Free YouTube Transcribe

Video transcript

How to Use Seedance 2.0 on Venice Studio (Full Guide)

Venice · 5,124 words · 24 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

— Venice Studio: What's Inside the Interface

0:00The Venice Studio is here.

0:01And in this video, I'm

0:02going to show you how it works

0:03and we're going to

0:03make a fun little project.

0:04You can access the Venice

0:05Studio in the sidebar here.

0:07Just click Studio.

0:08And what you'll see is that we have

0:10individual workspaces

0:12for generating images,

0:13editing images, generating audio,

0:16generating video with an optional cue,

0:19which we'll get into,

0:20editing videos into movies

0:22on a timeline, add in titles,

0:25transitions, subtitles, text.

0:27And then finally, you'll see a library

0:28of all your media

0:29assets that you've generated

0:31that are being stored on

0:32your browser through Venice.

— Why Seedance 2.0 Changes Everything

0:34And the most exciting part about the

0:36Venice Studio right now

0:37is Seedance 2.0.

0:38So we're going to use the

0:39most powerful video generation

0:41model on the market here

0:42to make a fun little video.

0:58[MUSIC PLAYING]

1:01[MUSIC PLAYING]

— Using AI to Write Your Video Treatment

1:26So to get started on making a video

1:28harnessing all the power

1:29of the Venice Studio, first

1:30we need some reference images.

1:33Well, we're going to need an idea.

1:34So in order to do that, I'm just going to

1:36head over to the Venice

1:36chat real quick.

1:37And I'm going to use the newest open

1:39source model GLM 5.1.

1:42Help me create a video

1:44treatment for a 30 second cinematic

1:49trailer with multiple shots and one

1:53protagonist character.

1:56I need a full

1:57description of our character.

2:00Let's say a science fiction cinematic

2:02trailer, of course, right?

2:04I need a full description of our

2:05character for a reference image

2:08to use as a reference image prompt.

2:11Then I need image generation prompts for

2:13the scene backdrops.

2:15And then I need video generation prompts,

2:18which reference the character

2:21and the scenes.

2:23First, let's discuss the story.

2:27OK, so you don't need

2:28to do this step, but it

2:29is going to be the quickest way to get

2:31highly detailed prompts

2:32and just understand

2:33what you're trying to do.

2:34This is called a treatment.

2:35So you can see it's

2:36got some questions for us

2:37to get as deep and specific as we want.

2:40I like this seed A idea.

2:42A lone scavenger on a frozen toxic planet

2:44finds a buried alien structure.

2:46Sounds pretty sci-fi to me.

2:48Great.

2:49So now we have our treatment.

2:51We've got a shot list

2:51and our audio beat sheet.

2:54Now, these are not going to

2:55be the actual prompts we use.

2:57That's coming next.

2:58But the first step is just understanding

3:00what do we want to create.

3:01So now GLM 5.1 is asking if

3:03we want to make any changes.

3:04I said it looks good.

3:05So let's now continue and create the

3:07character reference image

3:08prompts.

— Reference Images: The Secret to Character Consistency

3:09Now, why do we need

3:09character reference image prompts?

3:11Let me show you here.

3:11If we go back to the studio, I'm going

3:13to open the studio in a

3:14new tab while that loads.

3:16And when we start

3:17generating with Seedance 2.0,

3:20we're going to use

3:20reference to video, R2V.

3:22And R2V means we add reference images.

3:26Now, this is not the

3:26same as a starting frame.

3:28This means we add a

3:29reference image, and we can tag it

3:32as image one runs through

3:34the deserted spaceship, which

3:38is image two, for example.

3:40So these are reference images, which

3:42we're going to use

3:42profusely in this video.

3:45Now, Seedance 2.0 is also available

3:48with starting and end frames, as I'm

3:50sure you're familiar with, where you just

3:52upload the first frame of the video

3:54and or the end frame of the

3:56video, add in your prompt,

3:57and it will generate the video like that.

3:59Or you can simply use Seed

4:00Dance text to video, T2V.

4:03And all you do is pop in the prompt.

4:05You type it out, and it'll generate it.

4:06But you won't have any

4:07control over what it looks like.

4:09So we generate prompts to

4:11generate our reference images

4:13so we can control

4:14what it looks like so we

4:15know what we're getting into

4:15before we spend the credits

4:17generating videos.

— Generating Multiple Character Angles

4:18So now we have all the prompts here.

4:20The first thing we're going

4:20to do is take this prompt,

4:22head over to the Studio to the Image tab,

4:24and then we're going to

4:26use multiple models here.

4:27I'm going to use Nano

4:28Banana 2 for one of them,

4:30and then I'm going to

4:30use, let's see, ZImage Turbo

4:33for the second one.

4:34We're going to generate two images here.

4:36It looks like it's

4:36going to cost us 10 credits.

4:39And let's send that off.

4:40Cool.

4:41So now we have two options here, Nano

4:43Banana 2 version and ZImage

4:45Turbo.

4:45Those are both pretty good.

4:47I want to go with Nano Banana 2 just

4:48because I like all these

4:49extra details, and I think this will

4:51be interesting to see how Seedance takes

4:53those details as well.

4:55Now we can see we

4:55have a session going on.

4:57So if I want to start generating

4:58different images or new scenes,

5:00then I would just start a new session,

5:02and I can revisit this in the studio,

5:04which is very convenient.

5:05So now I'm going to take our main image,

5:07and I'm just going to click Edit Image.

5:10And it's going to

5:10take us to the Edit tab.

5:12We're going to generate a

5:13total of three reference

5:14images for the character.

5:15So this is like a side angle here.

5:17Let's say change the

5:19angle to be a rear portrait.

5:24And I'm just going to

5:25stick to Nano Banana 2 here

5:26because it does a good job.

5:28But once again, if we want, we can see

5:29that in different models,

5:31maybe to save credits, maybe

5:33just to see what's different.

5:34There's the rear image.

5:36We'll come back here.

5:38We'll edit this again.

5:39We'll say change to a 90

5:42degree side profile angle.

5:44And that's really as simple as it is.

5:46We get each angle of the character.

5:48So the video generation model knows

5:51what it looks like when the

5:52character is moving around

5:53the screen.

5:54Boom.

5:55And you'll see up here now, I can

5:56reference this other session.

5:59Make sure you save these images.

6:02So now we have our character.

6:04We'll head back to this

6:05tab. We've got that done.

6:06Now we're going to

6:07need our scene backdrop.

— Creating Scene Backdrops That Don't Confuse the Model

6:08So it looks like we have three

6:09backdrops for our five shots.

6:12So we're going to take that prompt.

6:14We're going to start a new session.

6:17We paste that there.

6:18I'm just going to

6:18stick to nanobanana 2 now.

6:20I want to make sure we get

6:21the 16 by 9 aspect ratio.

6:23We only need 1K

6:24resolution, but as you can see,

6:26this is all the same settings you would

6:28do in the regular Venice Chat

6:29when you're generating an image.

6:31So there's our first

6:32backdrop. Pretty neat.

6:33We do not need to edit that.

6:35So now we'll generate the second one,

6:37which is inside of the

6:38pitch black alien cavern.

6:41Cool. Got the light shining in.

6:43Now we'll get our final backdrop.

6:46Paste that in.

6:47That's the word I thought I

6:48had selected, nanobanana 2.

6:50Oh well. Cool.

6:52We even get the character down there.

6:53I'm actually going to

6:54remove the character.

6:55So I'll edit image.

6:57We'll say remove the

6:59person from the clip.

7:02Okay. So now we have our setting image

7:04with no character there.

7:05We want to make sure there's no people or

7:06other things we're going to reference

7:07inside these scene images.

7:11Now we have all of our reference images,

7:13and it's time to generate the videos.

— First Video Generation with R2V Mode

7:15So we head over to the video tab, and

7:17we're ready to start generating

7:18and having some fun with Seedance.

7:20So we're going to make sure we have

7:21access to all of the Seedance models.

7:23I'll just search for Seedance here, and

7:25now we only have Seedance models.

7:27But you can obviously use

7:29any model that Venice offers.

7:32Open source,

7:33uncensored, anonymous, private.

7:35There's all kinds of video models here.

7:37There's 75 it looks like at

7:38the time of the recording.

7:39But Seedance honestly is

7:40maybe the most fun at the moment.

7:42So we are going to choose

7:44reference to video here.

7:46And I'm going to go ahead and just use

7:47fast for sake of this tutorial.

7:50It will also be cheaper

7:51than using the regular version.

7:53So now let's find this prompt. It's a

7:56wide cinematic shot.

7:58So we paste this prompt in, and we'll see

8:00our lone scavenger reference

8:02is going to be needed as

8:04well as the frozen planet.

8:06So here in reference

8:08images, I'll just attach them.

8:10So we'll say a lone scavenger, image one,

8:13in a bulky frost-covered

8:14survival suit and amber visor

8:15trudges slowly towards the camera through

8:17a fierce blizzard on a frozen planet.

8:20Image two. Simple as that.

8:22So we have lots of detail here. It's a

8:24wide cinematic shot, swirling snow.

8:27The guy is trudging slowly towards the

8:29camera, and the camera is

8:30moving backwards slowly.

8:32We got some lighting definitions and 24

8:35frames per second cinematic motion.

8:37And here it's saying this

8:38only needs to be four seconds.

8:40I'll just make it five seconds.

8:42I want audio, but I'm going to make sure

8:44we also say no music

8:46because we're going to generate music at

8:47the end for all of it.

8:49We'll say no music, sound effects, okay.

8:52So now we shoot that off.

— The Queue System (Keep Working While It Renders)

8:53Now here we can see that

8:54this is in the queue over here.

8:56And this is what makes the Venice Studio

8:58fun, is once these are processing,

9:01we can keep going. We don't have to wait.

9:04So now we'll take this

9:05medium shot of the discovery.

9:08We'll pop that in. We'll

9:09change the reference image.

9:11And I'll even add more

9:13shots of our guy here.

9:15So for this next shot, I'm actually going

9:17to show you using Cling,

9:19because it will help you understand why

9:20we generate all the

9:22different angles of our character.

9:24So here in standard R2V. So here we

9:27attach our reference image,

9:28and then we go here to the

9:29plus, we go to references,

9:31and here Cling allows us to add elements.

9:33So this is going to be

9:34a full image reference.

9:36So we add the frontal image, we add the

9:39side image, and we add the back image.

9:42So now I can see all the different

9:43angles. And Cling really excels at this.

9:45We could also add more

9:46reference images if we want.

9:48So this next shot we're

9:49going to generate with Cling.

9:50We'll compare it to Seedance, say if we

9:51want to regenerate it

9:52with Seedance after.

9:53The scavenger in the Warren Survival

9:55Suit, now we tag element one.

9:57And I was wrong, it looks like we're

9:59still in the glacier scene.

10:02And then we'll tag image one there, the

10:04alien archway embedded in a glacier.

10:05The camera slowly pushes in towards the

10:07character in the archway. Cool.

10:08Now we'll set that to be another five

10:11seconds, and we'll shoot that off.

10:14So now let's take a

10:14look at our first shot.

10:18Look, it's got the anchor there. Perfect.

10:21That's exactly what we wanted.

10:22It's wide, the camera's tracking

10:25backwards as he walks.

10:28I almost want it to be a little bit

10:29longer, but I think it'll be okay once we

10:31edit it all together.

10:32So next we're on to our third shot, and

10:35this is where we enter into the cavern.

10:37So I'll paste that in, and I'm going to

10:39go back to Cling, and I'm going to go

10:41back to Seedance here.

10:45We'll paste the prompt. We have two

10:47reference images we're going to use here.

10:49One is him walking through,

10:51the other is him being inside.

10:53And then we'll make sure we've got our

10:57character reference in as well.

10:58So over the shoulder shot from behind the

11:00scavenger, we'll say that's image three.

11:03They step through the obsidian archway in

11:07image one into a pitch

11:09black cavern, image two.

11:11See how this goes here,

11:12guys? It's just tagging things.

11:15Suddenly a harsh bright white utility

11:16spotlight clicks on from their shoulder,

11:19from, should not confuse it with these

11:21lack of genders,

11:23suddenly a harsh bright white.

11:23Cutting through the darkness will make it

11:25to the next one in dust, the camera pans

11:26slowly to follow the beam of light.

11:28We'll make sure this

11:29is another five seconds.

11:31Shoot that off.

11:32And now we also have our Cling video.

— Iteration Reality: Fixing Glitches and Retrying Shots

11:38Eh, I think we're going to try that with

11:40Seedance. I don't

11:41really like how that looks.

11:42So over here, you'll see these buttons.

11:45Share on social feed, download, share as

11:46a gif, regenerate, delete,

11:48or reload original draft.

11:49So I'll reload original draft here and

11:52it'll put everything back here.

11:53Now I can change the model to Seedance

11:562.0 fast, RTV, and we'll make sure we

12:01have our character in there.

12:03And we have to change that

12:04from element to image two.

12:06Stop walking, stops walking in the

12:08background, a massive.

12:10So let's try that again.

12:11And now we see the queue is starting to

12:13add up part of the fun

12:14of the Venice studio.

12:15Now let's get to our next prompt.

12:18This is going to be six seconds.

12:20Close up on the scavengers,

12:22Amber tinted visor, image two.

12:25Oh, it looks like I accidentally selected

12:27too many models here.

12:29So here's that shot again,

12:30regenerated with, there we go.

12:33He sees it.

12:34He wipes his visor kind of.

12:35And I'm just going to make that longer.

12:38I'm going to make it eight seconds.

12:40We'll shoot that off.

12:41Now let's come back to video four.

12:43Close up.

12:45He is inside the cavern.

12:49Image one.

12:50Let's do that eight seconds as well.

12:52Remember, if it's too

12:53long, you can always cut it.

12:55It's a little frustrating when your shot

12:56is just a second or two

12:58short, like it feels cut off.

13:00Now we've got our queue running.

13:01So let's get the final shots here.

13:04Epic wide shot from behind the scavenger,

13:06which is image three now.

13:08Stand on the edge of a precipice.

13:09Okay.

13:10So we need our final scene here.

13:14That's three.

13:15That's actually image one.

13:18A sweeping camera pan reveals the whole

13:20subterranean world lighting up.

13:21Great.

13:22So let's just do that for 10 seconds so

13:24we can really fill in the

13:25space and explore that world.

13:27Great.

13:28And now it looks like we

13:28did this one more time.

13:30This looks more.

13:31Oh,

13:31a little bit of a glitch here.

13:43This is tough.

13:45It's just what happens sometimes.

13:48So we're going to reload this draft and

13:51let's see how we can fix it.

13:53First off, I'll remove the second

13:56reference image of the character from

13:58behind the scavenger three

13:59as he steps through the obsidian archway

14:02in image one into a pitch black cavern

14:04coming through the darkness from the swim

14:06address particles as the

14:08cavern in image two is revealed.

14:12So we're going to make sure it starts out

14:14here and then he walks through the gate

14:16and it just becomes pitch

14:18black until the light comes on.

14:20So let's give it 10 seconds here.

14:22Just I think sometimes the models do

14:24better with more time to

14:25kind of tell the story.

14:26And we have some more

14:28shots to check out here.

14:35No.

14:44I like this one.

14:45OK, so now we're fast forwarding a bit

14:48for redoing the shots.

14:56Pretty neat.

14:58But I want it to be longer.

14:59So let's reload it.

15:01And we know that shot looks right.

15:02It just needs to be longer.

15:04But let's just go full 15 seconds here.

15:06And we have one more shot.

15:09Extreme close up of the scavengers eyes.

15:12So let's tag image two.

15:14Seen through the scratched amber visor,

15:17the reflection of the golden light in

15:18their eyes suddenly shifts

15:20and strobes through terrifying deep

15:21crimson red subtle camera shake.

15:23And we forgot to say no music.

15:27OK, so this one for 10.

15:29Shoot that off.

15:30OK, here's another one.

15:32Forgot to say no music in this one.

15:34But I kind of like it.

15:40My problem is he's

15:42facing the wrong direction.

15:43So let's do that one again.

15:45And let's make sure he's

15:46facing the right direction.

15:47So we just say image two instead of

15:49having both to confuse the model.

15:52And that was 10 seconds. Cool.

15:55Shoot that off.

15:55And now I know I forgot to

15:57say no music in most of these,

15:59but we can always cut it out.

16:01And if we need sound

16:02effects, we can also generate those

16:03when we head over to

16:04the audio tab after this.

16:08This is what we wanted.

16:10Walks through.

16:11OK, well, good enough.

16:15I don't know how, I guess

16:16it would like fade into black

16:18or there's a big flash on screen.

16:20We can actually add that later as some

16:22kind of a sound effect.

16:24But I think this will do for now.

16:26And I also like how we

16:27didn't get any music.

16:28We just have the echo in

16:29claustrophobic atmosphere.

16:32I like the sound effect change too.

16:33This is a really good shot. Cool.

16:36So I'm almost done here.

16:37I think when these three shots will have

16:39almost everything we need,

16:41we can go through and filter out the ones

16:42we're not going to use.

16:43This one's good.

16:45This one he's facing the

16:46wrong way, so we'll delete it.

16:50This one could be optional.

16:51We'll keep it just in case.

16:56This is the one that was

16:57weird, so we'll delete it.

16:59This one doesn't make sense.

17:01We'll delete it.

17:02This will make our

17:03job a little bit easier.

17:04This one might work

17:05because he looks over to it.

17:08Yeah, we need them to

17:09acknowledge that gate there.

17:12OK, so.

17:14It's one of, I think, our final shot.

17:19Whoa.

17:23Oh my gosh.

17:25Yeah, OK.

17:26And this is the second to final shot.

17:30This is what we wanted, right?

17:34Not quite.

17:42Could work.

17:43Which one was this?

17:44Close up on the scavenger.

17:46I think this one we want to do.

17:47Yeah, we want to do it

17:49with this scene instead.

17:52So let's try that again.

17:54OK.

17:56Yeah, that's more what

17:57we wanted the first time.

18:05Let's just do this again with no music.

18:09So while that's all going, we can just

18:11acknowledge here that this whole

18:12workspace here is just so much more

18:14convenient than using the

18:15traditional Venice Chat,

18:16where you have to wait for it to finish

18:18and then you can do another one.

18:19Being able to cue everything.

18:20You can even turn

18:21notifications on in your browser.

18:23See?

18:24And yeah, you can see

18:26everything going on here.

18:27And then meanwhile, in your library, we

18:30can see all the shots here.

18:31So we'll have access to those wherever

18:33it's necessary, and it's easier than just

18:35downloading from one of

18:36our saved chats in Venice.

18:39Up here, we can see the video generation

18:41cue can be turned on and off.

18:43OK, we got our final two shots.

18:53Music ]

18:55Whoa.

19:02Music ]

19:07Not sure what we're

19:07going to do with that one.

19:09So we have enough now

19:11to start having some fun.

19:13And what we do need to do is go to this

19:14audio tab and make some music eventually.

19:17Like this is for a different project.

19:23But first we need to edit this all

19:26together, make sure it looks good and

19:27coherent, and then see how long of an

19:30audio track we're going to need.

19:32So let me set this up.

— Timeline Editing: Arrange, Trim, and Layer

19:33Here we go.

19:34Let's find our first shot.

19:35And I just click that and automatically

19:36get sent down to the timeline here.

19:38This is called the timeline.

19:40This is a non-linear editor.

19:42So you don't have to

19:43edit things in order.

19:44You can drag things around.

19:45Now we got to find what was the best shot

19:47with that we liked after this.

19:51So we can just demo these

19:53and see which one we like more.

19:54And now the video that's going to be on

19:56the top layer, that's

19:57what you're going to see.

19:58We're not going to see this one.

20:00Let's take a look at this one now.

20:11What we could do here

20:12is just shorten that.

20:14All you got to do is drag it in and out.

20:16There, let's see.

20:20I think we'll just leave this one here.

20:24Now we do want to notice that the sound

20:28is a little off because these are

20:30separate video tracks that

20:31were generated differently.

20:32So I can click this and I can click

20:34"detach audio" up here in the audio tab.

20:37"Detach audio."

20:38And what I can do now to just add a

20:40little bit of a smoother

20:41transition is we shorten this clip.

20:44We can also shorten that one.

20:46We'll detach the audio.

20:48We'll move these together so

20:50they kind of fade in and out.

20:51And even that one I can do fade out here

20:54a little bit at the end.

20:55And this one I can do

20:56fade in here at the end.

20:58And now let's listen.

21:02Yeah, so...

21:04Let's do it really quick.

21:11Yeah, so that's a bit smoother there.

21:13I like that shot now. I like it a lot.

21:15Okay, now he enters in.

21:16So I think we have

21:18this shot here for that.

21:20We'll drag it in at the end.

21:21And I see how we're

21:22working with layers here.

21:24These layers are going to be how we just

21:26make it as smooth as possible to watch.

21:29Now I'm going to go ahead

21:30and go to the audio here.

21:32I don't need to detach it, but

21:33I am going to just fade it in.

21:35And I am going to detach it.

21:37And what I'm going to

21:37do is I'll shorten this.

21:40And I'll bring that there.

21:41And then I'll make sure this is aligned

21:43at the end so it's still in sync.

21:44And then this is going to fade in.

21:47So now it's not going to be as weird to

21:49suddenly have the sound

21:50when there wasn't sound before.

21:52Like this, there's no

21:53footsteps or anything.

21:54But now we...

21:57We'll even fade that out.

22:00We'll fade this in quicker.

22:02That's called, I guess, an L-cut where

22:03you hear the shot that

22:05comes in before you see it.

22:09So we got the footstep

22:10here before we see it.

22:12Okay, now we might want to cut

22:14that shot. It's a little long.

22:15But first let's just see

22:16what we're working with next.

22:17He walks in.

22:19We're on shot four now.

22:21Close up on the

22:21scavenger's amber tinted visor.

22:24I think that was... was it this one?

22:35Okay. I don't know if

22:36that's the one we want.

22:37If we go back to the video editor, we can

22:39see the prompts here.

22:40I think this was another one.

22:44I do like this one. I think this one

22:47makes the most sense here.

22:50Even though this is a cool shot.

22:53Now I can zoom in and out here to make

22:55this a little easier.

22:56That's not the one.

22:58That could work there.

23:00But I think...

23:02Yeah, it's this one.

23:03Okay, so let's take a

23:04look at how these would fit.

23:06The problem is we don't

23:07have a transition to here.

23:08So what we might want to do is cut here.

23:13See, we need him to walk.

23:15Yeah, we need that one.

23:16So we go here.

23:20Let's even...

23:25We can get away with this.

23:26Remember, it's a teaser trailer.

23:28So this doesn't have to be absolutely

23:30perfect because the music and just the

23:31vibe will kind of change it.

23:38[Music]

23:46Good.

23:50All right, and then our final shot.

23:56We're going to start this shot at the

23:57end, like right there when

23:59his eyes start to change.

24:00And we can even make this one shorter

24:02here because we kind of get the point.

24:08[Music]

24:23Cool.

24:24A little fade out.

24:26So now as we can see,

24:27we have about 35 seconds.

24:30That's a good length here.

24:31It's suspenseful.

— AI Music That Hits Your Exact Timestamps

24:33So we're going to head

24:33over to the audio tab.

24:34But actually, first we're going to go

24:35back to our chat here.

24:37We say we need a prompt

24:40for generating the score.

24:42It's 35 seconds in total.

24:45And I'm just going to explain to GLM 5.1

24:48how the first 15

24:52seconds are like the intro.

24:54And then the next 14

24:58seconds is the buildup.

25:01And then the final whatever, 66 seconds

25:03is going to be the crescendo, which we

25:05kind of already have here.

25:08So we'll say the first 10 seconds is the

25:12first 15 seconds, which in the next 14

25:16seconds are the buildup.

25:19And the final six seconds

25:21should be the crescendo.

25:26Okay, cool.

25:27So now I'm just going to

25:28copy all of this, to be honest.

25:30We'll head back here.

25:31We'll go to the audio tab, paste that in.

25:34I'll remove that title.

25:38We'll go to the

25:38settings, force, instrumental.

25:41Yes.

25:42We'll go to this orchestra.

25:45We want it to be 30,

25:47we'll do like 38 seconds.

25:49So there's a little

25:50bit of a boom at the end.

25:52And we just have all the detail about

25:54what this is going to sound like.

25:55We'll say at 36, the

25:58massive wall of sound.

25:59So let's shoot that off now and see how it sounds.

26:00So I generated it twice.

26:04Let's listen to one of them.

26:08Got that.

26:10Our beat buildup.

26:11That's exactly what we want.

26:13That's exactly what we want.

26:30The crescendo comes a little early.

26:32So here's what we're going to need to do.

26:34I'm going to need to do that again.

26:36We're going to copy the

26:37prompt, put it back in there.

26:39It's going to need to be like 42 seconds

26:41because it needs to trail off.

26:45And that's what it's factoring in there.

26:46So I'll also change the

26:47time here to like 42, 43.

26:51So now.

26:58Wow.

26:58So right at 36 seconds, it hit that.

27:01It hit it.

27:02So we can just use natural language to

27:04create anything we want here.

27:05So let's save that.

27:06Okay.

27:07We'll head back to our movie editor and

27:09here let's find our media.

27:12We'll click audio.

27:15Let's find the one that was 42 seconds.

27:18It's this one.

27:19So now let's just see how it lines up.

27:57Okay.

27:57So it drops like a few seconds later.

27:59So let's see if we can expand that.

28:04Oh, yeah.

28:16Okay.

28:16We have the score now.

28:18That's cool.

28:19And I think even in this

28:21one, we had audio in it.

28:23We can just remove that and make sure all

28:25the audio is still there.

28:26They were probably not

28:27even going to really hear it.

28:28We could turn this down a little bit.

28:30If I select the track,

28:31I can lower the volume.

28:33And we're also going to need some sound

28:34effects because of the little kind of,

28:37I don't want to say glitches, but just

28:39the nature of making AI generated videos.

28:43They're not going to be perfect.

28:44So as we already saw, it was right here.

28:47We should put a little sound effect there

28:48and even a little

28:49glitch, like a visual glitch.

28:52So we're going to put

28:53a sound effect there.

28:54And then also here, when it just changes

28:56that, we're going to try

28:57to add something there too.

28:58And it could just be a simple transition.

29:00If we find something that's not too

29:02cheesy here, it could

29:05just be a crossfade.

29:06If you just drag that in there and see.

29:11See, right there.

29:12That already did it.

29:13We don't even need a sound effect there.

29:16And in fact, it took a

29:17little bit of time away.

29:19So we've just got to put that there.

29:20Put the crossfade back.

29:22Cool.

29:27Excellent.

29:30And then hopefully.

29:41And one thing.

29:44Right here is where we want to cut it

29:47here, split it, and

29:48we'll just move it over.

29:50So it's like, let's see, we'll put it on

29:53another layer and we'll fade this one out

29:56and we'll fade this one in.

29:58So now, even though the timing wasn't

30:01perfect in the generation based on how

30:02we're editing, we can

30:03still get that effect.

30:06Just like that.

30:11Perfect.

30:12You don't even notice.

30:13Okay.

— Sound Effects with 11 Labs

30:14So last but not least, we do want a

30:17little sound effect for when this

30:20happens, when that opens up.

30:22So we'll come back to our audio tab.

30:24We'll do a little sound effect and we'll

30:26say, sudden science

30:28fiction flash as portal opens.

30:31And we can just kind of

30:32see what happens with that.

30:34We only need one second maximum.

30:36We'll stick to 11 labs here.

30:38Let's see how that sounds.

30:40And I'll go ahead and

30:40generate multiple versions here.

30:42So there's 11 labs.

30:45Good enough.

30:46Not quite what I had in mind.

30:49That's not distinct enough.

30:51Let's say sudden ethereal,

30:53whoosh flash as portal opens.

30:58In fact, I wanted it to be one second,

31:06but actually that's okay because it'll

31:09just trail into the next shot.

31:10And I think that'll

31:11actually look pretty good.

31:12Let's go back to the video editor.

31:14So we're putting a sound effect right

31:15here and we'll just see how it makes it

31:17look cleaner, even

31:18though it's still not perfect.

31:19This is how we can get away with things

31:20when we're video editing.

31:23Little sci-fi swoosh.

31:26And if we really wanted to get funky, we

31:28could split this right here and we could

31:31put a transition right there that maybe

31:34did something with colors or something.

31:37Like right there.

31:40We could make it a shorter transition.

31:42Let's see.

31:49Oh cool.

31:52That's how we do it.

31:53So we have a transition down there and an

31:55audio sound effect to go with it.

31:59And we have our video.

— Final Export and Trailer Reveal

32:01We could put a little

32:01text here at the end.

32:03We could just write, you

32:05know, alien apocalypse four.

32:10We can change the font.

32:15Look over here.

32:21There we go.

32:22That's the name of the

32:22movie alien apocalypse four.

32:24We can actually cut this at like, we'll

32:26just cut it at 40 seconds.

32:28Great. So now we're done

32:29and we can click export.

32:32Now you can switch from

32:3316 by nine for YouTube.

32:35You can do shorts.

32:36You can do a custom resolution.

32:37So I need anything you

32:38want to do in the studio.

32:40You can you can share to

32:41the social feed after this.

32:42I am going to click export and we're

32:45going to get the final MP4 file that then

32:48we can use for whatever we need.

32:50Download it.

32:52Okay. So here's the final

32:53file open on my computer.

32:54Let's take a look.

33:04We've created a whole world here.

33:06I love the consistency of the settings.

33:09That's fun.

33:15Mysterious, suspenseful, magical.

33:30Shaky camera.

33:32Seedance 2.0.

33:33Ladies and gentlemen, incredible.

33:35All available on Venice.

33:37Check out the Venice studio.

33:39Super powerful, super fun.

33:41And share with us what you're working on

33:42in the Venice discord.

33:43Hope to see you there.

More from Venice

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.