flow-videos・📹

[proud forum]

Poseidons wrath

[frigid geyser]

any tips on getting drawing timelapses from start to finish? Usually i get something like this lol

[frigid geyser]

i got this one the other day with "a cozy cabin line drawing timelapse in ms paint" which is kinda close

but not rlly from scratch

[frigid geyser]

oo nice, prompt?

hmm

hmm a lil better

[iron barn]

Swapped it monochromatic ballpoint pen

[proud forum]

Just wanna know how an AI does this

[iron barn]

Removing the artist's hand would probably make it more fluid and less moments with the pen as magic wand

[proud forum]

Hand-drawn style

[proud forum]

This turned out SUPER GOOD

[viral bough]

damn

yeah overall veo is better,

though did u turn on the inspiration mode for that?

oops 👁️ looking forward

atm i want to compare how Wan adheres to the prompt versus Google Veo.

The enhance prompt feature in Wan completely transforms the original concept of my prompt. me no bueno

Oof

try the image to video one

I love the style of the veo one, its aesthetic

[radiant totem]

I loveit 🤗 ,could you share the prompt?

[remote saffron]

Whisk Animate... so technically not VideoFX but just playing.

[surreal yoke]

Whoah!! What's the prompt??!! Do you mind sharing?

Thank you!

[viral bough]

I wonder if wan model wasn’t trained to .. blend realism and something that doesn’t exist irl/practical effect

Wan really said 🧍‍♂️

I noticed Veo tend to do track camera motion even when I didn’t prompt it

Btw u can redeem the credit check in for today , which is 50 in wan

But I actually kinda like the image to video model on wan

Mine 0 now 💀

Oh well at least I didn’t sctually spend money on it-

[viral bough]

Or maybe use imagefx 👀

[viral bough]

Oh ok nvm

Are those fail generation or , still generating

Either way , I had to stop using chrome because chrome freaked out whenever I open videofx library

Microsoft Edge actually does not lag at all when I open the library page

[proud forum]

Omg Veo 2 left me on such a cliff hanger. Why does the generations have to be 8 seconds long 😭

HOWWWWW

I swear to god I think this AI takes like youtube videos of someone playing fortnite and inegrating it into the video but AI style and this is Veo 2 - FAST mode

Wait really??

I discovered that Veo 1 had 60 second videos

Ohh so they are less quality

Okay

[compact dune]

this happened to me once, but they took a donut.

[manic hedge]

an actual heron? 😮

[compact dune]

no, a pelican 😂 close enough

[manic hedge]

pelicans eat anything XD scary dinosaurs

[stray vector]

Made 100% with VEO, the physics are getting good, I test the outer limits and that data is almost their on those figment movements 👌😎🔥

https://youtu.be/EWtpUkniukU

Yeah longer form content is typically only needed for contextual b-roll and dialogue shots. Everything is between 5-10sec. Pushing more artistic long shots to 20sec. If you think about that and watch a show or film now you’ll notice it. My hope is focus turns to editing clips. SORA is perfect length and has an editor. If Google had that they’d be unstoppable, I’d make a whole film right now, lol.

Exactly

lol thanks bud, my next one is gonna be wild, I have 4 more coming

All diff concepts

What’s your X so I can follow you if I don’t already!

I saw that today and I’m gonna def try it asap

That’s what I would do. Often times developing from scratch is faster and you just pull code you want and put it in the new version.

Plus they’re layout needs major update, so there’s lots of reasons to start fresh lol

Neg prompting would be ideal

I use it a lot

It’s good for what we’re doing at first but it’d be better to get to crossover with the apps too and have more UX/UI comfort

I’m a huge prompter so visibility could be better lol

My prompts can be pages

[chrome tangle]

i am back to veo!

lets see its been a while

[quick river]

The prompt👀
[Truck right,Pan left,Tracking shot][Truck right,Tracking shot]

[viral bough]

I like how veo didn’t follow anything that has pan or tracking shots in your prompt structures but it did follow right, left, and right movements. That’s interesting

@muted quartz remember when you were struggle to ask Veo to generate sunset thingy

From my experience, Veo followed my prompt accurately in only one out of four generated videos, though not always

I wonder if Veo was intentionally designed to generate videos that way or if it independently decides which words from the prompt to use

That’s true,

I wonder … if I could try use different language in prompt

Or “reverse psychology”

https://tenor.com/view/spongebob-spongebobsquarepants-gif-9146238

[bleak bridge]

Me too. I'll put in some effort reword or rework a prompt, but if the model can't seem to do it, it's easier to change my plans than it is to fight with a model. Generative AI models just have some weirdness from time to, and it's generally seems to be more fruitful to sidestep their blind spots than to get fixated on trying to make them do exactly what I had in mind to begin with.

[viral bough]

In conclusion,

[bleak bridge]

Of course it is productive in our role as testers to downvote and comment on cases when the model has hard time with prompt--let the engineers sort it out if there is a real problem behind the scenes. From that perspective, it's good when we run into roadblocks, even if pushes a little against our hopes as creators.

[novel basin]

Try this :

Hyperlapse animation. Starting with a single dot multiplying rapidly, transforming into fragmented visualizations of: AlphaFold protein structures forming, AlphaZero chess game in motion, Imagen AI-generated landscape art appearing, scrolling AI code. The final frame must synthesize into a vibrant, artistic representation of the Google AI logo. Fast-paced, dynamic, visually stimulating

[quick river]

Modified version

[novel basin]

Thanks for trying. Seriously, they're mind-blowing.

[proud forum]

I built that minecraft build

[proud forum]

I didnt actually build the minecraft build

[rocky basin]

Ok?

[viral bough]

Veo 2 can produce any type of video, dynamically embedding your custom text into the visuals.

[sly glacier]

Well to be fair it's not just veo. I've been doing these good morning slogans for awhile. Prompt coherence

[viral bough]

Just yesterday, I used Veo 2 to put together a birthday greeting video for my mom. And she loved it ,

Translates to “happy birthday , mom!”

[sly glacier]

Nice

I haven't been making many videos lately. Been stuck in Google studio.

[grim oar]

GTA 5 veo 2

[grim oar]

gta 6 by veo 2

[uncut cosmos]

Wow so realistic, just a few mistakes

[grim oar]

New Delhi in GTA 5 I would play this version of the game

[floral harbor]

prompt?

[floral harbor]

looks like favela in brazil lol

[uncut cosmos]

Impressed

[long crypt]

I did edit in after effect for this using vid generated by VideoFX, inspired by Evangelion and Seraphim by Jonas Pfeiffer

[manic hedge]

guys,

I was making these birds eating pizza,

....and at one point the ai messed up, merging the emu with the kestrel, and now I've created a new animal species

[uncut cosmos]

Do u pay 0.5$/sec to do that 😭😭

[grim oar]

house in gta 5 in a poor area

he even made children in gta 5 they are not in the original game if anything and I have a question how can veo 2 create things in games that are not in real games? what is the secret?

[manic hedge]

prompt? I like the pov look

[olive plaza]

still haven't got access....that's crazy

the fact that google gave to these guys who are charging this much for 5 second hahaha...and not launched globally shows how much they care....dystopic world!

[paper egret]

https://youtu.be/8adPRJ2_GGA?feature=shared

Completed. Enjoy!
🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥💯💯💯💯💯💯💯💯💯💯💯💪💪💪💪💪💪💪💪💪💪💪

[manic hedge]

I don't think prehistoric men looked like they just went to the barber, but cool idea and nice audio!

[grim oar]

this is very cool shanghai in gta 5

[fluid orbit]

Realistic, first-person, handheld, jittery camera video of [your scene description]. Raw, unedited footage

[cinder horizon]

damn!

[proud forum]

Why are people in the videos walking away from the thing I want them to walk towards

or facing away from it

[mossy sparrow]

What was your prompt for this one?

[paper egret]

I know lol I tried to make it to where at the end he was making a knife out of a stone so he can cut his hair if you can tell. It shows it in the video. But yeah I realize a lot of things lol The best I can get it

[viral bough]

The Gemini logo, was that part of the veo generated too?

[sly glacier]

@viral bough yeah it's their watermark.

[quick river]

Scary year ahead of us!

[sly glacier]

@quick river yeah some fierce competition...and I know MJ been cooking behind scenes for a minute

[uncut cosmos]

Is that veo2 ???!!!!

[sly glacier]

Yes

[signal pumice]

A tight medium shot of a majestic snow leopard crouched on a rugged, snow-covered rocky ledge, its piercing blue eyes locked onto the camera. The shallow depth of field blurs the distant, icy peaks in the background, while individual snowflakes drift gently through the crisp mountain air. A soft rim light from the overcast sky highlights the leopard’s thick, spotted fur, capturing the delicate frost clinging to its whiskers. The camera lingers, allowing the raw intensity of its gaze to create an intimate yet powerful connection.

[dire barn]

looks better than the actual game itself

[torpid rose]

A tight medium shot of a majestic snow leopard crouched on a rugged, snow-covered rocky ledge, its piercing blue eyes locked onto the camera. The shallow depth of field blurs the distant, icy peaks in the background, while individual snowflakes drift gently through the crisp mountain air. A soft rim light from the overcast sky highlights the leopard’s thick, spotted fur, capturing the delicate frost clinging to its whiskers. The camera lingers, allowing the raw intensity of its gaze to create an intimate yet powerful connection.

This channel can generate videos?

[viral bough]

usually user post the prompt in #flow-tips-n-tricks・✏️

also ill give that atry

@torpid rose freepik veo image to video

[torpid rose]

ok

[grim oar]

yes that's true it looks cooler than the real game

[torpid rose]

Are these videos generated by Google VideoFX?

[quick river]

Beautiful shots!

[viral bough]

It’s generated using Freepik Image to Video Google Veo2

[strange stirrup]

10,000 B.C. AI Filmmaking (4k)

[long crypt]

wait did you make this in Google AI Studio?

[muted solar]

Exploring Veo 2's understanding of shot types. Experimenting with more moodboards like this, feel free to make any suggestions.
https://x.com/che42_/status/1902246166118486273

[sly glacier]

Yes

[stray vector]

lol at my next music video 🤣🔥

[manic hedge]

I really love that dreamlike alien scenery. Really reminds me of '80s videogames for Amiga

[grim oar]

funny glitch how a city appears out of nothing in veo 2

I have a question, what number of different seeds can be entered in veo? Well, I mean the seed under the video generation settings

[quick river]

Looking forward to seeing it

[quick river]

Not human.

[daring vessel]

anyone have a way to access veo 2 ?

[manic hedge]

he looks like he's wearing planet earth. is that a polluted planet god?

[viral bough]

sighting 🦊

[stray vector]

All VEO 🔥

[viral bough]

i thought this was wednesday adam

https://tenor.com/view/jenna-ortega-wednesday-wednesday-addams-netflix-gif-27293940

[sly glacier]

Lol...indeed it is. Just my adaptation

[viral bough]

OH

🧠 big brain me today

[proud forum]

It's possible if you got command blocks

[proud forum]

Would you guys eat all this for 1 penny

[viral bough]

Yours are even obvious omg , the artifact noise

[shell jolt]

😭

[viral bough]

And this too!

The artifact noise thing (idk the actual term so pls don’t blast me)

[dire barn]

wonder if its a bug or something

that didn't happen a few days ago

[viral bough]

Yeah , I noticed it started like today

Or yesterday , at least during the happy st Patrick day today’s special

[dire barn]

is it the same thing on fast mode

[viral bough]

Ok yeah the more I rewatch the other generated video , the more I can’t unsee the noise

[viral bough]

I’m not sure

[viral bough]

Was this fast mode?

[shell jolt]

yes

[viral bough]

Ah-

Let me check my other fast mode generated

This was on normal mode

I think the model is.. having hiccup

[dire barn]

yeah that looks really bad

scrolling up it looks like its been happening since yesterday

or two days ago

[viral bough]

@compact dune can you guys take a look at this ? Sorry if I ping you out of nowhere Georgia 😣

Ok maybe I should make a VideoFx bug thread about this so that it doesn’t get buried

[proud forum]

No I did that style on purpose

[viral bough]

@shell jolt feel free to share your video too in the thread https://ptb.discord.com/channels/1172568727942860810/1352132020368572437

[shell jolt]

yup

[compact dune]

thanks for making the bug report! 🪲 appreciate it, we’ll take a look as soon as we can

[viral bough]

No worries! Take your time Appreciate your hard work

[viral bough]

VideoFX early access users' dreams during sleep be like

[sly glacier]

Lol

[proud forum]

Do you think Veo 2 controls our dreams? 😲😲😲😲

I dont get the artifact sound effect thingy

[uncut cosmos]

Wdym artifact noise

[old frigate]

🐺

[hybrid burrow]

ThonDaria Queen of Valor Action Image (AI) Comic Doc B-roll Carma Carma Comic Con HubCity MS A little bit of everything in this one.

[quick river]

She's beautiful!

[manic hedge]

is it possible to save the image in image2vid?

[long crypt]

can you show me how? I thought it could only do images

[sly glacier]

@long crypt it can. I used image to video elsewhere

[long crypt]

oh ok I thought you done image to video right inside Google ai studio

[sly glacier]

No not yet

[thick mulch]

https://x.com/maxescu/status/1902742535618888025\

Alex Patrascu (@maxescu) on X

AI video creation is very disjointed today.

But there's a glimmer of hope.

And it's called Whisk by @labsdotgoogle

The new early-access Animate feature is spectacular!

It's using Veo 2 image to video natively:

1.

X · Mar 20, 2025, 3:22 PM

[quick river]

This means you get studio level production control. Hailuo did the same a few days ago [image model], and it's wonderful, but in my opinion, veo is the quality.

Scene setup plus elements= Whisk.

[thick mulch]

Yeah! It's super easy to create a story now in the same environment with the same characters

[manic hedge]

Aw, I'd like to do more videos of that guy but I can't re-generate him in image to video

[rugged crystal]

Veo2, tex to video, music and voice -over made with Udio, hope you like it. https://youtu.be/u-Oq00UX6Bk?si=jVqVMtFRSe6Sjoqq

[quick river]

Just beautiful!

[rugged crystal]

thanks!

[grand zenith]

Amazing... can you share the prompt? Grazie

[knotty marsh]

Just release this one made essentially with Veo2 in combination with other tools. 😊

https://youtu.be/-vABdeB3C74?si=i4s4aGj68Bmiu9HK

[quick river]

Lipsync & upscale are top-notch!

[knotty marsh]

Thank you! 🙏

[glossy radish]

Waiting waiting ✋️ until next Christmas I don't have chance to try videofx 😭 😭 😭 😭 😭 😭 😭 😭 I sended my wait email 2 month ago

[thick mulch]

You can give it a subject (or more), a scene and a style

That's a lot more than other tools

I managed to get the same 'room' more or less the same

[proud forum]

It feels like I'm creating things that don't exist into the real world with this tool

[proud forum]

• With sound

[gritty blaze]

https://x.com/Arkatech/status/1902924212550987814?t=WvnFUvADNvzGGEkh7AdFjA&s=19

Damn Arkatech (@Arkatech) on X

WAR machine

@labsdotgoogle
#AIvideo #AIgenerated #DeepLearning #MachineLearning #VideoArt #animation #motionpictures #ai #art #reels #CreativeJourney #arkatech #hybrid #arkatechfilms #Anime #vfx #google #veo2 #digitalcinematography #cgi

X · Mar 21, 2025, 3:24 AM

[viral bough]

A full wide shot presents a powerful anthropomorphic male buff snow leopard, in a futuristic monochromatic streetwear ensemble, standing tall against a minimal gray studio background. White combat boots anchor his avant-garde outfit. His bright blue eyes are narrowed in fierce determination, whiskers bristling, pointed ears swiveled, and muzzle pulled back in a challenging grimace, exposing sharp teeth. Natural sunlight floods the studio, creating dynamic interplay of light and shadow accentuating his musculature and outfit textures. The camera slowly pans around the snow leopard, highlighting the intricate details of his outfit and the intensity of his expression. A subtle zoom emphasizes his powerful physique as he subtly shifts his weight, showcasing the dynamic energy of his pose. Every strand of fur, leather strap, and parka weave is exquisitely detailed.

The flickering issue is still present, hoping it gets resolved soon

[grim oar]

and the quality and physics in the video itself hasn't become worse than Cross Myrtsaniye?

[viral bough]

The what now

[viral bough]

Also, it straight up ignored “camera pans” and “subtle zoom” from the prompt. Interesting

[proud forum]

Doesn't do well with panning, it only knows how to instantly cut to different aspects of the scene

Like a music video

[proud forum]

The object permanence is INSANE

[bleak bridge]

I'm continuing to practice with short videos with new music, in this case DavidArchuleta's "Crème Brûlée"

As usual, I generally like Veo 2's physics, but it doesn't nail "cracking" creme brulee with a spoon, which, to be fair, is probably very hard.

https://youtube.com/shorts/rMYVaKb2sEw

[dire minnow]

What's the prompt? And how was it implemented?

[viral bough]

OOOH STYLISH AND PRETTY ❤️

[grim oar]

cool of course he creates minecraft so far there are no competitors in this sphere for veo 2 and in general besides minecraft he makes other games very cool

[grim oar]

the city is showing off with the click of this weird object i dont think any other model could make this video like veo 2

[cinder horizon]

it's even showing the rendering effect... wow

[proud forum]

kamehamehAAAAA

[bleak bridge]

Continuing to have fun, here's a remix of Greek myths with MARINA's song "Cupid's Girl"

https://youtube.com/shorts/Dih0o_5mrtg

[carmine shell]

Creating clips for the purpose of looping and using them for ambient videos on youtube. I love these results.

[quick river]

Ty, yes.

[viral bough]

Amazing

[copper fjord]

Love how it can do a complete consistent sequence in one shot

and some surreal explorations:

and its really good at VFX 🙂

[glossy radish]

Yall are flexing Video fx in front of me (guy still in waitlist) 😭

[copper fjord]

This one with T2I2V

[bleak bridge]

Another little music short:

https://youtube.com/shorts/F7umCY7O1c0

[polar hollow]

What the hell 😂

[proud forum]

Oh good god

[proud forum]

What people in the past thought cars would be like:

[alpine vigil]

Those rear tires look like they would be a pain to replace

[proud forum]

The car got inflated

[remote saffron]

i just love the details of the simple shots. Like you can see homes reflected in the paint as it fly's off

[knotty marsh]

When are you cookin something for us brother? haha! Love these!

[copper fjord]

Still in exploring and prompt engineering mode, brother! 😃 Soon :9

[proud forum]

I think I accidentaly created a loop with this one

[quick river]

The stock footage industry is absolute.

[quick river]

Ok, let me get ready.

[quick river]

Yes, in post.\

I have access to all of them, this is my go to, followed by hailuo, then runway.

My guess is it’s the prompting.

The devil woman.

[quick river]

I think it's the source image [feels like cgi to me].

And if you want more fidelity, t2v is definitely the way to go.

The dynamics 💯

[quick river]

Aerial View of Rocky Tropical Coastline.

Yes, few mins ago I guess.

That was the prompt.

Or, you mean something else?

[proud forum]

Holy this came out better than the last one

[quick river]

The server crashed😠

[rotund gulch]

A futuristic tiki bar interior. Sleek, dark chrome replaces the wood, reflecting dynamically adjusting neon lights. Woven light fixtures are replaced with geometric, glowing panels. The bar's counter is a seamless, polished surface with holographic stools. Behind the bar, bottles of liquor are displayed in illuminated glass cases. Instead of carved tiki masks, stylized holographic projections of futuristic tribal figures flank the bar. Small, chrome tables with glowing bases replace the wooden tables and candles. A holographic display replaces the framed picture. The overall atmosphere is a blend of tropical themes and advanced technology, with a cool, vibrant color palette.

[dusk solstice]

Whats your prompt?

[mossy sparrow]

Love this shot! What was your prompt?

[proud forum]

`An intense anime-style scene unfolds as a powerful being with glowing golden eyes spins through the air, their long, tattered cloak trailing in a fierce whirlwind. The camera tracks their fluid, calculated movements, circling upward as golden sparks and crackling arcs of energy radiate from their body. The atmosphere hums with raw power, distorting the air around them.

The being halts mid-spin, levitating with arms outstretched, their hands gripping a massive, pulsating yellow orb that throbs like a miniature sun. The orb's surface swirls with blinding currents of golden plasma, sending shimmering rays of light cascading across the darkened sky. Tendrils of energy lash out, crackling and illuminating the surrounding storm clouds.

With a sudden pulse of force, the figure unleashes a concentrated yellow beam from the orb. It tears through the air like a blazing comet, leaving a golden trail in its wake. The recoil sends shockwaves rippling outward, distorting the environment. In the same instant, countless smaller yellow orbs erupt from the main sphere, streaking outward in every direction. They hover for a brief moment before exploding into dazzling bursts of light, each detonation accompanied by a low, resonant hum.

The camera trembles under the intensity of the blast, lens flares scattering across the screen. Dust and debris spiral through the air, glowing with reflected golden brilliance. The powerful being remains suspended, their glowing eyes locked onto the destruction below, the lingering crackle of energy illuminating their fierce silhouette`

[viral bough]

No way it can generate those genshin impact game style

[proud forum]

Yep

But im gonna try to be more specific on the "Anime-style" part

I should put "2D Anime-style"

[viral bough]

that is an epic otter omg

[chrome tangle]

text to image to video

[viral bough]

Text to Image to Video too

[magic drum]

Aadding a little bling with some quick shopping by the shore on a sunny day. 🦋

[barren barn]

I was just about to check but then I realized it's not saved 🤪

But it was something minimal like "POV falling from tall building towards pavement. Cinematic. Minimalist, dim lighting,"

[proud forum]

The power from the persons surroundings allows him to walk on air peacefully

Huh?

Oh the "person" is the person who is walking on air

[quick river]

How many prompts are allowed now? I’m only getting 2!

[proud forum]

You can edit it in the settings

[quick river]

Sorry, I meant concurrent prompts, I’m only seeing 2 running at once, sometimes it accept 3. Is that the new limit?

[proud forum]

No idea

[quick river]

I make packs, so 4 prompts was perfect.

[bleak bridge]

The new UX certainly seems to speed my workflow--I liked being able to make these shots in parallel.

https://youtube.com/shorts/QaOW2DlWMjg

[proud forum]

WOAH!!!

[frigid geyser]

spoiler bc of the strobing

[remote saffron]

Love the sporradic cuts

[frigid geyser]

tyty, prompt was "Adobe Effects Showcase, a cool mix of actual footage and digital manipulation through bold color additions and accents. Mix of pigments and textures interacting in satisfying ways with a mix of club dancing and funeral in the rain, film art. Dark glitch art"

[alpine pine]

VEO2's new UI is pretty cool

It would be better if could upload own diagrams, this update has a much better character effect!

[raw torrent]

I LOVE the new VEO2 UI!!!!!

[uncut cosmos]

Nice

[uncut cosmos]

🙂‍↔️

[alpine pine]

Thanks for enjoying. The new UI is truly fantastic!

Looking forward to adding some more sound features and extensions!

[viral bough]

Text to Image to Video

[barren barn]

Used Veo 2 for a lot of this short I'm releasing soon (+Kling + MiniMax). Here's the trailer. Will post full film Wednesday.

[viral bough]

the way i instantly recognize which one is veo generated (the moire effect) 😅

but overall that looks dopeee

[barren barn]

Ya I didn't notice it as much until i did this vertical version and everything's zoomed in lol

[viral bough]

but if you have time , i think they fixed the moire effect now, maybe u could get it generated one last time

i assume the bloody hand one was generated with minimax?

[barren barn]

Ah it's not worth the trouble tbh, especially on the wider screen. Good to know tho.

Bloody hand was flux pro ultra + kling

[viral bough]

ahh wait im surprised it passed on your part, mine automatically refused to do image to video generation cuz it detects blood

[barren barn]

The other unnerving bloody shot I have in the film tho is flux pro ultra + minimax. Kling & MiniMax have usually been good for me on the less censored stuff. Minimax particularly.

[viral bough]

have u tried wan? @barren barn

[barren barn]

Ya a lil, but I just don't have a use for it yet until it grows a little more maybe. Is it beyond the 16fps or whatever or are people still motion smoothing?

[viral bough]

its 30fps i think if that answer ur question (?)

[knotty marsh]

Arrow shot were usually never working perfectly! Great job on this one!

[grim oar]

Man, and the new feature with the creation of video by image Cool hope that soon it will be possible to create videos and on their own images and thank you for removing the flicker

[grim oar]

if they soon give the ability to create videos based on your own images, then this will be the best neural network for creating videos

[proud forum]

Domain Expansion.. (I think)

[viral bough]

🙈

[proud forum]

🌗

[compact dune]

i need a coffee made by a puppy. that's the dream

[proud forum]

Awww hes so cuute

🥺

[strange stirrup]

just seeing this. soooo beautiful!!

[glass zealot]

A little too much caffein? 😄

[gritty blaze]

https://x.com/Arkatech/status/1904623594677662193?t=nrhjnSK-wjALYCtiEs55Ew&s=19

Damn Arkatech (@Arkatech) on X

"Family Time"

@GoogleLabs
#AIvideo #AIgenerated #DeepLearning #MachineLearning #VideoArt #animation #motionpictures #ai #art #reels #google #CreativeJourney #arkatech #hybrid #arkatechfilms #Anime #vfx #veo2 #whisk #whiskanimate #digitalcinematography #cgi

X · Mar 25, 2025, 7:57 PM

[bleak bridge]

Nice music, setting, and visual consistancy.

My music experiment today: a close up, and an extreme close up!

https://youtube.com/shorts/26yGMGGNIqI

[viral bough]

They’re just excited to serve the pawstomer ☕️

[strange stirrup]

ooo nice. what were your videofx prompts? (i've been meaning to give a listen to the new japanese breakfast album 🎵 )

[bleak bridge]

I start with an initial concept, which I then refine with a prompt to include things that I want in all my video clips--for example a color palette along the lines of the 60 30 10 color rule, and the action in the middle of the frame (because I am sharing this as a YouTube short, so it'll be cropped to the middle.) In this case I ran my "prompt improving prompt" with the new Gemini 2.5 Pro experimental model on AI studio. (I updated my prompt improvement prompt to specifically mention I was using VideoFX, in case Gemini 2.5 could benefit from this, with the Jan 25 knowledge cutoff.

Here is the final prompt:

Photorealistic, cinematic, high detail. Mood is somber, intense quietude, melancholic longing. Colors follow a stark chiaroscuro palette: 60% deep, cool-toned shadows and indistinct dark blues suggesting night or a dimly lit room; 30% pale skin tones harshly illuminated from a single, unseen side light source; 10% brilliant, wet highlights on the eye surface and the forming tear. Extreme Close-Up (ECU) lens, positioned slightly low, angled slightly up towards the character's right eye. The character, a young adult with sharp androgynous features, intense dark eyes, messy dark brown hair, pale skin, wearing a simple dark grey knit sweater, is visible from the cheekbone up. Their eye is wide, fixed just off-camera, conveying profound, still sorrow. In the exact center third of the frame, a single, crystalline tear slowly wells up, catching the harsh light and trembling precariously on the lower lash line. The skin around the eye is taut with restrained emotion. The environment is an out-of-focus dark interior, shapes barely discernible. Camera is completely static, locked off.

Anyway, hope you get a chance to enjoy the new Japanese Breakfast album!

[strange stirrup]

soooo cool. thank you so much for sharing your process @bleak bridge !

[bleak bridge]

Sure thing! Typically I'll just run the refined prompt, and only edit futher manually if I don't like the results. For example, if I really wanted more shadows, and darkness, I'd have rewritten the prompt to push for more of those, but I think the video works as is.

[fallen kernel]

Whisk animate 🔥

[dark nova]

Up, up and...

[alpine pine]

https://x.com/Eric520CC/status/1904712689840173070

Metaverse | Eric (@Eric520CC) on X

🎬Movies : Desert Dance
Dancing in the desert, Exploring the mysteries of life
Videos: Veo2
Voices: Hailuo AI
Edited in: CapCut
All clips are generated by VEO2 Text To Video.
#Veo2 #hailuo_ai @GoogleDeepMind @GoogleLabs @Hailuo_AI

X · Mar 26, 2025, 1:51 AM

https://www.youtube.com/watch?v=K3HzgeQPslA

[true pike]

how many parallel renders are allowed?

[viral bough]

SO IF YOU CARE TO FIND ME LOOK UP TO THE WESTERN SKY

[grim oar]

why doesn't video generation from images work for me? loading reaches 10 percent and an error appears

[uncut cosmos]

@sonic geyser

[uncut cosmos]

@short anchor

[grim oar]

Am I the only one who thinks that the quality has become a little better than it was?

[proud forum]

Yes and No.. i think

[crystal shore]

Does anyone know if there’s a limit to our generations? I’ve been creating some crazy cool generations but I’m not sure if there’s a limit

[viral bough]

https://ptb.discord.com/channels/1172568727942860810/1346912264187285597 someone actually manage to hit limit on videofx

but we dont know how many videos till it reach your quota

whisk animate only have 100 credit per month

[crystal shore]

Oh that’s sad, I’m currently only using it in whisk animate because I didn’t get access to VideoFX. Appreciate it though!

[viral bough]

yeah u might need to be mindful with how many vids u generated in whisk animate, tho it will reset it monthly

[crystal shore]

True, thanks for letting me know

[viral bough]

mine is like at 90 credit left right now

u can check yours through inspect element network log btw

[crystal shore]

I probably used almost everything 💀

[viral bough]

cue hackerman gif

oh i mean you will get notice saying you hit your limit if you actually finished up all your credit

or you can manually count how many vid u generated so far= credit consumed

[crystal shore]

Do the image generations count as generations or only when I click “Animate”?

[viral bough]

nope

[viral bough]

only video generated

[uncut cosmos]

@somber hound someone do something

[grim oar]

is there a video generation limit in veo 2?

[barren barn]

Just dropped my short that I used Veo 2 for a lot particularly for those inserts, cutaways, and crowd shots.

After his mother's death, an emotionless accountant visits a strange comedy club promising "guaranteed laughter." But when he can't stop laughing, he discovers the club's dark promise was more literal than he thought. An AI short film about the grief hiding behind a smile.

https://www.youtube.com/watch?v=4E3NJUjZTsU

[barren barn]

wow image to video just dropped for me 👀

[viral bough]

Wait what

[barren barn]

ya new ui and everything

[viral bough]

Huh oh

I thought you said actual image to video is released

Cuz the one we got right now is still text to image to video haha

It’s the interface that had major update

[barren barn]

oh wait nevermind, just new ui. i saw the "+ image" button and got excited

[viral bough]

HAHA it’s alright I feel ya, I was excited at first too

[barren barn]

nice to finally have all the generations saved i thought they were gone forever

[crystal shore]

Is there any way to get access to it if I have whisk animate?

[quick river]

Awesome job, seriously, great work!

[barren barn]

Thank you 🙏 it was a lot of work haha

[spring dagger]

this is incredible .. absolutely mind blowing .. so well conceptualized and edited

[uncut cosmos]

Who does tho

Huh

[compact dune]

hi, I'm one of the mods / part of the Google Labs team. sometimes messages get through our filters and we sometimes miss messages -- we are human beings and that means we are not perfect 🙂 we're bringing on more mods soon as well. let me know if you have any questions.

as a reminder, this channel is solely reserved for sharing videofx generations. please bring all other chatter to #flow-general・🌊 to keep this space reserved for showcasing gens. thanks!

[uncut cosmos]

Sorry for disturbing, i understand your situation but it was an invitation and i pinged 3 mods to remove it, don't worry about that. Thanks. if you need any help let me know, i am ready to work with you.

[bleak bridge]

"While lots of people are playing with OpenAI selfies, I'm testing Gemini 2.5 Pro with my 'Turkey Test' - challenging it to write metaphysical poetry about a bird. Here it is, recited with VideoFX illustrations, and followed by a bit of NotebookLM commentary…

https://youtu.be/MagWnkL14js?si=ywCvQQY12Kruh6aZ&t=54

[manic hedge]

WAIT it's out? oh

[manic hedge]

oh heeyyyyyyy

[uncut cosmos]

Is Gemini 2.5 out ?

[viral bough]

Yeah it’s out

[lucid juniper]

Help! I got the email to try Whisk animate, but I'm getting "Access Denied" when clicking on the link!

[uncut cosmos]

On advanced plan only ?

[viral bough]

From the looks of it yeah

[uncut cosmos]

Nvm I've got access already to it thnx

Not in the app tho

[compact dune]

hey! please add this to #1319422603001200731 with any screenshots/extra details you have so we can look into this! thanks!

[sudden goblet]

Whisk animate

[compact dune]

i'm gonna forward this to #whisk-gallery・🖼 so more people can see it!

[copper fjord]

Great quality shot!

some surreal shots I made using paintings from my wife and daughter, pu them into chatGPT and let it make prompts for me

[compact dune]

oh wow! neat idea to try that out — the results are sweet

[copper fjord]

Thanks Georgia! Yes, its a great experiment. The results look kind of different from the input images in terms of the exact style, but all the elements are there

Veo 2 is so good at surreal scenes, love that!

and so many different animation / Illustrations styles (based on a drawing of my daughter)

and of course it shines on cinematic magic as well 🔥

[compact dune]

I was going to ask how these compared to the originals your family made! thanks for sharing

[proud forum]

This new Roblox game is OUTSTANDING 🤯

[radiant totem]

https://x.com/imaginedremms/status/1905015933359591917?t=Q5ZoWBXBrsCEToB1XHbpeA&s=33

imaginedream (@imaginedremms) on X

🌸 Some anime videos made with #Veo2 ,It's really amazing what it's capable of, the best video generator. @GoogleDeepMind @GoogleLabs

X · Mar 26, 2025, 9:56 PM

[manic hedge]

getting to make it look more real

[manic hedge]

merged 2 results and made this. I'm almost tempted to add audio and claim it's a scene from a Dredd cartoon!

[obsidian juniper]

you can use it on aistudio.google.com for free

[exotic minnow]

Slow vertical tracking shot: The camera moves upward, revealing "Tekzo", a 10-year-old boy with spiky electric-blue hair and glowing cyan eyes, standing in his secret high-tech garage. He gazes in awe at his cyber-tech glove, which emits neon-blue energy trails. Floating holograms display futuristic schematics around him. The dimly lit garage flickers with monitors, scattered gadgets, and robotic parts, while sparks of electricity and digital particles swirl in the cyberpunk atmosphere. Animated movie quality.

[viral bough]

Please stop spamming on every channel

[exotic minnow]

whay?

[manic hedge]

an update:

it will make the image, but refuse to animate it

and, also........

........why TF is she blushing anyway????

I'm starting to understand why it refuses to make kids. Its mind is in the gutter. Everytime I ask for a lil' mischevious catgirl talking in a circus, somehow she's blushing

[uncut cosmos]

Yea found it

[knotty marsh]

is this whisk? What do you mean by merge 2 results?

[uncut cosmos]

Any way to use veo2 for free ?

[viral bough]

Technically , VideoFx early access user can use it for “free”

[manic hedge]

I mean I cut and merged two of the eight seconds results from Veo2. It's not Whisk as it's glitchy today

[uncut cosmos]

I don't have early access sadly

[viral bough]

Did u get whisk animate

[uncut cosmos]

Lemme try vpn

[viral bough]

POV of a pedestrian walking through Victorian-era New York City’s

[uncut cosmos]

Just got access(it was in my mail already) thnx

[olive bluff]

got access to whisk animate

what i cant generate

it says error

[viral bough]

go to #whisk-general・💭 to discuss this further btw

[olive bluff]

actually i did made a video

[viral bough]

using whisk animate?

[olive bluff]

yes

[viral bough]

yeah theres a proper whisk channel for that btw

#whisk-gallery・🖼 #whisk-general・💭

[barren barn]

my first i2v generation for the new short image from whisk

[proud forum]

It feels like Veo 2 got faster

[viral bough]

My first image2video generation too, (image generated from halfmoon)

[bleak bridge]

My first image to video generation, from a Midjourney image of a tabletop RPG character.

[quick river]

The last update is top-notch!

[bleak bridge]

Another fun image to video attempt, this time from an image I had commissioned for a tabletop RPG I've published.

Enjoyed the output, but did send a report into Google about unexpected behavior. What I sent was, "Trying the new image to video feature, I asked for a continuous shot in the prompt, and got a shot with a number of cuts, and also with the characters shifting from illustrated (as with the starting image) to photorealistic."

Still think this feature is super cool--and am also cool with encountering something like this, because it means I actually get to feel like I'm finding anomalies and giving (hopefully useful) feedback.

[viral bough]

I also sent the same feedback like yours! hope fully they resolve this

[bleak bridge]

The more we test and send reports, the more data they'll have to work with!

[viral bough]

Used my own drawing that i did and use it on img2video , and im loving it 🥺

[quick river]

Impressive!

[bleak bridge]

I also tested an early sketch:

[sly glacier]

Still need to allow other options of aspect ratio

[viral bough]

i just love how the model adapt to our linework

[old frigate]

Sick 🔥

[tranquil locust]

This is helpful! Can you share the image and text prompt?

[quick river]

Interesting! It can also work as a subject reference, kinda like what Gemini Studio does.
"Take just the face, build a new scene, and have him play guitar in it."

[bleak bridge]

Sure, here is the image. I had closed my Labs tab, so I can't see my old prompts (although I can see the "use prompt" for new generations.) My recollection was the prompt was something simple like "continuous shot, zoom in to space opera heroes flying their ship"

[old frigate]

img2vid is 🔥

[bleak bridge]

Works with photos nicely.

[fallen kernel]

Wow I just discovered image to video

Any limits?

[proud forum]

This one has better audio

[bleak bridge]

Just wanted to see how the model would do with a painting style that was less modern--in this case a small detail from the Qing court version of the classic painting Up the River on the Qingming Festival

[manic hedge]

some testing. source image, ideogram 2.0

[proud forum]

So um I was going to type a prompt along the lines "A pig flying in the air", but then I made a mistake and accidentally pressed Enter and it left the "A" at the start of the prompt and this is what it generated. Pretty impressive though.

[viral bough]

Oh this is actually a great idea, never thought of using non 16:9 image on 16:9 canvas

Hence making the character/subject consistent

It’s mesmerizing 👀

[wraith rose]

awesome, short and sweet and sweet and build out🧡 thank you very much! love it🧡🐺🌑

whoa lol

that just needs a x-wing fighter 😜

[quick river]

Nice one!

[proud forum]

A pack of wild dogs

[viral bough]

img2video (image generated with halfmoon)

[viral bough]

Text2Img2Video (imagen3)

[tall geode]

Insane!

[hollow crest]

It’s so good omg

From my niji・journey image

[hollow crest]

Lol

[grim oar]

О привет ну как тебе функция генерацию видео по твоему изображение?

[tall geode]

Имба

[grim oar]

Да согласен

[viral bough]

img2video (image generated using halfmoon)

[carmine sigil]

Img2 Jesús

[viral bough]

i’m tryna do this “Time Machine Series” thing
basically using google veo videfx to generate like... gopro-style pov videos but set in old timey eras

heres Harlem street scene during the 1920s Harlem Renaissance era.

Ancient Rome around 80 AD

[viral bough]

this looks like cosplay convention going on, and it’s the only gen so far that didn’t sneak in some random modern tourist with a laptop bag or backpack

Medieval Paris – Notre-Dame Area (1200s)

[viral bough]

The 24fps really made it look naturally realistic

[rotund wagon]

image to video makes everything perfect 🙂

[sly glacier]

They need to fix img2video. Still has errors. I only made 3 vids with it.

[rotund wagon]

what errors?

[sly glacier]

Not able to finish the generation process. Almost every image is declining

[viral bough]

in my case it failed to generate bcs of the auto prompt it did when we submit it

like one of mine was a studio ghibli artstyle image, but i didn’t even prompt it to animate like studio ghibli style or anything, and it still failed. i checked the network log, looked at the prompt it generated for me and surprise surprise… it actually wrote “studio ghibli” in the prompt lol.

[sly glacier]

Lol

[viral bough]

and yeah i do know tools like imagefx whisk and videofx will straight up refuse to generate anything if the user prompt has “studio ghibli” in it lol

[sly glacier]

It's all messed up. Never had the problem in freepik

[viral bough]

the safety filters they put on the videofx tools feel a bit stricter now tbh

[sly glacier]

I use normal images tho

Cars for instance

[viral bough]

I mean we don’t know how the tools itself interpret what you wanted , it definitely has something to do with their safety filters

[sly glacier]

It's bugged

[viral bough]

I’m not saying your image input was wrong lol that’s not what I was saying

[sly glacier]

Plus the ratio shouldn't be 16:9 only by default

[viral bough]

DOPE

[sly glacier]

Using 16:9 ratio?

[viral bough]

I mean to be fair

Yeah I agree

[sly glacier]

That's a problem I don't use that ratio. Most of my images out of thousands

[viral bough]

The reason u can do it on Freepik is because Freepik autocrop your image to 16:9 lol

Which is what they’re working on right now

[sly glacier]

They need auto crop then

[viral bough]

Yeah they’re working on it lol

It’s being posted in #1319424187730755636

[sly glacier]

Lol it's wild the main feature i can't use that's free

[viral bough]

maybe voice your complain properly in VideoFx bug thread tbh

[sly glacier]

My library has a bunch of invalid generations

[viral bough]

like I understand your frustration but at this point it won’t help both users and the team either

W I D E WOLF

[sly glacier]

I ain't worried since it been mentioned. I'm always using other tools while even typing this

[viral bough]

I think someone here did like uhh

Let me find.

[viral bough]

This one ☝️

[sly glacier]

Eh...I don't want to speak on that

I'm in a split

[viral bough]

From what I’m seeing they put image that’s not 16:9 ratio in a 16:9 canvas size

And it can help with making consistent subject too!

You mean your masterpiece art 😎

[bleak bridge]

New music Friday, so the chance to play with new songs to make YouTube shorts. Here: Kesha's "Yippee-Ki-Yay"

The new uploaded-image-to-video feature really saved me here. I couldn't get VideoFX to create a good video of a smartphone burning in a fire, but that was a one-shot task in Midjourney, and VideoFX did a solid job animating, once it could see where I wanted it to start from.

https://youtube.com/shorts/PRWWiyfXD00

[viral bough]

Very good!!

[raw torrent]

New day, new car. I2V.

[manic hedge]

somewhat cartoonish aliens

more realistic version of the last one, with added limbs

[bleak bridge]

Another short music video where upload-image-to-video cut down the time when a pure text prompt struggled with framing the horror/humor in the final shot.

https://youtube.com/shorts/UcNUgbrTT88?feature=share

[proud forum]

Perfect Veo 2 AD

Here's another one

[proud forum]

If it wasn't for the text but still good!

[grim oar]

I have a question, will credits for video generation be added for veo 2?

[viral bough]

Eventually yeah 🙂

[viral bough]

They've already added credit(100) for Whisk Animate, which will reset monthly once it's fully used up.

Just saw the new string update, something like “With the Google AI Advanced plan, you’ll get access to Veo 2.” And “Manage Membership” from VideoFx source page.

So this whole thing’s getting bundled into one big plan , like how Google already dropped the native image tool through Gemini, and soon they'll roll out codename Robin for video generation in Gemini too.

[grim oar]

I meant about veo 2 which is in google labs

[viral bough]

yeah they will eventuall add it

[grim oar]

I really hope that they won't add credits and that this will be the only free neural network for generating videos

[viral bough]

Oh believe me, the only reason it’s “free” for us early access folks is because they want to refine the tools and gather feedback from us before rolling it out to the public. Which makes sense

[proud forum]

They will add what?

[viral bough]

#flow-general・🌊 let’s move to here so we don’t bury other video here

[strange stirrup]

these are wild!! thanks so much for sharing

[proud forum]

Veo 2 on phone is a joke

[shell jolt]

💀

[dire barn]

i want one now

[shell jolt]

hahaha thats fire 🔥

[proud forum]

Guess what it is. Hint: It's not snow

[viral bough]

W I D E

Img2Video (image generated with image_gen tool from chtgpt)

img2video ( image generated with halfmoon)

[crystal shore]

That looks amazing

[viral bough]

but it glitched out when he start walking huhu

i guess thats the downside of fast mode

[crystal shore]

Maybe you should try adding an end frame for it not to glitch out

wait idk if Videofx has that feature

[viral bough]

only start frame for now ahaha

[crystal shore]

Theyll probably add it later haha

[viral bough]

love how it didint disturb the text in image

also another try

[crystal shore]

Same, the text worked pretty well. Even though the ones in the corner are all wrong from the beggining

[crystal shore]

That run was really cool

[viral bough]

the jump cut was something i didnt prompt it to do but it looks good at the end

[viral bough]

well! i didnt prompt it to be cartoonish water ,

[lusty river]

Pls give excess 🥹

[sly glacier]

Ooh our names are now blue

[proud forum]

AI Minecraft Decart gameplay 😄

Curious on how Veo 2 knows EVERYTHING about Minecraft

[grim oar]

I also ask myself the same question. Maybe he was taught by watching Minecraft videos.

[proud forum]

Or searching up minecraft and making the videos 😮

[raw torrent]

I2V experiments with mural

[raw torrent]

The Getaway Car

[alpine pine]

Very Cool

[viral bough]

img2video with my own drawing, and 🥺 ❤️ it transformed into 3d style

[fallen kernel]

Did you prompt for it?

[viral bough]

Same

[viral bough]

I didn’t specifically wrote “3d” it was more of general “animate this”

Tho it was just one out four video that it generated in 3d

[supple patio]

The ai probably thought you meant animate as in digital animation

[tall geode]

Truly, the Veo 2 is the best model for video and img-to-video. (It even added a real interaction of the created object with the environment of the input image) 🤯.No matter what anyone says, Google surpasses everyone in multimedia by three light years!

[final pendant]

https://whatnot.com/s/JARIpRFa

[flat quail]

is anyone else getting "oops something went wrong" with veo image to video right now?

[bleak bridge]

Quoth the Raven, 'Let's Disco.'

https://youtube.com/shorts/F5LXhatxdHs

[raw torrent]

T2V

[proud forum]

Both have different sounds and music

[raw torrent]

I2V We're working on it! 😂

[proud forum]

Who likes my new AI Generated Minecraft House??

[bleak bridge]

Hmmm, that's odd, the song was added directly to the video clips inside the YouTube app, so there shouldn't be any issues there. I've just been able to see it out three different devices, and YouTube Studio is saying that the watch rate is typical.

Can't think why you recieved that message, unless it's a blip or oddity related to how Discord is being invoked on your device. Weird that you screen is showing "watch it on YouTube" on what appears to be a YouTube screen.

[raw torrent]

When I click the actual YT link in the Discord msg it takes me to YT and the video plays. But if I click the arrow to play within Discord, it shows that msg. Also, it's not the first time. Just thought you might want to know.

[proud forum]

My discord is lagging like I'm in the Sahara

[viral bough]

@novel basin

Instead of organizing furniture

He snapped himself to dust

[raw torrent]

I2V with a Volkswagen Rabbit as a reference image.

Born in 1975, Rusty the VW Rabbit has so many stories to tell, like that time when Otter and Snail saved Parrot!

[proud forum]

This AI is too sentient. It made the Earth fully no problems

[viral bough]

The scene opens with a dynamic handheld shot, mimicking a real-life documentary style, focused on a busy construction site bathed in natural light. The camera follows an anthropomorphic realism but animated face, male Shiba Inu, dressed in a realistic construction worker's outfit—complete with a bright orange safety vest, a weathered hard hat, and sturdy work boots. He strides confidently through the bustling environment, carrying a blueprint roll in one hand. Dust motes dance in the sunbeams as he surveys the ongoing work. Other construction workers, both human and animal, are scattered throughout the background, operating heavy machinery and carrying building materials. The shallow depth of field blurs the distant activity, keeping the focus firmly on the Shiba Inu as he pauses, unfolds the blueprint, and points towards a specific area of the unfinished structure with a gloved paw, seemingly giving instructions. The camera subtly shakes, emphasizing the immediacy and authenticity of the shot.

[compact dune]

thanks for sharing your prompt! love to see them

[crystal shore]

Using your same prompt, amazing result 🔥 . "The camera opens with a shaky cam flying low over a frostbitten mountain ridge during a whiteout blizzard. Emerging from the snow, a burly anthropomorphic polar bear in a bright red arctic rescue uniform trudges forward with urgency. His breath fogs up his cracked goggles. He scans the horizon with a thermal scope, snow whipping at his face. A beacon blinks weakly in the distance. Behind him, a sled drone with emergency supplies hums quietly. As he locates the survivor buried under the ice, the wind dies for a moment—and you hear only the sound of a heartbeat and crunching snow.

[viral bough]

Omg that looks amazing

[crystal shore]

Thank you! Yours too

[raw torrent]

🦦 🌸 🗺️ 🌴 "The scene opens with a dynamic handheld shot, mimicking a real-life documentary style, focused on a busy tropical beach bathed in natural light. The camera follows an anthropomorphic female otter, dressed in a realistic tropical vacation outfit—complete with a bright floral top, sarong, straw hat, and flip flops. She strides confidently through an outdoor market, carrying a cotton market tote with a blooming flower logo. Dust motes dance in the sunbeams as she surveys the bustling fruit and vegetable market. Other travelers, both human and animal, are scattered throughout the background, shopping in the outdoor market and carrying their market finds. The shallow depth of field blurs the distant activity, keeping the focus firmly on the otter as she pauses, unfolds her tourist map, and points towards a specific area of the market with her paw, seemingly pointing out her destination. The camera subtly shakes, emphasizing the immediacy and authenticity of the shot."

[viral bough]

Ahhh it totally keeps that same style, even with a different animal! 🤩

[crystal shore]

Oh wow, first time I see 3 different shots in one shot, amazing

[viral bough]

I am loving these

[crystal shore]

Thank you!! Your prompt works so well!! I have generated 30 videos like these alerady 😂 and all of them look amazing

[radiant totem]

This so cute and beautiful

[compact dune]

really love this style, the lighting, the mood -- super charming 🐂

[compact dune]

I appreciate every time you make these

[sly glacier]

I just like doing it. A challenge for me to create a new one differently

[agile fulcrum]

Can it create videos in loop?

[crystal shore]

Thank you!

[crystal shore]

Appreciate it! I really liked it too

[stray vector]

My newest music video made with VEO 2 🔥🎥

[magic drum]

Outtakes from my last film shoot 🤣

[proud forum]

Oh my GOD the quality of the first 2 scenes of my story is SO FREAKIN GOOOD

[proud forum]

https://www.youtube.com/watch?v=BdSj-iJQBj0&ab_channel=Veo2Movies

[crystal shore]

My first video with sound 🔥

One is made with Veo2 and the other is a different model, which one do you guys prefer?

[viral bough]

This one is veo for sure

What model was it on the first one?

[crystal shore]

Sora

Veo looks more realistic in my opinion

[rocky basin]

What are you on about, that Sora one looked way better 😂

Sora for sure, I would say Veo 2 for most gens, but for that one, Sora has the win

[rocky basin]

#MESSAGE_REMOVED_BY_MODERATORS#

[viral bough]

cool

[viral bough]

A slow, majestic camera glide reveals an extreme wide shot of Saturn’s glowing rings stretching across the cosmos, shimmering with icy particles and cosmic dust. Amid this surreal expanse, a lone ballerina in a flowing silver-white tutu pirouettes gracefully on the translucent curve of the rings, her movements fluid and ethereal. The shot captures her in perfect contrast against the deep void of space, Saturn’s immense, shadowed surface looming behind her like a divine spectator. Stars flicker in the background, and a soft lens flare glows as distant sunlight refracts off the ice. Subtle motion blur trails her limbs as she leaps and spins in slow motion, her delicate pointe shoes touching the ring’s surface with impossible lightness. The camera, set to a high depth of field and wide-angle lens, floats steadily backward, enhancing the scale of the planet and the dreamlike isolation of her performance in the silent, endless expanse of space.

[viral bough]

I just realized there’s two sunlight in the video

[gusty isle]

https://youtu.be/F6IEvla0OCI?si=yIDXLb2OQ4gNxiup

[crystal shore]

That’s why I said Realistic not “Better.😂 look at the composition and shot of the Veo and then compare it to Sora, I would say it is more realistic

[crystal shore]

I love the 360 that it does 😂

[chrome tangle]

happy to see everyone is having fun with the model

[cloud obsidian]

Prefer Wicked Games one for the colors & the sound 👍

[crystal shore]

Thanks for the feedback

[crystal shore]

Appreciate it

[raw torrent]

I have different “feelings” for each of them.

For the first one, it feels like I’m watching from the first person pov of maybe a tourist with a camcorder filming a monkey (because of the way it zooms in and out). Although, because the monkey is doing such a great job playing the guitar, it makes me feel like this tourist might be imagining things. For the next shot, I’m expecting to see what the tourist is going to see next.

For the second one, it feels like I’m watching a movie about a monkey, and the monkey is the star. The monkey seems more realistic because of the way that it’s not really playing the guitar, more like it might be mimicking a human it saw doing that. Also, the rock seems more realistic. For the next shot, I’m expecting to see what the monkey will do next.

[rocky basin]

Mate, Sora looked way more realistic

[pseudo cedar]

Google has a whole youtube and this is the best they can do. Img to vid seems mid

[crystal shore]

Hey, thanks so much for your feedback! I really love how you described the different vibes you’re getting from each scenario, it’s exactly the kind of insight I was hoping for.

For the second scenario, your movie analogy is spot on, that’s exactly the feel I wanted, with the monkey as the main character.

Appreciate you taking the time to break it down like this!

[abstract rose]

Anyone else having issues with Veo today? I haven't been able to get an output in almost an hour.

[bitter shard]

@abstract rose I keep getting "something went wrong" for the past hour as well. Is that the same issue you are having?

[abstract rose]

yeah

[eager crown]

Im getting an error as well "Opps, something went wrong"

[quick river]

Yep, they’ve already been alerted.

[stiff hinge]

Only now I think of this since it's available now. Let me see what I can do.

[proud forum]

ChatGPT made this for my channel logo believe it or not

[viral bough]

img2video Fast Mode (img generated using halfmoon)

[viral bough]

img2video using my own drawing again (Fast Mode)

Tried img2video on regular mode with this same sketch, it turned my sketch into a fully colored drawing even though I never asked for that 😅 it really caught me off guard

[bleak bridge]

Very nice!

[sullen pivot]

https://www.reddit.com/r/kingsvilletxnsfw/s/3ozoDVDiiA

[proud forum]

Your drawings are AMAAAAZING. Even Veo 2 agrees!

🍎 Stay healthy!

[bleak bridge]

Interesting!

[bleak bridge]

New Music Friday, so my usual chance to practice on themes I can't prepare for. While photorealistic videos seem to be very appealing to people generally, I personally love things that are a bit more conceptual, like this origami protagonist.

(note, YouTube shorts with licensed music can sometimes not play in Discord, but should play on YouTube directly)

https://youtube.com/shorts/U84y1bu3Jlk

[chrome tangle]

what is wrong with this prompt!??

[sudden goblet]

Beautiful!

[chrome tangle]

still says its bad!!

why is footage a cencored word?!?!

[viral bough]

the underline? i dont think thats the word that triggers the filter..?

is it img2video or textto video @chrome tangle

@sudden goblet also thanks!!

[chrome tangle]

text2video, the underline is just a spellchecker

i removed "footage" and the prompt now works

[viral bough]

oh

wait so, the initial idea is walking on moon right

[sudden goblet]

Because we never went!

[viral bough]

let me try something

[sudden goblet]

just kidding haha, idk thats weird

[chrome tangle]

i wanted first person POV footage walking on the moon

[chrome tangle]

like this video i made with veo

[viral bough]

its generating

[chrome tangle]

i tried other prompts with "footage" and it seems to work, but not on the moon

[viral bough]

mine worked tho it was like one video didnt generated

i only did 2 output

[viral bough]

trying out yours

[chrome tangle]

seems to work, idk why "Moon" in cencored

[viral bough]

a very long prompt (i used prompt assistant)
`Real-life, raw, amateur footage, shaky handheld camera, first-person POV shot
Subject Focus: A first-person view through a spaceship window focused on Earth below
Subject Movement: The camera slightly shifts and jitters with natural hand movement as the person gazes out at the slowly rotating Earth
Subject Details: The scene shows a curved glass spaceship window with subtle reflections, edges of the window frame in view, and occasional flickers of onboard control panel light glinting on the glass
Scene: Outside the window, Earth is half-lit, with visible atmospheric glow, moving slowly in the blackness of space; faint stars speckle the background while the inside of the ship remains dimly lit
Camera Language: Handheld first-person POV with subtle natural shake, medium framing, mimicking human eye level and head motion
Lighting: Low light from inside the spacecraft, combined with the soft, ambient glow from Earth’s surface and rim light from the sun in the distance
Texture: Scratched glass window, smooth metal framing, ethereal Earth cloud patterns, faint noise grain from low-light capture
Atmosphere: Intimate, contemplative, and slightly awe-struck, with a raw, unpolished realism that evokes a sense of isolation and wonder
Environment: The interior of a spaceship in orbit, dark and minimally detailed except for the viewport; Earth dominates the external view, suspended in the vast silence of space

A first-person, shaky handheld video capturing the raw and unfiltered experience of looking out a spaceship window at Earth, with subtle head movement, dim lighting inside, and Earth slowly turning beneath the stars, seen through scratched glass and accented by low glow from both interior instruments and distant sunlight.`

[viral bough]

btw i found fast mode generated the exact thing u wanted, the regular mode , only 1 out of 4 did "pov" style

Fast Mode

[viral bough]

i find it weirdthat it cant do moon , cuz it generate moon when i use imagen 😮

but not for videofx

Shaky, real-life, handheld camera footage from a first-person, point-of-view perspective looking out of a spacecraft window. The camera subtly pans across the view. The bright sun flares intensely as the vast, blue Earth slowly rotates in the distance, dominating the shot. A thin layer of the atmosphere is visible, shimmering slightly. Parts of the spacecraft interior are visible at the edges of the frame, like a dark, metallic frame of the window and a glimpse of control panels with flickering lights. The overall lighting is bright, reflecting off the glass, with some lens flare. The amateur footage shows the point-of-view perspective of an astronaut in a blue uniform, with only parts of the sleeves, hands holding onto the window frame, and their breath fogging the glass.

[agile fulcrum]

With a single prompt?

[viral bough]

This is a fascinating one , the prompt assistant gave me a character name that i didnt even ask for, but it generated very well done.

Raw, shaky handheld camera footage from inside a spaceship, showing a first-person perspective. The camera slowly pans across the window, revealing a breathtaking view of Earth. Sunlight glints off the window's surface, creating lens flares that dance across the frame. The Earth hangs in the blackness of space, a vibrant blue and green marble swirled with white clouds. The camera wobbles slightly, emphasizing the amateur, raw feel of the footage. Through the reflection in the window, we see the faint silhouette of the astronaut, Dr. Aris Thorne. The faint outline of a second astronaut, mission specialist Lena Hanson, can be seen next to Thorne. Their faces are obscured by shadow, enhancing the POV perspective and focusing attention on the awe-inspiring view.

[agile fulcrum]

Hey @viral bough how I convert image to video in whisk animate, where to upload the image in object scene or style?

[viral bough]

oh u cant do whisk animate on your "own" uploaded image

its only whisk generated image that can be animated

[agile fulcrum]

This is video FX right?

[viral bough]

also #whisk-general・💭 if u want to ask more about whisk related

yeah its videofx

[chrome tangle]

no, many diffrent videos and prompts

[sleek pendant]

Wes Anderson cable news

[viral bough]

im surprised u can generate anything involving "wes anderson" in ur prompt, mine refused it

[worldly cape]

I am getting some great consistency with just prompts right in google shorts. https://youtu.be/sO15GKnPfnA?si=fS8v6TKJrJN1YIZN

https://youtube.com/shorts/54usKEm-XFE?si=iqlKVmcyklkwR6wc

[sleek pendant]

Yeah he wasn't mentioned by name

Full prompt: A meticulously designed, pastel-colored cable news studio set. A male and female pundit sit side-by-side at a vintage wooden desk, both facing the camera directly. They are engaged in a heated, rapid-fire debate, but with deadpan expressions. The man wears a perfectly tailored tweed suit with a too-short tie. He has a slender, slightly lanky build, with perhaps slightly unkempt hair and a thoughtful, rather than athletic, appearance. The woman wears a bright, symmetrical dress and perfectly coiffed hair. She has a slim, perhaps slightly petite frame, with a stylish but not overly-muscular physique. Vintage microphones sit in front of each of them. Beneath each pundit, a static, centered chyron appears. The man's chyron reads "POINT" and the woman's reads "COUNTERPOINT". The chyrons are in a simple, sans-serif font with a neutral color palette, such as a light beige or pale gray. Behind them, a whimsical, diorama-like depiction of the Capitol Building is visible.

[worldly cape]

Those look amazing.

[viral bough]

ahh understood

[sleek pendant]

Best img2vid result I've gotten. Uploaded a close up cell phone pic of a little Buddha figurine.

[viral bough]

was this on fast mode or regular mode ? 😮

[sleek pendant]

I believe in regular. Haven't done much img2vid. Back to testing after getting access to FX today. Is there no vertical scale for FX like on freepik?

[viral bough]

as for now nope,

but u can upload non 16:9 image , and it will pop out a cropping tool , which it will ask u to crop the image

[lusty cape]

How to generate in Veo 2 in Gemmini?

[rotund gulch]

A dimly lit, smoke-filled room from the 1950s (beatnik vibe). The walls are plastered with torn astronomy maps, old typewriter pages, and philosophical quotes in fading ink. A single, flickering bulb hangs from the ceiling. Our protagonist sits at a cluttered desk, furiously writing in a notebook, surrounded by empty coffee cups and cigarette butts. Suddenly, the camera starts to _slowly zoom out_ and the room begins to dissolve like sandcastles in the rain, pages disintegrating, maps tearing apart, until we're left with just a contemporary, empty, sterile apartment space — same desk, but now blank, cold, and modern. Transition seamless, like memory evaporating

[proud forum]

I GOT ACCESS TO WHISK ANIMATE :DDDDDDD

[rotund gulch]

A futuristic night cityscape with neon signs reflecting off puddles in the street. A solitary figure in a hood walks slowly, shoulders burdened with the weight of the world. Shadows shift in and out, dancing with menacing shapes. The scene pulses in rhythmic slow motion, matching a heartbeat-like bass

[proud forum]

Oh wait it didnt give me the "Animate" button thingy

[compact dune]

are you using the link within the acceptance email? that custom link is the only way to try Animate at this time

[proud forum]

Ohh thank you so much

[compact dune]

no worries! feel free to share your gens in #whisk-gallery・🖼 too 🙂

[quick river]

Why red color!

[surreal yoke]

Impressive. First frame generated from Midjourney v7

[viral bough]

Realistic handheld footage from the passenger seat of a car shows a chaotic scene unfolding in a sunny, modern city. The camera shakes slightly as the car navigates the absurd situation. Through the windshield, a giant, fluffy Maine cat is visible atop a gleaming glass skyscraper. The perspective emphasizes the colossal scale of the cat. The cat is playfully tilting a colossal ceramic bowl, from which a massive waterfall of creamy white milk cascades down, partially obscuring the view. Cars are sliding uncontrollably on the slick surface, their metallic bodies reflecting the sunlight, the scene captured through the milky-streaked windshield. people are seen panic running towards from cat, The daylight illuminates the scene with a bright, cinematic quality, casting long shadows from the skyscraper and highlighting the absurdity of the milky deluge. The backdrop of the city showcases pristine modern buildings and a clear bright sky.

[viral bough]

The camera pans across a colossal, multi-tiered cake, its scale so vast it resembles a surreal landscape of confectionery wonder. The entire cake is covered in a billowing expanse of fluffy white whipped cream, sculpted into gentle slopes and peaks like a dreamy cloudscape. Scaling this creamy terrain are actual human, real-life-sized pastry chefs, each meticulously adorned in pristine white uniforms and tall, elegant hats. With focused intensity, they navigate miniature ladders, their movements precise and purposeful. Each chef wields delicate piping bags, crafting intricate swirls and patterns across the whipped cream surface. Some carefully place edible sugar pearls that catch the light like tiny jewels, while others use minuscule spatulas to smooth sections of the frosting to perfection. The scene is illuminated by a soft, diffused light, casting a gentle glow that emphasizes the delicate texture of the whipped cream and the detailed features of each pastry chef, from the focused expressions on their faces to the fine stitching on their uniforms. The background is intentionally blurred, creating a shallow depth of field that keeps the focus tightly on the chefs and their delectable landscape.

[viral bough]

this is a fun one
A first-person perspective shot using a shaky GoPro with an ultra-wide lens captures the surreal scene. The camera is handheld, emphasizing the miniature world inside a pint glass. Inside the glass, a realistic tiny lake teems with life: small fish dart through the clear water, and miniature birds flutter and perch above the water's surface. Tiny animals, including rabbits, deer, otters, and snails, are seen relaxing near the lake's edge, their minuscule forms adding to the dreamlike quality. The pint glass rests on a wooden table beside a plate of real, human-sized food, creating a stark contrast in scale and emphasizing the bizarreness of the tiny ecosystem. The lighting is bright and natural, highlighting the vivid colors of the miniature landscape and the textures of the surrounding environment.

[viral bough]

The camera smoothly glides in for a close-up, capturing a realistic, sun-drenched scene. A woman, appearing as if from a classic film, vigorously pushes a vintage, red-and-chrome lawn mower across a small, perfectly square patch of vibrant, green grass. Sunlight bathes the lush grass, making it appear extraordinarily healthy and well-maintained. This peculiar grassy platform floats peacefully in a vast, naturalistic space filled with softly glowing stardust. The woman, in her light blue dress with delicate white polka dots, radiates a nostalgic, mid-20th-century charm. She wears spotless white gloves, her face set with determination as she meticulously mows her tiny, isolated lawn. The stark contrast between the cozy scene and the infinite backdrop creates a playful and charming atmosphere, emphasizing the delightful absurdity of the situation. Softly illuminated clouds drift around the floating platform, adding to the dreamlike quality of the scene, all under the warm, natural glow of the sun.

[sudden goblet]

Midjourney has really fallen behind its competitors, im interested to see if they will still be in operation a year from now

i think multi modal models are becoming a must

[viral bough]

i heard many ppl saying it struggle with arms finger and such which ihave yet to see those results yet

[sudden goblet]

I can confirm, it looks like a slight update from V6 overall, which when V6 came out it was SOTA, right now with the recent bangers from google,openai,.... Its already outdated

[viral bough]

really hope google can catch up...the native image on aistudio is promising

[sudden goblet]

yep i think they will honestly, looking at the video they are already ahead and with a few tweaks their native image gen will get there

[crystal shore]

It should’ve been v6.2 or v6.5 definitely not v7

[sudden goblet]

Agree!

And for 1+year of waiting... definitely not worth

[crystal shore]

Insane

And I don’t understand how so many people are defending this 💀

It was really disappointing imo

[sudden goblet]

Haha for many people MJ was their first image gen model, I think they are just being fans honestly, but it will hit them later

[crystal shore]

MJ was my first too haha, I used to love it so much

Sad its not good anymore

[sly glacier]

I see the mj discussion...

[viral bough]

yea michael jackson is very iconic

🙂

[crystal shore]

https://tenor.com/view/conanwolf-gif-15824252101317178951

[sly glacier]

Yeah the real version

[viral bough]

will the real version of mj stand foward

[crystal shore]

that looks so cool

[viral bough]

ok we gettin off topic nowLOL

let me try generate mj in veo