GR#188 - Erde Scene - Pixel Artwork Scenery

[||||]

Something I recently pixeled showing no WIP (I probably should have posted a couple days ago). I like to just work on things continuously sometimes from start to finish and at my own high pace.
I realize it's odd to post something not sketchy here but here it is anyhow,
(You never know what might happen).

[PixelPiledriver]

Can you explain this?
Is the blue stuff the tops of trees?
Clouds?
Is it just mean to be somewhat abstract?
It's fun to look at, but I don't really know what it is.

[jams0988]

I'm reading clouds. You can see the slightly greener trees poking through.
I had to look a few times, though. The clouds were reading like tree tops at first to me, too...
What's really confusing for me is the perspective. It feels like the horizon should be a lot lower? It feels like the ground is curving steeply upward. It's a little disconcerting!

[||||]

Yeah. the blue stuff is 'low-flying' clouds, or fog. Like the ones you see in N. California (at least that's what I got the idea from seeing; in the wine country areas particularly), or more so in China I reckon; the misty sort of clouds seen in lots of old ink paintings. The clouds are more blue than they'd really be though. Yet then again in this fantasy world the light could be more blue... There are three elemental moons in orbit and the Ice moon shines blue.

So, here's  a labeled version to clarify what I intended. It's a lady overlooking a valley at night which is mostly enshrouded in clouds:

[PixelPiledriver]

The blue isn't really a problem.
It just seems overly thick and lumpy.
Which is why I asked.
They look like solid objects.
Instead of seeing whats through the fog in various thicknesses, like an variable atmospheric perspective, I just see the fog as a thing.
No time for edits.
Maybe look at some photos.

Stuff
Thing

[wolfenoctis]

Just want to quickly add some very rough thoughts on composition and horizon line:

I also thought it would look cool if the sword had a faint glow.

Hope it helps 

Edit: Also not sure about colors just used some from your piece

[Manupix]

I must say I was confused as well. The blue clouds not only don't read as clouds, but they don't have any sense of scale, distance and perspective (in the sense of being a rather horizontal, vertical or other, expanse). One reason is the evenness of the texture scale.

The tree foliage texture is weird too; the tree trunk doesn't contrast enough against the bg.

Wolf's edit raised interesting points. I'd go a different way though. What I like best in your piece are the upper left corner (landscape and sky) and the lower right (character and tree trunk), but they're lost in the fog and foliage. Maybe you should get them close together and make the canvas smaller.

[Drazelic]

I worked off Wolfenoctis's post a bit with the character. (mostly was too lazy to draw a character on my own lol) In my version though the primary lightsource isn't the sword but rather the starry sky as a whole.

[||||]

This doesn't address the lumpiness of the clouds but it's a different composition; raised the foreground, tree and character. Gave the clouds more depth (just a quick transform, would have to repixel over them). Just a quick edit based on some of the things shared here:



Will likely do a total rework on this, an Erde Scene Ver. 2; but as I mentioned just posting to see what comes up for now.

Pre-post- update
(just before posting I see Drazelic's Edit; looking good)..
 definitely a lot of new ideas coming in for Ver. 2 already. Thanks guys
(although you're really changing the character design)

[Mr. Fahrenheit]

Last update helps immensely. Keep working on it and dont give up 

[Ryumaru]

You rendered without first having a solid foundation. For the foliage, you need to think of clumps and forms, not individual shapes you can AA to look like leaves. Sometimes even with a good start you can fall into this over-individualization. Same thing goes for the segmentation of the tree which neither looks realistic or cool, but as disfigured and lumpy as the clouds.

Your palettes and technique for the most part is good, but I've seen a trend in your work of losing the whole when it comes to creating convincing forms that exist in space by poor shape design and over-rendering.

Your latest edit is a step in the right direction; you are getting good critique here and you have the ability to execute.

[RAV]

Ryu's advice is true. over-rendered is a strange word though. even over-detailed can be misleading. For example I think the problem with the clouds is, that it has the sort of detail definition that suggests it's something else than clouds, that is to say most of all it's wrong-detailed. Altogether a lot of the detail seems to be made up than instrumental to identity of object. Maybe it's a problem of working from memory instead of reference. Clouds don't have so strong surface, if anything complexity would be more in silhouette. Shades are soft and also large in volume, since the interaction with light is very diffuse. Also take into account what the lighting conditions overall mean for what detail is visible, like with the leaves of tree. and "compositional detail", maybe the shading detail of the far clouds also contrasts too much with that of the mountains.

And yet it looks pretty nice, keep up the good work.

[cels]

I have a question, more than a criticism:

Isn't it a bit problematic when the sky in the upper left corner is more interesting to look at than the landscape below, which seems to be kind of where our eye is meant to be leading? I mean, this kind of composition reminds me of the typical picture where a person has climbed a mountain and is looking down at the sprawling landscape before him / her, usually with some sort of building or settlement or special feature in the landscape as the focal point. So the eye is drawn to the character, who is looking at that feature in the landscape, and our eye follows his / her gaze to the horizon.

In this picture, there's a character who appears to be looking at just... a blanket of clouds.  And then our eye is rather drawn to the stars at the very top of the image, instead of the clouds.

Wouldn't it be more natural to place an object of interest, towering over the clouds or flying through the clouds? Or, if that shiniest star is supposed to be an important feature, move the entire canvas higher, so we see less of the foreground and more of the sky above?

Again, I don't know much about this sort of stuff and how to draw good scenes that guide the eyes naturally through different points of interest, so this is very much an open question, rather than a criticism.

EDIT: Actually, Drazelic did make the clouds more interesting to look at, because they're so dramatic that they appear like crashing ocean waves. Whereas in ||||'s version, it's a rather featureless carpet of lumps. So I guess that's why Drazelic made the change.

Spoiler

[RAV]

The mind-blowing vastness can be spectacular to behold.

Also relevant is this classic painting from Caspar David Friedrich - Wanderer above the sea of fog.

[cels]

The mind-blowing vastness can be spectacular to behold.

Also relevant is this classic painting from Caspar David Friedrich - Wanderer above the sea of fog.

But surely, one of the reasons that painting works well, is because there's nothing distracting in the upper 10% of the painting? And the mountains all create lines that guide the eye towards the focal point, the "wanderer".

[RAV]

I do think that a scene can be interesting overall, without a particular object of interest / focal point. The dramatic vastness is then of exhilarating interest, if you're coming from crowded space like a dungeon. Maybe the impact is increased by how well we can empathize with the depicted character experiencing it with or for us. In wolfenoctis and drazelic's version, the character is a bit more prominent by various means, and the tree less clearly dominant. Also of note is the ambiguous game of power that drazelic's tree plays, with the retreating trunk to the side away from character, and coming back mightier from above, adding to the interest of the spectacular view, but not at the cost of character.

[||||]

You rendered without first having a solid foundation. For the foliage, you need to think of clumps and forms, not individual shapes you can AA to look like leaves. Sometimes even with a good start you can fall into this over-individualization. Same thing goes for the segmentation of the tree which neither looks realistic or cool, but as disfigured and lumpy as the clouds.

Your palettes and technique for the most part is good, but I've seen a trend in your work of losing the whole when it comes to creating convincing forms that exist in space by poor shape design and over-rendering.

Your latest edit is a step in the right direction; you are getting good critique here and you have the ability to execute.

Thanks for the post..  That tree; So true about the texture.
 I really learned something important in all this.. never work too hard rendering something without first either putting a lot of thought into it or sharing compositional sketches. Those clouds took a long while and yet  they're useless.. a waste of hours.  (I actually may recycle them for a transparent layer in another scene; so possibly they will serve some other purpose)

Cels: I don't know either. I didn't want anything specific as a point of interest or focal point in the background. I wanted something mysterious and dark.. but not too dark, ie the stars breaking through the veil of blackness, the semi happy colors, etc.
I actually considered putting a tower in the background but decided against it because I felt like it's been done so many times. I can think of a few games even where the object or destination is seen in the background as the character looks upon it. Although yours does look pretty cool peeping out over the clouds there.. hmm decisions decisions.

What are the divisions in those two edits of Drazelic's example for?

Rav: Glad you said you think a scene could be interesting sans a particular object of interest, cause there really was no intended focal point beyond the ledge in the foreground.. just an open scape with many possible points to look at.

I may crop it to the size in the quick edit I posted although I'll fully redo the clouds and tree. the main issues were the horizon line and the textures.. hopefully with those two things corrected I can make the image clear. Although I'm curious on peoples thoughts on focal points and whatnot.

[RAV]

I think the feeling of what's better depends on the "narrative" context. If this is the introduction screen to an adventure, maybe it makes more sense to have a destination point in far away view, the character tight and ready to go. But if this is the conclusion to defeating that tough dungeon, this epic sense of freedom at last is all it needs. In that moment the character would relax, loosen up cloth and hair.

[cels]

I really learned something important in all this.. never work too hard rendering something without first either putting a lot of thought into it or sharing compositional sketches. Those clouds took a long while and yet  they're useless.. a waste of hours.  (I actually may recycle them for a transparent layer in another scene; so possibly they will serve some other purpose)

I did the same thing with this piece:
http://www.pixeljoint.com/pixelart/64366.htm

Didn't know anything about composition, didn't know how to draw clouds, didn't know how to adjust light and colours in a landscape scene. Started rendering too early and spent so much time drawing and redrawing those clouds.

Cels: I don't know either. I didn't want anything specific as a point of interest or focal point in the background. I wanted something mysterious and dark.. but not too dark, ie the stars breaking through the veil of blackness, the semi happy colors, etc.
I actually considered putting a tower in the background but decided against it because I felt like it's been done so many times. I can think of a few games even where the object or destination is seen in the background as the character looks upon it. Although yours does look pretty cool peeping out over the clouds there.. hmm decisions decisions.

Well again, I don't think that you really need a tower. I agree it's a cliché. Same as an approaching ship (or in my case, above, an approaching airship). But I do think you need something that is a bit interesting to look at, or just leave it featureless enough that the character becomes the only focus and the horizon serves only to create a sense of emptyness.

But again, I hardly know anything about this, so I'm just raising the question to get an answer from the educated artists here.

What are the divisions in those two edits of Drazelic's example for?

Rule of thirds. A rule of thumb that can be useful in some cases. In Rav's painting above, it seems to be applied vertically, but not horizontally, for example.

http://en.wikipedia.org/wiki/Composition_(visual_arts)
http://en.wikipedia.org/wiki/Rule_of_thirds

[Ryumaru]

You rendered without first having a solid foundation. For the foliage, you need to think of clumps and forms, not individual shapes you can AA to look like leaves. Sometimes even with a good start you can fall into this over-individualization. Same thing goes for the segmentation of the tree which neither looks realistic or cool, but as disfigured and lumpy as the clouds.

Your palettes and technique for the most part is good, but I've seen a trend in your work of losing the whole when it comes to creating convincing forms that exist in space by poor shape design and over-rendering.

Your latest edit is a step in the right direction; you are getting good critique here and you have the ability to execute.

Thanks for the post..  That tree; So true about the texture.
 I really learned something important in all this.. never work too hard rendering something without first either putting a lot of thought into it or sharing compositional sketches. Those clouds took a long while and yet  they're useless.. a waste of hours.  (I actually may recycle them for a transparent layer in another scene; so possibly they will serve some other purpose)

Cels: I don't know either. I didn't want anything specific as a point of interest or focal point in the background. I wanted something mysterious and dark.. but not too dark, ie the stars breaking through the veil of blackness, the semi happy colors, etc.
I actually considered putting a tower in the background but decided against it because I felt like it's been done so many times. I can think of a few games even where the object or destination is seen in the background as the character looks upon it. Although yours does look pretty cool peeping out over the clouds there.. hmm decisions decisions.

What are the divisions in those two edits of Drazelic's example for?

Rav: Glad you said you think a scene could be interesting sans a particular object of interest, cause there really was no intended focal point beyond the ledge in the foreground.. just an open scape with many possible points to look at.

I may crop it to the size in the quick edit I posted although I'll fully redo the clouds and tree. the main issues were the horizon line and the textures.. hopefully with those two things corrected I can make the image clear. Although I'm curious on peoples thoughts on focal points and whatnot.

Yes. Learning this lesson can be quite disheartening. And it may happen more than once, it has for me. Something simply stated right is always better than something highly detailed wrong. Look at henk nieborg's work and you will notice the powerful forms he creates with only 5 colors, rarely any AA, and he never loses the big idea to finnicky details.

These hours are not wasted, however. You've lost some now to save many in the future. Getting the image right will be all the more rewarding. I suggest color reducing some images of clouds and seeing the shapes that are created. It may give you a
Better idea of how to design more natural, form provoking shapes.

[astraldata]

I suggest color reducing some images of clouds and seeing the shapes that are created. It may give you a
Better idea of how to design more natural, form provoking shapes.

I haven't considered doing something like this before, but it really seems to be a good way to study how to represent more complex forms in pixel art. Nice to hear a technique like this suggested by a skilled pixel artist -- I'm sure many folks might feel like such a thing is 'cheating' in a sense, but really, studying how pixels actually cluster from natural subjects sounds like a great technique.