Transform Your Videos w/TemporalKit 🔥
ฝัง
- เผยแพร่เมื่อ 19 มิ.ย. 2024
- TemporalKit is still one of the best tools to edit your videos with AI! Almost no deflickering and perfect refining. This is a combination of Img2Img and EbSynth. We are going to edit together a video step-by-step! 💙🤩
💬 Social Media:
[Discord] / discord
[Patreon] patreon.com/IntelligentArt?ut...
[Instagram] / lacarnevali
[TikTok] www.tiktok.com/@lacarnevali?i...
____________________________________________________________________
🤙🏻 Learn More:
/ membership
/ lauracarnevali
📌 Links:
TemporalKit GitHub: github.com/CiaraStrawberry/Te...
FFmpeg: ffmpeg.org
Stable Diffusion Art: stable-diffusion-art.com/vide...
EbSynth: ebsynth.com
DaVinci Resolve: www.blackmagicdesign.com/prod...
00:00 Scope of the video
00:36 Requirements
01:02 How to install TemporalKit
01:22 Install FFmpeg
01:53 Install EbSynth
02:20 FIRST STEP - TemporalKit
02:35 Settings: Slide
03:24 Settings: Height resolution
04:18 Settings: Frames per keyframe
04:23 Settings: FPS (Frames Per Second)
04:47 Settings: Target folder
05:41 Settings: Batch run
06:25 Settings: Split
07:13 Settings: Max key frames
07:19 Settings: Border key frames
07:37 Output
08:33 SECOND STEP - Img2Img to edit your video style
09:09 Model used
11:37 THIRD STEP - EbSynth within TemporalKit to generate keyframes
12:59 FOURTH STEP - EbSynth to generate the final video
14:01 View/Combine w/DaVinci Resolve
#aiart #stablediffusion #generativeart #stabilityai #stablediffusiontutorial
#videoanimations #temporalkit #ffmpeg #ebsynth
![[Full Episode] MasterChef Junior Thailand มาสเตอร์เชฟ จูเนียร์ ประเทศไทย Season 3 Episode 3](http://i.ytimg.com/vi/A1XXYbV048k/mqdefault.jpg)
perfection. It’s like the moment I clicked on the video my brain 🧠 immediately started to relax and melt
Goose bumps run through my body. As if the most delicate cloud falls over me,
and touches my head and runs along the back to the very waist, and rises along the back back to the back of the head.
And I am so smiling all over because I love you
LOL!!!
That's really neat. Does it also give good results to use a higher denoising strength to try getting a much different video? Like if you wanted her wearing armor in the video instead?
very difficult, you would need to work a lot on the keyframes
Great information with thorough explanation. Thank you.
ah cool. I built something like this, but haven't gotten around to making it an add on. My computer is a potato, I haven't tested it on a decent gpu. the program is fully automated, translating a raw video into another one. Basically automating all these steps. but my videos are pretty inconsistent.
hmm. I see you use line art. and Img2img... maybe I should try that. My keyframes are too far apart in their latent space for me to want to use my software.
No matter what I try Im not getting anything in keys folder
AFAIU, Low Denoising reduce the importance of the prompt (and LoRA), so you can't really re-interpret as another character for instance, right?
The denoising strength impacts the denoise you apply to the input image to generate the output. The lower the denoise, the lower the changes applied to the input image. This will impact Lora and the prompt importance indirectly. Indeed, what impacts the prompt and lora the most is the CFG.
So i get to the very end with no issues at all, but the final crossfade mp4 is black and when i hit play nothing happens. any ideas? ive completely started from the beginning multiple times but the end result is the same black crossfade mp4 file that wont play.
Love to give it a try, but I don't think my PC will be able to handle it lol
Can try with Colab or Runpod, or Diffusion Hub/Think Diffusion
Laura im on Thinkdiffusion and when I run the cell of temporalkit show me an error on the output window, only says Error doesnt say why :(
what do you have on the terminal?
I have a little mess qiththat, running in thinkdiffusion do I see reflected whats happening on the terminal of my mac? or the terminal of think diffusion?
@@LaCarnevali
Excuse me ms.Laura, but where do you download the necessary controlnet models,
my tile images appear as a disgusting blob.
But when I do single images, they come out beautiful....what am I doing wrong.
yeah tiled doesnt work very well with most of models. Do you mean batch though?
@@LaCarnevali yes for the batch function.
Is there anything you recommend that works well for backgrounds/scenery only?
Great video! My keys folder remains empty. Any one else have this problem and know a fix?
try pulling the last update for both A1111 and temporalkit
Laura shot some daily life vlogs :P
Ciao, credo sarebbe interessante un video su xome creare un checkpoint o almeno le idee teoriche alla base
th-cam.com/video/xholR62Q2tY/w-d-xo.htmlsi=8yKObGKHiZz3h7hY
The video is a little bit inconsistent ... how to resolve this ?
what exactly? :)
@@LaCarnevali Thank you for replying me. The inconsistency is from frame to frame. I tried your approach with face. It is quite good. however, when I apply that on body and clothes, the frames changed a lot, in details.