this is the video I wanted to see. Please go on with this topic, camera projection would be great next topic. Then we could go even further and cut geometry with projected generated image using “marching squares” algorithm.
Segmentation Control Net is based on special colors, there is google spreadsheet somewhere on the internet, so if we fetch this information and assign key colors in Houdini, this would be strong controlling of SD output.
On my opinion your videos are unique and if you reach “kritische Masse” of videos on this topic and when more people open AI for themselves, your channel will group exponentially.
Normal maps dont work properly. It works vague on relatively simple thing like apple or some single item but not in general nor on some exteriour scene. It would be nice to fix that somehow, since proper normal maps give best controll, better than depth map
Unlock with Patreon
this is the video I wanted to see. Please go on with this topic, camera projection would be great next topic. Then we could go even further and cut geometry with projected generated image using “marching squares” algorithm.
Segmentation Control Net is based on special colors, there is google spreadsheet somewhere on the internet, so if we fetch this information and assign key colors in Houdini, this would be strong controlling of SD output.
On my opinion your videos are unique and if you reach “kritische Masse” of videos on this topic and when more people open AI for themselves, your channel will group exponentially.
Normal maps dont work properly. It works vague on relatively simple thing like apple or some single item but not in general nor on some exteriour scene. It would be nice to fix that somehow, since proper normal maps give best controll, better than depth map