Iris

[LaCan]

DING DING DING DING DING! We have a winner!

Well, the winner was known as of post #31. It just needs some time to get widely acknowledged

[Black Spot]

I'll be back at the weekend.

[bill618]

I'll be back at the weekend.

If you’re reshooting reference photos, I would strongly suggest using a 50mm lens (if you have one), if you want that natural-eye-view of the scene. Even better, if you have a zoom, start at 50mm and determine how close/far you need to be from the subject to get what you want. If you feel the subject is too small in frame, step a little closer while increasing the focal length a tad (*a scene through a lens has the full range of increasing focal lengths out to infinity starting with the one it is set at). Don’t crowd the subject at 50mm just to fill the frame, you’ll just increase the perspective distortion effectively producing a wider lens effect at an otherwise comfortable subject-to-lens distance. You might even want to position a large enough cubic object, or two, in the scene to make quick work of creating a perpective grid, if you think you’ll need one.
But then you may know all this already.

EDIT: Thanks Cider. I forgot to mention 50mm lens on a full frame dslr. You’ll have to find the 50mm equivalent based on the crop factor of your camera sensor, if it’s not full-frame.
I'm still in the mind-set of 35mm film. I'm old(ish). Most of my good junk is on slide film.

[Cider]

If you’re reshooting reference photos, I would strongly suggest using a 50mm lens (if you have one), if you want that natural-eye-view of the scene.

Note that bill618's advice is the usual suggestion for full-frame cameras. It works out to something like 27mm on a Rebel, because of the smaller sensor.

[bill618]

2) Figure's perspective is ambiguous. We are not sure where the viewer is standing.

You may not know where the viewer is standing. The geometry of the linear perspective space is pretty obvious to me as well as the general xyz coordinates of where my eyes are in both images. This set space is the same whether she is sitting on an idealized Platonic solid or laying flat on her back on the Living room floor with her legs plopped over an ottoman. Where would you guess the viewer is, if she were sitting? Where do you think the horizon line is? There are plenty of visual cues staring right at you.

The z planes of the cube are parallel to the plane of her upper back:





I’ll get to your other points later. I’m balls to the wall busy, atm.

PS–I think you’re doing a good job helping folks out in this forum, so far.

[Griffonnage]

I'm quite sorry Bill618, but I agree with Giacomo and Lacan about the view angle issue. I think the initial references point of view and the one in the painting are differents and all the issues come from that. No need to start a war

Actually your sketch here do not works : the floor do not supports the lady, she is floating. No way you can stand this position, just try it on your chair, without your knees or the foot on a ground/chair/support if you don't trust this. Currently I see her standing using as only support a small part of her butt. The feet, the right arms, none of them support the body. You definitly can't be that relax.

edit : by the way and looking again at it, even if she was lying on the floor I really have a problem with the center of her body, it miss a part, like half an head-size I think and I'm not sure this actual perspective can explain it...

[Aceoko]

Would her right foot being held up by the ground help in anyway? I don't know, I find the discussion about this a little pointless.



Does it really matter how realistically her body is positioned and weighted down considering she is posing(posing causes unnatural positions) amongst a bunch of flowers anyway?

[bill618]

Griffonnage–The more controversy this supposed “wishy-washy” wonky pose generates, the more I like it.

Just think, the heel of that 'floating' right foot could be firmly pressing into the cheek of a severed head.
Who says she’s relaxed–or has to be?

The main problem with perspective, let alone how camera lenses influence it, is the inadequate understanding of it by too many artists.

What you’re looking at is a reference cube surrounding the woman. The cube and the box she is sitting on are reference objects defined by and defining the perspective grid space of the image. There is no established floor there, yet.

The perspective space of the post# 65 image has nothing whatsoever to do with any reference photos posted. The one and only purpose for the quick sketch I made was to show a nude rendition of Black Spot’s latest iteration for loose reference for whatever it was worth. I wasn't, and am not, set out to reinvent BlackSpot's painting. I used her image as an underlay to conform fairly close to what she had. Yes, the subject has a short trunk (well within the anatomical tolerance range of a healthy person). No, she’s not a pinup model. Does she have to be?

[Griffonnage]

She do not have to be a pin up I agree but why she wouldn't ?
If it was intended to do a painting that break the rules of classical painting I would say nothing. But here we have an antique inspirationnal painting, classical pose and mood attempt, definitly there are no interests to do a portrait and do not respect the model anatomy. I understand it's no easy to redo a full body on traditionnal canvas, but doing something with one intention will create some rules.
In my opinion, it would be really sad to have two message in that painting. The distorded perspective/anatomy is not intended, so it do not have to exist, that's my opinion.

By the way I'm not sure the body perspective is working with the perspective cube, she sould be really really more distorded to fits it. So... doing change to do change, why do not just let Black Spot takes a new reference shot and work in peace from it ? Simplier is better !

[bill618]

I don't know, I find the discussion about this a little pointless.

Yes, it is by and large pointless. But there are a few of us here that relish in such anal retentive, mental gyrations, for the heck of it.
Then when I sit down and do art, all of the disparate seeming nonsense I fill my head with, blends smoothly into a tacit sea of warm creative consilience. [cough]

[Black Spot]

I've just got an average digital camera, nothing fancy. I originally stood on a chair to take the reference shot. I'm trying to get my model back to reshoot, but in the meantime I've made some dough to construct a manikin. As for the painting, I've sanded it down and gesso'd over the figure - the paint was getting mile to thick.

[bill618]

I've just got an average digital camera, nothing fancy.

If the camera has an unmarked zoom, just stay in the mid (low middle range) zoom extension to try and keep the angle-of-view as ‘normal’ as you can.

While you’re waiting for the model, you could also take a look at posespace.com to get some ideas for a sitting pose.

*nsfw alert*
http://www.posespace.com/img/contact...10contacta.jpg

[LaCan]

Bill, I think your diagram just illustrates my point.

I've got nothing further constructive to add to the discussion until the next iteration arrives.

[bill618]

Bill, I think your diagram just illustrates my point.

Hm. That's not possible, if you say you’re unsure of where the viewer is.
You can’t begin to create a scene in accurate linear perspective without first knowing exactly where the virtual viewer is located. The angle of view is the only thing that is variable. And before you make the claim that the perspective space in my diagram is wrong I’d like to add that it is based on a 3d simulation of the figure using an accurately proportioned 3d human model to match the pose and perspective (seen through a 50mm lens). I threw the 3d scene together to verify the fact that the location of the viewer is far from a mystery, as you imply.

[LaCan]

Hm. That's not possible, if you say you’re unsure of where the viewer is.
You can’t begin to create a scene in accurate linear perspective without first knowing exactly where the virtual viewer is located. The angle of view is the only thing that is variable. And before you make the claim that the perspective space in my diagram is wrong I’d like to add that it is based on a 3d simulation of the figure using an accurately proportioned 3d human model to match the pose and perspective (seen through a 50mm lens). I threw the 3d scene together to verify the fact that the location of the viewer is far from a mystery, as you imply.

Sure, you can scaffold the working perspective around the pose, even few of them that may look somewhat believable I'd suspect. But that would not make the gesture working. There is unfavorable back and forth between the perspective and the gesture. That was my point. Something always boggles the eye. It may be subtle, but it's noticeable. Don't know why are you getting so defensive about it.

[LaCan]

I’d like to add that it is based on a 3d simulation of the figure using an accurately proportioned 3d human model to match the pose and perspective (seen through a 50mm lens). I threw the 3d scene together to verify the fact that the location of the viewer is far from a mystery, as you imply.

Would it be a problem for you to export the whole scene into *.obj and post it. I'd like to see how the body is positioned in space.

[Black Spot]

Now guys play nice. One thing that hasn't been mentioned is that the pose had the model slightly leaning forward and her backside is slightly sunk on the springy flowers. In my manikin I've exaggerated it as I don't want her to be a pinup in the traditional way. My camera is a Cannon Powershot A490.

Never made home made salt dough before to play with - it's not great to model with. So after an hour, using a teaspoon, I took it into PhotoPlus to soften the edges.

[Hexokinase]

Is there a particular reason why you're sticking with that pose and that camera angle? What qualities of the model (both appearance and personality) are you trying to emphasize? What sort of mood are you going for? Are there any classical paintings that have a similar feel to what you are aiming for? If you were to describe the intended impression on the viewer using 5-10 keywords, what would those keywords be?

My personal grievances are that: a) the pose lacks clarity of intent, and b) the chosen camera angle isn't the most flattering. I trust that you did a series of thumbnails to figure out the pose and composition before shooting the first batch of reference photos, but is this truly the best direction?

[Black Spot]

Never had that idea in your head and wonder how far you could push it? There is no intent, it's purely decorative. Lip service to ancient Greece, as were a lot of classical pictures being early porn dressed in respectability. Getting out of my comfort zone. Isn't that what learning is about? If you don't try, how the heck are you learn. So it might end in a horrible mess, but that doesn't mean I haven't learnt anything in the process.

Goddess Iris - sister of the harpies, messenger of the gods, fetches water from the river Styx for Zeus. Makes rainbows (shoot me if I put one in). May or may not be a virgin - might have married Zephyus and be the mother or Eros or might not. Didn't take side in the Trojan war but might have told Helen's hubby about her running away with Paris. There are loads of conflicting stories about her. My model is very sweet, I don't think Iris is.

[Velocity Kendall]

"I’d like to add that it is based on a 3d simulation of the figure using an accurately proportioned 3d human model to match the pose and perspective (seen through a 50mm lens). I threw the 3d scene together to verify the fact "

to be fair poser for example makes horrible reference, the people carry no weight in their frames and always look awkward. but maybe youre using a more sophisticated ragdoll than that..

[Hexokinase]

Never had that idea in your head and wonder how far you could push it? There is no intent, it's purely decorative. Lip service to ancient Greece, as were a lot of classical pictures being early porn dressed in respectability. Getting out of my comfort zone. Isn't that what learning is about? If you don't try, how the heck are you learn. So it might end in a horrible mess, but that doesn't mean I haven't learnt anything in the process.

Goddess Iris - sister of the harpies, messenger of the gods, fetches water from the river Styx for Zeus. Makes rainbows (shoot me if I put one in). May or may not be a virgin - might have married Zephyus and be the mother or Eros or might not. Didn't take side in the Trojan war but might have told Helen's hubby about her running away with Paris. There are loads of conflicting stories about her. My model is very sweet, I don't think Iris is.

That's fair - I'm all for painting according to your vision and painting for the sake of stretching your boundaries. But how do you know this is the best approach for posing and composition? And just so that I'm clear, is the idea in your head to depict this pose, this goddess, or this goddess in this pose?

Granted my knowledge of Greek mythology is limited at best, after reading your description and looking at the original painting, it does not scream 'Greek Goddess Iris' to me. Sure there are distinctly Greek elements present, from the garment to the pillar, but what tells me that there is divinity here, and not an ordinary woman? How often would a mere bystander be looking down upon a Goddess? What in this picture should tell me that this is Iris, and not any other goddess? I like the inclusion of the vase, and think it is key in this regard. I don't understand the link between the abundant flower bed and Iris (again noting my limited knowledge behind the mythology).

It's great that you've decided on Iris's personality, at least to the extent that she certainly isn't sweet. If there are several conflicting stories about her, I would stick with one, and use that sense of clarity to guide the gesture, the vantage point, and the composition.

I hope my ramblings were at least of some use. Should you decide to stick with this exact pose and vantage point, might I suggest sculpting all the elements you intend to include in the painting to ensure the perspective is undoubtedly consistent.

[bill618]

[LaCan]

@bill618

I was hoping you'd post a 3d model, not just the picture of it. You could have at least included a sitting box and a bounding box, like you did for the drawing, so we can compare things.

Nevertheless, this snapshot is useful. Though it still lacks a lot of articulation and natural grace, it got rid of some of the acute oddities from the initial sketch. I do hope you see that. Let me just point out couple of those, interpreted using the perspective you last proposed:

1) Extra high elevation of her right ischium, resulting in an unlikely pelvic tilt. Anticipated counterbalancing tension in the upper torso caused by this is missing too. Her heinie is in much more attainable position in 3d rig.

2) Seeming position of her left knee in the transverse plane suggests an uncomfortable adduction of the femur, with knee reaching over the medial saggital plane. Not what most humans do when sitting, even if curled up in atypical ways. The effort is laborious, not graceful.

I'm glad however that 3d pose is actually converging into gesture I did in post #45 (which is not perfect either). Compare them. If you eliminate that uncomfortable forearm supination, and let her right thigh rest down naturally - it's basically the same thing.

Anyway, I think any additional chewing on this pose is just beating a dead horse.


@Black Spot

I didn't realized up until recent discussion that this is supposed to be a depiction of a goddess. I thought it was just an allegorically themed ordinary portrait. I'd certainly go with a low horizon if goddess is the main theme, as Hexokinase already hinted.

Here's a very nice painting of Iris by Guerin. It might be a good point of departure for your image, in terms of atmosphere and pathos. I certainly wouldn't be ashamed to be heavily influenced by it.
You probably saw it already if you did some research, but here goes nonetheless:

[Velocity Kendall]

the contrast between the 3d thing and the painting is.. marked.

[bill618]

LaCan--Dude, your’re gonna turn your grey matter into a heap of ashes before your 30th birthday. I’d suggest having a virtual drink over at the Lounge forum (or a real one then go visit) to loosen up.

I figured throwing this disembodied graphic companion to something I said earlier (scroll up to the red devil smiley) might send you off on a wild goose chase.

That 3d model is not the model I used in the exercise to find Waldo’s eyes in the scene. Since that was a quite simple exercise, requiring only a few key landmarks, and a few minutes, I used a quite hideous segmented model. I ain’t postin’ it—unless you show us some of your good stuff.

The left hand? That sir, is the ‘Talk-To-The-Hand’ hand she is palming at the critics on the sideline.

So, that model is a Gumby-bent ZBrush model I quickly twisted up for a 'funny'. Sorry you ran down the field the wrong way with it. I should have just modeled Gumby.

…and no, my dear Watson, the plane cleaving the pelvis into two bilaterally symmetrical halves does not intersect the left knee. I at least made sure of that last night before bedtime.
The anatomical crits on that Gumbyfied bend job made me smile.
–The Lounge I say, go now.

–The Guerin Iris is splendid.

[Velocity Kendall]

"I ain’t postin’ it—unless you show us some of your good stuff."

who me? or lacan? what is good stuff? id like to see the model. if it really is very hideous, maybe its not accurate enough to model the subtle stuff LaCan means..

"Sorry you ran down the field the wrong way with it. I should have just modeled Gumby. "

i dont understand this part. or the wild goose chase part actually. seems like a cop out. heres hard evidence to back my ideas; haha i tricked you its a 'funny'..? bit odd that.
seems like a missed opportunity to really make your case stick which i assume is what you want..?

[Griffonnage]

Seriously what's with this attitude ? Look down on others just because they ask for your 3d references ? It feels we just have a troll here that don't want to quit... really hate that.

[bill618]

"I ain’t postin’ it—unless you show us some of your good stuff."

who me? or lacan? what is good stuff? id like to see the model.

"Sorry you ran down the field the wrong way with it. I should have just modeled Gumby. "

i dont understand this part. or the wild goose chase part actually. seems like a cop out. heres hard evidence to back my ideas; haha i tricked you its a 'funny'..? bit odd that.
seems like a missed opportunity to really make your case stick which i assume is what you want..?

No, no, I wasn’t referring to you after the ‘Poser freak’ line, nor before. Just letting you know (in response to post# 80) that it wasn’t a Poser model, since I didn’t reply to your post.

*The lines gone. I guess it was confusing.

[Velocity Kendall]

Ah gotcha.

[bill618]

Okay, this mole hill keeps getting brought back into the light. I got a little time to play. Here’s the rag doll setup. You don’t need a finished ZBrush sculpt with 50 hrs of work in it to get the camera/eye position in the sketch scene. We’re not going down to half second tolerances here. Blocks for the body and limbs would suffice. You don’t even have to go this route to approximate closely where the viewer’s eyes are in the scene to develop a perspective grid space, even based on the wonky drawing with it’s anatomical peculiarities. All you have to do is look at it—with an understanding of perspective. Contrary to what’s been stated as a problem with the image "Figure's perspective is ambiguous. We are not sure where the viewer is standing."
The viewer is at the #1 position, not 2, not 3 or any other position, obviously. Again, the general #1 position of the viewer is not a mystery. It is easily determined by just looking at the sketch.
Yeah, her left thigh is sore, as evidenced in the sketch.






That’s it for the mountain sized mole hill. I’m off to enjoy some rapid eye movement, and fellas—stay the hell out of my dreams!