01-20-2014, 09:41 AM
Samszym said it:) I mean, it's literally some kind of invisible force making my drawings more "normal":) It's like "ok, let's draw this stylized thin anime girl" and the "force" is like "oh no you don't, we are going to have normal waist, legs are going to be thicker, knee is going to be moderately sized and shins aren't going to be longer than thighs". And then you get some anorexic kid with bulging knees and pronounced bone lines:D
I think it's about constant learning and application and also studying of some good examples of exaggeration. I think even tracing could help sometimes, even though I haven't tried it. I suppose it could literally force you to "unwillingly" trace the original artwork and go like "ooh, these hip bones are reaaally low hur dur" :D
For example, if you squint your eyes, you'll see that (on the screenshot) the jacket and the shelf behind it form almost a single value mass, which could be put down as such, for starters. And later, when you are breaking down the jacket, you would see that it consists of two rough value parts which are pretty close in value. And, for example, that the figures behind are closer in value with the cloth and shadows to the left and above and that too could be marked down as a single value mass when you start and then it could be broken down later. This way you are more likely to avoid having "value jumps" or putting down incorrect details.
Of course, in the anatomy study or something else than what you are trying to achieve in this Pacific Rim piece, your approach should be more 3d and more informational etc.
A good practice for turning off the "scumbag library" is to turn the composition upside down and paint that. It messes with your recognition, so you are like "ok, this here is this shape and this value and it's next to this and I don't have the slighest idea what i'm looking at" and not like "ooh look, a chair, I know chairs".
I think it's about constant learning and application and also studying of some good examples of exaggeration. I think even tracing could help sometimes, even though I haven't tried it. I suppose it could literally force you to "unwillingly" trace the original artwork and go like "ooh, these hip bones are reaaally low hur dur" :D
Quote:And finally, I started a study of a pacific rim screenshot...when I made it ready for upload I already saw a million wrong things. Likeness always takes a long time in to develop in my paintings...XD Overall I think I still use colours that are too light. I should do some value studies first maybe...and: does anybody have advice on lips and colour? I just never seem to get it right. Everybody looks like they are wearing lipstick all the time....My advice (and others may disagree) would be to try and analyze the image like a neutral composition - so there are no "lips" or "people" as such, only surfaces and shapes that have their value and color info. Our brain, if unchecked, automatically calls for our library of things and stuff and then we draw from that and we draw "lips no. 14" and stuff like that, which are lip shaped and lip colored and often look like our own lips (as we know those the most). While they might be totally obscured or merged with the neighboring value surface or green or something like that. Try to lay out the image the least partially as you can, squint your eyes and merge all the similar values.
For example, if you squint your eyes, you'll see that (on the screenshot) the jacket and the shelf behind it form almost a single value mass, which could be put down as such, for starters. And later, when you are breaking down the jacket, you would see that it consists of two rough value parts which are pretty close in value. And, for example, that the figures behind are closer in value with the cloth and shadows to the left and above and that too could be marked down as a single value mass when you start and then it could be broken down later. This way you are more likely to avoid having "value jumps" or putting down incorrect details.
Of course, in the anatomy study or something else than what you are trying to achieve in this Pacific Rim piece, your approach should be more 3d and more informational etc.
A good practice for turning off the "scumbag library" is to turn the composition upside down and paint that. It messes with your recognition, so you are like "ok, this here is this shape and this value and it's next to this and I don't have the slighest idea what i'm looking at" and not like "ooh look, a chair, I know chairs".