Need some quick help with something...

I went with A. I liked the looks of it better.

If we're just supposed to judge the terrain and not the image quality, I'm going with A. Random peaks and dips in B doesn't look too hot compared to the relatively diverse terrain of A.

Sorry if it breaks your poll but I also went A, it just looked more visually diverse.

@fluxwavez: @mooseymcman:

Would you say that given the other differences between them like file size and time invested, that A is still preferable?

And about the random peaks and valleys, that is something that is definitely going into my "disadvantages" pile, most of the terrain the generator seems to spit out looks like that, a big part of my report so far has been:

"Here are the good parts of procedural generation, but here's why you will always need an artist to touch it up".

@bollard: It's fine, I was already writing about how my hypothesis has been disproved.

@serhulse: If you can speed the process of terrain formation by 1,000x using method B, then it's probably best to use that method in every possible circumstance. However, it's also probably best to use a handcrafted approach for the most important and visually relevant segments of terrain, which might be where your mention of artists touching up the procedurally generated approach comes in.

I dunno why I judged the two based on 'realism,' but I did.

B looks like a believable terrain in a crazy rough mountain area, A looks made up and silly.

The poll only asks which is better. I chose to look at it from a perspective of which can humanity develop better. The rigid peaks and valleys in B would be a much higher cost to build housing or roads on while the smoother topsoil and flood plains in A would make for easy land development. I went with A.

@serhulse: I guess it depends on what it's being used for, but yes, I'd still say that A is preferable.

@mosespippy: That's certainly an interesting way to look at it, and also very valid, I chose the word 'better', and told you to make a snap judgement, as I wanted to keep it as subjective as possible since I already have most of the objective proof I need.

So far I feel this will make for a more interesting report than I first thought.

@singingmenstrual: I agree, there is some thing about A that looks a little unreal. I think how smooth it looks does not help, B looks craggy and therefore more natural.

A

I can actually see things with A.

I voted A, though I have to say that I am not a good judge for this. I have a bit of experience with terrain generation and B looked familiar... It brought me back some unpleasant debugging memories.

Edit: And I gotta say that it's not perfect to compare a supposedly badly sculpted terrain with a somewhat well-designed terrain generation algorithm. At least I think it's relatively well designed, since it's definitely too smooth for a purely random height map and there's no egregious flaws in that image.

I picked B because it was bumpier.

@audiobusting: The OGRE engine is doing things behind the scenes to the model, I've only used OGRE for about 6 months so I'm not fully up on it but it is definitely doing LOD.

Also the textures (which are awful) are covering up some of the nasty stuff, just looking at the wireframe you can see it isn't as smooth.

I voted B because I like jagged mountains.

B looks better for me.

A looks like melting ice cream, and B looks like a shitty plastic terrain you would get packaged with a bundle of Star Wars figures, so i guess B is better in that it actually kinda looks like mountains.