I’m not the OP, but you don’t want to fully create the entire scene and only then look for angles, because you’ll just waste time building stuff that’s gonna be either outside the view or really small.
Also you might want to tweak some placements to improve the composition.
So, I’d assume, that they either had a (mostly) complete composition from the start or blocked out a scene with gray-boxes and then looked for angles, maybe tweaking some stuff
Pretty much what four@lemmy.zip said. I usually start by making the foreground terrain, find a good angle which frames the terrain nicely and add a few assets to find the composition I’m going for and roughly block out where things are going to be.
Here’s an early render of the scene with just a few buildings and completely different lighting:
Once that’s all locked down, I start building from there. The next thing I did was start masking and sculpting the terrain, added the path and fixed displacement, then started building up assets slowly (fences, extra buildings, vegetation) to fill the empty areas, trying to emphasise parts of the image based on composition guides (thirds). Sometimes the camera will change very slightly during this, but not more than a few degrees at a time.
Then it was lighting, volumetrics (clouds, fog, fires), particle effects (birds, snow, embers), smaller assets (shrubs, fences, barrels / shields / skulls). After that it just becomes lots of tiny tweaks to colour, render effects (dispersion, rainbows) until I’m happy.
This was a few iterations ago before many of the last tweaks were added over the final 20 or so hours
I’m not the OP, but you don’t want to fully create the entire scene and only then look for angles, because you’ll just waste time building stuff that’s gonna be either outside the view or really small. Also you might want to tweak some placements to improve the composition.
So, I’d assume, that they either had a (mostly) complete composition from the start or blocked out a scene with gray-boxes and then looked for angles, maybe tweaking some stuff
Cool - yeah was curious if they were in the ‘model building’ mindset or the vantage mindset, but this helps, thanks!
Pretty much what four@lemmy.zip said. I usually start by making the foreground terrain, find a good angle which frames the terrain nicely and add a few assets to find the composition I’m going for and roughly block out where things are going to be.
Here’s an early render of the scene with just a few buildings and completely different lighting:
Once that’s all locked down, I start building from there. The next thing I did was start masking and sculpting the terrain, added the path and fixed displacement, then started building up assets slowly (fences, extra buildings, vegetation) to fill the empty areas, trying to emphasise parts of the image based on composition guides (thirds). Sometimes the camera will change very slightly during this, but not more than a few degrees at a time.
Then it was lighting, volumetrics (clouds, fog, fires), particle effects (birds, snow, embers), smaller assets (shrubs, fences, barrels / shields / skulls). After that it just becomes lots of tiny tweaks to colour, render effects (dispersion, rainbows) until I’m happy.
This was a few iterations ago before many of the last tweaks were added over the final 20 or so hours
Wow, awesome! Thank you very much for this insight into your process- the results are amazing!