Running real-time 1080p video generation and editing on your own (Dreamverse OSS release)

https://redd.it/1tpfbrl
@rStableDiffusion

Microsoft Lens - Non Turbo with 5 CFG (ComfyUI)

https://redd.it/1tpfrvt
@rStableDiffusion

The Hunt 2: Z-Image Turbo - Flux.2 Klein 9b - Wan 2.2

https://redd.it/1tphl5l
@rStableDiffusion

Wan2.2 continues to outperform LTX2.3

Wan 2.2 \(sound by LTX 2.3, 1 shot at a time, 3s each, no redo\)

LTX 2.3 \(4 shots, 4 prompts in 1, no redo\)

LTX 2.3 \(4 shots, 4 prompts in 1, no redo\)

Wan 2.2 \(sound by LTX 2.3, 1 shot at a time, 3s each, no redo\)

Setup: storyboard prompt and keyframes by chatgpt, from start to finish \~ 30mins for the entire storyboard video (including waiting for the image from gpt).

https://redd.it/1tpjgi6
@rStableDiffusion

Using depth maps and weight noising to get better character LoRAs
https://redd.it/1tplsmr
@rStableDiffusion

A Wan 2.2 post-training Quant . 1 model instead of high + low
https://redd.it/1tpcm59
@rStableDiffusion

How to upscale image How to upscale a noisy concert photo and fix a blurred face using a high-res reference image?

https://redd.it/1tpvej6
@rStableDiffusion

Tried custom lora for anima base 1.0 and its absolutely amazing.

https://redd.it/1tpx4kg
@rStableDiffusion