/ic/ - Artwork/Critique

>>2193950

>>2193950

An edge is described using value change so the question is irrelevant, In my experience every fundamental principle needs an equal amount of attention paid to it so no, edges don't trump values.

Listen, edges are value changes, objects are defined by planes and those planes may have a hard transition, like a cube, or a soft transition like many parts of the human head, so edges and value are equally important, looks for hard and soft edges in this painting and how they relate to value changes.

>>2193958
>>2193985

i think you both didn't get my point. i'lll elaborate when i'm back home.

>>2194005
I don't think you understand whatever point it is you want to articulate.

>>2193950
>Do edges trump values? It feels like the brain is wired to spot form on the basis of edges than value.
Nope. In grayscale, you can describe form with value changes but only one type of edge (hard for example); you cannot describe form with edge changes and only one value. Edges build off of value just as color builds off of value.


>Could it be that edges should get much more attention as a fundamental principle in art? Or am I wrong to assume that they don't? Because nobody ever emphasized edges to me.
Edge control, color control and value control are all fundamental skills in representational art. Chances are you were looking to lackluster resources if you've never seen any emphasis given to edges.

That said going by Schmid's Ala Prima, to assess errors in a painting from life you'd check to see you have the right shapes, the right values, the right color and the right edges in that order. Edges are something to be concerned with in the refinement stage, whereas values are addressed early on in the block-in.

>>2193950
Values are fundamental. Edges are refinement. The brain is wired to spot silhouette when it comes to form - values will always be a clearer, more immediate indicator of silhouette from that background than edge will be. That's why black and white lineart that focuses only on edge and not value through line weight and/or hatching will always look kinda shit, no matter how well-composed, detailed, and clean it is, whereas a good picture can be made from strong values and very little focus on edge.

Is edge control and refinement an important thing to learn in art? Yeah. Does it trump values? FUCK no. Save that stuff until after you've mastered your values.

>>2194243

Dude you don't know what you are talking about. Lineart has nothing to do with this.

I think OP might be confusing edges and drawing. Both are related. If you have good drawing you can have bad values and it will still read. In his table example the topleft one which I assume is him showing bad edges is actually just bad drawing, the table is wonky. Mullins has talked a bit on this, about how if the drawing is right the values can be wrong. He uses HDR photography as an example. That being said, to completely ignore value as a tool for readability would be silly in 99% of cases.

>>2194259
I'm not that anon, but I think he has a good idea of using line art for comparison and his points are valid. Line art is art without value, so is pretty relevant to the discussion. It literally is just a delineation of the edges. That being said, he did cherry pick an example of a drawing that was likely done in preparation for a painting since it has no variation of line weight. Obviously when every edge is treated the same it will not read as well as it could. It would be better to find an example of line art that uses different line weights and in some areas forgoes line entirely to lose edges (though it may be hard to find an example of this that has no value at all, probably there will be some occlusion shadows or some hatching somewhere).

>>2193950

Okay I'll use this to try to explain my point. Give me a minute. Just posting this already out of impatience.

>>2194283
Might be a better idea to use master paintings as example as they will not require time for you to paint and they can probably get the point across better since they will be done by someone with very high level skill and decision making. Like you example is pretty rushed, so actually has sloppy edges and values which then will likely muddle what point you are trying to make.

>>2193950
Well in order for there to be an edge there needs to be a value change. So by definition a value change would be more important. They are pretty related though.

A relevant image is the famous optical illusion of Einstein and Monroe, in which up close the detail/edge hierarchy takes precedent and it looks like Einstein, but from a distance the value hierarchy overpowers the edges and it looks like Monroe:
http://images.lifeandstylemag.com/uploads/posts/image/58167/marilyn-monroe-albert-einstein-eyes-vision.jpg

So obviously both are important and can sway the read of the entire image. But the fact that values are readable at a distance and the fact that the edges in the close view still has value changes means that values would be the stronger of the two when it comes to importance.

>>2194289

Yeah, would be ideal. But to eliminate the differences I'll just whip up this little scene here.

>>2194295

>>2194295

I think it's an interesting effect and could be used effectively to create some "resistance" in your image. it invites the viewer to a closer inspection while simultaneously undermining the three-dimensionality of the objects in some ways

>>2194306
>all edges lost

But they aren't lost. They're just soft. The objects look like fluffy things instead of like smooth ones.

>>2194295
>>2194306

So now looking at it... yeah it seems like when you go completely balls to the wall with the values, any overlapping edges struggle HARD to tell your brain how stuff is sitting in space. Arguably it reads less than the ultra soft no-edge spongy example that still retains basic values. I think that's sort of what a 3/4-blind grandpa would see when he takes off his 4lb glasses.

But I'll try one more with values that are still very 'wrong' and ignore how light would actually work, while retaining some consideration for the shapes.

>>2194307
Good point. I think you definitely see guys like Sargent do this, although much more subtly. Unless I misunderstood your post, I'm not sure.

There. This is much better. I'm not sure I'd still agree with my initial claim but this shows the extent in which edges hold the read together even when you have 'incorrect' values and treat each object as if it were lit completely differently.

>>2194311
Not sure what you mean.

>>2194335
They sort of look like velvet now. Also did you read my post >>2194293 and look at the link? I think that image encapsulates what you are exploring. View it from up close and then from across the room (or blur your eyes really strongly).

>>2194338
Also you haven't really explored all there is with edges and values. Their relationship can be quite interesting and there are many ways to handle things. You can for instance use only hard edges, but have pseudo soft edges appear by the way you use values. Pic related only has hard edges, but the shapes are very carefully laid down (as are the values): https://congotronicartfxambassy.files.wordpress.com/2012/08/albertomielgojodiesmith.jpg

>>2194340
Here's another interesting use of edges. Lots of hard edges but they made the hard edges lost at the same time inside of the forms. And it relies of careful drawing to still read well. Very interesting stuff!

you can describe form with edges and have a low range of values. this is advantageous when creating value structures where multiple objects appear with the same value range i.e. background objects, crowds of people. defining a separate value range for each person in a crowd would break the composition/distract from focal point. edges take precedence in that case.

low contrast(value) also supports color identity but thats treading towards my taste and is not actually relevant

couple of examples; we've all seen it done before

Bump.

Thread has potential for interesting discussion regarding use of edges. OP seems to have abandoned ship though.

>>2195300

i'm here i'm here. i just want to take the time to think about the recent posts (which are 10/10 replies imo) and right now i'm in a bit of a pickle that demands my full attention. client work and the ex gf will be outside of my flat in 3 hours. shit's fucked, i haven't even shaved my balls yet. appreciate the bump though.

>>2195324
Well you let us down anon. And you are screwing yourself over by toying with fire (an ex). Let your pubes go wild and embrace the loneliness, then come back to us and to art.

>>2193950

My goodness, gem of a thread by /ic/ standards. I agree with a lot of the points already made and to be honest, I don't know exactly what you're looking for if your question hasn't already been answered. But, edges do not trump values, since edges are essentially just the progression of values by either visual mixing (noisy brush texture) or by smooth gradation.

In the end it's all just representing what you see. Edges are a component of values, but they're also a part of draughtsmanship and technique. In those regards, they're important, but they don't "trump" values. Edges can be used to wonderful effect though, like how a hard edge between two similar values can make them feel separate, or a soft edge between two different values can make them feel closer together. Also, harder edges tend to arrest the eye, whereas soft passages let you skip over stuff - but again, this goes in tandem with drawing focus in value contrast. You can see a ton of both happening here >>2194572.

>>2193950
our of those two, values are far more important.
even your images are proof of that, because they all read about the same, and that's because they have the same values.

think in terms of contrast. value is what makes your image read in the first place. it is also the thing thar regulates the impression your edges make.

also it's like some of the other anons mentioned, you are indeed thinking about drawing here.
for example: if you have a good drawing, you can use whatever values you want to support it, it'll still read bcause the drawing itself is descriptive.

also about this one >>2194295: it would be much less readable if you didn't have the patterns and the local values with some of them.