How To Make Realistic-Looking AI Films

In this video, I am going to show you

how to create realistic looking AI films

using one of my favorite platforms which

has emerged in recent times. Here is

what we are going to be creating.

[music]

He is a prince by blood. [music]

In the land of the red collar, the man

who will be king lost Everything

[music]

[music]

[music]

power or riches.

Why must I choose? Then leave this

place.

You will never wear [music] a crown

here.

Your kingdom lies under the hooves of

your horse.

[singing]

He met a man who could turn into a woman

and he was willing to help.

I will give you the world, son of man.

But for every land you owe me a blood

debt.

100 horses,

100 cattle.

Quasida.

It's done.

Beware,

the earth is hungry.

If you do not stop to eat, the earth

will eat you.

He conquered many lands

and for every land he conquered he

planted [music] his bloodline like seeds

in the soil. [sighs]

But he never stayed to watch the

harvest. Master,

we do not want to fight.

We want [music] to offer you these

weapons called guns.

[music]

I accept your gift.

From this day, your people will forever

be our playmates.

[music]

Those of you who want the guns can have

them. [music]

I would continue to fight with my spear.

>> [music]

>> The Ashanti have a far larger Amy.

We have enough land.

Let's go back.

>> There is never enough.

We much forward.

>> We have fought these brave men for too

long.

They humiliated us in the past, but

today we shall end the war.

>> His body can only be penetrated by a

stone from the river.

[music]

>> [music]

[music]

[music]

>> What is this place?

My leader, [music]

we are at the river.

Call it beep

[music]

for I have become a heavy load.

>> This is the legend of the

[music] brave men smiler diura chapa.

[music]

Just few months ago creating shots like

this was virtually impossible.

Especially for this shot where the guy

turns around and we are able to see his

face. If you stick with me till the end

of this video, I am going to show you a

new tool that came up that makes this

possible. Before the end of the video, I

am also going to show you some new

additions that they have done exactly as

at the time where we just finished

creating this video. The platform we use

to create this awesome AI film is

Hicksfield AI. Once you go on Google and

just type in Hicksfield, just select

this option and it will take you here.

[snorts] Now once we are here in

Hickfield, let's just go up here. As you

can see up here, we have Cinema Studio

and then all we did was to click on it

and then it opens this menu. So down

here is where you write your prompt. And

when you press this plus icon, it gives

you the chance. And then you can click

this upload images button here. And then

it's going to give you the chance. In

our case, we selected our character

sheet which we already created earlier.

Then we clicked on open. It's going to

give you this warning. Just go ahead and

click on agree to continue using. And

now the image has been brought in. So

let's go ahead and write a prompt here.

All right. And down here we can control

how it looks like. That's the aspect

ratio. And this would also give us

control over I mean the resolution. And

now when we go to the right hand side,

this is where the amazing stuff are. Why

do I say so? Now when we click on it, we

would realize that we can change

different aspects of the camera. Now we

can select the type of camera we want to

use to produce the images or videos

overall. In this case over here, we can

change the lens. So as I keep scrolling

my mouse, we can see that the numbers

keep changing. And right now I've left

the 35, right? And then this is the type

of lens that we are going to use. In my

case, I've left it at compact

anamorphic. Later on, I'll come back and

change it to something like the helios

swirly lens because I like those swirly

bokeh that you see at the edges of this

images. And so, we select the type of

camera here. And as we keep scrolling,

it will give us different forms of

cameras that we can choose from. And

then to our extreme right over here, we

can also change the aperture of the

lens. So quick recap, we can select a

camera, select a type of lens that we

want to use, and then also select the

focal length of the lens that we want to

use and then also select the aperture

that we want to use. This is absolutely

amazing and as at the time of recording

this video, I haven't found this

anywhere on any other platform just yet.

At this point, let's go ahead and bring

in the character sheet which I had

already created earlier. And all we need

to do now is to go ahead and select our

character sheet that we have over here.

We can see all the different angles. So

right now, let's write a prompt. So the

man is walking through

an African or let's say an As ancient

African market with many sellers selling

red cola. Edit it a little. Don't mind

the spellings over here. So 2K

resolution is fine. I don't want to bend

through too much credits over here. And

then it's currently on 16 by9 over here.

So I can just click on generate and it's

going to generate this image based on

the camera settings, the lens and all

that we have chosen. And voila, when we

click on it, this is what appears. The

image looks really, really cinematic. I

like the blur in the background, the

blur in the edges. It makes it looks so

realistic. So let's come back here

because I want him to dress white. So

dressed in white clothes with many

sellers selling red color, right? So we

just added white clothes. So in this

case, I really want to be able to change

the clothes he's wearing to hide so that

later on when we get to the war since he

will now um you know wear the other one.

So he should not be holding the spear,

right? So [snorts] we are taking off the

spear as well so that he's just going to

dress in white and walk through the

market or standing in the market. And

this is the image that we get.

Absolutely amazing. Once we click on it,

this is how the image actually looks

like. And when we click on this three

dots over here, it's going to show us a

range of options, right? Or we click

here, watch next. It takes us here

multi-shot. And this is a gamechanging

feature for me guys. Once we click on

this multi-shot, it is going to generate

different angles. And if you have

followed this channel, it is one of the

things I do a lot using Nano Banana on

platforms like LMA because I try to get

these things for free as much as I can.

[snorts] Even though this is a paid

tool, this just simplifies the process

and streamlines everything. So let's

click on multi-shot and what it's going

to do is that it's going to create the

different angles and it's going to be

amazing. So we give it some time and

after processing in a while voila

different angles of the same shot guys

and is just amazing. Look at the first

image, the second, you know, the third

one with him walking this way. And then

this one where he's touching the red

collar, I mean, and this one where he's

standing from the side. And the other

one from the back. Simply amazing, guys.

So, let's come back here. Let's change

the aspect ratio. In fact, I want to

take it to a 21x9 that is kind of like a

cinematic one. So, let's click on this.

And then the movement. Once you click on

this movement button and you can select

different type of camera movements. So

you don't need to describe it. You just

need to select it and it is going to do

it for you. This is a well thoughtout

platform, right? So let's just go ahead

and write the man stands silently

looking. Okay, let's make some

adjustments. Yeah, so the man stands

silently and looking. Right. So let's

generate the shot over here.

>> [snorts]

>> And it is actually in progress. And it

is done. Right. So once we come here to

try to play the video, we just click on

the play button over here. And this is

what we got.

Really amazing. And one thing about

Hicksfield I've noticed is that the

object permanence is really good. So the

lens aesthetic and everything stays. So

once we click on the button down here,

we will be able to download the video,

right? So basically, you just go through

any of the angles that you want. You

click on the footage or you click on

that picture and then you can click on

animate or download. In this case, I'm

just going to click on animate here.

Then it loads it onto the timeline for

you to be able to generate your image.

So in this case, let's select the

movement that we want. So I am going to

select this uh dolly out where the

camera moves in. So the man stands

silently and looking. Just a simple

prompt. Let's click on generate and then

give you some time. is going to go into

Q then it will start um it will start to

generate and this is what we have and as

we can see the AI has forgotten how this

man looks like and that is the problem

with AI videos as at the time of

recording this video AI easily forgets

right however there is a new king in

town there's a way we can solve this

without any major hassle it is called

cling 3.0 and we can use it all on this

Hfield platform. On the Hixfield

platform, we just come up here. Once we

click on video, it is going to give us

this particular menu and once we scroll

through, we will see Clink 3.0. And once

we click on it, it is going to open this

particular menu. Now, what is special

here about Clink 3.0? We going to find

out shortly. Now, just like every basic

AI, it has start and end frame. And if

we come down here, we can select the

model, making sure that we are actually

on cling 3.0. And down here, we can also

select the duration. One amazing thing

about this model is that we can actually

generate clips more than 10 seconds or

within this particular model. So over

here is where we write our prompt. So

let's go ahead and add this particular

image where the man's back is turned

towards the camera. And I'll show you

one amazing thing how we got the face.

So over here is this green button. I

call it the next level button or the

game changer button. Once we click on

this multi-shot, it is going to give us

the chance to be able to generate

different angles of the same video. It's

going to do the editing and everything

for us. In our case, for this first

example, I want to show you how I got

the AI to remember exactly how this man

looks like. Right? So we come down here

to something called elements and in

elements is going to allow us to be able

to create a character where the AI

understands how the person looks like.

So we are going to build a whole model

of our character. So once we come in

here, we're just going to click on the

plus icon and it takes us to this

particular menu. And in this menu, we

can click on the plus icon here. So

we're going to upload his face, the

front and the back. So let's upload the

front first over here. And at least you

need more than two images to make this

work. Let's give him a name. So let's

say

Jaka, right? That is the name. And then

we have to describe who he is. Right?

But before that, let's go ahead and add

the next image to it. So let's find an

image of uh his back, right? And then we

describe him. So a tall African man,

right? And we just click on generate.

And after some few minutes, it has

generated the character here. And to use

him in any sort of generations that

we're going to do, we just so click on

it. Make sure it has been ticked. And

then down here, we just click on use

element. And the element is going to be

loaded in the chat box. Now we can type

anything and the AI will remember who

this man is, exactly how he looks like.

So when we click here, we can actually

adjust the duration of each particular

shot. how it is going to look like.

Right? In our case, it is just because

we want the AI to remember the face.

Now, let's go ahead and write a prompt.

The camera is pushing in

from the back as Jaka turns

around looking at activities in the busy

market. So, all we need to do now is

just proceed and click on generate. And

now it's going to process. We give it

some time and truth be told as at the

time of making this video these things

can drag. It can be a little slow which

can be a little bit annoying but of

course it was a new tool as at the time

of recording this video. So probably

over time the pressure on it is going to

reduce and generation times are going to

be a little bit faster. So, let's give

it some moment to still keep processing.

And then after a while, when we look to

the right, we can still see the prompt

that we use in creating the video. And

it is ready. Once we play it, he's going

to turn around. Does it remember him?

Perfect. Perfect.

The AI remembers our character exactly

and how he looks like. For comparison,

we need to go back and look at the

original AI models we were dealing with

before this point, what it created. So,

let's play them side by side. And as you

can see, the one at the top doesn't

clearly remember who this character is.

It's even showing us, you know, a fair

skin guy. But that is not who the

character is. But as we can see, Clink

3.0 O does an absolutely

great job by making sure that our

character is our character and this has

been a major issue in AI films. Now it

has been resolved with this cling 3.0.

So it means our character can run

around, he can turn around, he can do

whatever he or she wants and we can

still have him looking like the same

exact character. So, what I'm going to

do here is that I'm just [snorts] going

to come and then click over here at the

first frame. And then I'm going to

upload this particular image of the army

matching, right? Just to give it

context. Now, let's have it there. And

then we are going to go turn on our

multi-shots. My most favorite button in

the world right now, guys. And then I'm

going to go to the first shot. [snorts]

And we're going to write a description.

Right. Let's keep adding more shots. We

add another shot down here.

And then we can keep adding more. So

what I'm going to do is that I'm going

to reduce the timing of the first one

from 9 seconds to probably let's say

somewhere 3 seconds there about right

for now let me leave it at five. And

then when I go down it's going to give

me the chance to keep creating more

shots. So in this case each shot is

going to be 33 seconds eventually.

So let's go back up here.

Let me still reduce this further to

probably 3 seconds and that will allow

me again to add more shots which means

at this point every shot is three three

seconds throughout. So let's come to the

first shot. I'm going to write thousands

of soldiers marching across the savannah

dust rising like a storm. So and then we

go to the next shot. I just write Japa

throwing his spear. It lands in the F.

The map overlay shows territory turning

red. So now let's go ahead and add Jakpa

so that the AI will remember who Jaka

is. But remember the only Japa we have

right now is the Japa wearing the white.

Now we need Jakpa the warrior. So we now

need to go back and build a character

model of him in that costume holding the

spear. So let's go ahead over here.

Click on elements and [snorts] then we

click on create new element. Right.

Click on the plus icon over here. Then

let's select his full view holding the

spear. And then we go back to windows.

Then crop just his side view as well.

And then once we are done, we can go up

there. Trim here a little. Then we go up

and then we just save it as a separate

file, right? Using just the normal

Windows um picture viewer, right? Give

[snorts] it a name to full size. And

then we can just click on save. Now

right from there we come back here. We

click on plus and we can find that same

image import it in. And once it's

imported let's write a name for it. All

right. Jakpa with spear. So that we can

it will serve as a distinction between

that and where he's wearing the white.

So let's write a description. An ancient

[snorts] African

warrior. Right? Just something as simple

as that. Right? Let's click on generate.

Give it a while. And now the AI is going

to understand how this jackpot with a

spear looks like in with his full body

all in view. Now it is ready. Let's

click on it over here and then we click

on use element. So when we come back

here, Japa has been added. So [snorts]

it is at the wrong position because we

are writing and we want to mention his

name, right? So let's pull it and send

it up here. So now it is going to read

Japat throwing his spear. it lands in

the F right so a map showing or a map

this map overlay I want it to be a

separate shot so let's select it cut it

and then we take it to short three so

that it will just be a shot on its own

so a map overlay shows territory turning

red on the fourth one we write

silhouette of swords clashing against

the sun setting

and then we move on to the fifth one I

already have the information so I paste

here Japa placing a younger man, his son

on a wooden son on a wooden stool,

right? So we take off the japa because

we want to use select the actual japa

from here and it comes in. So we

rearrange it to the top as well. And if

you take a look, these are all the

descriptions that we have given for all

these shots that we intend to get at the

end of the generation. So we just go

ahead and click on generate. It is going

to start processing and then we give it

some time. After a while it is done and

looking at all the number of shots it

has edited it the sound and everything.

So this is no editing it has pieced

everything together and it does

understand context. This looks very very

interesting. Now the main issue here is

this particular shot where [snorts] the

spear kind of disappears. I don't know

exactly what that is, but I'm suspecting

that to a certain extent because this

thing understands context and the fact

that I have explained that this guy is a

very powerful African warrior. Probably

I am just assuming that maybe it just

assumes that this guy is that powerful

that his spear can disappear and all

that. Don't quote me on that. I am just

assuming. But if that is a glitch as

well, it looks interesting. One other

thing that is worth highlighting that I

know most of you are wondering is how we

got the characters to talk. That is the

lip sync or the dialogues that are

happening in the piece. Now we achieved

everything still within Hickfield

Studio. Now when we come here all I did

was that we use the same platform and

then just added the text in the process

by just making sure that sound is turned

on whilst generating. Let's say in this

particular case let's click on the image

icon and then we come to this grid here.

Let's select this particular one and we

click on animate. Now we want to get

this character to talk. So I've written

the man opens his eyes and asks a

question in his native northern Ghanian

accent. He speaks slowly in a weak tone.

What is this place? Then space call it

for I have become a heavy load. And so

if you look at this prompt it is divided

into several places. I have the first

part where I'm describing the action.

what is he doing? Which is the man opens

his eyes and asks a question. Right? And

now I went ahead to now describe what

type of voice who is he. So I now said

in his native northern Ghanaian accent,

right? He speaks slowly in a weak tone.

So I'm also describing the tone at which

the voice should speak or the voice or

the dialogue should sound at that

particular moment in time. And then I

now go ahead and put what he's saying in

quotations. It's not really necessary,

but sometimes just to be safe, you can

put it in quotations. That is exactly

the words that you want him or her to

say. Right now, let's go ahead and just

click on the generate button. Give it a

while as usual. And then it's going to

generate. And then once it is done,

let's play it and listen to it.

>> What is this place?

Call it Gbby Bipe

for I have become a heavy load.

>> Amazing. Just exactly at the time at

which we completed or we were just about

finishing this project, Hicksfield

updated the Cinema Studio. So, we now

have Cinema Studio 2.0. And oh boy, they

cooked as always. They've made some very

little adjustments. That may look little

but they are very monumental and I like

Hicksfield for how they keep innovating

and thinking rationally about the actual

problems of AI filmm and how they solve

them. So if we look back over here at

the top now it's been updated to Cinema

Studio 2.0. And so when you click on it,

it works just exactly like we were using

it earlier, right? Same similar menus

and all that. But a new addition here is

we can select the grid that is the

number of iterations that we can see on

one grid that is different angles and

all that and they've also incorporated

the at button here remember so now that

I've clicked on it we can see all the

japa characters that we had now when we

click on the video icon here it takes us

to video we can [snorts] select single

or multi-shot right or we can also

select manually or manual shot

generation. So once we click on here as

usual that is where we select our aspect

ratio. Then quality here will allow us

to select 720 or 1080p.

And over here we can select or upload

any sort of character that we want to

generate a video of. And of course the

movement button has also been moved

here. And then we have speed ramping. Do

you want to have normal footage? Do you

want to speed it up? You can select

different different options over here.

linear and all that. So, we just take it

back to custom and auto. Over here is a

duration and we can increase the length

of the videos that we wish to generate.

And this is simply amazing and

unbelievable. And of course, our

generate button is over here. Let me

know if there are any other platforms

out there who offer this because for me

this is the most thoughtful platform out

there for now that I have discovered

which thinks about AI filmm and the

problems we are currently facing and

proposing and putting out features that

actually solves these problems that we

are all bedeled with when making AI

films. If you have any further

questions, let me know in the comment

section below and I'll be addressing

them. And if this is your first time of

seeing my face, welcome. This is African

Amaze channel. Over here, I do talk

about cinematography, photography,

editing, and all its related branches

and testing out all the new AI tools and

creating awesome content with them. If

all this sounds interesting, I do much

appreciate it if you do. Click on the

subscribe button. Please make sure to do

turn on the post notification bell so

you are notified anytime I drop awesome

content like this right here from the

continent of Africa. If you want to know

more about Hickfield, the cost and all

that, I have attached links in the

description of this video. Kindly go

check out Hickfield. I cannot recommend

this platform enough. And if you are

here and you do not have money to do

this, if you are looking for free

alternatives or how to do this for free

even though your quality is drastically

going to drop, click here to watch this

video as I share with you how to use

free AI tools in creating awesome

content as well. Until next time, as

always guys, keep practicing film

making. Peace out.