Hair request: Princess and Calamity Rose from Angel Corps (1 Viewer)

[snusnumrik]

Intrsting stuff to discover about Beaver.

Yet it always goes this way awesome niche artist often stop drawing due to pressures.
Even if they try to keep their professional and hentai artist idenitities separated people notices the style similiarities.

It is sad.


Tolive made Olivia hairs
http://www.sdtmods.com/index.php?topic=2206.0

But unfortuantely she is not one of the characters I enjoy.


@stuntcock

Awesome animations! Especialy the fact the rest position of the line on the neck is a cruve. Great job.
I'm totaly fine with the knotted grommets. Do as you see fit. You should slightly change the highlight to accentuate the more rapid curvature of the hat.
Also it is fun to ee the hat go between her butt checks.

I gave a look at some of the physics mods you are making on the loader section. There are tons of stuff!
It's great but one question rises: Are they resources heavy? The game is already slow on some computers when tons of cum goes around.

Anyway I made a new version of Princess hair.
I'm not satisifed, but I'm too tired to work on it anymore. If you want to work on it I will post the svg.

 

[stuntcock]

You should slightly change the highlight to accentuate the more rapid curvature of the hat.

If you mean "redraw the boundaries because the hat tends to hug her back instead of her shoulders" then yeah - that's on the to-do list. If you mean "the highlights should shift in real-time as the hat moves" then nope.avi.

It's great but one question rises: Are they resources heavy? The game is already slow on some computers when tons of cum goes around.

FPS degradation occurs almost exclusively because of graphics. When you decide to run a bukkake scene, the game needs to add lots of extra layers (each with partial alpha-blending). That stuff adds up. sby wrote a mod (named "strandlimit") to deal with that specific problem. Strand physics is actually more difficult than hair physics, because each strand must constantly check for collisions with many different elements in the scene: face, breasts, torso, etc... Hair doesn't collide with anything, which makes it "cheap" to include in a scene but means that it tends to clip through anatomy -- the physics mod gives artists some tools to avoid clipping.

The CPU cost of the actual hair-physics math is trivial. But there are two complicating factors:

• because it's a mod, the code integration and optimization isn't great. One commonly-invoked execution path uses three function calls when it should use only one. Therefore memory overhead and seek-costs are a bit higher than they ought to be. It's not really an issue unless you were trying to simulate thousands of hair strands simultaneously.
• hair segment rotation is implemented recursively. This is fine for normal strands (<20 segments). If you tried to use it to simulate a thousand-segment rope (real-time bondage physics???) then you'd see memory bloat.

I'm not satisifed, but I'm too tired to work on it anymore. If you want to work on it I will post the svg.

There's no rush.

I offered to work on the Bubblegum hairstyle for selfish reasons - it will make a good sample/demo for advanced hair physics.

• it has a lot of messy "flyaway" strands which can be independently animated
• it requires complex layering choices when allocating the static and dynamic elements
• it includes a hair ornament which should exhibit displacement (very little, but nonzero) during gameplay

The Princess hairstyle is much more well-behaved; it's a great candidate for the standard dynamic-hair techniques. Since it doesn't really require (or benefit from) physics modding, there's no need for me to get involved early. I can wait until you're happy with the visual details.

 

[stuntcock]

The sling bikini is technically feasible. I've created a very crude physics-animated version of the "front half" (shoulder to breasts to pelvis). It's possible to create the whole thing as a single elastic garment, but I didn't bother because:

• laziness
• the front half suffices as a proof-of-concept
• if I fill in the rear portion, then pride will force me to script in the interactions with the wrists (deflecting the rear strap inwards when her hands are in the "back" position and then bouncing back into a taut shape when her hands move away). That's a timesink that I don't need right now.

Screenshots

Spoiler

Design View. It's just a bunch of short segments. What matters is "where they're anchored on the body" and "how freely each segment is allowed to bend or stretch" and "how the physics system stitches the sections together."


In-game. This is only a single "strap" - the actual garment would consist of two (animating independently). For the sake of visibility, I've put everything on a high layer. The actual garment would pass under the shoulder and leg.


Scale Adjustment. The garment maintains its shape around the body, but the straps become thicker as their anchor-bodyparts (nipples, torso, buttocks) grow. In the actual garment we could accept this effect ("bigger breasts deserve wider straps") and blend it into the design (avoiding the messy overlapping seams that you see in this screenshot). Alternatively, we could simply cancel it out.


Adjustment Cancellation. The straps maintain their default width, regardless of body size. However, note that the upsized nipples are now poking through. We could probably fix this by fine-tuning the anchor point. Alternatively, we could deliberately widen the strap at the breast (as in the Princess design) and taper it towards the pelvis and shoulder.

The right-hand screenshot shows that the scale-adjustment is not perfect. We're "relying" on the elastic pull effect from prominent breasts; if we slide the breasts to zero we won't have enough outward force -- the strap kinda fits around the upper ribcage but it clips through the pelvis and floating ribs.

This would be a very low-priority thing to fix. Honestly - who would want to put a sling bikini on a flat-chested character? PEBKAC.


Flexibility. The garment is anchored to the nipples, so it automatically follows them around. Even if the breasts get significantly moved or reshaped by a mod. The straps will also fit in with breast replacer mods -- assuming that the mod author used the template correctly.

Animation Samples

Spoiler

Discussion
Breasts have two animation modes - natural and clothed. Clothed is more constrained - it reflects the supporting and stabilizing effects of a bra. I've used the natural setting here, because I doubt that this sling bikini would provide any real support. It's very easy to change this setting, though, so we could provide a "less bouncy" variant if anyone wants one. You can also adjust breast firmness directly via mods.

The challenge at this point is artwork. We need a vector-trace of a sling bikini.

The animation process is destructive - it involves chopping the original image into small sections. Hence, it isn't really possible (or useful) to finish the animation and then draw the artwork on top of it. It's quite difficult to do anything with a physics-animated garment (even retouching the highlights becomes tricky).

If someone is willing to trace the Princess bikini - including the highlights and shadows and the various jewelry baubles, and making a reasonable guess at the portions between the legs and behind the shoulderblades so that the whole thing is contiguous - then I'll assemble and animate it. If not, then I'll get to it eventually. But it might not be for several months; I've done a lot of coding recently and now I have a documentation backlog to deal with.

 

[snusnumrik]

It's awesome. I'm so glad you are working on it.

For big tits the titties bounce out of outfit from time to time but it is not a problem since with such massive melons you expect them to be hardly contained.

The nipple can be easily resolved by drawing a nipple on the outfit.

For the drawing:

I'm willing to make the vecto of the Princess outfit but right now I have exams.
I will try to post soemthing here in a two weeks time.



I also saw you are working on the leash for the girl. I find such a thing incredibly arousing.

Please continue the work on it :)

P.S. what happened to the cowboy hat?

 

[stuntcock]

I will try to post soemthing here in a two weeks time.

There's no rush; worry about your exams first.

The nipple can be easily resolved by drawing a nipple on the outfit.

I had the same idea, but there's a catch. Ideally, we want to inherit the actual shape of the underlying nipple rather than just drawing a predefined indentation into the bikini -- because some users employ custom nipple mods (and because the scaling of nipple size is not identical to the scaling of breast size). There are two viable approaches: one involving masks and extra layers, the other involving animated "shading" patterns.

I was planning to implement the bikini first (stretching it forward slightly to cover the nipples) and then provide samples of the two approaches to see which one is more appropriate for the costume (or "more believable during animation," or "more sexy," or "more CPU efficient," or whatever). We could then create version2 of the outfit which fits more tightly around the breast and displays an appropriate amount of nippleage.

P.S. what happened to the cowboy hat?

I found a much better technique for animating it. It's based on the body-conformal approach shown in the "sling bikini" example. Because the behaviour is determined by body proportions (rather than an arcane set of fine-tuned physics coefficients), it can be used with a wider variety of body sizes and Animatools positions.

If you've looked at the mod's page on the Loader Imports board, you probably saw a few of the screenshots in which the hat gets oriented bizarrely or simply clips through the body. We managed to get a nice animation sample, but it wasn't really release-worthy in that state.

Anyways, the new technique may be useful to other modders, and the hat is a good example because it has several different geometric constraints. I really ought to document the whole process, but I've been procrastinating because tutorializing a mod project is pretty time-consuming - I need to take lots of screenshots and keep careful notes. Then again... I haven't actually seen much interest from other modders yet, so maybe I'll just implement it and worry about documentation later.

I also saw you are working on the leash for the girl. Please continue the work on it

It's pretty much done now and will be released in a few days. I'm waiting for two things:

• AnimtoolsV14 may allow users to adjust the guy's hand position during scene setup (and maybe also to move it during gameplay)
  • I expect that the mods will be compatible, but there are a few possible coding tricks (or features, such as the "third hand" idea) which could render leashes non-functional. I'll need to run some tests and possibly release a quick update.
  • Seeing the Animtools stuff in action may give me some new ideas - or it may inspire other users to make suggestions for additional Leash features.
  • If AnimtoolsV14 supports hand movement during gameplay, then I'm definitely going to investigate the possibility of mouse-driven physics effects. You move his hand onto a physics-enabled item in the scene. You left-click (of Shift-click or press F11 or whatever). The nearest eligible segment becomes anchored to his hand. You move the hand away, and the physics system simulates the "drag" and "tension" effect. Alt-click to release; the item relaxes back into its normal position.
    • Thus, you could have interactive versions of leash-grabbing, hair-pulling, clothes-yanking, etc... Which would be more immersive and more fun than going into the UI and switching from "Leash (dangling)" to "Leash (held in right hand)." Breast-groping and nipple-tweaking would not be included - that's a different physics system.
• As mentioned in the Leash thread, I don't really intend to improve the visual quality. I'm not an artist, and this stuff was intended as a proof-of-concept for the physics.
  • However, I am willing to rig and animate any leash/collar stuff that the SDT community decides to vector-trace.

 

[snusnumrik]

Sorry for the wait.
I have not forgotten. It is the usual RL stuff bla bla that eats my time.

 

[elputocristo]

:( :'(