Hello peeps,
So I'm not really sure if this was discussed before or not but was there ever a request to have Corona more animation "friendly" by working on reducing the scene parsing times?
Right now each time a new frame loads you need to go through that same ol, same ol scene parsing calculation phase. While that is in a way to be expected there are other path tracers out there that load the scene just once and so each subsequent frame has like a 99% reduction in parsing times. The other render engine I used that did things this way was a GPU based renderer so I'm not sure if maybe thats a pro of the way they can quickly access things stored in VRAM whereas with CPU rendering that access might be a bit slower :)
With the other renderer scene parsing times were like 30 second for the scene to load but then each next frame only took like 2 seconds to load up and it was already rendering. Now imagine the animation you are working on is 1000 frames long and you can see that we are talking lots of hours of compute time spared :)
I know there are some scene parsing improvements already listed on Trello but I'm not sure if the same thing I'm talking about. Also, sorry if this has been suggested before :)
Any thoughts perhaps?