Revision request for Mineur's Yugito hair import (1 Viewer)

[mryouallrock]

Here's the link to an image of it: http://www.sdtmods.com/index.php?action=dlattach;topic=76.0;attach=2967;image.

Well, for revisions, there's probably no need for reference shots, since there already is at least one version of the hair import to refer to, but here are some anyways, just to be on the unsmiteable side:

Spoiler

http://img3.wikia.nocookie.net/__cb20130808122624/naruto/es/images/c/c8/Yugito_libera_una_cola_del_Nibi.png
https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcS4oJyDsA8XVRDOnIWBSKMxUHbZD8zTObFnAvHpdP24fAAyWZHs
http://image.blingee.com/images16/content/output/000/000/000/528/394193518_830004.gif
http://fc02.deviantart.net/fs71/i/2012/016/0/5/hidan_x_yugito_by_lesya7-d4mitxb.jpg

Now, the main thing is, I'd like the revision to be a .swf file. I usually don't mind hair imports being .png files, but in Yugito's case it just doesn't look right because of her tails hair not having movement animation when moving her head. And before you say things like "But that's how all .png files work; hair doesn't have animation", just note that I mean the tail itself, not the hair import as a whole, not being animated that makes it look off.

Good luck, as always, to whoever is up to the revision and thank you for your consideration!

 

[stuntcock]

Now, the main thing is, I'd like the revision to be a .swf file.

The word that you're looking for is "dynamic." An SWF hair import simply means that the elements are vector-drawn, which gives the hairstyle a smaller file size and finer detail (less pixellation) but makes it slightly more costly to render. It's also more difficult to modify an SWF than a PNG, unless the artist releases their source files.

Mineur works in raster-drawing applications, so someone would need to trace it before it could be animated. Are you willing to do that?

Since Mineur is still active on the forum, it would be polite to get his permission before making changes to his work. Have you done so?

The complexity of the animation work depends on the level of fidelity that you want.

• do we need to animate the entire hairstyle? Or should the top portion be rendered as a solid "helmet" which is attached to the ponytail at the nape of her neck, with only the ponytail itself being animated?
  • Sub-question: imagine that we have a complete dynamic version of the hairstyle. I flip the character upside-down so that she's standing on her head. How much of her headband should I be able to see?
• does the hair-wrapping pull the ponytail into a curve (as shown in the PNG), or should it fall in a straight line down her back?
• should the wrapped portion of the ponytail exhibit greater rigidity (with the unwrapped hairtips being much more prone to "flicking" around)? Or is the whole thing equally flexible?
• how much work should be put into testing the hairstyle against body-clipping? Assuming that clipping does occur, what's your preferred mitigation approach?
  • make the ponytail shorter
  • rotate the whole ponytail clockwise (away from the body)
  • make the ponytail heavier (and/or more rigid)
  • pull the ponytail root backwards (away from the neck) so that it's less likely to collide with the body
  • other (please specify!)

 

[mryouallrock]

Now, the main thing is, I'd like the revision to be a .swf file.

The word that you're looking for is "dynamic." An SWF hair import simply means that the elements are vector-drawn, which gives the hairstyle a smaller file size and finer detail (less pixellation) but makes it slightly more costly to render. It's also more difficult to modify an SWF than a PNG, unless the artist releases their source files.

Mineur works in raster-drawing applications, so someone would need to trace it before it could be animated. Are you willing to do that?

Since Mineur is still active on the forum, it would be polite to get his permission before making changes to his work. Have you done so?

The complexity of the animation work depends on the level of fidelity that you want.

• do we need to animate the entire hairstyle? Or should the top portion be rendered as a solid "helmet" which is attached to the ponytail at the nape of her neck, with only the ponytail itself being animated?
  • Sub-question: imagine that we have a complete dynamic version of the hairstyle. I flip the character upside-down so that she's standing on her head. How much of her headband should I be able to see?
• does the hair-wrapping pull the ponytail into a curve (as shown in the PNG), or should it fall in a straight line down her back?
• should the wrapped portion of the ponytail exhibit greater rigidity (with the unwrapped hairtips being much more prone to "flicking" around)? Or is the whole thing equally flexible?
• how much work should be put into testing the hairstyle against body-clipping? Assuming that clipping does occur, what's your preferred mitigation approach?
  • make the ponytail shorter
  • rotate the whole ponytail clockwise (away from the body)
  • make the ponytail heavier (and/or more rigid)
  • pull the ponytail root backwards (away from the neck) so that it's less likely to collide with the body
  • other (please specify!)

Well, this gives me a better idea of how to be more specific in future requests of dynamic hairs. Speaking of which, thank you for clarifying that little aspect.

As for your questions, here it is:

-Only animate the ponytail
-For your sub-question: all of it
-In a curve
-Equally flexible
-Make the ponytail heavier

As for getting Mineur's permission, for one thing, he said he mentioned being done with Naruto characters, therefore, not being active when it comes to those any more. Secondly, wouldn't permission for modifying his work be necessary for the future creator of the revision, rather than for the requester of the revision?

 

[stuntcock]

-Only animate the ponytail
-For your sub-question: all of it

This seems contradictory. Here's a ponytail-animated-only hairstyle flipped upside down. Most of the hair is a solid helmet, so we'll never see the entire headband.


Here's (a very rough sketch of) a fully-animated hairstyle flipped upside-down.


Note that the second image shows the whole headband. Upside-down is the "extreme" case and probably won't occur during gameplay, but you would see the frontal edge of the hair slide back and forth across the headband as the character moves.

The second approach involves more work for the modder, and may require extended hair physics (depending on the amount of animation fidelity that they're trying to achieve). You're free to stipulate this as part of your Request, but it means that you're less likely to find a modder who's willing to undertake the project (and/or you may be waiting longer because there are simply more "hours of work" inolved in the second version).

As for getting Mineur's permission, for one thing, he said he mentioned being done with Naruto characters, therefore, not being active when it comes to those any more.

If you post a Revision request for a hairstyle whose artist is inactive since 2012 then I would simply go ahead and do it. Mineur is an active member of the SDTMods forum, so I would at least attempt to get his permission before I started working on this project. It minimizes the risk of duplicate work (in this particular case that's a negligible risk since he's stopped drawing Naruto stuff, but the principle still applies), and it gives them a chance to say "oh waitaminute I have a high-resolution multi-layer PSD file; I'll send you a copy."

Hell, even if someone released a mod under a permissive license (such as CC-BY) I would still attempt to give them a heads-up before creating a derivative work. It's a simple matter of courtesy.

Secondly, wouldn't permission for modifying his work be necessary for the future creator of the revision, rather than for the requester of the revision?

Please try to understand how the forum works. "Artists" are the most scarce resource, followed by "programmers" and "skilled testers." Ideas, suggestions, and requests are abundant.

When you're making a Request, it serves the common interest if you do as much planning and preparatory work as possible. Finding high-res reference images, sorting out permissions, specifying animation preferences, drawing mockups and sketches to clarify your requirements, etc... This work usually isn't especially difficult, but it does take time. In this way, artists and programmers can spend their time on creative projects, more shit gets done, and SDT players have more cool mods available in their game. But it also serves your own self-interest, because an artist will look at your Request and think "this guy has already anticipated the challenges and come up with workarounds - I can feel confident about committing myself to this project." A well-prepared Request is more likely to be fulfilled.

Imagine that I contact Mineur regarding this hairstyle, and he says "absolutely not." (He wouldn't because he's a nice guy, but just play along.) I've spent a half-hour thinking about this project and communicating with you; that time is now wasted. I'm going to be angry at you, and I certainly won't help out with any of your Requests in the future. You could have avoided this hypothetical problem by asking Mineur's permission before posting the Request.


You skipped one question: Are you willing to vector-trace the hairstyle? This is a fairly important point, because vector-tracing isn't fun. It's similar to the preparatory stuff that I described above -- your Request is more likely to succeed if you're willing to do the boring-but-necessary tasks on your own.

 

[mryouallrock]

I can't say I disagree with you. I'll make sure to ask the creators of the original imports from now on before requesting a revision.

As for the vector-tracing, just know that I'd do it if I knew how. Seriously, the best "modding" I can do when it comes to SDT is creating custom dialogues, which just about ANYbody could understand.

Nevertheless, although I'm repeating myself, thanks for pointing out some other ways to be more specific in my requests. For dynamic hairs, at least. Which is probably the only kind of revisions I may ask for, but then again, who knows.

Finally, I'd be completely fine with the hair in the first picture. As you said, it's unlikely the second one would be needed in gameplay, so I'll just go for the somewhat bigger chances to get the request done.

 

[stuntcock]

As for the vector-tracing, just know that I'd do it if I knew how.

Three months ago, I didn't know how to do vector tracing. I learned how by tracing an SDT hairstyle. Because I liked the design and wanted to see it animated. When you express doubt and hesitation, it suggests to me that you don't really care about this design very much. And if you don't care about it, then why should I?

Vector drawing isn't technically challenging until you start getting into fancy stuff (like multiple overlapping gradients with partial alpha blending) but the Yugito design is strictly flat-colour so it's a straightforward path-and-fill job. There are many guides and tutorials available. It doesn't require a drawing tablet, fine penmanship, artistic skill, or colour-recognition. Zero lines of code are involved. English-language proficiency is not necessary. The software is free. Afterwards, you can add "Familiar with vector drawing techniques" as a bullet point on your resume. The only reasons why someone can't do vector tracing are physical disability (in which case I apologize for unfairly ranting at you), lack of a computer and internet connection (in which case "how are you posting on the forum?"), or extreme laziness (in which case "fuck you, parasite").

Like everyone on the SDTMods forum, I'd love to see more stuff being created. If forum members were willing to step up and provide vector-traces, then we could see new (or improved) hair and costume SWFs being released more frequently. Everybody wins!

Therefore I'm going to keep bitching whenever someone asks for a mod but hesitates at the "trace the artwork" stage.

 

[Pim_gd]

Vector tracing, is that just opening an image in gimp and adding lines on a new layer?

 

[Faceless]

Honestly, for SDT hairs, it's generally sufficient to throw the image into a vectorizer.

 

[stuntcock]

Honestly, for SDT hairs, it's generally sufficient to throw the image into a vectorizer.

Sure, if you're in a hurry and you just want to get it done. But I'm in the opposite position - once someone delivers a vector file, I need to spend time and effort animating it.


Which means that I get to deal with all of the minor flaws which arise when auto-vectorizing a pixellated raster image (admittedly, the flaws would be less noticeable if we were working with the PSD instead). Either I correct these flaws by hand (which brings us back to the original question of "why am I supposed to care about this when the requester doesn't?") or I ignore them and animate as-is (in which case the result looks lousy and I can't really take much pride in my work).

Also, auto-vectorizing converts all of the lines to fills. This isn't ideal, because fills are subject to ScaleY stretching due to dynamic-hair elasticity, whereas lines can be configured to maintain a constant thickness. Same problem as above - I can either manually re-trace the fills back into lines, or watch the border grow/shrink inappropriately whenever the hair moves around.

 

[Faceless]

Ah, but lines have their own complications. Lines handle scaling differently from fills, so when given two identical objects, where one is a line and the other is a fill, when scaling or zooming out the line will appear thicker than the fill. Since all the actual SDT sprites use fills for outlines, this introduces an inconsistent look.

 

[mryouallrock]

Well, if the vector tracing is going to be left to me, then since it's gonna be the first time I ever get to vector-tracing, noob questions such as "In that link you provided in quote to my comment where I said I didn't know how to vector trace, there was this image with a hair import (apparently) and a head template. If I'm going to be doing this manually (which seems to be the case so far, for the sake of a job well done), I might want to have it slightly easier. Is there any way to get that same template to give me a hand with the whole thing? I figure you wouldn't have to vector-draw that yourself, since you're likely to mis-size it from the actual head in SDT, and what then? You will have drawn it for nothing, since based on a poorly sized head template, the vector-traced hair won't go in size with the head in SDT, which is what matters. " may give you even more to bitch about.

 

[stuntcock]

Is there any way to get that same template to give me a hand with the whole thing?

Templates are here -- but you don't actually need them unless you're drawing a new PNG file, or creating a new SWF import. I'll certainly provide any advice or assistance that I can - I'm happy to see more people getting involved in modding work :)



The speculation about whether the hairstyle will fit properly onto the head is great - it shows that you're thinking ahead! In this case, though, it won't be a problem:

• the PNG file already fits onto the SDT head, so a traced version of it will also fit onto the SDT head
• once the hair has been traced to vectors, we can very easily resize or skew various parts of it without loss of quality. Any minor fitting errors can easily be corrected; re-tracing will not be necessary.
  • Similarly, if I discover that certain parts of the hair tend to clip through the body during animation, then I can shrink them down or rotate them away.
  • Note that I'll try not to make such changes in this particular case, because you've indicated a preference for "make the hair heavier" as the clipping-avoidance strategy.

If you have a Flash license then you can open the FLA template file, find the head reference, import the Yugito PNG into the appropriate position and size w/r/t the head reference, and then trace the hair in Flash. This isn't necessary, but it might be useful practice if you intend to make your own mods in the future. When the work is complete, you can upload the FLA file and I'll start animating it.

If you don't have a Flash license, then you can import the PNG into a new canvas in Inkscape and trace the lines in that environment (completely ignoring the "missing" head). When all of the curved lines (aka "paths") have been traced, you can fill the spaces between them with the appropriate colours (via the Eyedropper tool). Then you can upload the SVG file - I'll import it into an FLA file, fit it onto the head, and start animating.



Ideally, we should start with the PSD (Photoshop project file) rather than the PNG. It's probably higher-resolution (which means that it's easier to trace due to less pixellation), and it may include additional detail via layers (such as the full ninja-headband). The movement of dynamic hair tends to expose details in underlying layers; if we have more details to show, then we'll be able to give the hair a greater range-of-motion.

• Mineur was able to provide a PSD file when I asked him about a previous project, so hopefully he'll have one for this hairstyle also. Just mention it when you're asking his permission to adapt the hairstyle. If he can't find the old PSD file then we'll proceed with the PNG instead.
• If you get the PSD but you don't have access to Photoshop then just share the file here. I'll extract the full-resolution raster images of each layer and you can trace whichever ones you'd like to include in the dynamic hairstyle (e.g. ponytail, forelock, headband).
  • You have some "leverage" here w/r/t project scope. If you find vector-tracing to be very frustrating and you just want to get it over with, then you can trace the PNG and I'll animate the ponytail. If you're willing to do more tracing (e.g. separate layers for headband, forelocks, etc) then I'll put in more animation details - to ensure that all of your tracing work will be visible during gameplay.

Please feel free to ask additional questions. At any point, it's completely valid to pause (especially if you're uncertain about whether the work you're currently doing is correct or productive), upload a screenshot (or a WIP copy of your trace), and ask for feedback. If you get bored or tired then feel free to put the project and hold and return to it later. You can take as long as you want with the tracing work.

The actual path-tracing stuff is pretty straightforward once you've spent a few minutes with the tools, but there are other aspects (such as overlapping layers in Inkscape) which can be more tricky. Feel free to ask questions here (or lookup tutorials on Youtube or whatever) until you feel comfortable with the task.

 

[mryouallrock]

Is there any way to get that same template to give me a hand with the whole thing?

Templates are here -- but you don't actually need them unless you're drawing a new PNG file, or creating a new SWF import. I'll certainly provide any advice or assistance that I can - I'm happy to see more people getting involved in modding work :)



The speculation about whether the hairstyle will fit properly onto the head is great - it shows that you're thinking ahead! In this case, though, it won't be a problem:

• the PNG file already fits onto the SDT head, so a traced version of it will also fit onto the SDT head
• once the hair has been traced to vectors, we can very easily resize or skew various parts of it without loss of quality. Any minor fitting errors can easily be corrected; re-tracing will not be necessary.
  • Similarly, if I discover that certain parts of the hair tend to clip through the body during animation, then I can shrink them down or rotate them away.
  • Note that I'll try not to make such changes in this particular case, because you've indicated a preference for "make the hair heavier" as the clipping-avoidance strategy.

If you have a Flash license then you can open the FLA template file, find the head reference, import the Yugito PNG into the appropriate position and size w/r/t the head reference, and then trace the hair in Flash. This isn't necessary, but it might be useful practice if you intend to make your own mods in the future. When the work is complete, you can upload the FLA file and I'll start animating it.

If you don't have a Flash license, then you can import the PNG into a new canvas in Inkscape and trace the lines in that environment (completely ignoring the "missing" head). When all of the curved lines (aka "paths") have been traced, you can fill the spaces between them with the appropriate colours (via the Eyedropper tool). Then you can upload the SVG file - I'll import it into an FLA file, fit it onto the head, and start animating.



Ideally, we should start with the PSD (Photoshop project file) rather than the PNG. It's probably higher-resolution (which means that it's easier to trace due to less pixellation), and it may include additional detail via layers (such as the full ninja-headband). The movement of dynamic hair tends to expose details in underlying layers; if we have more details to show, then we'll be able to give the hair a greater range-of-motion.

• Mineur was able to provide a PSD file when I asked him about a previous project, so hopefully he'll have one for this hairstyle also. Just mention it when you're asking his permission to adapt the hairstyle. If he can't find the old PSD file then we'll proceed with the PNG instead.
• If you get the PSD but you don't have access to Photoshop then just share the file here. I'll extract the full-resolution raster images of each layer and you can trace whichever ones you'd like to include in the dynamic hairstyle (e.g. ponytail, forelock, headband).
  • You have some "leverage" here w/r/t project scope. If you find vector-tracing to be very frustrating and you just want to get it over with, then you can trace the PNG and I'll animate the ponytail. If you're willing to do more tracing (e.g. separate layers for headband, forelocks, etc) then I'll put in more animation details - to ensure that all of your tracing work will be visible during gameplay.

Please feel free to ask additional questions. At any point, it's completely valid to pause (especially if you're uncertain about whether the work you're currently doing is correct or productive), upload a screenshot (or a WIP copy of your trace), and ask for feedback. If you get bored or tired then feel free to put the project and hold and return to it later. You can take as long as you want with the tracing work.

The actual path-tracing stuff is pretty straightforward once you've spent a few minutes with the tools, but there are other aspects (such as overlapping layers in Inkscape) which can be more tricky. Feel free to ask questions here (or lookup tutorials on Youtube or whatever) until you feel comfortable with the task.

Well, I'm certainly glad everyone is fine with noobish questions now. Which leads to me needing the feel to ask this: so I open a canvas template and import the Yugito .PNG file. By "tracing the lines in that environment", do you mean simply vector-drawing on the outlines of the newly-imported hair? Or using the hair as a model and vector-drawing it from scratch? If the latter is it, don't expect me to finish it too soon; I intend to make it look good, and since it's my first vector-draw, it's likely to take some time.

Sorry if this is obvious, it's just that when it comes to new things I'd rather have it spelled out than guess.

 

[Faceless]

Why would you do it the second way? The entire point is to get an image that looks exactly the same as the original raster except as a vector, so just trace the outlines and replicate the fill colors.

 

[mryouallrock]

Well, I sure took a long time to get to asking this, but anyway, before accusing me of being a lazy asshole or whatever, I had trouble finding the time to vector trace as I had some other things to take care of IRL, which I won't go through in detail.

Again, even outside of being a newb, this may be a stupid question, but what's the name of that tool you use so that you paint using other tools with the colour you clicked on previously? I figure that'll be useful in my little project here.

P.S. Now that the outline vector-tracing is over, I assume that now I just use the tool whose name I don't know and which I asked for above, select the colours of the hair, headband and so on and just use the fill tool to colour the spaces inbetween vector lines, right?

 

[stuntcock]

what's the name of that tool you use so that you paint using other tools with the colour you clicked on previously?

Eyedropper (F7) to grab colours from the reference image. Fill Bucket (Shift+F7) to fill in the empty spaces. The fill logic is pretty similar to raster-drawing programs (e.g. MSPaint), but it might misbehave for non-obvious reasons:

• if you have two nodes at the exact same position, when the program expected you to merge those nodes, then the fill pattern may become strange
• if your paths are self-crossing then the fill may extend farther than you expect

If you run into problems like this, don't spend any longer than 1-2 minutes per fill. If you're stuck on one of them, just leave it as-is -- it will be simpler for me to fix it afterwards (and I may be able to explain the root problem so that you don't run into it again).

P.S. Now that the outline vector-tracing is over, I assume that now I just use the tool whose name I don't know and which I asked for above, select the colours of the hair, headband and so on and just use the fill tool to colour the spaces inbetween vector lines, right?

Yes.

It's possible to adjust the colour of the vector lines ("paths") themselves, but that's not crucial in this case because they're all black (which Inkscape uses by default).

It's possible to adjust their thickness, but again it's not crucial because they're all identical. if the artwork included a lot of thick outlines as well as a few fine strokes then I would expect the vector-trace guy to capture those details.

It's possible to create gradients instead of hard-edged patches of shadow and highlight. This requires additional work and it isn't strictly an "upgrade" -- it will often make the image look worse, and it can degrade the game's framerate. Gradients also makes the animator's job more difficult. In this case, the hard-edged shadows are suitable for SDT's art-style and there's no need to change them. The hair could gain some three-dimensionality and visual appeal by adding highlights, but that would be difficult because the character is consistently drawn without any highlights -- we'd need to draw (from imagination) instead of tracing (from a reference image).

Finally, it's possible to specially designate some of the fills within the image so that they become eligible for RGB-shifting. The actual configuration work must be done in Flash, but you can include "hints" in your vector-trace file (to ensure that the Flash guy designates the correct fills instead of guessing). Not necessary here, because hairstyles don't do RGB-shifting.

 

[mryouallrock]

Well, I certainly took my time and, while I'd say the vector tracing is fine, the results concerning the filling may not be as great. Complete noob's work, but anyway, here's the link: https://www.mediafire.com/?d1teb36mriijex2

Sorry for taking so long, especially since all it took was a little dropping and filling. It's just real life can't get enough of me nowadays.

 

[stuntcock]

the vector tracing is fine, the results concerning the filling may not be as great

There's a minor problem here: the fill edges are ragged and they don't meet. It's as if they were drawn in using a paintbrush tool or a "Select by Color Range" operation. This is bad for two reasons:

• there are "holes" in the hairstyle. If we brought it in-game in its current state, you'd be able to see the pixels of the girl's skin (and/the background) poking through the hair
• creating these fills involved unnecessary work. The fills could have been created by applying the Paint Bucket tool directly onto the "empty space" between the vector-traced lines. It would then have filled this entire space, leaving no gaps or ragged edges.

This problem will not block the project. I can correct the fills with the Paint Bucket tool in Flash; it's only a few minutes of extra work. It's a bit silly, though - it would take me the exact same amount of work if you had sent me a "naked" vector file containing all of the lines and zero fills. And you could (presumably) have done that a month ago :P



There are a few flaws in the vector lines themselves - places where the outline shows a noticeable "break" or doesn't match up with its surroundings. Such flaws are unusual in a vector drawing, because we can always zoom in and refine the details to an effectively unlimited degree. This may be the result of bad advice -- we repeatedly told you to trace the lines exactly from the reference image.

In the future, please feel free to scan though your work and make minor adjustments (or just trace the lines "inaccurately" in the first place) for the sake of roundness, smoothness, etc. If you believe that the original artist made a mistake (or was a bit lazy) in a particular spot, you're not required to imitate the original mistake. Here are some examples of edits that I made:





I'm in the middle of another project at the moment, but I should be able to begin animating this hairstyle on the weekend. I'm pretty sure that we'll need to use extended hair physics in order to fulfill the requirements (curved resting shape, increased rigidity for the wrapped portion) but I'll keep you informed. I'll post the WIP stuff in this thread, since there are a few other people who have participated in the discussion and they may also want a chance to offer feedback and suggestions.


On the "plus" side: MG has a new Loader version (release date TBD) which makes extended-physics hairstyles much more convenient for SDT users. You no longer need to worry about your mods.txt file; you simply load the hairstyle itself (by clicking the "Add SWF Mod" button, or by including it via sby's moreClothing) and the physics library will automatically begin working when it's needed.

So - it's nice for you because mods will simply "work" without a lot of technobabble and setup. And it's nice for me because I can freely use extended-physics stuff (to achieve greater animation fidelity) without worrying about whether I'm creating headaches for users.



In the meantime: I just checked Mineur's thread and noticed that we don't have a charcode for this character. If you'd like to spend a few minutes choosing the appropriate skin tone, eye color, attitude, makeup, etc... then we can release it alongside the hairstyle. It's not mandatory, of course -- just an extra little task if you're bored (or if you really like the character and want to ensure that people use the "correct version" of her in-game).

 

[mryouallrock]

the vector tracing is fine, the results concerning the filling may not be as great

There's a minor problem here: the fill edges are ragged and they don't meet. It's as if they were drawn in using a paintbrush tool or a "Select by Color Range" operation. This is bad for two reasons:

• there are "holes" in the hairstyle. If we brought it in-game in its current state, you'd be able to see the pixels of the girl's skin (and/the background) poking through the hair
• creating these fills involved unnecessary work. The fills could have been created by applying the Paint Bucket tool directly onto the "empty space" between the vector-traced lines. It would then have filled this entire space, leaving no gaps or ragged edges.

This problem will not block the project. I can correct the fills with the Paint Bucket tool in Flash; it's only a few minutes of extra work. It's a bit silly, though - it would take me the exact same amount of work if you had sent me a "naked" vector file containing all of the lines and zero fills. And you could (presumably) have done that a month ago :P



There are a few flaws in the vector lines themselves - places where the outline shows a noticeable "break" or doesn't match up with its surroundings. Such flaws are unusual in a vector drawing, because we can always zoom in and refine the details to an effectively unlimited degree. This may be the result of bad advice -- we repeatedly told you to trace the lines exactly from the reference image.

In the future, please feel free to scan though your work and make minor adjustments (or just trace the lines "inaccurately" in the first place) for the sake of roundness, smoothness, etc. If you believe that the original artist made a mistake (or was a bit lazy) in a particular spot, you're not required to imitate the original mistake. Here are some examples of edits that I made:





I'm in the middle of another project at the moment, but I should be able to begin animating this hairstyle on the weekend. I'm pretty sure that we'll need to use extended hair physics in order to fulfill the requirements (curved resting shape, increased rigidity for the wrapped portion) but I'll keep you informed. I'll post the WIP stuff in this thread, since there are a few other people who have participated in the discussion and they may also want a chance to offer feedback and suggestions.


On the "plus" side: MG has a new Loader version (release date TBD) which makes extended-physics hairstyles much more convenient for SDT users. You no longer need to worry about your mods.txt file; you simply load the hairstyle itself (by clicking the "Add SWF Mod" button, or by including it via sby's moreClothing) and the physics library will automatically begin working when it's needed.

So - it's nice for you because mods will simply "work" without a lot of technobabble and setup. And it's nice for me because I can freely use extended-physics stuff (to achieve greater animation fidelity) without worrying about whether I'm creating headaches for users.



In the meantime: I just checked Mineur's thread and noticed that we don't have a charcode for this character. If you'd like to spend a few minutes choosing the appropriate skin tone, eye color, attitude, makeup, etc... then we can release it alongside the hairstyle. It's not mandatory, of course -- just an extra little task if you're bored (or if you really like the character and want to ensure that people use the "correct version" of her in-game).

Put in more simple words, this extra task needs me to pick the skin tone, eye colour etc. then generate the custom save in the Modding tab?

 

[stuntcock]

Put in more simple words, this extra task needs me to pick the skin tone, eye colour etc. then generate the custom save in the Modding tab?

Yes.



The initial fitting wasn't great. Several areas didn't align properly against the Head reference, and the hair was unreasonably thick - it would have completely swallowed the male character's hand. Such a hairdo might be acceptable on a dolled-up fashion model, but this is a freaking ninja who spends her time kicking ass. The hairstyle was also much too tall - it gave the impression of an enormous forehead.



I fiddled with the envelope to make the hairstyle fit more closely around the skull. The altered shape leaves some bare skin behind the ear - this is intentional. The behind-the-ear strand will be animated, so if we give it a very close resting position then it may swing CCW all the way through the ear during violent motion. I'll fine-tune the position after the first animation pass. The big question is whether the mass of the wrapped ponytail will suffice to hold this strand taut, or whether it will need additional (hardcoded) stiffness to keep it properly oriented.



Some of the vectors have been extended/extrapolated and then split out onto separate layers. This was done to provide overlap in the artwork, so that we can animate the various hair strands without leaving any visible gaps. The detail/fidelity in these "hidden" areas is fairly low, because they'll be exposed very rarely (if ever) during gameplay.

I left the front of the headband cantilevered partway outwards rather than fitting it tightly to the skull. It leaves a flaw in the physicality of the design, but this flaw occurs in the original PNG hair file and in all of the reference images that I could find. It's presumably an intentional decision by the character designer -- to prevent the headband (and its metal plate) from becoming obscured by the character's hair.