Firstly I want to say that Photo AI is incredible. I’ve been having some great results doing a similar up-scaling process using Stability AI but Topaz Photo AI is truly a pro-sumer workflow compared to Automactic1111/ComfyUI which makes me feel like a developer.
The Topaz Photo AI results I’m seeing are beyond what I could hope for. So a big thanks to the team and to everyone who is sharing their results with Topaz to train the models.
So here’s my idea for something different…
I would absolutely drop another $200 on an app that allowed me to train lora models and render entirely new poses of myself using off-the-shelf safetensors from Huggingface or CivitAI.
Can I do this myself with Automatic1111 and ComfyUI? Yes, not an issue. I also use Chainner, but these apps are designed by developers with a million sliders. It takes a lot of effort to wrap your head around them.
Anyone who uses Automatic1111 and ComfyUI is likely having fun, but the workflows are at worst unwieldy and at best exhausting.
I contrast this with Photo AI which is truly a pleasure to use. So I’m just sitting here wishing I had an app called Topaz Playground that allowed me to replace Automatic1111.
Just my 2 cents.
