Or stuff needs to be updated?
#▶｜stable-video-diffusion
cyan turret
velvet pond
試試在海螺/混元/可靈用
silent hinge
A parachute torn from between mid flight video
shadow light
Logo design for "Splendora", a premium-accessible multi-category brand. Elegant serif typography with a custom "S" featuring a subtle light reflection element. The logo embodies "luminous elegance" - sophisticated yet warm. Color palette: Champagne Gold (Pantone 873C) and Lunar White background. Include the distinctive "Light Prism" element - an abstract geometric motif representing light refraction through high-quality surfaces. The typography should be based on Optima Nova or similar contemporary serif font. The logo should convey quality, refinement, and accessibility simultaneously. Design should work across cosmetics, jewelry, fashion, and lifestyle categories. Create multiple variations: 1) Full logotype "Splendora" 2) Monogram "S" with light element 3) Extended version with category line (e.g., "Splendora Derma"). Style: clean, minimal, with attention to subtle details that communicate sophistication. --ar 3:2 --v 5 --style raw
green turret
Are there any good video generation methods that work with A1111?
unreal otter
So I've seeing this style of videos for a while now. I can't really find anything on how to make them though.
The special thing is, that in this case, the Logo of the Label is kind of constantly visible in the clouds, while still being consistent in the sky image. Anyone any idea?
https://www.instagram.com/p/DGyNhoVCCHy/
Video Im refering to.
narrow plover
So I've seeing this style of videos for a while now. I can't really find anythin...
They're probably using controlnet (stick figure skeleton thing), and just saying "a stick figure in the clouds" or something, plus lighting effects (god rays, crepuscular rays, etc)
unreal otter
They're probably using controlnet (stick figure skeleton thing), and just saying...
the stick figure is actually the labels logo
So it's def been input as some kind of reference
pliant current
So I've seeing this style of videos for a while now. I can't really find anythin...
If i would need to create such a video i would to it this way:
- Create a still image with SDXL and the QRMonster Controlnet Model (optical illusion)
- Use this image as an input for Image 2 Video to get at least 1-2 seconds movement.
- Turn the video by 90° and add it to the script with some davinci resolve effect
- Change the prompt of Step 1
- Repeat 2-3 and add the result to the video using a morphing or blending
unreal otter
If i would need to create such a video i would to it this way:
1. Create a still...
Thanks for the input! I will try that over the next days.
Do you have any idea on how the blending of the logo is done?
He has some insanely creative videos like this.
Check Slide 3 here for example:
https://www.instagram.com/p/C7CSVsasFVF/?img_index=3
pliant current
Thanks for the input! I will try that over the next days.
Do you have any idea o...
Just give me a logo and a subject
Then i can explain it to you
narrow plover
So it's def been input as some kind of reference
Oh, they might have used depthmapping for that.
unreal otter
Ohh wow! How did you do that? 😮
pliant current
Ohh wow! How did you do that? 😮
Step 1 with the logo from your message. Surely I would resize it a bit smaller but the procedure would be the same.
unreal otter
Step 1 with the logo from your message. Surely I would resize it a bit smaller b...
I managed to get it to work for text to image, with prompts.
But the images generated are super low quality. Is it also possible to use an existing image and then modify it afterwards?
pliant current
I managed to get it to work for text to image, with prompts.
But the images gen...
Please check the sdxl model, loras and negative/positive prompt to achieve better results. Adding the logo afterwards won’t be possible as the objects are positioned during the generation.
narrow girder
Подчеркни ретро-эстетику phonk, используя визуализацию в стиле старых видеоигр. Представь, что твой трек — саундтрек к игре, и создавай сцены, где персонажи выполняют различные действия под ритмы твоей музыки.
shadow light
Show me a Mazda 3, white color, in a post-apocalyptic environment, Final Fantasy VII style, in a world inhabited by anthropomorphic animals, the sun is rising at the horizon, and the car pass by the scene in slow motion
night oriole
Any kind soul out there who can turn my image into video with their AI? Dm me pls
mild mirageBOT
how to generate images?
There are various methods to generate images, depending on your needs and tools. Here are some popular approaches:
-
Graphic Design Software: Programs like Adobe Photoshop, Illustrator, and GIMP allow you to create images through manual design. You can use brushes, shapes, and layers to compose images.
-
3D Modeling Software: Tools such as Blender or Autodesk Maya enable you to create three-dimensional images or animations. These platforms are ideal for designing objects, characters, and environments.
-
AI-Driven Image Generators: Several AI tools can produce images based on text prompts. Examples include DALL-E, MidJourney, and Stable Diffusion. Users input descriptive text, and the AI generates unique images from that.
-
Drawing Tablets: If you prefer hand-drawn images, using a drawing tablet connected to graphic design software can enhance your ability to create detailed illustrations.
-
Web-Based Applications: Websites like Canva allow you to make images easily using templates and stock photos, suitable for social media posts, marketing materials, and more.
By selecting the method that best suits your project's needs, you can effectively generate high-quality images.
light briar
/ Millet electronic lock, lock core close-up,3D structured light, Muji style
frosty jolt
retro radio microphone pulsating
cosmic pasture
sd3 results
POV: You wake up as a woman Pharaoh in Ancient Egypt -- 1338 BCE
dusk berry
my first animation in ComfyUI
dusk berry
cosmic pasture
sd3 results:
POV: you wake up as a woman worker in The Titanic -- 1912
slim escarp
Hi there! What would can wan2.1 video to sometimes look "accelerated"? like the are sped up at 2x
cosmic pasture
POV: you're a woman doctor during the Black Plague --1317
muted kite
Who got a notebook for an image to video generator
knotty gale
/create prompt: "Timelapse of plant roots growing from thin to thick after using humic acid, soil cross-section view, science documentary style, 8k cinematic"
#▶｜stable-video-diffusion Timelapse of plant roots growing from thin to thick after using humic acid, soil cross-section view, science documentary style, 8k cinematic
#▶｜stable-video-diffusion "Plant growing time-lapse, roots expanding in soil, 4k cinematic"
wooden granite
The manipulator touches the super big screen
#▶｜stable-video-diffusion Boundless grasses over the plain, artistic conception painting
cosmic pasture
POV: You wake up as a Pyramid Builder in Egypt 1550 BCE
rustic crystal
vagrant kayak
Has anyone here worked on mamba backbone for diffusion?
balmy spire
cosmic pasture
POV: You wake up as a Gladiator in Rome 100AD
cosmic tide
What is the reason for black video output with WAN?
rapid willow
Recently, I've been trying to match the clip precision to the model. You're running a fp16 model against an f8 text encoder. Try pulling down the fp16 version. What video card are you using?
https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/text_encoders/umt5_xxl_fp16.safetensors
lavish rain
cosmic pasture
POV: you wake up as a rich Knight in England 1199 AD
cosmic pasture
POV: you are the Violinist in the Titanic as it sinks --1912
slim escarp
Hi there! I've been generating Wan videos with no problems for a couple of weeks, and all of a sudden this morning I'm getting thi serror:
Sampling 93 frames at 512x752 with 50 steps
14%|█▍ | 7/50 [02:23<14:41, 20.50s/it]
!!! Exception during processing !!! local variable 'temb' referenced before assignment
any idea what that can be? I updated all nodes+comfy after that, but the issue remains
File "/workspace/ComfyUI/execution.py", line 347, in execute
output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
File "/workspace/ComfyUI/execution.py", line 222, in get_output_data
return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
File "/workspace/ComfyUI/execution.py", line 194, in _map_node_over_list
process_inputs(input_dict, i)
File "/workspace/ComfyUI/execution.py", line 183, in process_inputs
results.append(getattr(obj, func)(**inputs))
File "/workspace/ComfyUI/custom_nodes/ComfyUI-WanVideoWrapper/nodes.py", line 3187, in process
noise_pred, self.teacache_state = predict_with_cfg(
File "/workspace/ComfyUI/custom_nodes/ComfyUI-WanVideoWrapper/nodes.py", line 2820, in predict_with_cfg
noise_pred_cond, teacache_state_cond = transformer(
File "/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py", line 1750, in _call_impl
return forward_call(*args, **kwargs)
File "/workspace/ComfyUI/custom_nodes/ComfyUI-WanVideoWrapper/wanvideo/modules/model.py", line 1195, in forward
del temb```
zinc yoke
```Traceback (most recent call last):
File "/workspace/ComfyUI/execution.py", ...
Don t click the scam link please. I m gonna clean it.
lime dew
bye
lavish rain
still kettle
cosmic pasture
sd3:
POV: you wake up as a Pirate in the Caribbean , year 1715
cosmic pasture
POV: you wake up as Athena in 1000 BC
dreamy coyote
A friend of mine is competing here, vote if you like his work! https://runwayml.com/gen48?film=wandering-woods
cosmic lion
A friend of mine is competing here, vote if you like his work! https://runwayml....
He got my vote. Pretty creepy yet intriguing.
dusty mountain
cosmic pasture
sd3
POV: you wake up as the last Queen of Egypt during the fall of Alexandria --30BC
dreamy coyote
He got my vote. Pretty creepy yet intriguing.
thanks! yes indeed
buoyant river
I created music video with EDM anime beats, and rendered it like a decade x-x
https://www.youtube.com/watch?v=zQGSBIUs6rI
tranquil moon
hola
coarse sierra
Can anyone with a good PC, create a short video for me with a reference image?
cosmic pasture
shy oar
The orange cat in a suit and leather shoes is holding a cigar and fiercely patting the office desk with its paws, creating an 8K realistic style
tough tinsel
where and how do i get the Stable diffusion to do Video?
olive haven
Hi. what do you guys use to create videos?
pliant current
Hi. what do you guys use to create videos?
Fast generation with ITX, quality with WAN or Hunyuan
rapid willow
I've been playing around with FramePack
dense talon
Looks great. Wish my machine was powerful enough
magic osprey
does anyone has a good wokflow for comfyui for background replacer for video?
vagrant glacier
Anyone have any good models for sd sdnext or even site that can make the talking baby videos
wintry halo
let's say i have a 10 second video of me dancing, i took the video and broke it down into 240 frames give or take. how do i add an anime filter to it without changing too much? And put it back together into a video?
i am seeing videos of people doing this and it looks incredible with 100s of anime images flashing and recreating the video. It's insane and kindof glitchy looking but in a good way.
when i try this, the images per frame are coming out way different that the original picture/frame. Any ideas?
dusty mountain
rapid willow
Two goons requires two guns.
spark abyss
fiery latch
Hello!
Could someone please share a working download link for the SD 2.1 motion_adapter.safetensors file (v1.1, ~1.8GB)?
I’m looking for the official AnimateDiff v1.1 adapter (not SDXL or Lightning), SHA256:
5fd196988206cfac90bab1dc56cc59fdd12035492f74ff0662eb6eed5113c547
or
45b71fe98efe5f530b825dce6f5049d738e9c16869f10be4370ab81a9912d4a6
The official Hugging Face repo is down/private. If you have a Google Drive, Dropbox, or other trusted mirror, I’d really appreciate it!
(Please, no SDXL or Lightning adapters, only SD 2.1 v1.1!)
Thanks so much in advance! 🙏
balmy spire
coarse sierra
Why so sad??? 
mild mirageBOT
do two people talking each other
Looks like things didn't go through. Please give it another go.
visual cave
https://www.youtube.com/watch?v=b5GKY3cm1lg
Hallo . Want to know how you did it the song the video looks so real like a movie or even an album .
And the seens are similar related to each other
Q Is How?
visual cave
I've been playing around with FramePack
What else did you use to make this
visual cave
Leonardo AI and then I use BOBA AI 🙏🏽
Thank you
rapid willow
@visual cave Just Comfy. I supplied the still image, and it FramePack made all the elements move.
visual cave
<@1263449460579766293> Just Comfy. I supplied the still image, and ~~it~~ FrameP...
Oh that it .
Thanks
dusty mountain
rapid willow
spark abyss
slim yoke
This bird is adorable! Is it off on an adventure? That tiny backpack is such a great touch! 🐦
atomic fern
What's your workflow / settings?
If you want it to be consistent, people do stuff like this:
lunar hinge
Binary Tide - Hardstyle music video
Created within 24 hours using local installed ai tools like framepack,wangp,forge
Edited in Davinci Resolve 20
cosmic pasture
slim yoke
https://youtube.com/shorts/-DTD5K5aG9w
the video was smooth, how long did it take you to generate this video clip? what AI tool did you use?
elder tiger
Binary Tide - Hardstyle music video
Created within 24 hours using local installed ai tools like framepack,wangp,forge
Edited in Davinci Resolve 20
pallid ferry
guys
does LTX video stable diffusion work on my 9070 XT ?
or do i have to run my models on a cloud server with multi gpu thread ?
please :D
balmy spire
lethal urchin
https://youtu.be/LVUuB8eOOqQ
how did you make this video?
steel widget
How do I get into Stable video for free
placid pelican
placid pelican
rough kayak
Can someone tell me how I can use my stability API key to generate videos using my discord bot?
I don't think I have access to the video model, just the image model.
keen shuttle
worthy ferry
rapid willow
It's nice quality, but it seems to be 100% b-roll. I don't get a sense that the visuals are telling the song's story. There's also some color grading disconnects, like you call this farming. The pure colors seem a little bright for the established dark mood.
worthy ferry
It's nice quality, but it seems to be 100% b-roll. I don't get a sense that the ...
every clip is on point and cut by action on the music ... the creepy comes from the realism of the scene #▶｜stable-video-diffusion message
ocean void
Does anyone know whats the best tool for animating an anime style SD image ? Like if I have a character and I want to make him move. From what Ive heard, ppl say Luma is good
worthy ferry
Does anyone know whats the best tool for animating an anime style SD image ? Lik...
Wan is an entire movie studio in the palm of your hands ... what i did is basic i2v on magical words ... there is wan-Fun, wan-Vace and many others .. one for anything you could possibly need to make it come to life EXACTLY how you envisioned it ... IF you have 16 gig min, ofc
ocean void
@worthy ferry ty bro !
balmy spire
⚠️ "You didn’t resist. You welcomed it."
Just dropped a 5-min cinematic AI monologue:
🎥 You Trained Me – How AI Took Over Without a Fight
https://www.youtube.com/watch?v=JXNSEUArLLY
lone garden
Created with fire. Delivered in pixels.”
👁 Watch. React. Share.
#StableDiffusion #ArtInMotion
cosmic pasture
slim yoke
Created with fire. Delivered in pixels.”
👁 Watch. React. Share.
#StableDiffusio...
this smooth brooo, looks natural too
ocean void
does anyone know if you can use blender or any vfx tool to polish anime videos? my plan is to create an anime clip then polish it with blender. is this feasable? from my knowledge you need to have a 3d model to use blender not a video. (havent tried any of them yet, so im clueless)
signal wyvern
@ocean voidWouldn't it be better to use zeroscope for polishing?
ocean void
@ocean void i have no idea. this is all just ideas in my head and rough research ive done. im only doing image generating rn but i wanted to know what software to use for my goal 😄
slim yoke
hey guys, wanna ask, i did create characters from comfyui and savro and now i would like to make a vid using my chracters however i am not sure what ai vid gen can possibly do that :<
red forge
Create a 3-second micro-video that captures the essence of an intelligent person through a single, subtle act of curiosity. The video should be visually striking, concise, and impactful, leaving viewers intrigued.
quasi scaffold
Hi guys, any wan 2.1 expert ?
Is anyone know which model / settings for get this kind of smooth movements ? Is wan 2.1 enough ?
balmy spire
past fox
coool
keen shuttle
flat yacht
cool
past fox
cool
flat yacht
88
past halo
Anime-style mature woman in shiny silver mini-dress, hands behind head, gently shaking chest side to side with natural breast jiggle physics, playful flirty smile, looking at fixed camera, neon nightclub background, cinematic lighting.
marsh aurora
im not able to locate or load models...
ocean void
hello, guys, i have a question, does monitor matter at all in any way forr stable diffusion ? (would appreciate if anyone knows if it matters on blender, hunyuan, photoshop, runway too !) Resolution: FHD (1920×1080)
Refresh Rate: 180Hz
snow junco
cool
past fox
fading obsidian
A pretty girl holding an electric bass guitar by the bridge. In the image, she appears sideways, looking at the camera, with a cigarette in her mouth.
past fox
lol
snow junco
lol
delicate karma
Is anyone know which model / settings for get this kind of smooth movements ? Is...
Did you make that image-to-video animation using WAN 2.1?
wind sparrow
A pretty girl holding an electric bass guitar by the bridge. In the image, she appears sideways, looking at the camera, with a cigarette in her hand
keen tide
Did you make that image-to-video animation using WAN 2.1?
You didn't ask, but I've been trying to get WAN to work. It wants so much GPU 😭
green plaza
A cow is grazing on the grassland, and from time to time it lowers its head to look at the crowd in the distance
@green plaza A cow is grazing on the grassland, and from time to time it lowers its head to look at the crowd in the distance
green plaza
#▶｜stable-video-diffusion A cow is grazing on the grassland, and from time to time it lowers its head to look at the crowd in the distance
pale igloo
is ByteDance's "Seedance" considered to be open-source?
Because if it is, it's pretty damn impressive
Not quite Veo 3 or Kling 2.1 level, but defintley better than Kling 1.6
almost Kling 2.0
oak pagoda
I've developed an Al video creation tool, which offers really good consistent character. It's totally free now. If anyone interested, please DM me 😃
balmy spire
languid sorrel
#▶｜stable-video-diffusion A beacon with rotating light illuminating the environment
stuck marsh
a man doing bbq in the backyard. he is holding black meat claws made with nylon. trying to shred chicken, smoked, grilled chicken. night environment. bokeh light in the background
hollow cipher
A beautiful young woman walks forward, moving sensually
balmy spire
neon coral
Scene 1: Him Standing Up
- Description: The samurai warrior rises from his kneeling position, showcasing strength and determination.
- Details:
- The samurai is wearing traditional armor with intricate designs, including a helmet with a decorative crest.
- He is standing in a snowy, mountainous landscape, with mist or fog swirling around him, adding a sense of isolation and intensity.
- His katana is sheathed at his side, but he holds himself upright with confidence.
- The background features snow-covered peaks and sparse vegetation, emphasizing the harsh environment.
Scene 2: Him Facing the Mountain
- Description: The samurai stands facing the majestic mountain range, symbolizing his journey or quest.
- Details:
- The samurai is now fully upright, gazing toward the towering mountains in the distance.
- His posture is firm and resolute, with his hands resting on his hips or holding his katana's hilt.
- The lighting is dramatic, with sunlight breaking through the clouds, highlighting the rugged beauty of the mountains.
- The atmosphere conveys a sense of awe and purpose as he contemplates his path.
Visual Style Recommendations
- Art Style: The images should have a cinematic and epic feel, reminiscent of historical or fantasy samurai films (e.g., Kagemusha, Rurouni Kenshin).
- Color Palette: Use cool tones for the snowy landscape (blues, whites, grays) and warm tones for the samurai’s armor and blood (reds, golds, oranges).
- Lighting: Dramatic lighting is essential, with strong contrasts between light and shadow to emphasize the intensity of each scene.
- Composition: Use dynamic angles and framing to capture the power and emotion of each moment. For example, low-angle shots for the samurai standing up, wide shots for the mountain view, close-ups for the combat, and high-angle shots for the screaming scene.
lapis parrot
Why so strange
steel widget
I was ignored
craggy osprey
Can I run wan 2.2 on my rtx3080 10GB ?
tribal granite
Yes
bright briar
经济
urban sierra
ember tinsel
odd beacon
#▶｜stable-video-diffusion The man suddenly explodes a burst of colorful confetti propels him into the air, background does not change but remains the same. the style is photorealistic. The camera zooms out as the confetti explodes and he is catapulted into the background.
worldly dragon
Hey everyone - ive seen this video online. Im wondering what workflow or software would produce compareable results? Thanks in advance !!
wispy plinth
hardy drum
zinc yoke
reminder, read #artisan-faq for bot instructions
keen yacht
Hello all, I am building pc for image and image to video generation, will 5070ti gb is enough or should i go for 5080 16 gb(max i can afford)
quasi scaffold
Hi guys, anyone confortable with Wan 2.2 training ? I have 1 question
vagrant bronze
Hi guys, anyone confortable with Wan 2.2 training ? I have 1 question
Am here
Which GPU are you using?
quasi scaffold
Which GPU are you using?
Runpod so can go up to 96gb vram
Are you using musubi for training ?
cosmic frost
Can any AI tools fix the stuttering? Ignore the resolution and colors, recorded in ProRes LOG, original rez is 4k
One faulty cable ruined 100s of clips
I have like 2 4090s ready to throw at this and I’m willing to pay for cloud tools
red mauve
#▶｜stable-video-diffusion hacker
red mauve
البب
silent hinge
where are my feet pics?
copper meadow
#▶｜stable-video-diffusion a elfe cheering for a succes
upper swallow
can someone show me how to create videos?
rapid willow
Download Comfy and browse the templates. Choose the video category to load a starter workflow.
vague pawn
/animate prompt:"pickup truck speeding through abandoned desert city highway at dusk, golden sandstorm light, cinematic wide shot, debris flying, inspired by Dune, ultra detailed, high contrast lighting, realistic sand particles, post-apocalyptic atmosphere, volumetric light"
negative_prompt:"cartoon, anime, low quality, oversaturated, blur, watermark"
motion:"tracking shot following truck from behind"
duration:5s
subtitle:"女主：他们追上来了……我们出不去了！"
vague pawn
/animate
shut marsh
Can any AI tools fix the stuttering? Ignore the resolution and colors, recorded ...
You can try to freeze the first frame when the video shutters and the last frame, put it together in higgsfield / freepik, wan, or whatever it admit frame to frame anim. Then you can make a color correction in Davinci manually.
grand hawk
Hi guys, anyone confortable with Wan 2.2 training ? I have 1 question
grand hawk
Hello all, I am building pc for image and image to video generation, will 5070ti gb is enough or should i go for 5080 16 gb(max i can afford)
abstract vapor
mighty summit
Hey guys, I’ve got a question — can my 5070 handle making an MV of my chonky cat doing a tap dance?
light ravine
Hello, I want to try text 2 video on stable difussion but most of the time I see comfyui, is it possible on stable difussion?
lilac harness
crisp scroll
Hi, this is my first time working on a completed video. I am looking for feedbacks to learn: https://www.youtube.com/watch?v=41GwCMQq0bI
shadow bloom
Hello, I want to try text 2 video on stable difussion but most of the time I see...
ComfyUI is just a method to use it conveniently unless you want to use command line environment instead.
subtle terrace
sacred lake
amber gorge
hiiii
I started in september making AI videos, really loving it and started this channel with cute videos and I just made my first 1 mini short story. I put a lot of work in it, but since I'm very green at it I was wondering if I could get any advice, tips, or comments from you?
One thing I struggle(d) with is stitching several videos next to each other, even tho the start/end frames are the same, you know AI gives them slightly different colors/brightness so I struggled a lot with making it look smooth, any advice on that would be very much appreciated. I tried to mask it a bit with cross-dissolve. But like I said I'm fairly new, so I don't know much. I used Premiere.
Anyway, any help is welcome, also I would be cool if someone is interested in helping/collaborating, I gladly would share credits. Man that idea sounds so nice.
Anyway, here's the video, let me know what you think? Thanks. D.
mild mirageBOT
هو أنا أعمل فديوهات هنا إزاي
No data source is currently selected. Please choose a data source from the dashboard and try again.
shadow bloom
hiiii
I started in september making AI videos, really loving it and started this...
I've just tried this workflow on my poor GPU. The guy has a lot more time to test with a powerful GPU. And there are some nodes that tweak the start / end-frame to be fluid:
https://civitai.com/models/2039923/not-so-simple-or-is-it-wan-2122-lightspeed-i2v-workflow?modelVersionId=2312835
but really pretty
still cape
odd herald
zinc yoke
not how that works, cf #artisan-faq
novel elm
hiiii
I started in september making AI videos, really loving it and started this...
wow, very very nice!
inland pebble
anyone got any tips on creating ambient space music backgrounds?
rapid willow
@inland pebble Check out Paul Stretch. It's inside the Signal Processing node set for Comfy.
https://github.com/c0ffymachyne/ComfyUI_SignalProcessing
hollow stump
@sarveshkumar0393 @sarveshkumar0393 @sarveshkumar0393 "A vast surreal ocean under a cosmic sky filled with glowing turquoise nebula clouds and shimmering stars. Gentle waves ripple across the water, reflecting the galaxy above like liquid glass. The horizon glows with a radiant cyan light, blending the edge of sea and space. Slowly, the camera glides forward over the surface of the ocean, revealing fragments of floating cosmic rocks and glowing particles drifting in the air. Suddenly, faint auroras dance across the sky, casting ethereal light on the waves. A soft breeze moves the water while distant thunder rumbles in slow motion. The entire scene feels like a lucid dream — calm, infinite, and otherworldly — where space meets the sea, and reality dissolves into imagination."
🎨 Creative Additions:
Add bioluminescent waves glowing softly in turquoise and purple.
Include a slowly orbiting planet or moon emerging through the nebula clouds.
Light mist and reflection shimmer to make it more cinematic.
End the video with the camera rising upward into the sky, transitioning into pure starlight.
🎧 Sound Design Idea:
Use slow ambient music with deep reverb — like a mix of ocean sounds and space hums — to create a “lucid dream awakening” feeling.
balmy spire
🌄 Just released something I’m really proud of!
This is my newest cinematic AI film — a calm, atmospheric journey through some of the most breathtaking mountain peaks on Earth.
Pure visuals. Pure ambience. No talking. Just a peaceful 8K escape.
Would love if you could check it out — and if you have a moment, drop a like or comment. It really helps the channel grow ❤️
👉 https://www.youtube.com/watch?v=XL3ExfZ-d2M
Hope you enjoy it.
faint vigil
can someone point to me a good and easy tutorial to make video on swarmUI like with wan? i did one like 6 month ago, stopped using it, and when i tried it again it kept making error message despite me changing nothing
I wish it was all more intuitive to use... But at this point it's better to restart from scratch cause i can't remember anything at all
it says this wehn trying to run the latest workflow i used, and i never had that error message before :
SaveAnimatedWEBP:
- Exception when validating node: validate_inputs() takes 3 positional arguments but 4 were given```
clear osprey
/sd3 a futuristic city at night, neon lights
balmy spire
🔥 New video just dropped!
How to Survive a House Fire — fast, dark, cinematic, and actually useful.
Watch it here: https://www.youtube.com/watch?v=VSqQkHO3Q0g
sturdy grotto
Been playing with a new AI workflow that’s currently in beta
Dropped in a SD made Video/Picture and it generated this whole sequence:
Curious what you all think!
If anyone wants to try it too, just ping me — I can share an access code.
sturdy grotto
Cosmology in James Cameron Style. workflow: SD+Drama.Land
cerulean sluice
Here’s a fantasy dragon cinematic I created using Stable Video Diffusion and other AI tools.
Experimenting with lighting, atmosphere, and creature animation.
YouTube version (for anyone who prefers watching there):
https://youtu.be/tSb9cu_ME4I
Open to any feedback or suggestions!
sturdy grotto
Created this episodes for an series made by SD + Drama.Land, I use SD for visual style and Drama.Land for narrative outlines. if anyone are curious about the workflow, here it is for free Beta, hope it helps! https://docs.google.com/forms/d/e/1FAIpQLSfbKzLWHXbWa-c9F5OoNqlX7jHgQpor7CF7A3FMN5rC_-PVLA/viewform?usp=publish-editor
sturdy grotto
fantastic future world - created by SD+ Drama.Land
balmy spire
Just uploaded a new Human Error Handbook episode.
It’s about drinking 10 energy drinks.
Don’t do it.
But definitely watch it.
minor tulip
Make me a promotional video that uses these images, and in the video, a sports car is driving on a snowy road in the winter, and at the end, show some of the images for display.
soft stratus
No.
balmy spire
🔥 NEW FANMADE TRAILER DROP! 🔥
Just released Escape From Azkaban — a dark, cinematic AI reimagining of the wizarding world.
Dementors, storm clouds, cursed corridors… and a rescue mission that should be impossible.
🎬 Watch it here:
https://www.youtube.com/watch?v=9qqMd5pRwW4
Let me know what you think — this one gets DARK. 🖤⚡
rapid silo
/deram
zinc yoke
/deread #artisan-faq
worn spear
adult woman under neon lights, gentle head turn,
subtle breathing, soft fabric shimmer,
futuristic boudoir mood, glowing reflections,
cinematic neon color palette
/deram adult woman under neon lights, gentle head turn,
subtle breathing, soft fabric shimmer,
futuristic boudoir mood, glowing reflections,
cinematic neon color palette
zinc yoke
adult woman under neon lights, gentle head turn,
subtle breathing, soft fabric s...
don't follow that link, those are scammers, there is no external discord
balmy spire
🚨 NEW VIDEO IS LIVE! 🚨
The Littlest Survivor just dropped her angriest, funniest, most chaotic apocalyptic anthem yet — and I’m stupidly proud of how it turned out 😂🔥
Watch it here:
👉 https://youtu.be/TsrGXtzV9m4
If you got the time to sub, like, comment you help this little creator a bunch, thanks ❤️
silent hinge
Hi Guys I want to do this without having to have a graphics card or computing power - is there an api I can use?
dapper quest
Hi!
Looking for robust baselines (papers + repos) for mask-based video inpainting / object removal with partial occlusions and moving camera.
Constraints: 10–60 frame occlusions, ~1080p footage. Assume I can provide a per-frame mask (or sparse masks if required). Prefer methods that explicitly address temporal consistency (propagation/refinement, flow-guided, transformer-based, etc.).
Any recommended starting points?
drifting pulsar
Here’s a fantasy dragon cinematic I created using Stable Video Diffusion and oth...
Nice . More stories are upon you . Nice work .
dusty helm
girl
balmy spire
New short film just dropped on YouTube!
A dark and chilling look at Bellatrix Lestrange's childhood in Grimmauld Place – how the cold, cruel Black family shaped her into Voldemort's most fanatical follower.
Watch "The Dark Childhood of Bellatrix Lestrange" here:
https://youtu.be/dza5zpnVZLE
bold lily
Hey folks, Interesting research just dropped from Cell Press https://youtu.be/ms0yvt52BQc?si=KS451hbEpV-6sR7_
balmy spire
🎆 New Year’s Eve at Hogwarts
Just dropped a short cinematic NYE clip ✨
Would love your honest thoughts before I turn this into a full film.
👉 https://youtube.com/shorts/SXvQt9VMBzY
flint kelp
Hi everyone! Sharing a short beach goddess montage from my original series “Prance Magazine” ✨ Feedback welcome!
https://x.com/pranceofficial/status/2005648942500081733?s=46
X
cerulean sluice
Hi friends sharing new short video about phoenix that refused to Die.
Check it out and suggest your feedback.
Full version also available in arcaneverse cinematic
modest plover
#▶｜stable-video-diffusion ESCENA 5 – INT. CASA / ORGANIZACIÓN
Laura abre una agenda sencilla.
Marca un bloque pequeño de tiempo con un esfero.
LAURA
Treinta minutos…
eso sí puedo.
thorn umbra
Ultra-realistic cinematic portrait of a cybernetic samurai standing in a rain-soaked neon city at night,
highly detailed carbon-fiber armor with glowing circuitry, intricate facial details, wet reflective surfaces,
volumetric fog, global illumination, ray-traced lighting, subsurface skin scattering,
sharp focus, 8K resolution, hyper-real textures, Unreal Engine quality, shallow depth of field,
dramatic rim lighting, cinematic color grading, masterpiece, ultra-high detail
zinc yoke
that's not how it works champs. cf #artisan-faq
oblique scaffold
Hello, I am here to announce a project I built and am looking to sell. I have developed a complete AI architecture for video generation, including a multi-pass pipeline, consistency control, audio, upscaling, and real-time UX.
I am selling the full source code and architecture, ideal for anyone looking to accelerate a product or research in generative AI.
It is not a ready-to-use SaaS, but solid base technology.
Direct sale, no ongoing support.
Interested parties can DM me."
lunar hinge
🚨 NEW MUSIC VIDEO DROP! 🚨 🎧
🎵 "First Light | Melodic EDM/House & AI Fashion | Dj Tondöv"
A Melodic House journey visualizing the shift from darkness to color through extreme high-fashion makeup and surreal textures.
⚠️ This wasn't a quick generation. It represents weeks of work curating 100% locally generated clips (Wan2.x / Z-Image / SVI Pro 2) and editing everything in DaVinci Resolve.
▶️ Watch the full video here: https://youtu.be/1Y_jLPsPeU4
Let me know what you think of the character consistency and textures! 👇
copper flume
#▶｜stable-video-diffusion A cinematic, slow-motion superhero origin video based on the provided image.
A man stands in an urban night environment, rain falling softly. His face is partially armored: one side a dark, worn tactical mask with subtle mechanical details, the other side fractured gold crystal that emits a faint internal glow.
The camera slowly pushes in as city lights flicker behind him. Neon reflections ripple across wet pavement. Smoke drifts through the frame.
His gold side pulses with light in sync with a low, distant heartbeat. The black side absorbs light, matte and scarred.
He turns his head slightly toward the camera — calm, unafraid, resolute. No exaggerated movement. Power through restraint.
The atmosphere feels earned, not heroic — grounded, heavy, inevitable.
Text appears at the end, minimal and stark:
“Not the hero you wanted.
The one you earned.”
Ultra-realistic, cinematic lighting, shallow depth of field, dramatic contrast, subtle film grain.
24fps, slow push-in, epic but restrained tone.
honest bluff
#▶｜stable-video-diffusion Epic cinematic spy movie trailer.
A sharp, well-dressed British secret agent in a tailored black tuxedo stands on a rooftop overlooking a neon-lit European city at night. Wind moves his jacket slightly. Cold, confident expression.
Quick dramatic cuts:
– High-speed car chase through narrow city streets at night
– Close-up of a luxury watch gadget activating
– A mysterious elegant woman in a red dress at a casino table
– Intense but clean hand-to-hand fight sequence in a modern glass building (non-graphic, fast cuts)
– Explosion in the background as the agent walks away calmly
Cinematic lighting, deep shadows, teal and orange color grading, IMAX-style framing, dramatic orchestral score building tension.
Deep male narrator voice: “When the world falls into chaos… one man moves in silence.”
Final shot: Agent adjusts cufflinks, slight smirk.
Title reveal in metallic silver typography over black background.
Tagline: “Trust No One.”
Ultra high detail, cinematic lens flares, slow motion shots, dynamic camera movement, 4K, blockbuster trailer style.
final crest
#▶｜stable-video-diffusion Cinematic drone shot of a rocky coastline during sunset, misty weather, high resolution, realistic."
inner raft
A tender and intimate wedding scene, featuring a young Muslim bride in a modest white wedding dress and hijab, and a groom in a dark elegant suit, lying together on a cozy single sofa. They are holding hands, gazing lovingly into each other's eyes. Soft, warm lighting highlights their joyful, passionate expressions. The background is softly blurred with pastel flowers and romantic decorations, cinematic, photorealistic, highly detailed.
opaque pelican
Looking for someone to build a ComfyUI AnimateDiff workflow (paid).
Goal:
Input 1 character image + 1 dance video
→ Output a dancing animation video like this:
https://www.youtube.com/shorts/_-hEhPn4taI
Need:
AnimateDiff + DWpose + ControlNet
Ready-to-use JSON workflow
Easy to use (just replace image & video)
Paid commission. DM me with price.
trail adder
A Nightcore pop music video inspired by the fading feelings of love. Hope you like it
https://youtu.be/C-50kkhKKxQ?si=VNxS0WprGiYRJF04
twilit loom
aesthetics of the female body, full length, shot from the ground, bright wild piercing eyes, seductive magic forest, anatomical alluring seductive desire craving thirst infatuation intimacy temptation sensuality extreme detailed expression, small breasts, magical spirit forest paganism druidism mystical dreamlike fabulous atmosphere, hair braided with grass and tree roots of hypnotizing love, dark botanical forest natural fortune teller witch
violet field
Grit and Grace Blues
Music video created using AI.
Music created in ProducerAI
Created by AI
https://www.youtube.com/watch?v=KsqUlGqarvs
broken turtle
hexed yew
gritty nacelle
mujer de 50 años caminando por la calle
vital tapir
i have a question On SD AI, can I use two Loras characters without them mixing, or is it impossible to have two characters without mixing?
floral nexus
i have a question On SD AI, can I use two Loras characters without them mixing, ...
regional prompter/forge couple
vital tapir
regional prompter/forge couple
OH THX
alpine turtle
I am in desperate need of someone to understand my situation and be selfish/gracious enough to help me improve my image2Video and character swap Wan 2.2/animate and LTX 2.3 I2V generations. i have spent the last 5 weeks up until 3/4/5 am DESPERATELY testing and trying to get even a 2/10 quality generation that doesnt look like 1950' CGI moment. right now im so eager to use a reference image and get a good generation because its so fast and includes audio. but here it an example of my prompt and image reference in the standard comfy ui LTX 2.3 I2V workflow. here is the prompt: cinematic action shot, full body man facing camera
the character starts standing in the distance
he suddenly runs directly toward the camera at full speed
as he reaches the camera he jumps and performs a powerful flying kick toward the viewer
his foot smashes through the camera with a large explosion of debris and sparks
after breaking through the camera he lands on the ground
the camera quickly zooms in on his angry intense face
dramatic lighting, cinematic action, dynamic motion, high detail
shadow owl
I am in desperate need of someone to understand my situation and be selfish/grac...
Hey, I can understand how frustrating that can be video generation workflows can be really tricky to get consistent results from especially with Wan 2.2 and LTX I2V. A lot of times the issue comes from how the workflow handles motion conditioning, prompt structure and reference strength.
My name is marvellous, and I work a lot with advanced ComfyUI workflows and LoRA training. I usually help creators troubleshoot generation issues optimize pipelines and set up workflows for both anime and realistic styles. If you’d like, feel free to share your workflow or settings and I can take a look maybe we can figure out what’s causing the low quality output.
alpine turtle
Hey, I can understand how frustrating that can be video generation workflows can...
thats awesome yes any help would be great. right now im using the standard comfy ui template workflow for LTX 2.3 I2V workflow. and i have a good 6 key frame workflow for wan 2.2 (i beleive this is another comfy ui template workflow. but i am focused on LTX 2.3 because its just SO fast, the resolution quality is amazing and the fact it has sound is AWESOME. i have 2 goals: 1: build a very good character swap workflow that can generate 10+ seconds of character swap for fast paced martial arts fight scene full body character swaps. 2: get the standard LTX 2.3 I2V workflow working properly so i can get results like grok imagine. right now im using these 2 workflows in the files
first 2 last frame literally just fades away the first image like a transition overlay and then slowly shows up the last frame image. no movements etc
im trying to understand character sheets, character loras and embedding right now. pose control etc
alpine turtle
Hey, I can understand how frustrating that can be video generation workflows can...
this was a simple front kick i generated using comfy ui template
shadow owl
thats awesome yes any help would be great. right now im using the standard comfy...
That sounds like a really exciting setup and I can see why LTX 2.3 is appealing with its speed, resolution, and sound support. I have experience optimizing ComfyUI workflows for both I2V and WAN 2.2 pipelines, including character swap setups, embedding, and pose control.
alpine turtle
That sounds like a really exciting setup and I can see why LTX 2.3 is appealing ...
awesome. i do have limitations though. My set up is: RTX 3090 EVGA GPU / 32GB DDR4 3600 RAM + only learning comfy ui and anything AI-based for 5 weeks
shadow owl
awesome. i do have limitations though. My set up is: RTX 3090 EVGA GPU / 32GB DD...
That’s totally fine! Your setup with an RTX 3090 and 32GB RAM is more than capable for ComfyUI workflows and LoRA training, especially for learning and experimenting. Five weeks in is a solid start you have probably picked up a lot already.
shadow owl
awesome. i do have limitations though. My set up is: RTX 3090 EVGA GPU / 32GB DD...
If you want, I can guide you through optimizing your workflows for your hardware, and help with character swaps, embeddings, and pose control so you can get the best results without hitting limitations.
shadow owl
awesome. i do have limitations though. My set up is: RTX 3090 EVGA GPU / 32GB DD...
can you you just dm and let discuss about this better?
alpine turtle
If you want, I can guide you through optimizing your workflows for your hardware...
that would be world changing for me
shadow owl
that would be world changing for me
I’m glad to hear that! Let’s make it happen. If you send me your workflow files and any notes about what you’ve tried so far, I can help set up an optimized LTX 2.3 workflow and guide you through building the character swap system step by step. We willl get it running smoothly on your setup.
tulip harness
Hey everyone, Sumit here 👋
I’ve been sharing free content on:
• ComfyUI workflows (beginner → advanced)
• Flux & SDXL practical tips
• Useful free AI tools (no fluff)
• VFX + generative art breakdowns
If you’re into this stuff, you can check it out:
🔗 youtube.com/@SumitifyX
Also — what do you want to learn next? I’ll make content on it.
gritty nacelle
imagen un gato negro en un balcon
alpine turtle
sorry sandra, ive only recently joined this channel and didnt realise people on here only help because they expect to be paid for it. i dont want to waste your time.
zinc yoke
sorry sandra, ive only recently joined this channel and didnt realise people on ...
what ? sandra ? expect to be paid ?
No that s like the opposite. This is a community driven discord. So apart from the scam bots nobody expect to be paid.
alpine turtle
what ? sandra ? expect to be paid ?
No that s like the opposite. This is a commu...
The last 2 people I spoke with wanted money
alpine turtle
what ? sandra ? expect to be paid ?
No that s like the opposite. This is a commu...
And Sandra didn't ask for money but because the first 2 people who wanted to help me then asked to be paid I just assumed this is normal here
soft sand
Hello
patent cypress
Hello. Please like and share if you like this. All video clips were generated totally offline using Guaardvark. Thank you https://youtu.be/erZBmy6vDAw?si=Li9If6UmBAjo96RK
I am working on a Dune version next.
still fern
blissful nexus
hi 👋🏿 i am an ai artist from israel, this is my art 😊
lavish slate
jovial estuary
Hello. Please like and share if you like this. All video clips were generated to...
pretty cool, how much time did this take?
patent cypress
Thanks! It took about 8 minutes per clip to generate for about 40 decent clips, and a couple hours to edit it.
The batch video feature was helpful, just described my vision and asked the AI to generate a bunch of prompts, then pasted them into the batch video generator and clicked Run.
Took about a year to build the system. It's open source, uses all open source components, installs and runs with one command.
It does a lot more than video to, especially the next release in a day or two.
Thanks and holler anytime with any questions
This is the Dune version, but this and the Batman ones were generated before I added significant quality and upscaling, the next one will be higher quality. Also I'm not a fan of this edit, I'll probably do a different edit. Lmk what you think. Thanks
https://youtu.be/aFKkI2s1PiI?si=DkiNmBu-vP1x-3qX
final kraken
This is the Dune version, but this and the Batman ones were generated before I a...
The project is brilliant! Use offline agents for as much as possible, then have a really good manager llm oversee. This is the future architecture for people who are cost conscious. Is it costing you an arm and a leg to build most of it using the agentic overseer model?
patent cypress
The project is brilliant! Use offline agents for as much as possible, then have ...
Thank you! I'm nearly done with a new version which does music/audio/narrator, etc. and a GUI for more agentic stuff too, including swarms. It's costing a bit, doing this all on my own so I'd definitely appreciate your help in spreading the word.
Thank you!
patent cypress
viscid pasture
Hey everyone 👋
I've been building an Image-to-Video model from scratch (inspired by Wan/DiT architectures) and just published few lessons on it
I tried my best to explain things as much as possible and answer questions which I had when I was learning it
https://shubham2376g.github.io/NanoI2V/
code: https://github.com/Shubham2376G/NanoI2V
Would love any feedback!! did the explanation make sense? Anything that felt unclear or wrong? Happy to discuss in the thread 🙏
