László Gaál
@laszlogaal_ • 2,560 subscribers
Shorts
Videos
Before you ask: yes, everything is AI here. The video and sound both coming from a single text prompt using #Veo3 by Google DeepMind .Whoever is cooking the model, let him cook! Congrats matthieu lorrain and the team for the Google I/O live stream and the new Veo site!
László Gaál2,751,652 views • 1 year ago
New day now Omni findings: it can translate audio (no original or translated text given in the prompt): - it keeps the background music intact - it adjusts the edit if needed. For example the japanese and spanish sentence during the creme close-up shot is longer, so it kept that shot longer and trims that edit point…
László Gaál161,935 views • 1 month ago
DŌKYŌ: A Three-Minute-Long Single-Take Short Film Made with AI We can agree that Bytedance Seedance 2.0 totally changed how people create videos. Omni mode enabled faster creation of short films, solo creators making feature-length films, and you can create a 15-second commercial in one shot with a single prompt. But it also opened the door to long, continuous takes, which were not possible. Usually the length of a generated clip is limited to 8–15 seconds. Of course some people tried to cheat around this limitation by doing last frame first frame generation and stitching, but it breaks the camera movement and the pace of the actions. With Seedance, it was very different: instead of fighting the engine I was able to think about what I would like to see, instead of what are the limitations. I planned the framing, the camera movement, the action, and built the three-minute sequence step by step. Of course, this means if you later realize that something is wrong three or four sections earlier, you have to redo everything that follows. Visually, I wanted to maintain a very cinematic, low contrast, grainy, soft look, siilar to an expired 8mm or 16mm film, so there was no upscaling, rather the opposite: softening the image in every possible way. #DreaminaCPP Dreamina AI
László Gaál19,660 views • 1 month ago
2026 = The end of upscaling? One of generative AI's biggest drawbacks was the resolution, and I'm not talking about the output resolution as 1080p would be enough for most of the projects. I'm talking about the "internal" resolution of how these models work, that makes small faces distorted in your generations, the ones that will make a wide-shot impossible because the small details will morph and of course with some models text rendering is not possible. Upscaling with AI was always a strange case, local upscalers sometimes lacked the quality compared to the closed source competitors, but those can be pretty pricey and the results are sometimes way too "creative". So I was very, very interested when Kling AI announced native, not upscaled 4K generation but wanted to have a good method test it. Again, I never missed the 4K output resolution with genAI, but I missed the nuanced detail rendering of the engines and the lack of that made it impossible to create for example very wide shots with tiny details. So I fired up some image models to generate stills inspired by Morocco and see how Kling behaves, now in 4K. Some notable moments: 00:22 - Skin and eye texture, the movements, the little coins in the headdress 01:18 - The reflection in the water 01:39 - The reflection in the eye, and the scarf textures 02:10 - Check all the details on the rails 02:20 - Water particles and leaves 02:29 - The rendering of the rusty parts of the rail 02:50 - How the waves change the reflection in the foreground 02:55 - The cloud movement in the background
László Gaál17,514 views • 2 months ago
My first “semi compelling screen-saver” in 2025, made with Google DeepMind #Veo2. It looked so real, I had to do a BTS for it so don't forget to watch it until the end! Everything is AI except the talking head with the strange accent in the behind the scenes part. When I will have more time I will write about the process and my thoughts, but for now enjoy the video, get ready for the tipping point of genAI video in 2025, and share the links in the comments!
László Gaál79,944 views • 1 year ago
No more content to load