let's build this scene out step by step. So the subject for my video is going to be a military helicopter with two soldiers visible inside the cockpit.
For action, we're going to be weaving at high speed through a twisting canyon at sunset, dodging debris
04:05
and hugging the terrain. The setting, well, it's going to be a wide cinematic view showing the full helicopter in flight with dramatic canyon scenery. Style.
We want this to be tense, highstakes action thriller with cinematic lighting, intense camera shake from the turbulence. And then finally,
04:22
audio. Want co-pilot one yelling, "We're not going to survive this." And co-pilot 2 shouting back, "We'll see about that." with pounding orchestral score and deafening rotor blades.
So, here's my full prompt exactly as I'd type it into Gemini. So, this follows uh all of the
04:40
five-part formula. So, V3 has everything it needs to create a cohesive cinematic scene for us.
So, now let's hit submit and give V3 a couple of seconds to do its thing. All right.
And now, let's take a look. We're not going to survive this.
We'll see about that.
05:00
Okay. And honestly, you know what?
That's not bad for a single prompt. I think the helicopter looks fantastic.
The scenery is beautiful. It looks really realistic. The people could be improved.
The dialogue's a little bit off, but you can obviously, you know, regenerate this as many times as you like, but as I say, I'm pretty happy with that. All right,
05:19
now let's switch gears a little bit and create an ad that you can use to promote a real product. So, the kind of ad I'm looking to create here is something that looks professional, like studio quality, something that looks like it came out of a, you know, an expensive brand campaign.
So, using the same five-step prompt formula we used earlier. Here's our ad step by step.
For subject,
05:41
we've got a premium pair of wireless headphones on a sleek rotating stand. The action, I want a slow reveal from shadow into light as the stand turns, highlighting every contour. The setting is going to be a dark, moody studio with water rim lights and a light haze in the air.
05:59
And for style, I want ultra realistic materials. I want cinematic lighting, subtle lens flares, and kind of that luxurious luxury vibe. For audio, I'm going to go with deep bass heavy music track with a confident voice saying, "Hear every detail." So, pulling all of that together, then here's
06:17
our full prompt. Now, let's go ahead and hit submit.
And now let's see what V3 comes up with.
06:32
Okay. Well, it gave us exactly what we asked for. It's pretty simple.
Uh I do like the music. The earphones look fantastic.
The headphones, sorry. Um yeah, very very happy with that. Not bad for the first try.
Now again, what you can do if you like is go through and, you know, add more prompts
06:48
and edit it and change it, you know, depending on what kind of style you like, but for a product promo, you know, made in a couple of seconds with AI, really impressed. So, we've seen it nail cinematic action.
We've seen it create high quality product ads. Another thing that V3 does
07:05
really brilliantly at is creating consistent characters from one scene to the next. So, in completely different clips, even in completely different scenes and settings.
And this character could be could be yourself, it could be somebody else, or it could even be a character that you
07:20
yourself have created from scratch using AI. So, you remember in the intro when you saw me walking on Mars, riding through the Wild West, ironing on the streets of Bangkok, and enjoying a cup of tea, even in the middle of a zombie apocalypse.
So, let me show you how to do it. And it's easier than you
07:37
might think. And it all starts with a photo.
If you already have a photo that you want to use, you can just go ahead and click add photo, upload it there, and you're good to go. I'm going to do things a little bit differently, though.
I'm going to create a brand new character image right here
07:53
inside Gemini. And then we're going to use that to create multiple consistent character videos in V3. We're in V3 video mode right now.
But instead of just typing our video prompt straight away, we're going to create a still image first because this is the photo that we're going to use to keep
08:09
our characters consistent across different scenes. So to do that, I'm just going to click the image option here under the prompt box. And that switches us over to Gemini's image generator, which is actually powered by Imagen.
And now we just go ahead and describe the type of character
08:25
that we want. Um, you know, you can be as specific as you like.
You can, you know, whether you want to add details about the clothing, expression, and style. That will all help V3 kind of keep that character locked in later.
So, for this demo, I'm going to make three completely different
08:41
Pixar style characters, just so you can see how versatile this really is. So, first up, here's my prompt.
a full body Pixar style 3D character of a curious young adventurer wearing a tan explorer's jacket, satchel, and cargo pants with messy hair and a big friendly smile,
08:59
bright colorful shading, large expressive eyes, cinematic rim lighting, and a white background. Next, we've got Pixar style 3D character of a quirky inventor. Mid30s, wearing round glasses, rolled up sleeves, and a tool belt with gadgets. Warm colorful lighting, exaggerated facial
09:19
features, large eyes, and a soft focus background. So, I used Chat GBT to help me come up with these. And finally, Pixar style 3D fox wearing a casual hoodie holding a coffee cup, cheerful expression, soft shading, bright colors, studio background with subtle gradient. And honestly, I like all
09:40
of these. I think they all look amazing.
And you can swap in any style or character you want. I've gone with Pixar style, but you know, if you want to go with ultra realistic or Lego, anime, fantasy, futuristic, cartoon characters, sci-fi, aliens, like whatever fits your videos. So,
09:58
I'm going to go ahead and I'm going to create a consistent character video with our young adventurous explorer here. I'll hover over it. I'm going to click the download icon to grab it in full size.
And now, let's switch back to video again from the prompt box. And that'll open
10:14
up a new chat, which is totally fine. And this is where we now add our photo.
And then we write our video prompt. So here's mine.
Adventurous explorer pushing through dense jungle foliage at sunrise. Golden light streaming through mist. Lens flares. Cinematic tracking shot from behind, then panning
10:32
to the side. Vibrant green tones.
Explorer says, "Wow, this place is amazing. Soft ambient jungle sounds with distant exotic bird calls.
Let's go ahead and hit submit. Let's take a look.
10:50
Wow, this place is amazing. And this is just absolutely perfect.
As you can see, it's the same face, the same outfit, and it's in the same style. And now, using this same image, I can go ahead and create as many consistent character videos as I like. So, I've gone ahead and made a few more
11:09
scenes with our little explorer here, just to show you how far you can push this. So, here's our little explorer in an alien desert.
Here he is crossing a snowy Arctic landscape. And here he is inside ancient temple ruins.
Same character, four completely different worlds. And V3 nails
11:30
the consistency every single time. And that is Google's V3 in action.
So now you know exactly how to create these cinematic scenes, these high-quality product promos, and these perfectly consistent character videos every single time from a single text prompt. But Google's next AI,
11:48
Genie 3, just takes things to a whole new level. Rather than just creating static images or videos, you can create interactive worlds you can change in real time. I mean, it just completely blew my mind.
So, watch this video to find out more about Genie3, and I'll see you in the next one.