Jump to content

Requesting Critiques on 3D Models: Dev Journal


DrGravitas
 Share

Recommended Posts

I must say your application of fur has vastly improved since when you first tried it out. It's looking mighty impressive now, well done!

The only thing I can think of right now if you wish to take the fur even further is if it's possible to mix fur of different colours together at various places? Yes, the fur does look good right now, but the fact that there's only one colour for each area makes it also look ... off. I'm mostly comparing to other foxes and wolves where you can see a mix of two or so different colours in their faces. Or just different variations of the same colour. If you can somehow implement this mixed variation, without it looking even more awkward due to it, then I believe it'd bring an extra level of life into your characters.

Also a side note that the nose kinda looks out of place with the pretty fur around it. Maybe if the texture had a grainy normal map it'd fit in better. ... yeah, I know I'm nitpicking right now, but I can't not see it when I look at Uniformly Directed Fur 02.

  • Like 1
Link to comment
Share on other sites

Thanks you guys for the ideas and suggestions!

Fur work this week was small because I finished working on some pose sequences which took most of the time. It was an especially short week due to the Easter holiday. When I came back, it was the perfect time to install my new 1 TB drive. I had been running low on space for quite some time. Seeing as this stuff is the fastest growing content on my drive, I decided to move the whole 60GB worth of Furry stuff over to the new F: drive! That meant I had to re-point many of the references of things like texture locations and such within the project. Since I already had to do that, I figured it was a good time to rearrange some of the stuff and flatten my folder hierarchy a bit. I'm glad it's finally done and boy is it nice to have all this new space and short paths!

The poses produced were as follow:

5533e16c0e5963d5c383dbea087ddff8a9d87ad9         5c71115a9536aa53cd98c318d141da75e301b03f

Nothing much to comment on about them, but I will point out that I've now standardized the alternate lip weight paints that enable me to do the lip-bite look Blythe had in the second one. I can switch between that and the normal, full deforming lower lip weights quickly and easily with a few scripts. Let me know what you think about anything with them!

 

Now, with these new posts, I decided I want to create a sort of work list to better organize what ideas I currently have, direction, and experiments. Hopefully this will serve me as an easy way to recall what I should work on!

Mesh

  • With the new scripts I devised, shape changes are a lot easier to integrate into the current design without having to recreate UVs and such. While this may assist topology changes too, those will remain somewhat more difficult. Thus, this section is devoted to changes regarding Rederick, Gravitas, or Blythe Mesh changes.
    • Shape:
      • As per Recel, Blythe's arms a bit wide in the shoulders and are lacking in muscle definition shape a bit. I'm not sure that I can do all of this without topological changes, but it's an excellent chance to see how far I can push this model for change.
      • Rederick's dick has been noted as connecting to high on his pelvis. Lowering it isn't much of a challenge with he new scripts
    • Topology:
      • In addition to the position of Rederick's penis, I have found the pinching brought on by how I connected the penis to be a bit problematic when sitting. It creates very strong creases and stretches polygons a bit too much. More over, I am considering alterations to the perineum as well to better facilitate resizing of it and more independence from the joint structure of the penis/testicles to give them more leeway in moving around. Finally, I need to consider further alterations to those joint structures; there's just something not quite "right" about how his genitals look when bent over or on certain other positions. Part of this may also be topology or even just the height adjustment required.
      • At some point I want to start working on developing Blythe's genitalia as well. Her's will be challenging in very different ways that Rederick's, as they must squash and stretch to conform to other objects as well as be flexible aesthetically to support the wide range of detailed appearances available when spread or together, etc. Lots of research is needed.
    • Shaders and Misc:
      • I like BerryBubbleBlast's idea of doing something for the nose. In the past, I have tried some textures and such for the nose, but was never really happy with it even though I did end up settling with a similar setup on the paw pads to give them some grain.
      • I am thinking of adding some plain black flat features drawn on the body, similar to what you might see in a cartoony look. There already exists such things around the eyes. It would help add some detail when they don't have fur, I think. It could also be useful for implying some sort of skin tight spacesuit.
      • I am considering ways to include an anus on the models. I would prefer to avoid topological changes and I'm thinking I might be able to pull it off with a texture-mapped layers in the shader network and perhaps some bump mapping. I would like to try out displacement technologies as well for this.
      • I'm thinking of experimenting with adding whiskers to their faces again. Not sure how I want to go about it, though. Might even have some kind of XGen thing.
      • Accessories (Collars, new swimwear, etc.) might add some variety, but more of that later.

Fur

  • Fur development continues! There's a lot of ground to cover here still. Once I have a strong package prepared and finalized with Blythe, I will move on to Rederick and Gravitas as much of the expressions and other work will be transferable. This will save me from having to keep going back and re implementing changes in the other two models every time I make a major improvement.
  • Color:
    • I have experimented with a ramp to add some variation to the fur color, but it hasn't been as useful as I'd like. I think BerryBubbleBlast's suggestion of giving individual hairs varied colors (as in the images) might be a better approach. Accomplishing this will no doubt require another set of expressions in XGen, but I also need a better way to control color choices that is more portable.
    • Now that I've gotten a better handle on keeping XGen previews under control, I'm going to experiment with complicated fur shaders again. I noted that I could use an expression map to specify weights in a MILA network, but even the simplest network slowed the entire program to a crawl. Perhaps now that won't happen. If so, then there will be no limit to how I color the fur! Anything I could apply to the model surface, I'd be able to apply to the fur itself. If that works, the be-speckled metallic shaders would provide an easy way to sprinkle in different colored hairs.
  • Maps
    • I really need to take some time to improve the length map, it's far too starkly contrasted in its height differences. I need a good way to paint gradients in 3D. Either I need to come up with a photoshopping process that eases it, I need to get better at painting, or I have some ideas for some really complicated (way more complicated than any I have now) scripts that could assist me in creating them automatically...

Scenes or Poses

  •  I have a story I want to tell. I'll need to develop entire settings completes with props and other assets to tell it, but I also have ideas for scenes I want to depict. General poses would also be listed here, but those tend to be more spur of the moment. Most of these will be very short descriptions, more as reminders to myself that I had them. The details will have to remain in my mind.
  • General Notes:
    • The story idea provides events out of order from which they actually happened. The actual timeline of events can be divided into an early period and a late period, separated by a central event with both periods themselves divided up by events, not all of which I have decided on.
    • Setting-wise, it is Sci-Fi and I have ideas on the certain flavor of hardness it exists as. It is not a world-hopping space opera but is restrained primarily to two settings, outside a few special instances. One is a very advanced space station and the other is a very limited spacecraft (both of which I mentioned a bit in the dancing piece from a few weeks ago). I won't get into all the specifics of these settings (or the special ones) here and now.
  • Poses:
  • Scenes:
    • Desync:
      • I have an idea for Rederick and Blythe to illustrate the disconnect between them in their lives, involving one and the other getting up for work while the other is sleeping or coming back and vice versa. Alternative ideas include images of one at work while the other is sleeping. The general idea is that there will be 4 locations and two images. One image will be Blythe sleeping and Rederick getting up/working and the other will be Rederick sleeping and Blythe getting up/working. A degree of symmetry is to be present in their general mode but the images should be differentiated enough in the details to not feel like a cheap cop-out.
    • Relating Gravitas:
      • I need a scene that better presents Gravitas' relation with Rederick and Blythe in two parts. From the early point in the timeline and the latter half. I have a few ideas. Chief among them depicts his multiple bodies in a sort of converted pool (like an Olympic swimming pool) with simple industrial machinery raising him out of it as Rederick and Blythe watch on. Dr. Gravitas in the early period is likely to be best characterized by being presented with numerous props (largely industrial and scientific) to emphasize vastness, complexity, and isolation. This is problematic as it will, obviously, require a lot of accessories and props as well as the setting and complex 3D technologies to implement many of the more important ideas. For example, using XGen's ability to mass replicate models just as it does hairs (like creating a forest from a set of template trees). Gravitas in the late period should be much more restrained in these presentations and I could probably start characterizing his relation there, even without backgrounds and settings.
    • Sexy tiem:
      • I kinda wanna do more sexy stuff with Blythe and Rederick. :3 It's not meant to feature heavily in the story, but it's fun to make!

Rigging and Scripts

  • This is one of the weakest areas for my models. Everything right now is FK only; no Inverse-kinematic limbs or anything. It is becoming desperately unmanageable as refining and perfecting poses becomes more time consuming. I need to get smarter with this stuff. The true saving grace of my current work are a handful of ancient scripts for writing out scripts that replicate joint positions/scaling/etc. But, these are cumbersome to work with and things are getting more and more complicated.
  • These will be rather huge undertakings but they're going to have to happen at some point.
  • Refurbish saved posed scripts:
    • I really need to implement a proper GUI interface for these scripts and abstract out the values from the operations and objects. For now, the only reason things on Rederick work on Blythe is because they have the same names for joints and stuff. This is a problem because when they're in a scene together, I can't use any of these scripts. I've gotten around this so far by developing them separately then bringing them together and refining by hand. This is slow.
    • By breaking out values, I could develop a system that compensates for name changes or duplicates and allows me to use any pose value set on any valid arrangement of joints or other components.
    • Managing this more complex system by hand-altering values will quickly become too complex. A graphical interface will help me to quickly and easily configure components and alter the way I save things to improve applicability across a range of scenarios.
      • For a simple example, I have Rederick with and without a dick. As of now, I need to remove dick joint rotation commands from scripts by hand if I want to run them for Rederick-without-a-dick, or Blythe for that matter. For the time being, I've simply been proactively compartmentalizing my script productions. There are scripts for parts or all of a pose and a separate script that handles the penis in various ways.
  • Annotation System
    • This idea ties into the one above and is especially important for facilitating the graphical interface. For these scripts to be able to do the more complex processing I require to speed up development, they need to know more about the nature of what they're handling. A well defined set of custom attributes serving as annotations on joints is one approach.
      • For example, using the penis problem from above, let's say I create a custom attribute called "accessory joint chain" or something. We apply this attribute on the root of the penis joint chain (where it connects up with the rests of the skeleton). Then, in the pose saving and or loading script, we set it up a routine that looks for joints labelled as accessories and populates a check box for them (complete with a name, possibly also from an annotation). This checkbox informs the rest of the script whether or not I want to include these joint chains in what I'm saving! Instead of trying to run the loaded script and hunting down why it didn't work, I'll have a list of all the things that might not be available on a model (for example, Blythe wouldn't have a penis joint chain or a partially completed test rig might not have a tongue) and choose not to try to update those when it runs, saving me time.

 

Other Assets (Props, Settings)

  • I really want to get these characters out of the void! I think a lot of why I feel so limited in coming up with ideas for them and poses is because they only have each others. The recent experiment with the creation of a sex toy led to several new and interesting ideas for pics. Even simple props seem poised to open up worlds of new ideas. But, I want more than just props. I have a story I want to tell. I need to develop entire settings completes with components. Given how slow I am to develop things, I am looking into "refurbishing" free assets from the internet; taking existing public domain objects and cleaning up or improving their shapes and stylizing their textures/shaders and general appearance to fit the sort of aesthetic that has developed around my work. It wouldn't do at all to have a big divergence in style between elements of the image! It all has to fit together as if they were part of the same world.
  • Settings:
    • The spacecraft already has a handful of miscellaneous assets. Some of the settings ideas I need for this space will require corridors, dense with infrastructure, a special (and simple) space involving a large sphere, and a specially designed unique living quarters space for Rederick and Blythe. I will probably be able to get the most mileage out of the living quarters one as Rederick and Blythe are the primary focus of the story. Better still, it's likely to be one of the simpler to build areas of the ship, save for the sphere-space.
      • The Living Quarters
        • This setting is differentiated from the rest of the craft by its amenities: it is a wheel-like design and the only place on the craft with any sort of artificial gravity, it is not covered in electronics and duct work, nor is it super tightly enclosed. It has comforts like bedding and paneling as well as views to the outside though special 3D videoscreens that for my purposes will work like direct windows. Showering/restroom facility are needed as well as other varied spaces. Not the simplest setting, but also not as cluttered with miscellaneous accessories and props like much of the craft.
      • The Sphere-space
        • A simple refurbished fuel tank interior with a bubble-sphere within it. By far the simplest setting, but I only have ideas for one or two scenes to be set in it. Getting the shaders right on the bubble will be an interesting challenge given what the bubble is to contain...
        • This might make a good first target for settings at some point, due to it's simplicity. It won't be able to provide too much to future settings, however.
      • Misc Settings
        • A viewing room largely for Gravitas, various corridor-spaces and bare-fuselage-like areas, and a couple special spaces involving biotech implements, kitchen/eating-space (ad hoc), and maybe some ship-exterior stuff. Much of this will be implemented to an array of independent accessories and props, backed by larger infrastructure pieces like trusses and such. My biggest inspiration with much of these spaces is the International Space Station (like these images), MIR, Space Lab, etc.
    • The space station of the early period of the story is a much more traditional sci-fi setting. Luxurious, big spaces filled with amenities and comforts (unexplained perfect artificial gravity for example) like a hotel in space in places and without all the exposed gadgetry and cruft of the spaceship.The nature of the station is an expression of the species that built it as well as fulfilling a purpose in research. It's to be form and function. My biggest inspiration here is to be Fermilab, for its combination of beautiful architecture with highly technical spaces, CERN, and other particle accelerator labs. Not every space is unique, however, it should also emphasize repeated applications of patterns and reuse of space to convey the limits of availability.
      • Blythe and Rederick's Apartment
        • Fitting to the station's all-in-one feel, employees are furnished with living spaces of varying quality. Befitting of their position, Rederick and Blythe have a particualrly impressive space taking a form not unlike a simple two-floor split-level apartment. I have a very specific idea of how this space is designed, much more so that most any other setting. The space is very open, with the second floor a sort of balcony-like space over half of the first, it has enough room for a modest bedroom, bathroom, open staircase to the first floor. It hang over the kitchen, and a utility space, while the open area composing the remaining half of the first floor consists of a living/eating area. The biggest challenge here will likely be measuring things out properly. It should probably be an early target.
      • Working spaces
        • Not many ideas here yet; generic physics labs, meeting rooms, desk space, and probably won't feature too heavily.
      • Ballroom
        • Yes, a ballroom. In space. It's a political thing. There are other oddities (like an Olympic pools in a recreation area) but I've already committed to the idea with the dancing piece from earlier. Probably should largely be built out of a general purpose auditorium-like space but, you know, for outer space.
      • Gravitas Areas
        • Some places to set him aside as an anomaly, making his spaces on the station similar to other areas but clearly oriented around him (industrialized, more misc props and exposed infrastructure). The best example would an idea I had mention above in Scenes. Elsewhere there is a recreation area with an Olympic-sized pool. In the Gravitas areas, the space is repeated again, but utilized not for recreation by reconfigured and utilized as a sort of spawning pool for growing and storing his bodies, among other things. Basically, there should be a hint of recognizable space with a clearly altered purpose. Few or now amenities and comforts or unnecessary things, possibly even hostile to navigate simply. Many of these assets should be re-useable to some degree with the Misc areas of the spaceship.
      • Technical spaces
        • Very few scene ideas here, but these are areas styled more like the particle accelerators themselves from the inspirations mentioned above.

 

Well, that got a lot longer than I expected! I have a number of other ideas floating around in my head, but for now I'll leave it to these. Comments on anything, or other ideas and suggestions, would be most welcome!

 

 

Link to comment
Share on other sites

  • 2 weeks later...

FUR! Yes, more fur. As suggested by BerryBubbleBlast, I have  started working on mixing in alternate hair colors into the fur. It took a bit of work to figure out, but the actual formula for doing it is a very simple case-statement with a random number generator whose seed value is the id of the primitive (hair) being evaluated. This ensure that the hairs that get a different color remain that color consistently regardless of the render, frame, motion, or stuff like that.

274fb80f6d011ec98e2e0bd217b70917a727e5ef

Still a lot of work left to make it look good.

I need to look further into real fur to get a better idea of color choices to make. I also want to make improvements to the way it chooses to change the hair color, so that it increases the chances of there being different colors closer to the boundaries of the fur markings. This will probably require adjustments to the existing fur marks, but it should make for a better transition between colors.

It also occurred to me that I never showed off the new Dr.Gravitas model! As I mentioned before, I have new scripts for copying skin-weights and I've devised a way to copy them across shape changes. This makes rigging models quite a bit faster when very little changes beyond shape and even greatly aided me when I added that penis to Rederick (which is a topology change, rather than just a shape change.) In comparison to that, Gravitas' changes are extremely mild, but I might as well show them off!

                    5445119fb707dc2dd8d7b79aea607e44bc5e7a7d             17b1cf1cb4729daf64fade44c8b047b741b39d04

Yes, that image is from February. I've had this lying around for a while now.

4a45c242256ac5592aa71f24a58d428e3529cd0b

A scrap comparing the silhouette of Rederick with Dr. Gravitas'

He even has his own custom fur markings, although I am not sure I like some of them. Probably will fiddle around with him more later, but for now I'm fairly happy.

Link to comment
Share on other sites

Fur was the focus once more. After considerable effort, I managed to create an alternative XGen expression for coloring the fur. It's a much more sensible concept, although the implementation needs some refinement. Their are several key differences:

  • Expression no longer outputs colors
    • Instead of outputting one of 3 colors chosen by hand, it chooses and outputs a number 1-3.
    • This is piped into condition utilities, which output 1 or 0 to their assigned MILA layer's weights.
    • This effectively lets me use shaders to color the hairs, rather than a single color choice, which has the potential to greatly improve the looks (and my ability to control to look) of the individual hairs. For example, I can alter the color of the hair when seen from certain angles or control specular highlights more precisely or maybe make them metallic!
  • New approach to the use of the fur-maps.
    • They are treated as the basis for calculating a color-selecting weight.
    • This will allow (once I revise the maps) for more realistic, gradual shifts in the density of fur colors at the boundaries of markings, rather than the sharp clean-cut lines of the current maps.
    • This is because each hair is a definitive color choice rather than a mixing based on the maps which was essentially replicating what the shader network was doing.
  • Integrates random variation in hair coloring.
    • Tieing in with the revised use of fur-maps, this new expression chooses hair colors at random weighted towards certain colors with a specified minimum chance for any given color.
    • This means that (for example) in the area of body fur color (blue in the case of blythe) there is a small (currently set to be 1 in 120) chance that the hair will instead be colored either lightmarks (white) or darkmarks (black). This results in a random speckling in of the other colors.
      • Currently there is no way to entirely prevent one or both of the colors from occurring in an area, but that could be accomplished with a additional map.
    • This is what enables this new expression to use the maps to produce gradual shifting in the density of specific hairs on the edges of the markings.
      • If you think of these greyscale maps as expressing a value from 0 to 1 (black to white) for the areas the light or dark marks should appear, then we can use that to calculate a percent of a range (0 to 120) that represents the chance of that particular hair color being chosen.
        • I have implemented this in a way that I derive the body as the opposite of the maximum of the whitemarks map value and the darkmarks map value.
        • The percents for those two maps are balanced against one another (as both maps can be 0 to 1 at any point and I wouldn't be able to coordinate the two maps to balance out to one even if I wanted to.) Meaning that if they are both one, then the chance that a given hair is light or dark is 50/50.
      • With this balancing of the maps, I can shift the densities of fur by using gradiented greyshifts in the maps to influence the chances of a given color appearing in an area.

 

There are still issues with the shadows and lighting of the fur that I haven't managed to work out. I haven't worked any further on the maps for the length, or revised the markings' to demonstrate the new color just yet. However, I have applied shaders to color the hairs now and all three markings are speckled in. So at least there's that.

0421 AltFur03_Close Post.png

No revised furmaps means no demo of the gradual shift just yet. The photons traced need some refinement; they are likely responsible for the blobby blotches in some of the areas. Photon tracing was surprisingly fast and the new approach didn't add much to the XGen overhead.

 

Next time, I may take some time for some anatomy work on Blythe. I've spent a bit of time working out an actual background scene and some props, but progress will be slow on that front and it will be some time before appreciable results are had.

Link to comment
Share on other sites

Normally I can give fair critique but when it comes to 3D that really isn't my domain. The only thing I can offer is my first-glance and opinion which, to me, everything just seems incredibly uncanny. Something about the eyes and structure of the face just feels off and unsettling but I can't exactly pinpoint why. Everything feels like it's trying so hard to mimic the look and feel of a human when they're the furthest thing from it.

  • Like 1
Link to comment
Share on other sites

1 hour ago, Photoshop Amateur said:

Normally I can give fair critique but when it comes to 3D that really isn't my domain. The only thing I can offer is my first-glance and opinion which, to me, everything just seems incredibly uncanny. Something about the eyes and structure of the face just feels off and unsettling but I can't exactly pinpoint why. Everything feels like it's trying so hard to mimic the look and feel of a human when they're the furthest thing from it.

Thanks for the feedback (and welcome to the forums!) You're certainly not the first one to mention the uncanny feel of the face, but so far I've not come up with any solutions to deal with it. I admit that I used measurements to aid in building human proportions; I simply lack the talent to pull off a stylized look.

Though, I am curious: Which of the 3 do you think is most affected by the issue:

                             Rederick                                                               Gravitas                                                                               Blythe

0319_WideEyedSadSmileCute post.png  0306 GravitasSmile Post.png0412 LeanSitSadConcern03_CloseNoFur WiP.png

 

And do you think that fur (even though it's still a work in progress) makes any difference or improvement in reducing the effect?

0409_CurlUp.png0414 LeanSitSad_Close_FurPhotons Post.png

 

 

Link to comment
Share on other sites

1 hour ago, DrGravitas said:

And do you think that fur (even though it's still a work in progress) makes any difference or improvement in reducing the effect?

I think the fur helps a lot with the overall look, because it makes them look less shiny. Keep practicing that fuzziness

  • Like 1
Link to comment
Share on other sites

8 hours ago, DrGravitas said:

Thanks for the feedback (and welcome to the forums!) You're certainly not the first one to mention the uncanny feel of the face, but so far I've not come up with any solutions to deal with it. I admit that I used measurements to aid in building human proportions; I simply lack the talent to pull off a stylized look.

Though, I am curious: Which of the 3 do you think is most affected by the issue:

                             Rederick                                                               Gravitas                                                                               Blythe

0319_WideEyedSadSmileCute post.png  0306 GravitasSmile Post.png0412 LeanSitSadConcern03_CloseNoFur WiP.png

 

And do you think that fur (even though it's still a work in progress) makes any difference or improvement in reducing the effect?

0409_CurlUp.png0414 LeanSitSad_Close_FurPhotons Post.png

 

 

It's a very close choice between Rederict and Blythe. Gravitas with the shading that's hiding some of his features makes it seem a bit more cartoonish which isn't as uncanny as the others. On top of the his eyes are actually okay. Seeing a lack of white in the eyes (like a lot of dogs) seems to really help remove that uncanny feeling

 

Fur wise yes, it does improve it a little bit because it looks more like his facial features are less proportionate than that of a human. Hell I think that's one of the main reasons that it feels so uncanny now. Your proportions are too close to human while trying to add on features that would (in my opinion) be best if they were exaggerated a little more.

 

Lastly thank you.

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...

Well, I didn't get to fur. I had previously started on certain additional features which then proceeded to take much longer than I would have liked. But I'm getting back to fur now.

The feature implemented was genitalia for Blythe. Seems only fair considering Rederick got his already! :P The vulva proved a fair bit more complicated to develop and rig than the penis/testicles, although the initial topology construction and integration into the model were much easier. This was largely due to my decision to ensure that it not impact topology beyond a certain area, strictly retaining those boundaries. It also worked out that I was able to include an anus in mesh with rigging.

Fully posed/shaded images will not be presented in this thread; I've left them to that other (black hole Red Lantern) thread (and, you know, my site galleries.) Instead only the scraps are here.

 

15_0929_MatureContent.png

(Scrap contains NSFW vulva/vagina and attempted erotic poses)

That scrap is actually a bit out of date. Later I would refine the skin weighting to remove some of the sharp deformation around the edges and re-position the anus further back. That said, these features could have been better done. After implementation, I found tutorials that would suggest my placement of the structures is a not quite right. More over, deformation can still be problematic and the whole thing is fairly under-detailed. Further detail could be implemented through textures, but I think I'll probably have to revisit the whole thing at some point (much as I am finding I will likely have to do for Rederick's junk as well.)

15_0929_AdultsOnly.png

(Scrap contains NSFW Male/Female sex with genitalia focus)

I discovered problems with the proportioning of Rederick's dick, relative to Blythe. His length means he penetrates deeply and his girth requires accommodations through the wire deformers, stretching the model to the point of creating pinching. However, the rigging does accommodate this while looking at least passable. The length is a problem only if I want it to be (not like that's actually visible, plus I can scale the shaft if I really wanted to.) The girth can be handled by scaling too, but I'm beginning to believe it's a problem more because of Blythe's proportions. I think she might need to have a wider waist (or at least the pubic triangle might need to be wider.

The rigging is also spectacularly complex, in comparison to Rederick's equipment, consisting of  5 joints and 6 wire deformers. Still not controls have been built so these remain operated directly. The skin weights won't mirror properly because the joint structures for controlling the open/close portion of deformation relies on two joints that share the same position in the center of the model. This means the structure has to be repainted every time I mirror the weights. There was also significant trouble cajoling my weight painting scripts into applying the normal weighting to the rest of the model, partly due to the decreased shoulder width of Blythe and some mix-ups on that front.

0421_ShoulderAlt_New ThumbnailCensor.png

Oh! I forgot to mention, immediately after my last update (on April 21st) I decreased Blythe's Shoulder width. It was very easy thanks to my scripts. Barely took an hour to get it back to fully rigged and shaded.

 

As far as fur is concerned, I plan to work on length and the smooth transition between fur color patches I previously described.

Link to comment
Share on other sites

  • 2 weeks later...

OK! I got back into fur this time!

Progress was kind of slow at first, so there wasn't really anything to show off last week. Things started to pick back up, but still the changes aren't super noticeable.

0524 Fur_KickUpPose ThumbnailCensor.png

Most importantly, I've dived into XGen's modifiers setup. I happened upon some documentation that suggested its possible to do clumping without guides (which I had previously dismissed as being dependent on guides, while my model is focused on expressions). After some work at it, I eventually figured out how to get clumping working without throwing a cryptic error about there not being any guides. Turns out the guides it was referring to can be more than just the guides the describe for controlling the splines; it can also be points which are generated specifically for clumping. Initially, I tried clumping on both the tail and the cheeks, but eventually dropped it from the tail after feedback repeatedly indicated how it looked too "wet".

0529_9Square TailClumpingTests Post.png

Various fur clumping tests on the tail. Click to enlarge.

Being wet was mentioned about the cheeks as well, but I think it still looks better than sticking straight out from the cheeks like it had been. I could always just say the character styled it with some gel :D

6280da92bc8d8a246df724a2c9589099ff7cbb8a

The cheek clumps also better highlight the new gradual change over in hair colors a bit better than earlier examples. This is partly because I also revised the maps used to calculate the hair color. I cleaned up most of them by had to eliminate the seams and mistakes. There's still some seams on the hands, paws, and most noticeably on the inner thigh/leg. I went by hand and smoothed out some of the transitions as well.

Color is still something of a question for me. I can't decide if I want to use a simple color shading or the PhongE colors I have been for fur. The PhongE takes longer, but I like the colors it does. I suppose I might be able to get those colors with the simple one, but I suspect it will be more uniformly applied. I fixed the weird shadows but there are still some ugly shadows caused by the extremely limited amount of lighting I'm using.

0524 Fur_KickUpPose_Face Compare.png

There are still several outstanding issues with the fur:

There are a few unexpected bald spots, where the hair bends so far it clips into the model, that I haven't managed to work out. I revamped many of the expressions to give better control over the bending and more variability, which looks good for the most part but also introduced those bald spots somehow. You can see an example on the face near the corner of the mouth.

The length expression has been altered to allow another map to make certain areas extra-long (I'm using it to add extra floof on the cheeks and tail.) But I still need to spend time refining the baseline length map and improve the transition in some areas. Specifically, I need to fix the fur hiding the lips. I might also need to use it to define a density map, so that areas with short hairs get more hairs and don't have those little sparse dots like you see around the eyebrows.

I need to add eyebrows on the black color map.

 

But the biggest problem of all is the render time! The faces above took 1 hour 45 minutes and over 2 hours each, with the body shot taking around 45 minutes for low anti-aliasing quality (which looks pretty poor) and around 2 hours for the aa settings I like (still a bit weak, though). It's clear to me that the expressions are a huge part of the time. Even with only 100 density (as opposed to 6000) it take about 7 minutes to render. The color choice expression seems to be the biggest culprit, but clumping isn't helping. It would be nice if I could find a way to bake these expression results out to a map and get the best of both worlds, but I've not seen any way to do that yet.

Link to comment
Share on other sites

Not as much work on fur this week, because I've finally started working on an environment to make scenes out with my characters!

The first one will be a pool area, with adjoining locker/shower rooms and other facilities. I plan to develop it in stages, beginning with a basic layout of all areas to get a feel for proportions and placement. From there, areas will be fleshed out further as needed. As such, currently everything is grey (no point in shading things until you've got most of the layout).

f884f994aa22dea0b088dd263d1f93ca494cc303

Check out this fantastic floor grating I've been working on!

While the pool setting's development was the major focus this week, I also decided to test out how easily I can setup fur in new scenes. Up until now all (except one other similar test) images of fur have been rendered through the same lineage of files from the same test setup.

0601 FurPoolTestError ThumbnailCensored.png

Generally speaking, this experiment went well. There were a few hiccups and mistakes but it took maybe an hour or so to iron things out all told. The one stickler was the shadowing. For some reason it is drastically worse in this test scene, even after double checking the fix I found before was still in place. It might be the lighting, but I feel like there is something else at work here. Not sure what.

As for the pool area itself, I have only a few sections blocked out so far:

0531_PoolAreaOverviewB.gif

Yeah, I reused the floor grates for some of the walls :P

The size of the pool itself is Olympic and calculated out by Rederick's height assuming him to be about 6ft 6 inches (or ~2 meters for convenience).

OlympicPool.gif

I think the key to developing believable and good quality spaces is going to be proportions and measurements. Digitigrade legs make for a slighty skewed height, but the biggest factor in getting them to look right early on turned out to be in keeping balanced proportions based on the head. Similarly, I've chosen a handful of anthropometrics (the body, the arm, and the paws, which simple box stand-ins) to serve as guides in building my spaces and the props that will inhabit them.

For example, the floor grate I've been going on about is about the depth of the paw (but not the height) brought out to the nearest Maya unit. I round it because that makes it easier to snap things together and calculate how many duplicates to generate.

Yes I Said Floor Gutter And Grating.pngHuzza I Dun Gud At Maths Fur Once.png

I may also eventually develop a simple texture for farther shots that include the grating, but for now I can do more interesting things with them as actual polygons.

The only other thing I've developed so far (not including a really dumb door frame) is a simple springboard/diving board. While examining the proportions for the board, I even had some fun pose ideas! I mocked up a couple of them during development, but didn't take them forward into actual renders.

0528 Springboard02 WIP.png

0528 Springboard03 WIP.png

I am particularly pleased with the bump-mapped rubber bits on the bottom and sides, but overall my shading could use work.

I am looking into commissioning some pool sign pictograms to use in textures around the pool. If it works out, I might end up using this idea to build more of my texture work. I'm rather terrible at that stuff myself, honestly, but it would really help make the area better. Aside from that, I plan to work out more props and finish blocking out the basic proportions and layout.

Link to comment
Share on other sites

  • 3 weeks later...

Seems like ages since I last updated this thread! It has very nearly been a month. Work has sadly kept me quite busy. So busy in fact that I over looked my silly little anniversary of starting on these models.What a pity, but as it turns out I was still working on something fairly comprehensive anyways. Might not be quite as big as last year, but I took a detour from the pool to the shower and combined fur and fluid physics:

0622 ShowerSinging ThumbnailCensor.png

 

Not the best results, but I'm happy that I at least explored the concept and got in some additional practice in asset creation. As you can see, it features fur clumping again but this time expanded to the full body to give an impression of wet fur. The fur is still plagued by minor issues including color seams and bald-spots. I have a theory on the cause of bald-spots: they seem to be from surfaces that trend towards concave. Not sure how to overcome that still.

 

BiFrost's fluids once again proved simple to set up but exceedingly difficult to take on the appearance I want. It is especially difficult for liquids that are emitted in particular fashions. Pouring out of a glass or faucet are simple, cumshots (and apparently shower head nozzles) are not.

0618 ShowerHeadTest01WIP.png

The majority of the lighting here is indirect, but this was causing extremely bright blooming light on the model and water. I created a much more basic shader for the water to work around some of the difficulties, but mainly I used light linking to solve this. I removed those elements from the normal scene lights and instead lit them with their own set of specially tuned lights.

0608 Gravitas WF_F2SFront ThumbnailCensored.png

Early lighting problems

The shower stall itself had a number of specially created elements, all of which I used MILA shaders to color. None of these have specifically crafted textures, but many of them have generative patterns from Maya. The gutters and floor tiles in particular owe a lot to those procedurally generated textures. The marble was a particularly complicated shader that took quite a while to get to where I was satisfied with it. I experimented a lot with scatter layers, but ultimately didn't include those.

0618 ShowerWater WIP.png

Another view of the shower space, with water effects. The "Glass" wall is actually translucent and show character shadows cast on it, although the effect is not as good as I'd like and I may be replacing it.

0609 Wall Mounted ShowerCaddie.png0611 control panel_Buttons.png0612 AirVentB.png0613 DotLight01.png0613 DotLight01_Mid.pngShowerWandTestRigB.gif

Various props developed for the shower environment.

Among the various props was a shower wand, which was particularly complex. Its hose is controlled by a large number of joints which are in turn controlled via an IK Spline rig. This makes it my first successful use of IK on this project! It's not exactly super easy to use, but its still a lot simpler to control than FK on all those joints. I had previously tried to implement such a system for the character's tongues, but found it unsatisfactory and unwieldy.

 

Unfortunately, I don't see July being much better in terms of development. Aside from a few other minor ideas, I will continue to develop the pool space, I think. But, I have a 2 1/2 week vacation over that time that I will be completely away from my computer. At this rate, I half-wonder if I'll have this pool area built before winter rolls around. -_-;

Link to comment
Share on other sites

Lots of little details and experiments this week. Most important among them was gathering some basic experience with Maya's Scene Assembly.

9665ee1d44d7ad2c78e5fef776472bbdb37f2bdb

With Scene Assembly, I can craft assets (in this case the public shower stall I developed),set them up in separate file,create a GPU Cache of them, and then produce a representation. This representation allows me to set either bare locator or the GPU cache to display in Maya through the viewport, rather than the asset itself (which when combined with many other assets as like the 27 showers here, can slow Maya to a crawl or even crash it.) With the right settings, Maya will swap out the cache for the final, fully shaded and textured, version when passing scene information on to the renderer. This keeps Maya's preview snappy and workable while allowing me to layout and alight all the scene elements! This will be very important to manage as scenes get more complex and I further develop areas like the public showers near the pool or even the pool itself.

With Scene Assembly, I can assign these varying quality representations and programmatically choose which will be used in the render! So for example, I could have the highly detailed and reflective shaders on the main focus shower that a character is in, while cheaper to render, lower detail and non-reflective shower stall are used in the background, saving rendering time.

 

0628_PoolLadderClimb WIP.png

Other developments include progress on fine assets like this pool ladder. I am still doing measurement work on it. It, along with others, will be necessary elements of the final pool environment. Not the most exciting developments, though.

0630 ShortyCutJustShort.png

A bit more exciting of a development would this newcomer here!

Because of the new Red Lantern sub-forum, I've decided to discuss that one in my Gravitas After Dark thread instead :3 As you may have noticed, I've gone back and censored my models. It is with a heavy heart that I adhere to the new old nsfw rules and go back to censoring Blythe's breasts and Rederick's dick. I may still link the NSFW off-site stuff here, but no more uncensored artistic nudity.

 

Progress next week is likely to be limited, if I update at all, due to preparing for my forthcoming vacation. But, I will probably mostly focus on the same stuff. As always, critique and suggestions are welcome!

Link to comment
Share on other sites

As expected, very little progress this week as I prepare for my vacation trip. July is going to be a low-productivity month.

0707 4Square_EvenMorePublicShowerTests Post.png

Click through to see the full size!

I squeezed in a few more Scene Assembly test renders and some optimizations. I now have two references for the shower stalls: a default one that's just a clone of the original, and one using lower quality shaders for more distant stalls. I also have those two types for some empty stall spaces that I put in to make the public shower space feel a bit more spaced out. The top image is roughly the same number of stalls as before (arranged differently) with the new varying stalls and empties. It rendered in about half the time and with more render features (Global Illumination and Caustics) enabled.

The bottom two images are of what will probably be the final layout of the stalls themselves. Really, it was mostly just a test of whether I could pass that many references through to the renderer. It worked out well!

0707 PoolLadderMockup02 Post.png0707 PoolLadderMockup04 Post.png

Aside from those miscellaneous tests, I also did some development on a single asset: a pool ladder, seen here with a speedo-clad Rederick. The pose is still a WIP, but you get the idea.

Link to comment
Share on other sites

  • 4 weeks later...

Hoo-boy, it has proven kinda difficult to get back into the swing of things, post vacation!

15_0929_MatureContent.png

NSFW Contains Penis: 4-Square examination of the experimental new shader look

0801 RederickSilkShader_Face_4m_47s.png0801 RederickSilkShader_Paws_5m20s.png

Standalone views of the Face and Paws so those who don't want to see dick don't have to. (Click for full size)

The majority of my effort was on a new shader, but ultimately it was found to be not quite up to par with the current metallic ones.

The key improvements:

  • Speed. This thing is super fast!
    • The image in the upper left corner took the metallic shader 5 minutes 28 second
    • The new one took a mere 1 minute and 3 seconds.
    • The speed up becomes much more apparent the closer you get to the figure.
  • Doesn't use the AO shaders (a first in the networked shaders)
    • Nor does it use anything Maya or old MR shaders, like PhongE, Ward, or Metallic Paint.
    • No second network just for Photons required
  • Better Eyebrows
    • More complex, better coloration
    • Bump mapping makes them not look so painted-on
  • Better Paws
    • The paws have been redesigned to look more realistic and cleaner
    • Eliminated the banding of the AO shaders
    • Changes the way the divots highlight at angles
    • Attempted bump mapping didn't work out
  • Better Nose
    • Added texture and more color
    • Attempted bump mapping didn't work out
  • Improved Penis
    • Vein now was bump mapping and some width. Better coloration
    • Body color fits in better while still standing out.

I will need to continue improving this if I want to use it, however. It just isn't on par with metallic shader just yet. At the least, I  expect to incorporate all the changes made in this one, except the body and light/dark marks parts. Those 3 main components still look much better on the metallic shader and make up some 90% of the model as well as the lion-share of the render time.

 

In addition to this, I fooled around with the idea of building a curvy, cute body type for Blythe. This also inspiried me to merge my 3 existing scripts for skin weighting into one mega script that does the whole process for me in a single file and is undoable. That should really speed up any experimenting with shape I do! But, it remains difficult to handle topology changes. I plan to integrate the binding and other painting scripts into this one so I can just keep jumping forward on improvements and have a fully bound, pose able and expressible model for each test. It's not super fast to run, though. It certainly could use some efficiency work!

0802 BlytheCurvyExperiement.gif

The experiment wasn't super successful, but I may pursue it further

It's clear the model's topology will fight against good-looking deformation on a curvy character. It's possible some adjustments to skin weight paining and precise vertex positioning may improve this, however. But, in some places there are very few verts to work with. I did not experiment on changing the shape of the face as that will be more difficult due to the needs of the face wires.

 

I also did some more lewd works this week, but that will have to stick to the Red Lantern! Next week I hope to dabble in the shader a bit more and get back to developing components for that pool scene.

  • Like 1
Link to comment
Share on other sites

Slow week :/

Aside from a silly little render, very little was accomplished as I struggled with design ideas for a lounge chair to populate the pool area. I am not so hot at furniture design and a tendency towards wanting this thing to be done with led to sub-par appearance. For now I'm considering it a WIP. I also mocked up some shaders to help me get a feel for what it would look like (and to seek feedback with.)

0811 LoungeChairColorMockup4 WIP.png

I can't decide what to do about the simplicity of the design. Especially that middle part.

I kind of have an urge to try my hand at fixing the last of the XGen fur issues again. But, no doubt that will slow down pool development. I did produce a combined version of the Metallic and experimental shader networks, but was unable to find a good way to replace the Metallic entirely. At least the other improvements have been included, I guess.

Link to comment
Share on other sites

Much of this week was spent on a bit of fun trying out doing palette swaps for people. But, I did manage at least one important result: I fixed the bald spots on my XGen fur!

0818 FurBitPoofyButNotBald Censored.png

Links to Weasyl with uncensored image

On the right is the bald-spot marked fur whose variable function (which adjusted the length and bend of the fur to be a tad random) used a varanoi function built into XGen's Expression language. On the left, a perlin function has been substituted. While it does clear up the bald spots, the perlin-based function also shifts more smoothly and less randomly (which results in a kind of wavy look to the fur).

The great thing about this XGen fur is that its look is defined through maps and math functions. That means with just a few texture map adjustments (and a few new ones for certain parameters) this same good-looking fur can be implemented on any of my character models!

Unfortunately, it also currently takes 7 hours to render even this basic image at the desired anti-aliasing settings.

I am considering trying my hand at geometry-based stylized fur again. But, there's also that pool stuff to be done...

Link to comment
Share on other sites

  • 2 weeks later...

With Blythe's bald spots fixed, I set about implementing fur for Gravitas. While it certainly took less time that all of Blythe's development to get him to roughly the same spot, it did take longer than I expected and his results still require a fair bit of refinement.

1e406ff178f2015a0fa0fb02b57deaa28366c20c

Title on the work is p accurate, tho

Not so happy with the outcome. There's something funky going on with the chest that I can't figure out and the hands are kind of a mess. I am kind of happy with the cheeks, though. The eyebrows have lost some expressiveness, however. I am considering trying a geometry approach again.

Aside from that, I produce a few new pool assets (and worked on some other *ahem* assets that are in the Red Lantern thread). Pillars! Columns! Something called an Atlantid!

0822 ShimmerPillar Post.png

The first pillar was looks good from a distance but not so much from up close and will likely require refinement.

cccfac17b4d0dd311f9e778e47563ebe2f6fe081

I am much happier with the design idea and execution of the second pillar, however! Although the shader could use some more work. It is an Atlantid, or Atlas. An architectural concept that comes from the mythological Atlas, depicted in some ancient buildings as decorative pillars holding up the ceiling, later adopted into a general concept. I have adopted it with my Lapsa form, though I think it might work better with a more buff body that Rederick's...

Link to comment
Share on other sites

This was a fairly productive week! Certainly, having an extra day open helped. I finished up a new character design sporting a new bodytype: Kean!

0831 CleanKean.png
Uncensored version in the Red Lantern, or on (nsfw) my Weasyl

I decided to try for a more muscular build to do it. While I can't say the results are perfect (deformation is touchy in the shoulders and neck area) I think I'm happy enough to roll him out. The arms come off a bit balloon-y but I ultimately determined that the topology just wasn't there to support better muscle-tone.

 

I also managed a trio of pictures with Rederick and Blythe!

 

I am very happy with this one. It's very close to the idea I had for it way back when I started modelling Blythe. The other one is a bit saucier and I can only post the second part here (In the Red Lantern, though...)

df8b4bfec8ea9e95e91451d447f56d32fac8ab7c

This has to be the corniest dialog I've ever written Still, I felt like trying out a fishnet top for Rederick to compliment Blythe's new leggings and, well, you'll just have to see the whole thing in the Red Lantern! :3

Or, you know, in (nsfw) my Weasyl gallery...


The fishnets themselves suffer from the same seam problems as Blythe's leggings. The UV seams just don't match up to realistic clothing seams and it would be incredibly difficult to get the laces to line up appropriately across the UV boundaries. I feel as though there could be a better way though. I am looking into one particular technique that create a sort of overlay on a model. This overlaid layer has its own UV set and thus I would be free to develop specialized UVs that are more suitable for developing quality fishnets, lace, and lingerie. ZBrush doesn't support multiple UV sets, but as long as I only pass it the model with on UV set at a time, it should be manageable. That said, this technique would also be a lot of extra work.

Link to comment
Share on other sites

Due to scheduling over the weekend, not as much was accomplished this week. After an idea was presented to me by Lio, I attempted to have my machine rendering a long turntable with fur over the time I would be out. Unfortunately, the fur is so slow that only 33 frames out of 150 had rendered over the period I was gone. Had I left it running, especially at a reduced number of cores, it still would not be done yet.

I am, however, quite happy with the pose I came up with and opted to render a still image version of it with a better camera angle. I may yet come back to the turntable idea, if only I can come up with a way to speed up the rendering without to great a sacrifice of quality.

0909 BlythePoleFlight Thumbnail.png

Uncensored nsfw version on Weasyl is linked.

 

One thing I've tried for reducing the time it takes, is to create the speedy, "Silky" shader I made for Rederick with Blythe's colors and maps. Tests have been inconclusive as to how much it helps fur renders, but it has other uses too. Likewise, I produced the same for my own personal model.

0909 4Square_BlytheSilky Thumbnail.png    70be460713d4a01595cc25903045275ef07a6d77

  • Like 1
Link to comment
Share on other sites

I'm finally back in the mood to work on the pool area! I had ideas to transform it from a relatively bland space to one that's more visually-interesting. It still requires a lot more work, but I have the basic layout of the left side complete now!

aa1e3c783832f7285d74d84a01c8cedb1a5a197b

It is now a tiered space with 3 levels, including a set of "box seats" that I plan to have some fun with.

I also had an idea for the high-dive platforms, to work them into a statue. Unfortunately, I'm not really happy with the preliminary result and I think I will need to find a new approach.

0914 KeanDivingBoard Test.gif

The experimental High-Dive Statue

0914 KeanDivingBoardHands.png

This was the pose developed for that high-dive statue

There is much work left to be done on the pool and many more assets to develop for it! For the time being, I'm going to concentrate on the space and maybe a few assets where they directly contribute to the general layout of the pool area. Various other assets that will need to be developed:

  • Jacuzzi (single person and 2-3 person variations)
  • End table for drinks and stuff, next to the lounge chairs
  • Decorations (more kinds of pillars, reliefs, various artsy crap)
  • Plants and vines (Simple bush-like stuff)
  • Miscellaneous Signs, and general pool stuff (ok, so I'm tired of trying to think of everything I need -_-;)
  • Locker room stuff:
    • Lockers
    • Benches
    • Utility light fixtures
    • Towels (Unfolded)
    • Hamper
    • Random Clothes and Personal Items
Link to comment
Share on other sites

  • 3 weeks later...

Dear, dear, dear. I've left this field fallow haven't I?

I completed a jacuzzi (renders contain nsfw material and are in the Red Lantern thread and on Weasyl)

I did another Kailum pose, this time for Rederick:

0921 RederickKailumPose06B PostB.png

And I completed a sliding glassdoor entry way into the box-seats area of the pool, as well as a little end table for placing drinks next to the lounge chairs:

0925 EndTableB WIP.png1001 BoxseatsEntryway04 Post.png

I am working on another wall as well as some decorations. I want to try to work out at least one room in the pool area so I can build out some scenes with it. Hopefully it will spur further developments in the pool-space.

I also spent some time fiddling with fur again, after adding my piece to DevilBear's big collab thread! I've tried some alternative versions and was surprised to find that one of the quickest and simplest ones tended to be the most favored. I am balancing this against my own stylistic interests to decide how I want to move forward with it, but I may be doing more fur renders in the future.

1004 GravitasFluffComparisonTimes.png

And yes! I'm still interested in opinions!

(Really large image - Click to expand!)

Link to comment
Share on other sites

  • 1 month later...

Stories of this thread's death have been greatly exaggerated. To be honest, I haven't kept up this thread as well as I have in the past because it started to become more cumbersome than helpful. It's nice to be able to look back, sure, but at the same time I put so much effort into this thread that at one point it was taking up almost 2 days worth of my effort (compositing, uploading, describing, and of course posting).

This was an amazing motivator for a long while, but I'd gotten bogged down in grand posts. When Discord helped me receive wider feedback, this thread suffered. Discord became more of a development space and as I focused on the long-term pool project, the need for well constructed WIPs with a fair bit of effort began to slow down my pace. So, this thread faded. But, I've found Discord is not quiet as good at keeping me focused. I tend to jump around a lot more with it. To fix that, I am going to try a revival of sorts.

This thread is being relaunched in a new format!

Rather than fairly consistent weekly posts, with tons of development details, I'm going to do posts as they come up (more like I have on Discord) and with less text details. Furthermore, I'm going to post more unrefined WIPs than final product showcases and less linking to my galleries.

 

 

In that spirit, here's the latest of my pool setting!

1201 WaterfallStatues_Far.png

Lighting is only basic tests, and the space is currently unshaded/uncolored.

 

The thread may change, but feedback/comments/suggestions will always be welcome!

Link to comment
Share on other sites

  • 2 weeks later...
  • 1 month later...

Working on block-segments for constructing a bar. Did a simple test with a figure. Not exactly happy with the decorative panel (shimmering gold) but they're designed to be replaceable. I initially tried a more complex marble coloring, but after suggestions by Feelwell, found that (mostly) white worked better.

0130 BarSegments01_Far02 Post.png

(Initial)

0131 BarSegments02_Far01 copy.png

(Current)

 

I'll try to develop a black marble later today

Link to comment
Share on other sites

  • 2 weeks later...
  • 3 weeks later...

I revamped the bar area into a more interesting architecture: It is now a suspended mezzanine bisecting the long, dull series of windows that was to be the exercise machines area. Moreover, I've created a depression below it that is to serve as a dance area! Some of the stairs are still not present and the dance area lacks much of anything right now.

0302 PoolBarArea.png

Link to comment
Share on other sites

  • 3 weeks later...
  • 2 weeks later...

What would a pool area be without some swimwear!

I've posted Rederick's speedo before, but a bikini for Blythe was a whole other challenge. It required a few extra strips for wear the bikini shouldn't stick to the skin, such as crossing the cleavage and some of the straps. In fact, all of the places that needed the strips were to deal with her breasts. The rest is painted on the same way the speedo was. Technically the strip across the cleavage is physically simulated. I was experimenting with bringing those systems into the rigging.

0326 BlytheBikini Post.png

As with other textures, adding making the fur work with it is a breeze! Well, it would have been, had I noticed sooner that a few parts were not absolutely, perfectly black. That caused the culling mechanism to draw some additional lines through the fur and took me quite a number of hours to figure out.

0326 BlytheBikiniFur Post.png

Link to comment
Share on other sites

  • 2 weeks later...

Segments of a python script script I'm working on that will allow me to copy weights of specific model parts. With it, I'll be able to just apply weights to the parts of models that are the same. That'll help speed up testing and work on new characters with different body parts (heads, legs, etc)

'''
	Insure only one object of a given type exists on the map. Otherwise throw an error.
'''
def addForUniqueType( obj, type, typeGroups ):
	#There can only be one mesh and we only want one mesh
	existingObjs = typeGroups[type]
	if len(existingObjs) == 0:
		typeGroups[type] = [obj]
	elif existingObjs[0] == obj:
		#Same object, no problem
		None
	else:
		# different object, which means two meshes were selected.
		pm.error( "Multiple " + str(type) + " selections. Only one node of this type may be selected. Nodes selected: " + existingObjs + str(obj) )
		
'''
	For pm.nodetypes.Mesh
'''
def meshSelectAction( obj, typeGroups, actions ):	
	addForUniqueType( obj, pm.nodetypes.Mesh, typeGroups )
	
	#Check children for skin
	typeGroups = sortByDataType( obj.getChildren(), actions, typeGroups, warn=False, warnRecursive=True, allowDuplicates=False )
	#Check parents for skin
	typeGroups = sortByDataType( obj.getParent(), actions, typeGroups, warn=False, warnRecursive=True, allowDuplicates=False )
	
	return typeGroups

'''
	For pm.nodetypes.Transform
'''
def transformSelectAction( obj, typeGroups, actions ):	
	mesh = obj.getShape()
	addForUniqueType( mesh, pm.nodetypes.Mesh, typeGroups )
	
	#Check children for skin
	typeGroups = sortByDataType( mesh.getChildren(), actions, typeGroups, warn=False, warnRecursive=True, allowDuplicates=False )
	#Check parents for skin
	typeGroups = sortByDataType( mesh.getParent(), actions, typeGroups, warn=False, warnRecursive=True, allowDuplicates=False )

'''
	For pm.nodetypes.SkinCluster
'''
def skinSelectAction( obj, typeGroups, actions ):	
	skin = obj
	mesh = skin.getInputGeometry()[0]
	
	addForUniqueType( skin, pm.nodetypes.SkinCluster, typeGroups )
	addForUniqueType( mesh, pm.nodetypes.Mesh, typeGroups )

	
'''
	Given an object, attempts to check its type and perform the necessary action or add to a group for return.
	Automatically handles lists of objects, breaking them up and adding to the appropriate groups, after 
	calling an appropriate action if any.
	
	By default, it will provide warnings if target type is found for a selection. 
	This may be changed with the warn boolean parameter.
	Whether recursive calls will use this warn value or the default (true) value is set by warnRecursive.
	By default, recursive calls will use the default warn parameter.
	By default, this will allows the same object to be grouped into a list, if it already is in that list
	allowDuplicates may be set to false and it will not add a object that already is mapped to its type.
	the allowDuplicates value is always carried forward recursively
'''
def sortByDataType( obj, typeGroups, actions=None, warn=True, warnRecursive=False, allowDuplicates=True ):	

	if isinstance( obj, list ):
		# iterate over the elements of the list and recurse for each
		for element in obj:
			if warnRecursive:
				typeGroups = sortByDataType( element, typeGroups, actions, warn, warnRecursive, allowDuplicates )
			else:
				typeGroups = sortByDataType( element, typeGroups, actions, allowDuplicates)
	else:
		for type in typeGroups.keys():
			if isinstance( obj, type ):
				acted, typeGroups = checkActAndReport( obj, type, typeGroups, actions )
				if not acted:
					if allowDuplicates or obj not in typeGroups[type]:
						typeGroups[type].append( obj )
				return typeGroups
		
		if warn:
			pm.warning( "Ignoring Selection. No targetted type found for " + obj )
				
	return typeGroups

'''
	For when you only *might* have a method to execute.
	Tells you whether you did or not
'''
def checkActAndReport( obj, type, typeGroups, actions ):
	#Check for actions to perform, if provided
	if actions is not None:
		if type in actions:
			action = actions[type]
			typeGroups = action( obj, typeGroups, actions )
			return True, typeGroups
	
	return False, typeGroups

'''
	Gets all targets (mesh verticies and skeleton root joint or possibly joints) and returns them as separately.
'''
def getTargets():
	typeGroups = { pm.MeshVertex : [], pm.nodetypes.Joint : [], pm.nodetypes.Mesh : [], pm.nodetypes.SkinCluster : [], pm.nodetypes.Transform : [] }
	
	actions = { pm.nodetypes.Mesh : meshSelectAction, pm.nodetypes.SkinCluster : skinSelectAction, pm.nodetypes.Transform : transformSelectAction}
	
	sortByDataType( pm.selected(), typeGroups, actions )

	# [:2] Only require verts and joints.
	for type, values in typeGroups.items()[:1]:
		if len(values) == 0:
			pm.error( "No targets selected for required type: " + str(type) )
	
	#TODO: I can't believe we still have to cheat for this.
	skin = pm.PyNode( 'skinCluster44' )
	#Ensure that the each MeshVertex object represents on a single Vertex
	targetVerts = typeGroups[pm.MeshVertex]
	targetVerts = descretizeVertices( targetVerts )
	
	return ( targetVerts, typeGroups[pm.nodetypes.Joint], skin)

'''
	Take a list of MeshVertices that may represent one or more vertices and
	make a list of discrete vertices, each MeshVertex represetning only a single vertex
'''
def descretizeVertices( targetVerts ):
	descreteVerts = []
	
	for vertexBundle in targetVerts:
		for vertex in vertexBundle:
			descreteVerts.append( vertex )	
			
	return descreteVerts

 

Link to comment
Share on other sites

t not sure how you're doing the fur texture, but i think a simple color map may not be enough. It seems that there needs to be a texture map underlying then the fur texture (if you're using a hair system) on top of it. Basically animals have multi layered fur. the color map makes sense in terms of it being "skin" but you do need a texture map above it which will probably give it more depth.

In Blender it's something similar to this

I know Zootopia has it's own custom built system for fur, but I think you may go a long way if you texture map the body, and then add fur on top of it.

  • Like 1
Link to comment
Share on other sites

49 minutes ago, QT Melon said:

t sure how you're doing the fur texture, but i think a simple color map may not be enough. It seems that there needs to be a texture map underlying then the fur texture (if you're using a hair system) on top of it. Basically animals have multi layered fur. the color map makes sense in terms of it being "skin" but you do need a texture map above it which will probably give it more depth.

In Blender it's something similar to this

I know Zootopia has it's own custom built system for fur, but I think you may go a long way if you texture map the body, and then add fur on top of it.

Thanks for the comment! I have multiple texture maps (which get converted into XGen's special .ptx contraption) driving multiple fur parameters: length, extra length multiplier, density (only occasionally used), clumping strength, and driving 3 hair-color choices (body, light-marks, and dark-marks, allowing for blending between the two.) XGen is a bit more generalized than a hair system, employing shaded splines controlled dynamically by expressions (and can even be tied into Maya's nHair systems for physical simulation) rather than simple textures.

The problem I'm faced with is the complexity of this solution weighs on render time and to combat this, (and difficulties in the expressions used to position the fur) generally leads to bald spots which the underlying "skin" helps obscure. It's not ideal, but it's the trade off I've had to make to address significant shortcomings in my map painting skills and desires to minimize duplicated work to speed up creation of additional characters while retaining comparable levels of quality between them.

That said, it might be worth considering a simple under-layer of short fur with little or no dynamic properties. Managing it may still be a pain, but it might nit impact the render time quite as much if I manage to do it without expressions (barring maybe some for color, although that's typically the slowest one).

 

Do you have any resources describing the variation in appearance between different layers of fur?

Link to comment
Share on other sites

Okay, Xgen...as I have stated I'm learning how to use 3D so I'm getting the terminology.

 

But with this example of the tiger, I'm seeing a map on the tiger before the fur.

 

 

what I'm noticing is that the texture map has fur on it. Then on top of it he's using Xgen to create the hairs, and then grooming them in pattern to the map.

I see it again here in this other tutorial.

So what you'll need to do with your characters is create a map that isn't straight "color" which is what you have now. You will need to paint and create maps that look like fur texture, overlay that with the Xgen system.

I think higher end computers can definitely do more renders and get away with it, but on the lower consumer end of the spectrum, we kinda have to cheat a bit to make it render faster and save time, while making it look good.

  • Like 1
Link to comment
Share on other sites

3 hours ago, QT Melon said:

Okay, Xgen...as I have stated I'm learning how to use 3D so I'm getting the terminology.

 

But with this example of the tiger, I'm seeing a map on the tiger before the fur.

 

 

what I'm noticing is that the texture map has fur on it. Then on top of it he's using Xgen to create the hairs, and then grooming them in pattern to the map.

I see it again here in this other tutorial.

So what you'll need to do with your characters is create a map that isn't straight "color" which is what you have now. You will need to paint and create maps that look like fur texture, overlay that with the Xgen system.

I think higher end computers can definitely do more renders and get away with it, but on the lower consumer end of the spectrum, we kinda have to cheat a bit to make it render faster and save time, while making it look good.

Ah, I think I get what your saying (I couldn't actually see the vids before as I was at work). You're saying that the skin shader shouldn't be the simple colors and shapes as they are now but textured to look more like fur with individual hairs. I think this is worth a shot. Although it will take some work to integrate into my current flow. 

My current workflow is driven by flexibility. Just as I have maps that define the various parameters on XGen, I use maps to drive a shader network that colors my model. So, the skin is not one texture, but in fact 10 separate greyscale textures plugged in as weights that drive the shader to apply colors and effects at render time. Construction of a more fur-like design would have to follow a similar approach in order to retain the flexibility I currently enjoy for tweaking colors, otherwise I'd find myself repainting the fur every time I want that to change.

There's also serious problems to overcome with seams for detailed fur paints. This is a much trickier issue for my models, unfortunately.

 

I might have to try giving this a shot with one of the experimental new chars I'm doing in my other thread!

Link to comment
Share on other sites

  • 3 months later...

The eye thing didn't really lead to shadows, but it did result in a new eye lens material.

 

In addition to the new script I spoke of in my nsfw thread, I produced another script that will be very helpful for my development. Whereas the other script was an extension of automatic skin-weight painting script to target specific parts of a model, this script alters the shape of the model itself. Specifically, I can take a model that resulted from lattice deformers applied to a base model, and replicate all or part of that deformation.

In effect, I can precisely recreate the alterations that generated my other characters from the base models. So, when I go to update topology on a model, I can quickly recreate my other characters with that new topology as well. I could also use it in combination with other scripts to mix and match pieces of models and potentially create a new way of developing characters.

This script will be very important as continue forward with model development! As a demonstration, I've used it to create sfw versions of my Kean and Yeltsin models, both of which previously only had nsfw versison!

597a796a2b2ba_0718YeltsinClean_CheerPost.thumb.png.01bdf540a813a7849599787e4d688b07.png597a797e9c462_0720KeanClean_ShrugPost.thumb.png.d535a8a44a971f0f3912a5dfdba04556.png

I'm really extremely happy to see this work out so well. I was able to recreate both characters in little more than a day's worth of work. Kean only took a couple hours, even. But most of all, I'm happy to see these characters finally starting to come into their own. Having sfw versions of them makes them feel more like real characters than ever before, rather the one-off joke design that Yeltsin started off as.

597a799b98b61_0727KYCleanHugPost02.thumb.png.6ec2100707255d654f7ee75a9aa102e9.png597a79b563513_0727KYCleanHugPost03.thumb.png.96e82e6436833eea4ab93efa1b38ebe6.png

Link to comment
Share on other sites

  • 3 months later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...