[SFW] Artist's Group Chat

• What you’re saying about the shape being simple resonates- I seem to make the same mistake with pretty much anything I fail to use a ref with- I think of everything in simple, malleable volumes, but have a difficult time making them actually look like anything but that. For example, Scootaloo and the clouds and the backdrop look like they’re made of the same thing in my latest image (following this post).
   
  -And, alright, I think I understand what you mean by lack of transitional information, but how could I provide it?
   
• Yeah! I tried something unusual that helped me better understand that perspective and placement is relative to everything else- the movement, the POV, everything, in my recent image!
   
• Ah, see, that gallery actually is really good, I’m not denying that, but it really FEELS like art.
   
  It’s almost cinematic. It’s dramatic, and there’s nothing wrong with that, I actually really do like that sort of stuff, and hope to make something really cool that might incite the same-ish sorta feeling. But my point is that not everything needs to be like that- an image could be a lot more subtle, where it feels like you’re looking in to a scene rather than looking at something that looks really cool design-wise, but while still using compositional techniques like the cactus for framing or value separation as a believable, natural lighting occurence.  

– What you’re saying about the shape being simple resonates- I seem to make the same mistake with pretty much anything I fail to use a ref with- I think of everything in simple, malleable volumes, but have a difficult time making them actually look like anything but that. For example, Scootaloo and the clouds and the backdrop look like they’re made of the same thing in my latest image (following this post).

 
I have a cure for that: stop thinking about them as 3D object for a moment, and look at them as 2D shapes. Not 2D clouds, not 2D trees; consider them as shapes. Then, just focus on making the shape interesting, by adding variation to it.
 

 
Of course, you have to study the kind of variation allowed in any given object. So for this, I’d suggest just look up pictures of what you want to improve on, and just paint/draw their silhouette. Only after you got the shape right would you apply your knowledge of volume and form to get correct light/shadow.
 

-And, alright, I think I understand what you mean by lack of transitional information, but how could I provide it?

However, I feel the pose could have been bettered by emphasizing further how the connection between Twilight’s chest and pelvis is twisting.

Also, considering that those clouds are touching the horizon line, Rainbow Dash might’ve gone quite a long ways.

@Grieffon

Like this? (On both counts, regarding the new one)

 
Yeah, that’s the idea. Though it’d be even better if you have the silhouette a controlled kind of variation, instead of random. For example
 

 
Maintain the variation in the shape (thin - thick, leveled - pointy, etc), but keep it look like something that can exist.
 
And thanks for the feedback! I’ll keep them in mind in the future.