agent

+0:25
reasoning

Examining the partially read HTML file reveals base64 encoded images, which might be preventing a complete read. Considering options: either read the remaining text directly, strip the image data to parse the HTML, or focus on specific line ranges. The text segment of interest spans lines 140 to 442.

  • read_file{"EndLine":442,"Path":"~/context/reflect/lowmarch.html","StartLine":142}

    input

    {
      "EndLine": 442,
      "Path": "~/context/reflect/lowmarch.html",
      "StartLine": 142
    }

    result, 4 lines

    Partial read (lines 142-442 of 442). Showing bytes 0-46080 of 156125 in this line range, truncated at the 46080-byte limit. To read the rest of this line range, call this tool again with the same StartLine and EndLine and ByteOffset=46080.
    142:   </figure>
    143:   <figure>
    144:     <img src="[redacted: data-URI prefix][redacted: base64 data]
  • read_file{"EndLine":442,"Path":"~/context/reflect/lowmarch.html","StartLine":146}

    input

    {
      "EndLine": 442,
      "Path": "~/context/reflect/lowmarch.html",
      "StartLine": 146
    }

    result, 298 lines

    Showing lines 146-442 of 442
    146:   </figure>
    147: 
    148:   <div class="box alt">
    149:     <h3>Can this route get there? My honest estimate <span class="g">guess</span></h3>
    150:     <p>It can reach a coherent, likeable cast that clearly belongs to one game, and one character at a time can go above that. I doubt code alone gets all 14 to Yin and Yang's face quality, because theirs rests on someone adjusting points by eye on a mesh built for it. Unicorn Overlord isn't 3D at all: it's painted parts. Nobody has measured this for a current model, so I can't give you a number. I should have told you this on 10-08, instead of <q>the look can be reached</q>.</p>
    151:   </div>
    152: 
    153:   <h3>What has been hard, and why</h3>
    154:   <div class="tbl w"><table>
    155:     <thead><tr><th style="width:15%">Area</th><th style="width:38%">What you said</th><th>Why it's hard</th></tr></thead>
    156:     <tbody>
    157:       <tr><td>Faces at an angle</td><td><q>20 degrees it looks cursed</q> (10-07), <q>too sharp</q> (10-12), <q>something still looks off</q> (10-30)</td><td>An anime face isn't a real solid. Studios say it can't fully work from every angle and fix it view by view.<sup><a href="#s1">1</a></sup> We can only nudge a face whose mesh VRoid decided.</td></tr>
    158:       <tr><td>Expressions</td><td><q>eyes are weird on the youth</q> (10-30), <q>we can't apply lashes like this like it's generic</q> (10-31)</td><td>VRoid's eyes, lashes and mouths are paint on fixed shapes. Blending samples breaks what the paint assumed, so 10-31 was a stack of workarounds, each for a number that differs per face (lesson 40).</td></tr>
    159:       <tr><td>Bases</td><td><q>I think the base model constrains us</q> (10-08), <q>I'm still made wondering about bases</q> (10-10)</td><td>The base sets the ceiling, and its limits only show once it's reshaped and dressed. See the list below. That fight ended well.</td></tr>
    160:       <tr><td>Hair</td><td><q>too sharp</q> (10-07), <q>seven pointed bangs</q>, <q>we really need to study how to do the hair thing</q> (10-08), <q>a slice down the middle</q> (10-30)</td><td>Our own notes called it the weakest part on 10-08, and it still is. A VRoid style is made for one head, and warping it moves its gaps. Our kit makes a fine lock, but nobody has designed a hairstyle you love with it. The study is still open (#735), and so is the slice.</td></tr>
    161:       <tr><td>Clothes, motion</td><td><q>pieces look floating</q> (10-08), <q>our pipeline isn't polished and iterations take long</q> (10-09)</td><td>We did these first, on bases later thrown out, then stopped for good reason. Nothing is known about how the current bodies hold up.</td></tr>
    162:       <tr><td>Seeing it</td><td><q>hard to explain</q> (10-10), <q>I can't explain it</q> (10-30)</td><td>You can see it and can't always name it. I can name things and see badly. My lash check passed lids you'd reject at a glance. In the one test I found, models judged visual results well below people (older models; mine is unmeasured).<sup><a href="#s25">25</a></sup></td></tr>
    163:     </tbody>
    164:   </table></div>
    165: 
    166:   <h3>Eight starting points in six days</h3>
    167:   <div class="chain">
    168:     <b>10-07</b><span>Hand-placed pixels: <q>they don't look anime styled to me</q></span>
    169:     <b>10-07</b><span>Code-built head: <q>20 degrees it looks cursed</q></span>
    170:     <b>10-08</b><span>VRoid: <q>the face is great now</q>, then <q>the base model constrains us</q></span>
    171:     <b>10-09</b><span>MakeHuman: bodies vary, <q>not anime styled</q></span>
    172:     <b>10-09</b><span>VRoid head on MakeHuman body: <q>the head surgery we do is kinda offputting</q></span>
    173:     <b>10-09</b><span>Other free bases tested (Quaternius, Blender Studio): none better</span>
    174:     <b>10-09</b><span>MakeHuman muscle carried onto VRoid: <q>didn't sculpt well</q></span>
    175:     <b>10-10</b><span>VRoid shaped by a MakeHuman cage: <q>this looks amazing now</q>, <q>it looks perfect</q> (10-12)</span>
    176:   </div>
    177: 
    178:   <h3>What kept going wrong the same way</h3>
    179:   <ul>
    180:     <li><strong>A check appears after you catch the fault.</strong> Elaine's arm through her chest (10-08), body holes (10-09), lashes (10-31). Each got a checker afterwards.</li>
    181:     <li><strong>I check in a view that hides the fault.</strong> Arms down hid her waist (10-08). The first rig sheet used the old pixel renderer (10-30). Mouths tuned on two bodies broke on the other twelve (10-31).</li>
    182:     <li><strong>A fix breaks what you approved.</strong> Saber's waistband (10-08), the heavy woman's thighs (10-12), the combined row's lashes (10-31).</li>
    183:     <li><strong>Numbers stand in for taste.</strong> Ratios from paintings stretched the faces (10-12). The heroic man passed every measure and still read as a stretched strong man. <q>Cohesion</q> became a 15% lash limit.</li>
    184:     <li><strong>Lessons don't carry.</strong> On 10-07 I wrote <q>authored shading beats computed shading</q>, with a forced shadow on the far side of the face. By 10-30 I'd set the face shading so the front stays lit and only the far jaw shades.</li>
    185:     <li><strong>Parked means you have to remember.</strong> <q>we still have repairs to do for some models, right?</q> (10-31). Mature faces have been waiting since 10-12 (#770).</li>
    186:   </ul>
    187: 
    188:   <h3>The nature of your feedback</h3>
    189:   <div class="two">
    190:     <div>
    191:       <h4>What you respond to</h4>
    192:       <ul>
    193:         <li>A whole character beating your doubt: <q>I'm sorry for doubting you. that saber was actually pretty good</q> (10-08). This is the high point of the record.</li>
    194:         <li>Seeing the climb: <q>I loved how we got better while going forward</q> (10-08).</li>
    195:         <li>A long fight ending: <q>it looks perfect, we made headway</q> (10-12).</li>
    196:         <li>A reference beside ours. One look at Unicorn Overlord's arms and you had it: <q>there's no muscle bulge</q> (10-09). They were right about an hour later.</li>
    197:       </ul>
    198:       <h4>What you reject</h4>
    199:       <ul>
    200:         <li>Anything stuck on or one-size: <q>surgery</q>, <q>scar</q>, <q>generic</q>, <q>seven pointed bangs</q>, <q>doesn't fit her body</q>. It's the same complaint each time: this doesn't belong to this person.</li>
    201:       </ul>
    202:     </div>
    203:     <div>
    204:       <h4>What you keep asking for</h4>
    205:       <ul>
    206:         <li>Where we're going: <q>where to?</q> (10-08), <q>what next?</q> (10-12), <q>where are we next headed</q> (10-30).</li>
    207:         <li>How far this goes: <q>see what our ceiling is</q> (10-07), <q>full confidence</q> (10-08), <q>I wonder how else</q> (10-07).</li>
    208:         <li>Fewer rounds: <q>guaranteed after a turn</q> (10-10).</li>
    209:         <li>Simpler: <q>I'm lost</q> (10-30), <q>I just need the sheets</q> (10-31).</li>
    210:       </ul>
    211:       <h4>What's hard to put into words</h4>
    212:       <ul>
    213:         <li>You reach for images (<q>cursed</q>, <q>scar</q>), draw an orange line on a screenshot (10-08), and ask for comparisons so <q>I can imagine it easier</q> (10-30). Your sharpest verdict all month was the blind pick between three rows (10-30).</li>
    214:         <li>In those moments I mostly named a cause and built: <q>I think I can name the "something off"</q> (10-30). I found no case of me making options just to help you find the word.</li>
    215:       </ul>
    216:     </div>
    217:   </div>
    218: 
    219:   <h3>Where my approach made it harder</h3>
    220:   <ul>
    221:     <li><strong>Too much, too soon.</strong> Costumes, cloth, a duel and enemies by day 2, on a base we later dropped. You called the retreat (10-09).</li>
    222:     <li><strong>No target frame.</strong> Nobody settled what the game shows: <q>where did we get the idea that it will be rendered so small</q> (10-08), <q>until we actually compose game scenes</q> (10-30). Everything is judged on T-posed tiles at four angles.</li>
    223:     <li><strong>You steered, I brought lists.</strong> Nearly every pivot was yours. My best contribution, the cage, came after <q>I wonder what's a way out</q>. I held no map.</li>
    224:     <li><strong>I oversold.</strong> On 10-30, when you said it wasn't what you expected, I said <q>the shading layers are what close the gap</q> and promised <q>you can see the gap closing</q>. You never said it closed. Your reference also moved from Unicorn Overlord to Yin and Yang, and I never asked about that.</li>
    225:     <li><strong>Research after failure.</strong> The chin shadow took three builds and a <q>scar</q> before you said <q>we might need to research</q> (10-30). When study came first (sleeves, idle, Saber), results jumped.</li>
    226:     <li><strong>Tools grew faster than art.</strong> The lessons are nearly all about meshes and caches. Almost nothing records your taste.</li>
    227:     <li><strong>Too much to read, and you as logistics.</strong> Long replies with ledger numbers, plus running a command for me, PC schedules, reopening pages.</li>
    228:     <li><strong>The hair report (10-31)</strong> answered your question and changed nothing on screen.</li>
    229:   </ul>
    230:   <p class="note"><strong>What I'd keep:</strong> the bodies and their fit checks, VRoid as the base, the painted renderer and its cast chin shadow, regression checks, the Saber armour kit, the one-lash rule, and blind rows.</p>
    231:   <p class="note"><strong>What this rests on:</strong> your typed messages in the two main sessions, the handoff, the portrait guide, commits, the ledger, sheets, and a fresh reader with no stake in the work, whose read I checked and largely agree with. I did not read comments left on review pages. My guess at what you want, for you to correct: a character you'd love, whole, doing something, like Saber, and someone who says plainly how far this goes.</p>
    232: 
    233:   <h2 id="pros">2. How professionals do it</h2>
    234:   <p>Numbers point to sources, read 2026-11-08. <span class="g">guess</span> marks my inference. Gaps: I found no HoYoverse primary source, and no studio publishes hours per character. The Guilty Gear GDC PDF, vroid.com, Polycount and ArtStation refused automated reads. I couldn't confirm the idea in my own lesson 03 that Guilty Gear stores geometry per angle.</p>
    235: 
    236:   <h3>Faces</h3>
    237:   <div class="two">
    238:     <div><h4>What they do</h4><ul>
    239:       <li>Make the three-quarter face work first. Straight-on and pure profile are rare, and if 3/4 works, the angles between tend to.<sup><a href="#s1">1</a></sup></li>
    240:       <li>Four things decide it: the cheek line, brow and eye depth, how far the mouth sits back, the jaw line. The cheek line sets sharp or soft and is called a <q>lifeline</q> for pretty characters.<sup><a href="#s1">1</a></sup></li>
    241:       <li>Common failure: squash the head front to back to slim the 3/4, which gives a <q>festival mask</q>.<sup><a href="#s1">1</a></sup></li>
    242:       <li><q>Breakage can't be avoided</q>, so fix it with bones or shapes.<sup><a href="#s1">1</a></sup> Gakumas has one shape that pushes the nose dot back so it vanishes in profile.<sup><a href="#s9">9</a></sup> Guilty Gear keeps separate battle and close-up heads,<sup><a href="#s3">3</a></sup> and in finishers the rig moves eyes, nose and mouth so the face reads from that one camera.<sup><a href="#s4">4</a></sup></li>
    243:       <li>Expressions don't transfer. Uma Musume authors named presets per character, because parts sit differently.<sup><a href="#s6">6</a></sup></li>
    244:       <li>The drawing is a guide, and copying it exactly isn't the goal.<sup><a href="#s1">1</a></sup> Arc trains new modellers for a year.<sup><a href="#s13">13</a></sup></li>
    245:     </ul></div>
    246:     <div><h4>For us</h4><ul>
    247:       <li><q>Too sharp</q> probably lives in the cheek line. <span class="g">guess</span></li>
    248:       <li>The four moves are small, so code can make them per character at 3/4. <span class="g">guess</span></li>
    249:       <li>One shared expression list is the wrong shape. Every character needs a tuned set, which is why the open mouths look alike.</li>
    250:       <li>If the game shows both battle and dialogue, plan on two faces. <span class="g">guess</span></li>
    251:     </ul></div>
    252:   </div>
    253: 
    254:   <h3>Shading a face so it reads as drawn; outlines</h3>
    255:   <div class="two">
    256:     <div><h4>What they do</h4><ul>
    257:       <li>Quality sits in authored control data, not polygons. Uma Musume: under 20k polygons a character. Its shadow mask, a painted map of where shadow may fall, has the biggest effect on quality and takes most of the texture work. Shape-based shading can't look anime.<sup><a href="#s5">5</a></sup></li>
    258:       <li><q>Where there is no vertex, you can't change the shadow shape.</q> A vertex is a mesh point.<sup><a href="#s1">1</a></sup></li>
    259:       <li>Guilty Gear paints a value on each mesh point: how readily it goes into shadow, and how thick its line is.<sup><a href="#s2">2</a><a href="#s3">3</a></sup></li>
    260:       <li>Normals are the direction a surface claims to face, which decides light or shadow. Faces get them edited by hand; hair and clothes get them copied from simple shapes.<sup><a href="#s1">1</a><a href="#s3">3</a></sup> Outlines keep a separate set.<sup><a href="#s2">2</a><a href="#s5">5</a></sup></li>
    261:       <li>Guilty Gear adjusts each character's light per motion, and per frame in cutscenes.<sup><a href="#s3">3</a></sup> Blue Protocol keeps light to shade at 7:3, because 5:5 went flat in backlight, and cuts the light's angle on skin by half to stop stray face shadows.<sup><a href="#s7">7</a></sup></li>
    262:       <li>Face shadow maps: paint masks at about 9 light angles and merge them. It's a community method, mirrored from one side.<sup><a href="#s10">10</a></sup></li>
    263:       <li>Outlines: a dark inverted copy of the mesh gives per-point control, while screen-wide lines are hard to tune. <q>Not showing lines where unintended</q> matters as much as showing them.<sup><a href="#s2">2</a><a href="#s7">7</a></sup></li>
    264:     </ul></div>
    265:     <div><h4>For us</h4><ul>
    266:       <li>This explains the <q>scar</q>: the shadow's edge followed VRoid's mesh points. <span class="g">guess</span></li>
    267:       <li>All 26 VRoid faces share one texture layout (lesson 29), so one designed shadow mask could serve them all, tuned per character. This is the best cheap next step for faces. <span class="g">guess</span></li>
    268:       <li>Moving light per shot is a cheap cheat once we have shots. <span class="g">guess</span></li>
    269:     </ul></div>
    270:   </div>
    271: 
    272:   <h3>Hair</h3>
    273:   <div class="two">
    274:     <div><h4>What they do</h4><ul>
    275:       <li>Bundles of locks with thickness. A study of 37,000 VRoid styles found 25 to 130 per style.<sup><a href="#s14">14</a></sup></li>
    276:       <li>Start from the drawing's outline and shadow positions. Keep only what's needed, and delete locks that get in the way from other angles.<sup><a href="#s11">11</a></sup> Too many fine strands make posing hard.<sup><a href="#s11">11</a></sup></li>
    277:       <li>Build the inside first, decide the bones then, and add the outer hair last.<sup><a href="#s17">17</a></sup></li>
    278:       <li>Hair can cost as much as the body: Gakumas 23k to 26k polygons, body 25k to 26k.<sup><a href="#s16">16</a></sup></li>
    279:       <li>Motion: root keyed, tip simulated, simulation off for dramatic moments.<sup><a href="#s6">6</a></sup> A hobbyist's settings for long cel hair: high stiffness, no gravity, to hold the drawn shape.<sup><a href="#s18">18</a></sup></li>
    280:     </ul></div>
    281:     <div><h4>For us</h4><ul>
    282:       <li>Our lock kit is the same family as VRoid's, so building hair in code is sound. <span class="g">guess</span></li>
    283:       <li>What's missing is design: a silhouette first, then few big dark shapes before any strands. <span class="g">guess</span></li>
    284:       <li>A dark inner layer is still the fix for the slice.</li>
    285:       <li><strong>Correction:</strong> my 10-31 report said to lay back, then sides, then front. No source here confirms that order.</li>
    286:     </ul></div>
    287:   </div>
    288: 
    289:   <h3>Clothing</h3>
    290:   <div class="two">
    291:     <div><h4>What they do</h4><ul>
    292:       <li>Folds are placed by hand for common poses, exaggerated, with shapes taken from animation rather than life.<sup><a href="#s11">11</a></sup> Revisions moved to pose shapes an animator dials in.<sup><a href="#s12">12</a></sup></li>
    293:       <li>Skirts follow the legs first and sway second. Getting the skin weights right (how much each bone moves each mesh point) beats physics settings.<sup><a href="#s18">18</a><a href="#s22">22</a></sup> Colliders on legs alone fail, and a keep-inside cage works better.<sup><a href="#s19">19</a></sup> VRM now has inside colliders.<sup><a href="#s20">20</a></sup></li>
    294:       <li>Nobody reaches zero clipping. Uma Musume adds special motions for big sleeves.<sup><a href="#s6">6</a></sup> Blue Protocol hides clashing parts.<sup><a href="#s7">7</a></sup> Dark Souls capes <q>intersect pretty much everywhere</q> and <q>nobody really notices</q>.<sup><a href="#s21">21</a></sup></li>
    295:       <li>Real simulation is for big items like capes.<sup><a href="#s21">21</a></sup></li>
    296:       <li>Fabric pattern shows in light and mostly vanishes in shadow.<sup><a href="#s24">24</a></sup></li>
    297:       <li>Armour: only forum evidence found. Why simulated folds look wrong: no source says so outright.</li>
    298:     </ul></div>
    299:     <div><h4>For us</h4><ul>
    300:       <li>Our folds are sine ripples that don't change with pose. A few named fold shapes tied to joint angles would suit code. <span class="g">guess</span></li>
    301:       <li>Our stiff skirt cone is close to the opposite weighting. <span class="g">guess</span></li>
    302:       <li>Grade clipping by what the game camera sees, not by millimetres. <span class="g">guess</span></li>
    303:       <li>Keep our cloth sim for capes only. <span class="g">guess</span></li>
    304:     </ul></div>
    305:   </div>
    306: 
    307:   <h3>VRoid, studios, and where code falls down</h3>
    308:   <div class="two">
    309:     <div><h4>What's known</h4><ul>
    310:       <li><strong>VRoid:</strong> hair guides don't survive export. Files hold merged, baked meshes. No scripting API found, and our driver is blocked by Windows until you allow it (#1105).</li>
    311:       <li><strong>Better base:</strong> none found that ships in a closed game. MB-Lab and CharMorph require AGPL, which can block closed-source games; Ready Player Me ended in January; Character Creator is paid and realistic.<sup><a href="#s32">32</a></sup></li>
    312:       <li><strong>Studios:</strong> modeller, texture artist, rigger, technical artist. A freelance listing prices an anime model at about 195,000 to 295,000 yen over 3 to 4 weeks.<sup><a href="#s31">31</a></sup> My estimate for a hero with rig: 3 to 8 person-weeks. <span class="g">guess</span></li>
    313:       <li><strong>Agents judging looks:</strong> in BlenderGym the best model's error was about 10 times a person's on shape edits and 5 times on geometry. Judging which result was better, the best model agreed with people at 0.66; people agree with each other at 0.79. These are older models.<sup><a href="#s25">25</a></sup></li>
    314:       <li><strong>Agents in Blender:</strong> good at scene work, poor at organic forms, proportion and weights, per mostly vendor reports.<sup><a href="#s29">29</a></sup></li>
    315:       <li><strong>Code-built garments:</strong> about 64 to 72% simulate cleanly, with no fine straps and nothing on armour or anime.<sup><a href="#s27">27</a></sup> Models do far better on a small spec language: 94% vs 65% in one test.<sup><a href="#s26">26</a></sup></li>
    316:       <li><strong>AI-to-3D anime:</strong> fused meshes or research-only weights, all trained on VRoid models.<sup><a href="#s28">28</a></sup></li>
    317:     </ul></div>
    318:     <div><h4>For us</h4><ul>
    319:       <li>Keep VRoid. Keep our spec language small. <span class="g">guess</span></li>
    320:       <li>Have code produce shadow masks, normals and line weights, not more geometry. <span class="g">guess</span></li>
    321:       <li>You stay the eye, but earlier. One CG studio keeps fixing the block model until it reads as the character.<sup><a href="#s30">30</a></sup></li>
    322:       <li><strong>A licence catch:</strong> our sample exports say <code>Redistribution_Prohibited</code>, <code>OnlyAuthor</code>, commercial <code>Disallow</code> inside the file. I checked two. It's probably the export dialog's default, since VRoid Hub lists samples A to C as open. Pixiv's A to Z help page has never been read; it blocks me.</li>
    323:     </ul></div>
    324:   </div>
    325: 
    326:   <h2 id="tools">3. Tools</h2>
    327:   <p><strong>Cost: $0. Nothing here needs buying, and I'd buy nothing yet.</strong> You said on 10-30 you're open to paid extensions. No paid hair tool I found has a documented way to script it, so they only help if someone styles by hand. Versions read 11-08, except where noted.</p>
    328:   <div class="tbl w"><table>
    329:     <thead><tr><th style="width:20%">Free</th><th>What it does</th><th>Suits us?</th><th style="width:19%">Licence, how alive</th><th style="width:10%">Say</th></tr></thead>
    330:     <tbody>
    331:       <tr><td><a href="https://docs.blender.org/manual/en/latest/modeling/geometry_nodes/hair/index.html">Blender 5.2.3 hair nodes</a>, Curve to Tube</td><td>Braid, Curl, Frizz, Clump; tube meshes ready for textures</td><td>Yes. Scriptable, already installed, no install needed</td><td>GPL. LTS to Jul 2028</td><td><span class="tag yes">try</span></td></tr>
    332:       <tr><td>VRoid Studio</td><td>Face and hair editor. Its sliders keep a face coherent, and it's the only source of mature faces</td><td>By hand, or through our driver once you allow it</td><td>Free; game use allowed per 10-09 notes. v2.14, Jun 2026</td><td><span class="tag yes">use</span></td></tr>
    333:       <tr><td><a href="https://github.com/pixiv/three-vrm">three-vrm</a></td><td>The standard VRM renderer</td><td>Cross-check our models outside our own renderer</td><td>MIT. 3.5.5, Jul 2026</td><td><span class="tag yes">use</span></td></tr>
    334:       <tr><td><a href="https://github.com/mikedh/trimesh">trimesh</a>, <a href="https://github.com/KhronosGroup/glTF-Validator">glTF-Validator</a></td><td>Ray and mesh checks; export file check</td><td>Yes, for clipping checks and an export gate</td><td>MIT; Apache-2.0</td><td><span class="tag yes">use</span></td></tr>
    335:       <tr><td><a href="https://github.com/saturday06/VRM-Addon-for-Blender">VRM Add-on for Blender</a></td><td>VRM in and out of Blender, by script</td><td>Only if the Blender trial earns its place. Installing it needs your go</td><td>MIT or GPL-3. 4.10.4, 4 Nov 2026</td><td><span class="tag later">later</span></td></tr>
    336:       <tr><td>Blender 5.2 Cloth and Hair Dynamics nodes</td><td>Node-based physics</td><td>One skirt trial, only if a rigged skirt fails. Marked experimental</td><td>GPL. In our copy</td><td><span class="tag later">later</span></td></tr>
    337:       <tr><td><a href="https://github.com/vrm-c/vrm-specification/tree/master/specification/VRMC_springBone-1.0">VRM 1.0 capsule colliders</a>, spring limits</td><td>Fewer collision spheres, possibly less hair through bodies. We're on VRM 0.0</td><td>A half-day test. Low urgency</td><td>Open spec. Limits landed Oct 2026, not ratified</td><td><span class="tag later">later</span></td></tr>
    338:       <tr><td>Our own shadow-map bake</td><td>A clean face shadow that moves with the light</td><td>No free Python baker found. QuickSDFTool is Unreal-only. Write it in numpy if a mask isn't enough</td><td>Ours</td><td><span class="tag later">later</span></td></tr>
    339:       <tr><td><a href="https://github.com/maria-korosteleva/GarmentCode">GarmentCode</a></td><td>Sewing patterns from code</td><td>Ideas only. One garment averages 30k triangles, more than a whole Uma Musume character</td><td>MIT. 2.0.1, 2024</td><td><span class="tag later">ideas</span></td></tr>
    340:     </tbody>
    341:   </table></div>
    342:   <p class="note"><strong>Rejected, free:</strong> official Blender MCP (running Blender by script is simpler); mcp-for-blender (four security flaws reported by a third party); UniRig (we have a rig); Abnormal (hand tool); CATS (on life support); Goo Engine (doesn't reach three.js); lilToon, Poiyomi, Unity Toon Shader (Unity only, read them for technique); MB-Lab and CharMorph (AGPL); StdGEN and CharacterGen (see above).</p>
    343:   <div class="tbl w"><table>
    344:     <thead><tr><th style="width:27%">Paid, all on hold</th><th style="width:22%">Cost</th><th>Why hold</th></tr></thead>
    345:     <tbody>
    346:       <tr><td><a href="https://superhivemarket.com/products/stylized-hair-pro">Stylized Hair PRO</a></td><td>$22, MIT (read 10-31)</td><td>The first buy if you want hair styled by hand. Anime locks, braids, curls. No scripting. A separate build per Blender version.</td></tr>
    347:       <tr><td><a href="https://bartoszstyperek.gumroad.com/l/hairtool">Hair Tool 4</a>, <a href="https://superhivemarket.com/products/anime-hair-maker">Anime Hair Maker</a></td><td>$52; $25 to $27 (read 10-31)</td><td>Hand tools. Stylized Hair PRO comes first.</td></tr>
    348:       <tr><td>Marvelous Designer</td><td>$39 a month or $280 a year</td><td>Hand tool for simulated clothes. I found no confirmed way to script it.</td></tr>
    349:       <tr><td>Lightning Boy, Anime Shading Plus, Character Creator 5</td><td>$30+; $30; $299</td><td>Blender 3.4 only per its listing; Unity; realistic and hand-driven.</td></tr>
    350:     </tbody>
    351:   </table></div>
    352: 
    353:   <h2 id="forward">4. How to move forward</h2>
    354:   <div class="box"><p style="margin:0"><strong>Restart on one whole character in one game frame. Everything else waits.</strong> Cost: $0, about 10 minutes of your time on most days.</p></div>
    355: 
    356:   <h3>What I'd change</h3>
    357:   <ol>
    358:     <li><strong>One whole character first:</strong> dressed, moving, in the frame the player sees. Range comes after. Saber is what won you over, and it was a whole person.</li>
    359:     <li><strong>Big shapes before detail.</strong> You approve hair and outfit as flat shapes, plus the light and dark pattern, before any fine work.</li>
    360:     <li><strong>You choose between rows.</strong> Two to four options beside your references, blind when it matters. You point, and I work out why.</li>
    361:     <li><strong>Per-character work over global fixes:</strong> a tuned expression set and a designed shadow mask per character, and hair designed from a silhouette.</li>
    362:     <li><strong>Study first, test the same day.</strong> One look at how it's done before the first build of anything new, and every piece of research ends in a test on our model.</li>
    363:     <li><strong>I hold the map.</strong> One short page: the target, this week, what's parked. And each week one plain line: how far we are, and whether I still think this route gets there.</li>
    364:     <li><strong>Two rounds on one complaint with no visible change, and I stop and say so.</strong></li>
    365:   </ol>
    366: 
    367:   <h3>What I'd ask you, one at a time</h3>
    368:   <div class="tbl w"><table>
    369:     <thead><tr><th style="width:52%">Question</th><th>What I'd assume if you have no strong view</th></tr></thead>
    370:     <tbody>
    371:       <tr><td>What do you wish ours looked like now? Five images.</td><td>Yin and Yang for faces, Unicorn Overlord for costume. Your reference moved, and I never asked.</td></tr>
    372:       <tr><td>What does the player actually see? Camera and size, in battle and in dialogue.</td><td>A side-view battle frame, plus one dialogue close-up.</td></tr>
    373:       <tr><td>Who's the first character, and what are they doing?</td><td>An original swordswoman on the slender body, in the Saber armour kit, doing one attack. Saber was the high point, and the kit exists.</td></tr>
    374:       <tr><td>What's this for: a game you'll ship, or finding how far a model can go?</td><td>Both, with the second driving. It changes what "done" means.</td></tr>
    375:       <tr><td>Is model-only part of the point, or would a person's eye on key decisions be welcome, if free?</td><td>Model-only for week one. No money implied.</td></tr>
    376:       <tr><td>Could you open pixiv's AvatarSample A to Z help page by hand? It blocks me, and we need it before anything ships.</td><td>Not urgent, but nobody has read it.</td></tr>
    377:     </tbody>
    378:   </table></div>
    379:   <p class="note">Small ones still waiting: the old man's fangs (I'd remove them), and allowing the VRoid driver through Windows (#1105), which any mature face needs.</p>
    380: 
    381:   <h3>A first week</h3>
    382:   <div class="tbl w"><table>
    383:     <thead><tr><th style="width:9%">Day</th><th>What I do</th><th style="width:30%">What you see</th></tr></thead>
    384:     <tbody>
    385:       <tr><td>1</td><td>The questions above, one at a time, played back to you. No code.</td><td>One page: references, target frame, the character. About 30 minutes of your time.</td></tr>
    386:       <tr><td>2</td><td>Put the current body in the frame on plain ground. Test the light and dark pattern: skin tone, a shadow down the side of the face, darker and simpler hair.</td><td>Three rows beside your references. Pick one.</td></tr>
    387:       <tr><td>3</td><td>Hair and outfit as big flat shapes, three or four options. The dark layer under the hair goes in here, which closes the slice.</td><td>Options in the frame. Pick or mark up.</td></tr>
    388:       <tr><td>4</td><td>Her face only: the 3/4 moves, a shadow mask, six expressions tuned for her.</td><td>Blind rows at 3/4 and in the frame.</td></tr>
    389:       <tr><td>5</td><td>Dress her. Folds placed for the pose, skirt following the legs, hair and cloth motion. One idle, one attack. Clipping graded from the game camera.</td><td>A short clip and a turning view.</td></tr>
    390:       <tr><td>6-7</td><td>Your verdict, then my plain note: how close, what's between, what more work buys. Then we decide: next character, more depth, or change route.</td><td>One page, one decision.</td></tr>
    391:     </tbody>
    392:   </table></div>
    393:   <p class="note">How we'd know it worked: you'd show her to someone, or say something like 10-08 again. If you don't, the route gets questioned out loud instead of patched.</p>
    394: 
    395:   <h3>What I'd stop doing</h3>
    396:   <ul>
    397:     <li>Fixing 14 bodies before one looks right.</li>
    398:     <li>Judging on T-poses on black.</li>
    399:     <li>Writing new checkers for looks. Checks stay for things with a right answer: proportions, holes, clearance, regressions.</li>
    400:     <li>Patching VRoid's painted eyes and mouths face by face without asking whether that's the right fight.</li>
    401:     <li>Reports that end with nothing changed on screen.</li>
    402:     <li>Ledger numbers and long replies during review. A sheet and under 100 words.</li>
    403:     <li>Saying something closes the gap. I'll show it, and you say.</li>
    404:     <li>Carrying 32 open todos. Close or merge the ten or so about retired characters and old dressed work, and keep the rest on the one page.</li>
    405:   </ul>
    406: 
    407:   <h2 id="sources">Sources</h2>
    408:   <p class="note">All read 2026-11-08 through a summarising fetch. I checked the main quotes from 1, 5 and 9, and the figures from 25, directly. Local evidence: <code>docs/portrait-guide.md</code>, <code>docs/vroid.md</code>, the handoff, session transcripts, and the VRM files themselves.</p>
    409:   <ol class="src">
    410:     <li id="s1" value="1">Arc System Works, modelling tips: <a href="https://www.docswell.com/s/ASW_Academy/5D2VY6-GG-Modeling">docswell.com</a></li>
    411:     <li id="s2" value="2">Arc System Works, toon lines: <a href="https://www.docswell.com/s/ASW_Academy/5LVY67-GG-Toonline-Eng">docswell.com</a></li>
    412:     <li id="s3" value="3">4Gamer, Guilty Gear Xrd part 1 (pages 1-2): <a href="https://www.4gamer.net/games/216/G021678/20140703095/">4gamer.net</a></li>
    413:     <li id="s4" value="4">4Gamer, Guilty Gear Xrd part 2 (page 1): <a href="https://www.4gamer.net/games/216/G021678/20140714079/">4gamer.net</a></li>
    414:     <li id="s5" value="5">Game Watch, Uma Musume models: <a href="https://game.watch.impress.co.jp/docs/kikaku/1366154.html">game.watch.impress.co.jp</a></li>
    415:     <li id="s6" value="6">Game Watch, Uma Musume conversation scenes: <a href="https://game.watch.impress.co.jp/docs/kikaku/1366170.html">game.watch.impress.co.jp</a></li>
    416:     <li id="s7" value="7">Game Watch, Blue Protocol, CEDEC 2020: <a href="https://game.watch.impress.co.jp/docs/news/1275034.html">game.watch.impress.co.jp</a></li>
    417:     <li id="s9" value="9">Game Watch, Gakumas, CEDEC 2024: <a href="https://game.watch.impress.co.jp/docs/kikaku/1617554.html">game.watch.impress.co.jp</a></li>
    418:     <li id="s10" value="10">Anime Shading Plus, face shadow map workflow: <a href="https://erichu33.github.io/ASPDocs/en/articles/face-shadow-map-creation-and-baking-workflow.html">erichu33.github.io</a></li>
    419:     <li id="s11" value="11">Autodesk cel-look modelling tutorial: <a href="https://area.autodesk.jp/column/tutorial/celllook-character-modeling/01-basics-preparation/">part 1</a>, <a href="https://area.autodesk.jp/column/tutorial/celllook-character-modeling/02-edit-rough-model/">2</a>, <a href="https://area.autodesk.jp/column/tutorial/celllook-character-modeling/03-detailed-modeling/">3</a></li>
    420:     <li id="s12" value="12">CGWORLD, Revisions: <a href="https://cgworld.jp/feature/201812-cgw245-rvsns.html">cgworld.jp</a></li>
    421:     <li id="s13" value="13">CGWORLD, Arc System Works feature (partial): <a href="https://cgworld.jp/special-feature/202508-asw.html">cgworld.jp</a></li>
    422:     <li id="s14" value="14">CHARM, a study of VRoid hair: <a href="https://arxiv.org/html/2509.21114v1">arxiv.org</a></li>
    423:     <li id="s16" value="16">Inside Games, Gakumas budgets: <a href="https://www.inside-games.jp/article/2024/08/31/158987.html">inside-games.jp</a></li>
    424:     <li id="s17" value="17">80.lv, long curly anime hair: <a href="https://80.lv/articles/modeling-long-curly-hair-for-a-3d-anime-style-character">80.lv</a></li>
    425:     <li id="s18" value="18">Zenn, a hobbyist's memo on hair and skirts: <a href="https://zenn.dev/link/comments/524655cd70e5a3">zenn.dev</a></li>
    426:     <li id="s19" value="19">Avatar makers on skirts: <a href="https://cgbox.jp/2021/11/09/howtomakeskirt/">cgbox.jp</a>, <a href="https://signyamo.blog/phys-bone_skirt/">signyamo.blog</a>, <a href="https://note.com/airgreen/n/na9eeaf74e6ed">note.com</a></li>
    427:     <li id="s20" value="20">VRM extended colliders: <a href="https://github.com/vrm-c/vrm-specification/blob/master/specification/VRMC_springBone_extended_collider-1.0/README.md">github.com</a></li>
    428:     <li id="s21" value="21">80.lv, cloth simulation for games (2020): <a href="https://80.lv/articles/cloth-simulation-for-games-difficulties-and-current-solutions">80.lv</a></li>
    429:     <li id="s22" value="22">MagicaCloth, preventing penetration: <a href="https://magicasoft.jp/en/prevent-penetration-2/">magicasoft.jp</a></li>
    430:     <li id="s24" value="24">CGWORLD, Soten no Ken: <a href="https://cgworld.jp/feature/201812-cgw245-soten.html">cgworld.jp</a></li>
    431:     <li id="s25" value="25">BlenderGym (GPT-4o and Claude 3.5 era; photometric loss): <a href="https://arxiv.org/html/2504.01786v1">arxiv.org</a></li>
    432:     <li id="s26" value="26">Design2GarmentCode: <a href="https://arxiv.org/html/2412.08603v2">arxiv.org</a></li>
    433:     <li id="s27" value="27">GarmentCodeData: <a href="https://arxiv.org/html/2405.17609v3">arxiv.org</a></li>
    434:     <li id="s28" value="28">AI-to-3D: <a href="https://github.com/hyz317/StdGEN">StdGEN</a>, <a href="https://github.com/zjp-shadow/CharacterGen">CharacterGen</a>, <a href="https://github.com/ShuhongChen/panic3d-anime-reconstruction">PAniC-3D</a></li>
    435:     <li id="s29" value="29">Agents in Blender, practitioner and vendor reports: <a href="https://www.mindstudio.ai/blog/claude-blender-mcp-real-world-performance">mindstudio.ai</a>, <a href="https://www.3daistudio.com/3d-generator-ai-comparison-alternatives-guide/blender-mcp">3daistudio.com</a></li>
    436:     <li id="s30" value="30">CGWORLD, Snack World (search snippet, likely page): <a href="https://cgworld.jp/regular/201805-c-olm.html">cgworld.jp</a></li>
    437:     <li id="s31" value="31">Prices: <a href="https://www.lancers.jp/menu/detail/1238820">Lancers listing</a> (search snippet only, two price tiers)</li>
    438:     <li id="s32" value="32">Bases: <a href="https://raw.githubusercontent.com/animate1978/MB-Lab/master/license.txt">MB-Lab licence</a>, <a href="https://raw.githubusercontent.com/Upliner/CharMorph/master/license.txt">CharMorph licence</a>, <a href="https://genies.com/blog/ready-player-me-shutdown">Ready Player Me shutdown</a>, <a href="https://www.reallusion.com/license/content.html">Reallusion</a></li>
    439:   </ol>
    440: </div>
    441: </body>
    442: </html>

agent

+0:48
reasoning

I'm thinking through how to approach this.