Roses on the fence

My latest project, managed to render it in time to submit to competition (still I don’t expect much)


IvyGen, particle systems, a lot of image textures, rendering and compositing in Blender 2.79, Cycles.
There isn’t color variation in the roses and leaves due to RAM limits, particles were too heavy to process so I used linked dupligroups which didn’t allow me to use Particle Info or stuff like that (but I’d really like to add that).

8 out of 10

esimacio, yep, I know why… It all seemed to be better in viewport render. I have to be more sarcastic that time.

The thing is that we clearly see there’s always the same rose shape…

Next time when you do a rose, when you do your first one, apply shape keys (5 or 6) to have different shape. It’s efficient and quick to do, it will normally do the job.

Other than that, it’s nicely done!

CurtYoung, thanks, there are 3 types of roses actually but yes, that’s apparently not enough, repetition is visible. And as those are different models I think it’s more resources hungry to do compared to shape keys… didn’t try. It could be a challenge though to figure out how to feed particle system with the same object but different shape keys active. I’ll keep that in mind.

when you do all the roses,
randomly select roses and press CTRL+P. You’ll have another object eg
Rose
CTRL+P you’ll get
Rose and Rose.001
Shape keys of Rose.001 won’t affect Rose.

Ctrl+P will parent selected objects to the active one. Did you mean to do that?

The only way I can think of is to use several shapekeys for the object and mix them. However this will either way mean creating several copies of the object to allow randomizing shape keys.

I don’t know, is it bad to have several copies of the same object?
Does it affect the CPU/GPU?

It affects RAM, that I can tell for sure. CPU? Well, I think not much - not more than maybe increasing render time for several percents (if rendering on it obviously).
It’s not bad to have several copies of the object. It might be a bit less efficient when optimizing which I wanted to get (and couldn’t).