No.73462
Been working on a TUI nhentai downloader. Can download whole artist/group galleries or individual galleries with it. Its not complete but at least it works for all intents and purposes. Still have to work on some error handlings though. Already downloaded 280GB with it.
Also been working on a image browser to go hand-in-hand with nhentai downloader. An app for viewing and searching/filtering your (presumably) large nhentai library.
No.73469
>>73467>low scarcityDoes this mean its abundant?
No.73471
>>73470Some/most programmers are really just copy/paste-ers. They hardly have any idea why they're doing something. Highly prevalent in Indians, I think.
No.73472
>>73469compared to drawing and even music making, yes, its abundant
>>73470>a surprising amount of people that actually work in software can't do something as simple as fizzbuzzi've seen lots of hobby programmers are better than them, who don't work in any programming job but have their own personal projects on github
every kids in the corner can code up a discord bot nowadays, while most employed programmers, especially those in non-tech companies, are below that level
No.73473
>>73472I'm not so sure every kid can do something like "code up a discord bot"- especially when the rise of smartphones has been driving down the computer literacy of the average child for the last ten years. I believe you that there are plenty of hobby programmers out there better than pros who actually got hired, but I still think you're vastly overestimating how many there really are, for some reason.
No.73475
>>73473smartphones are irrelevant if they have the motivation to learn (pcs are also useless if they don't), and video game modding and hacking and related communities contribute to that significantly, in fact lots of young people learn programming starting from that
lots of the "pros" have no motivation other then making money, so they only have very minimal skill
No.73477
>>73462I wrote a python script to download galleries from nhentai back when exhentai shutdown. Still works great I just tested it.
https://hastebin.com/sewifesifa.sqlThis is clean, simple, minimal python correctly written.
It can get a single gallery or get all the galleries from a search result (e.g. nhentai.net/search/?q=chuunibyou+demo+eng
)
No.73478
Think if I were to try and write a gallery downloader I'd figure out a way to automate going through my favorites on ex and downloading the best torrent. Though it seems like it'd be fairly difficult.
No.73479
>>73477I should add a "download from favorite" feature too. Though I don't have a nhentai account, I just have lists of artists/groups/gallery IDs. But maybe for someday when I decide to share this on github or something... Do people even have nhentai accounts?
>downloading the best torrentThat'd need the script to know what 'the best' is, which is very difficult, I think.
No.73565
dang that seems really tough to paint
No.73622
>>73621Post Script:
That's an Ophanim if anyone was wondering. Very kowai..
No.73721
>>73720but (hehe) here it is with a bunch of filters
trying to figure out how to get jiggle to look right, but even if it doesn't look right it's still kind of hypnotic
No.73723
>>73621I too am working on a novel, though mine's not quite so unique in it's plot. it's just about a bunch of humans living on board a giant spaceship after aliens attacked earth and they had to flee.
my favorite bit of lore I've written so far is that the spaceship only has necessary food stocked except for one brand of cola that was a sponsor of the project to build the ship. I've also taken to making up my own brand names like anime does, so it's called Doctor Cola.
No.73758
>>73757I guess if you can get importing example units (from vermintide 1?) working it'd be a first step in making the Total Warhammer units portable, but to take units from completely seperate genres and add them into the game seems like a really daunting task. Although I guess who knows, the units in L4D-like games are fairly simple so maybe a Total War unit fits into the class of units that'd actually be possible to have.
No.73759
>>73758The thing with total warhamemr units is that modeling and animation work is already done. From what I've seen of total warhammer the attack animations while they aren't quite as varied as vermintide there's enough of them to cover atleast a couple of light and heavy attacks for each unit
There's also some good tools for extracting the whole model, skeleton, animations and all directly from the source for total warhammer so getting those files aren't a problem. The texutres might be a tad low resolution for vermintide's first person perspective but that's a secondary concern compared to enemy variety.
I've actually got a decent grasp for how to implement proper emissive and transparent texture layers for models in game.
While tedious the implementation of custom ai for the custom units is mostly straight forward and a lot can be used from existing code since the source scripts for vermintide 2 are available online.
This is the tomb scorpion from total warhammer, I've got it to replace the vermintide 2 pig unit. From what i've been able to tell it's just like the normal pig unit except it doesn't have animations or a ragdoll. The game registers it as an entity that you can kill and hit and whatnot all that's missing is the animations. I should be able to do this with other units/enemies as well but I'm starting with the pig as it doesn't move and only has two states that it every uses; idle and ragdoll(death).
No.73760
>>73758More to your concern I’m really only thinking about going for 5 unit types. Two low health normal units, two high health elite/armored units, and one large health monster unit. Analogous units in vermintide now are the slave rat and clan rat, the stormvermin and plague monk, the rat ogre. Aside from the special enemies and a second monster/large enemy this is the setup for the skaven faction in vermintide 2. The unit roaster for total warhammer is more than varied enough to provide enough units for that sort of setup.
I thought about trying with vermintide 1 first but I’m not sure if anybody got animations working for that one either and I need ask for a steam key to get access to the sdk currently they aren’t available (as in a few weeks ago they said they ran out and need to ask steam for more) but that shouldn’t be a problem as I think a few other active modders in the community have a few keys on hand.
No.73904
>>73723How much do you have written in terms of actual story? Like word-count I guess?
I'm 22k words deep in my first story, and 20k words deep in the second one.
No.73913
Bought a 3D model from Booth. It was surprisingly affordable compared to other models I've seen. $5 compared to the average $40. Now I'm slowly learning how to use unity to import it into a game and stuff. Granted, most everything was pre-configured already. All I really need to do was translate some radial menu stuff, and pose the skeleton to get a nice screenshot of it :>
Not sure if anyone will care, but since I really like the creator and the avatar, here's the link to it:
https://booth.pm/en/items/2852497 No.73919
>>7390410 pages, 3,600-ish words, still in the first chapter. I have quite a bit of the story already planned though. it wasn't a story per se, but I have finished a book before that's about teaching people python.
No.73923
>>73919Right now I'm thinking on trying to have the main character go off into the next act of the story.
Like, all I literally have to do is write "and they set off" and then they set off.
What are your main characters like?
My main character is a bit of an asshole who learns to stop being a cunt.
No.73926
My mini album is almost finished.
pretty much only gotta mix now.
Oh fuck I cant mix!
No.73931
>>73923one's a genki girl who spends all her time making trouble for other people and being incessantly nosy, and the other is a no-nonsense former hikineet that had to stop being one cause of the whole "fleeing to outer space to avoid annihalation" thing. he's usually not fond of socializing but he's forced to work in the hangar of the ship with the genki girl so he appreciates everyone that helps him deal with her shenanigans greatly so he puts up with it. I already know where both of their character arcs are going to end up, essentially. the only other character that's around enough to count as a main is genki girl's dad, who's a gentle giant more or less, except for whenever he's trying to instill manners in his daughter.
No.74257
I haven't worked on my book any since last post but I have my own mini album in the works (though mine's not complex enough to need mixing) and I've slowly been teaching myself how to draw since I'm tired of being an artlet
>>73926time to learn a new skill, friend!
https://ledgernote.com/columns/mixing-mastering/audio-mixing-for-dummies/ No.74400
>>74256The game keeps rejecting this new data file, I might have to ask someone if it's possible. I wonder if they put in some sort of file check for the localization file to check for tampering since they're supposedly working on an international version.
Now I kind of want to learn Live2D, though. Wallpaper Engine made it look pretty appealing and it shares a lot of similarity with 3D animation.
No.75611
>>74400This is a massive headache, but I wonder if I can switch the free skin for the paid skins somehow. I did a quick switch and I was able to get to the character before the game crashed. This means that unlike the localization file which gets outright rejected (wtf?) tinkering is possible for these files.
Why the fuck would they block people from translating the UI?
No.75694
File:.png (589.31 KB,1280x720)
made a bunch of nonsensical threads on/jp/
No.77397
>>77396So I open it up in Blender on a hunch and this old 3DS Max version wasn't displaying the weight fully accurately. Again, I can't see a method or script to copy specific vert weights between meshes so I have to copy and paste these. 3DS Max was rounding these up, but did have the full proper weight stored somewhere...
sigh
No.77450
>>77415i thought you quit programming altogether
No.80888
>>77415Almost forgot I posted this here. You can mess around with it already. I think it might work online too but I've never tested it.
https://github.com/meifuku/AFGEngine/releases/download/v0.25/afgengine.v0.25.zip No.82462
Hi Kissuanons, I don't post here too often but I'm poking around to find answers for a problem I'm having.
I want to do a creative project. I have plenty of time and energy, but I forgot how to be creative. I used to have so many ideas as a kid/teenager but I didn't know how to execute them, nor did I have the time and focus that I have now. So my question is, how do I get the creative juices flowing? I'll take anything, tips, books, any resources or insights you guys have.
My medium is choice is drawing/art. My skills are fairly beginner but I realized that aimlessly studying abstract fundamentals (human anatomy, for example) won't get me as far as using them as tools for the true purpose of creativity. The picture isn't mine, just some eye candy, I'd love to reach a level like this some day, though..
Thanks guys.
No.82466
>>82462Here, the best thing to do is to just do it. Even if you feel tired and burned out, you keep pushing forward.
I usually smoke a bowl of pot (dude weed lmao) and it usually helps stimulate my creative juices (I write) and it helps me push forward.
In short just push forward niggy and slowly work at it.
No.82473
>>82462I usually start by just arranging some basic elements according to some idea of what I want to emulate. The idea is usually not so important, because at some point the elements I'm playing with will fall together and tell me how to procede further. Most of the time I end up with something only loosely related to the idea I started out with.
No.82477
>>82462Finding inspiration would be the best way, but it's a pretty abstract thing.
I really have no idea. For thousands of years people had religion, drugs, and/or undiagnosed mental issues that provided it. It's going to be an uphill battle if you don't have inspiration
No.82484
>>82462I'm currently trying to get into drawing, the one thing that drives me is looking at beautiful illustrations and thinking "I need to be able to draw like this". Has kept me going so far (2 weeks). Hope you find your 書きたい事 too, Anonymous.
No.82485
>>82462Do the first thing that pops into your head, even if you think it's the most worthless, nothing idea you can think of. It could even be a singular word like "door" or something. All that matters is that you have some kind of starting point. You'd be amazed how easy it is to come up with ideas once you have that.
No.82513
>>82484Out of curiosity, how do approach your practice?
No.82515
>>82514Actually, translating the title wouldn't make much sense since the books are going to be in Japanese. Here are the Japanese titles in text:
- 美少女作画 ゼロから学ぶプロの技 by フライ げみ けーしん 田中 寛崇 U35
- 増補 エロマンガ・スタディーズ by 永山 薫
- きらめく瞳の描き方 How To Draw Eyes That Sparkle by 玄光社
- マンガで使える 食の描き方 by ポストメディア編集部
- 女の子の人体パーツの描き方 セクシーに魅せる by うめ丸ぶんぼん
- かわいいファンタジーイラストが描ける本 職業, 種族のデザイン衣装144パターン by 藍飴
- 動物×人擬人化アイデアテクニックケモミミキャ
ラクターデザインブック by しゅがお
- アジアンファンタジーな女の子のキャラクターデ
ザインブック by 紅木春
- デジタルツールで描くキャラクターを引き立てる
髪の描き方 by スタジオ・ハードデラックス
- 美男子作画 ゼロから学ぶプロの技 by 秋赤音 えびも キナコ 佐木 郁 しきみ べっこ ホノジロ トヲジ
- スカート描き方参考帖（丈長め編） = Long skirt drawing method reference book by Rakugaki lab
- ファンタジー衣装の描き方 by もくり
- 物語を動かすキャラクターデザインとイラストの
描き方 by スタジオ・ハードデラックス
- メルヘンでかわいい女の子の衣装デザインカタロ
グ by 佐倉おりこ
- 才能はいらない イラストで食う技術 by ニリツ,Niritsu
No.82516
>>824621. After scanning in your picture, adjust the levels in an image editor. It will look so much better!
2. Use Ebony pencils. They sharpen so easily and they're nice and dark. And while you're at it get a kneaded eraser if you haven't already.
3. Just Draw by Feng Zhu will get your mileage up. Try tracing over a picture if you're having trouble with textures.
No.82521
>>82516how do you know he's drawing traditionally?
No.82522
>>82521I don't. I guessed based on the book he picked.
No.83099
>>82462 Hi guys, this anon here. I've made some progress on my quest and I'll share what I've learned. My big discovery was the importance of writing all your ideas down, and putting aside some hours of the day every once in awhile just for brainstorming. I write down everything that comes to my head that could potentially be something I want to draw: characters, objects, abstract ideas, whatever. I can get inspiration for this journal by just exploring fiction, the internet, anime, manga, reading books, talking to people, walking around: pretty much anything can be a springboard for a picture idea, since we are idea generating creatures by our very nature. Before I let them all pass by but now I write basically everything down and shifted my mindset to one of thinking in these terms. Also pinterest is a great place to find images of people, places, and objects.
I still wouldn't be able to turn this hodge podge of random ideas into an actual story with a theme, but I feel like I have some raw materials for making illustrations. I also discovered automatic/meditative drawing, which I recommend for stoking creativity and expectations-free drawing. It's a good warmup.
https://www.youtube.com/watch?v=MJYGFwGhHnA>>82484How has your progress been, Anonymous? I hope you are still going strong.
No.83216
I did a thing.
No.83527
>>83511That's a great idea in how simple it is despite probably being difficult. Add some music and make it a rhythm game
No.83632
>>83629Human eyes can only see in 512p.
No.83652
>>83651megu megu megu megu megu megu
megu megu megu megu megu megu
megu megu megu megu megu megu
No.83680
>>83663Don't play Melty Blood much, what's modded here?
No.83683
>>83680Many of those moves can't originally combo.
No.83874
File:k.mp4 (4.69 MB,640x480)
>>83683>>83680I guess it wasn't very crazy. This one is more obvious.
No.84488
>>83663>>83683creating a world where Sacchin can win!
No.84491
>>84488she's a top tier in melty already
No.84497
I have a feeling this is going to be a headache to do because I'm not good at messing with bones, but it's worth a shot so I'm going to see if I can make the cute twin girls into the ridable dog mount replacements in Monster Hunter Rise.
>>83705Really cool!
No.84498
>>84497>I'm going to see if I can make the cute twin girls into the ridable dog mount replacements in Monster Hunter Rise.Phrases I never imagined seeing on Kissu for 1000
No.84597
>>84498I tried to do it in Monster Hunter World but ran out of motivation. Thankfully the modding pipeline for MHR is significantly less of a headache to me, and I still have the naked body I spent so much time making for World so it's a lot less involved this time.
Anyway...
I decided to use the player body instead of NPC body. I thought the bones would be the same, but they're not. I really don't want to start messing with two different skeletons so I had to start over.
Positioning the hands is too much work so I replaced them with big fluffy cat paws from a Palico weapon which fits better anyway. Body still looks weird bent this way, though...
No.84600
>>84597>Body still looks weird bent this way, though...needs a saddle to hide the janky body.
No.86160
>>86158What's custom here? I haven't played the game so I can't tell.
>>86159Ah, good ol' exploding models. "Inverse pose matrix" sounds really complicated and not something I'm familiar with. With modding I really rely on smart people to do all the decoding stuff. I don't know how those guys do it, they're wizards.
No.86161
>>86160The waving part is an added in animation, I made it so that whenever the unit's state machine gets passed an "idle" event it just waves instead. I thought it was kinda obvious given how stiff the model becomes once it waves...
The matrix is kinda complicated. It's the transformation matrix that tells the game what a particular bone's rest pose(orientation) is in that particular bone's, "bone space". Bone space being a coordinate system with it's origin and orientation relative to a specific perspective at the bone in question.
I'm about 85% sure this is my issue as the particular file format that Fatshark(vermintide 2 devs) use in their workflow is a bsi and I've narrowed down the this issue to matrices called "inv_bind_matrix". From what I can tell it's supposed to be the bsi's version of inverse pose matrix.
No.86234
>>86161Wow, that's really cool. Animations are certainly a very advanced modding thing it's impressive when anyone can do it. Did it replace another one or did you manage to get a custom event/behavior in there somehow?
>I thought it was kinda obvious given how stiff the model becomes once it waves...I've learned never to make assumptions about this stuff.
That matrix explanation makes sense, I never knew there was a word for it as I'd just say "default position" or something, but I guess that's too simple.
No.86249
>>86234>>86234> did you manage to get a custom event/behavior in there somehowI'm technically replacing the idle animation by just intercepting the idle event call but given that I'm doing this by swapping state machines between the custom and vanilla one, it's really not all that hard to have the idle animation exist side by side with the waving one. I control the conditions for swapping with lua so in theory I could make this only occur when a player dies, or loads in game, ect.
>>862361000 bones seems a little much...
No.87331
Currently in the process of designing a slide rule. Model is mostly done aside from needing to "engrave" the actual marking and lettering onto the model. Likewise, some of the dimensions may need to be adjusted to improve tolerances.
I would be working on the engraving, however, after referring to some scale models, I've determined that the scale markings are FAR too small for practical reading and need to be made larger. So, right now I've been manually resizing each tick, one by one for each scale. The sizing of the numbering likely will need some manual adjustment as well, but I'll get to that when I get to that...
Once I have final production models ready, I'd be happy to provide the source files so anyone can 3D print their own. Be warned, however, unless you want to make a giant version your 3D printer will need to be capable of at least ~0.001 inch print resolution, otherwise the tick marks won't come out correct.
No.87332
>>87331Also: I fucked up massively. I was fighting scaling artefacts trying to get dimensionally accurate tick marks rather than working with the full-size version. Whoops.
No.87333
Still undecided on whether to include/design a case. I really want to make one, but at the same time it might look a little wonky, so... I dunno. I still need to work on the slide rule itself before I start thinking about a case.
I also really want to design a quick reference insert ruler as well, but that should be quite a bit easier. The 3D model would literally just be a rectangle. The harder bit is arrangement of things and what to include and what not to include. So, really, the biggest hurdle is just design. For example, should I make it double sided to accommodate more information?
No.87335
>>87331Cool. Is the moving part between the region from the B to C scale inclusive?
No.87556
>>87555with this the vermin will be too enamored to see the sword coming for their skulls
No.87668
>>87667Why would you need permission to distribute something like that?
No.87669
>>87668Using someone else's assets for your own stuff without permission is a pretty major no-no, it's like tracing art in 2D. Add in the paid mods fiasco and so many modders having patreons now (which rubs me the wrong way) and it's just disrespectful.
I have my Monster Hunter body I'm going to mess around with and import, and that's something I largely did on my own (building on Capcom's body and who cares about that..)
No.87689
>>87687There wasn't anything extra to see... It's just easier to see how absolutely fubar'd my print is at a glance by cropping it. The supports look all crumbly and bloated when they should be straight cylindrical supports, meanwhile my parts those supports were supposed to be supporting are nowhere to be found...
No.88362
>>88304Waaah this part cracked on the edge when I was trying to snip off the supports
;_;I think I need to see if I can alter the support density and support height so that it's easier to remove from the raft, and also so that no excess resin gets trapped inside the supports leaving globs that are un-snip-able.
No.88641
>>88640and this is a manual layer on top of it trying to balance it out and make it look more natural.
Black = full shininess like a robot soaked in wax
White = No shine at all like dirt
Nothing is full white or black of course
No.88675
>>88674Cast off my shackles from this mortal plane, and embrace the 2D VR world!
No.88709
Spent a few hours practicing soldering today so I'm prepared for when I have to start soldering a million bajillion things together for my SlimeVR trackers. Just as a little test for myself, I soldered two LEDs in parallel with resistors and then tons of little snippets of wire between each for extra soldering practice. My joints were pretty bad, but my LEDs did work when attached to a power supply! Gonna frame my stupid soldering attempt and tell my future children Thomas Edison made it.
Unfortunately, I didn't get any heat shrink practicing done because I neither own any lighter nor know how to use them. So, after my soldering I bought one at a convenience store but I'm really bad at getting it to light. Best I can do is get it to flicker, but most of the time nothing happens.
No.88763
>>87667>>88641perverted elden ring modding update
dickgirl warning!There's a weird texture issue on the
enemy, I don't know what's causing it but it's not supposed to look like that. Also I forgot to remove some overlapping body parts (feet most noticeably since you can sometimes see bare feet)I'll figure it out tomorrow maybe.
And yeah, I do have a morph for turning the penis into a vagina. I guess if I spend enough time with this I should try to overcome my fear of releasing a public mod...
No.88764
I'm writing a visual novel about a schizophrenic man's descent into madness as he fights invisible creatures from another dimension. It's really hard to write something this big. The sheer volume of text I need to write demands a lot of side-stories, subplots and fluff that I'm just not used to doing.
No.88780
Finally got all my components in today. It would seem my 3D printed parts are about 0.5mm undersized. So, I'm left with two options: reprint everything and potentially go through the pain of sanding those new parts down so they fit flush and hope and pray the components fit. Or, just press on and sand everything again so that the components fit.
I do not like either of these options. 3D printing really is a pain in the ass. All these weirdos who want 3D printing to go mainstream need to sort their shit out. If I can't figure out what the HELL I'm doing wrong, how the FUCK is some stupid norm going to understand what they're doing?
No.88815
>>88780It's too bad you can't make touch-ups like that in the 3D printer. Maybe that's what it needs to go mainstream.
No.88817
>>88815What I wrote was a little ambiguous. The 3D printed part is undersized so that the PCBs that need to go into it are too large to fit. This sort of thing would not be able to be fixed by putting it in a 3D printer because material needs to be removed not added.
One of the many annoyances when dealing with 3D printers is that they have a myriad of settings that are not immediately understood as being better or worse. For instance, would turning up the fan speed be better or worse? Is wall thickness better thicker or thinner? What's the best infill setting? How hot should the extruder be? When arranging parts to be printed does their distance matter? Does part orientation matter? And many, many more questions and settings. Another thing among the general annoyances is that the 3D printer filament can go bad! If stored somewhere with even a little humidity, the plastic may take on water, resulting in a failed print.
No.89273
>>89272Dunno. It's just a funny, recognizable name. There's nothing slime to it.
No.89276
Made two more that also work. Seems like it takes about 2 and half hours per one.
No.89295
They're multiplying...
I now technically have a full set of trackers, but of course I'm going to make more. Making a tracker with an extension was really painful. I spent an hour just going back and forth between reading what pins I needed to set high/low, unsoldering and resoldering, before finally reading a definitive answer. For normal people, it should be fine, but I chose a peculiar IMU PCB design. Anyways I also had a scare that my IMUs might have been dead because they weren't showing up at their I2C addresses. Fortunately, they were perfectly fine. There was just a short between some of the pins because I didn't trim the wire leads after soldering (for fear of needing to pull them out). For a while I also messed around with some connectors before deciding they suck and that I hate them. They were yucky PVC wires, and the ordering of the colored wires was all wrong from how I wanted; I tried pulling them out and reordering them, but they kept coming out irreparably broken compared to their initial, unbreakable strength. My wiring also was a bit chaotic. I needed to cram three (count em: 3!!!) wires into a single PCB through hole. That was hellish. My wires coming out were also less than ideal. I didn't consider cable length difference, and had 4 wires of the same length coming out, so one wire sort of holds back the rest (and it's the one that needed to be crammed into the through hole!!). I thought about extending it by splicing in a small section and heat shrinking it, but it would ruin the aesthetics of it, and cutting new wires and resoldering 4/5 others in their places was just too high a commitment for a purely aesthetic concern. Finally, I became concerned when the SlimeVR server reported my batteries staying at 4.3V and the charge controller not turning off at 100%. Turns out, the software wasn't accurately reading the voltage at all. My handy new multimeter reported that the batteries were actually at ~4.05V, not >4.3V. So, that's strange. What's especially strange is that that trend of being completely off, was consistent across 5 devices, meaning it's probably a software bug.
No.89296
Also! I tried my straps and they're junk! They feel alright and stuff, but they're too big to fit my little aux tracker, as well as not being a particularly good fit. So... I'll need to get some replacements that don't suck. One person showed off a neat concept I really like: quick attaching/detaching straps. Instead of working like a belt loop, they're more like an open hook. One side has a hook, and the other is securely looped and used for sizing. Then, the hook side can freely attach and detach with ease rather than needing to undo a belt every time you want to put them on.
No.89470
Not sure if this fits in this thread but I evaluated how feasible NVENC is for anime encoding. I found some quality benchmark results suggesting that NVENC can perform better than software encoders but none of them use anime content, so I did a micro benchmark myself.
I selected 2 1-minute segments from an episode of Aharen-san for evaluation, one is bright and low motion, another is dark with high motion to evaluate typical scenarios. They were encoded by several encoders (x264 slower preset, x265 slow preset, NVENC AVC, NVENC HEVC) at different bitrates (CRF from 17 to 28), and the result videos were tested for PSNR, SSIM, and VMAF scores.
My result showed that for both videos, the quality ranking are:
PSNR, SSIM: x265 > x264 = NVENC HEVC > NVENC AVC
VMAF: x265 > x264 > NVENC HEVC >= NVENC AVC (the margin is narrower for the bright and low motion video)
So the result showed that NVENC HEVC is on par with x264 at best, although it's 5x faster on my computer. For NVENC, using 2-pass ABR instead of QP slightly increases the VMAF score, but it doesn't change the ranking and quality decreases when bitrate is too low.
No.89496
To show that the rational numbers in (0, 1) take up 0% of the length of (0, 1) according to the Lebesgue measure, you need to be able to construct a sequence of open intervals with arbitrarily small total length containing all rational numbers in (0, 1). A simple and common way to do this is to make the rational numbers in (0, 1) into a sequence and take ε/2ⁿ around each rational in the sequence. Some of these intervals overlap, making it more difficult to figure out what the total covered length is. That's not really a problem; all that's needed for the proof is the upper bound on the length. But I wanted to figure out a way of doing it where I could calculate the actual total length covered by the intervals. This is what I came up with.
Start with the open interval (0, 1) and n = 2. Take the outer 1/n² of the interval, 1/(2n²) from each side, including the boundary points between the taken and untaken parts. Split the remaining interval into n²-1 parts and take the boundary points between the parts. In each of the n²-1 remaining parts, repeat the process with n increased by 1. Continue to infinity, and the set of points you have taken will be a union of disjoint open intervals which contains all the rational numbers in (0, 1) and has a total length of 1/2. If you start with an n other than 2, you get 1/n of (0, 1), but still get all the rational numbers.
No.89882
>>89496there isn't a single part of this I understand.
No.90115
>>90109Huh, are you following a guide or something? What method are you using to build? Looks like you're going polygon by polygon?
No.90121
>>90115Yep. I'm slowly following a tutorial. I wouldn't say I'm going polygon by polygon. More like, gradually shaping things out by moving and scaling vertices, and then adding more detail by creating additional edge loops. Where limbs and feet need to be added, they're extruded out from a set of vertices.
This is the tutorial series:
https://www.youtube.com/watch?v=5JA0f8ZgIw0I've had to go a bit off script because they're working on creating a female body whereas I want to make a male one. I think I'm coming along okay so far.
No.90124
>>90121Holy crap, 200k views. I forgot VR Chat was so popular. I really need to get into that so I can sell stuff to fellow perverts.
I'd probably go for a tutorial from a more traditional artist instead of a VR Chat person. I'd volunteer to share some tutorial videos, but just this one blender course from a private tracker is 40gb......
I'm much more of a sculpting person, I hate moving verts around like that, but there's no escaping it if you're doing extreme low poly stuff like that. Have you tried messing around in Sculpt mode? Blender's isn't the best, but it's pretty damn good.
Testing if I can upload this quality-destroyed and rescaled video. I'll probably delete it in a few weeks if I can since it's pretty big.
No.90128
>>90124You'll notice the number of views drops off pretty rapidly as people realize that stuff is beyond their ability... The next part has 78K, then 44K, then 26K...
>there's no escaping it if you're doing extreme low poly stuff like that.That's sort of one of the issues with VRChat, it's not necessarily that I'm planning on making something low poly -- I'll add more detail as I go along, and do some smoothing -- but when it comes to avatars, it's best that they're optimized and don't have an excessive number of polygons. A lot of people have insanely detailed avatars with more than 100K polygons, multiple outfits, heavy shaders, dozens of dynamic bones for hair or tails, and an over all size nearing 200MB! For Oculus Quest compatible avatars, for instance, there is a hard limit of 20K polygons, and a limit on dynamic bones, predefined shaders, and a recommendation of only one material. Needless to say, although I'd like to eventually end up with something with a fair amount of detail, I'd rather not fall down the trap of creating something insanely unoptimized either. I'd like other people to see me after all.
>Have you tried messing around in Sculpt mode?I haven't... I'm not very good at those sorts of more artistic detail work stuff. I have a Wacom tablet which might be helpful but it's mostly gathering dust when it comes to this... I suppose I can imagine details, but it's hard to transfer that to reality, at least for someone at a beginner level like me. Likewise, the whole not wanting to create a million polygons because I want something a bit more round or whatnot applies again, I think.
No.90132
>>90128>100K polygons, multiple outfits, heavy shaders, dozens of dynamic bones for hair or tails, and an over all size nearing 200MB!Wow, 200MB for an avatar. Yeah, that's absurd.
How does VR Chat handle this? Do you get a window to consent to downloading so much bloat? Are there rooms/worlds that bar that stuff? Most of what I've seen seems to be anime-style and furry stuff and detail really isn't needed there, but the people are probably running around with 4k or even 8k texture maps for hands and other stupid stuff. Optimization really isn't a taught skill unfortunately and you see it fairly often in modding, too.
Sculpting is something you can put off since you're dealing with such basic shapes, but it's also really fun to mess around with. Well, if you enjoy that kind of thing. I love it, personally, but there's also people that hate it and prefer moving verts around or typing coordinates in.
No.90148
>>90132>How does VR Chat handle this?Poorly. Due to there being no hard limits really being enforced as to how optimized avatars are, even with high end graphics cards, VRChat basically runs at 40 FPS in VR. Likewise, because of these avatar loads, VRChat caches worlds and avatars, so after a few days, you can pretty easily fill up 15GB just from avatars and worlds.
>Do you get a window to consent to downloading so much bloat?Thankfully, yes. There's varying levels of granularity in the safety settings, which allow toggling animations, shaders, SFX, etc. per user "trust level" (somewhat arbitrary measure based on a combination of playtime, uploaded content, friends, etc.). There's also global settings for setting the maximum avatar size to download, as well as the minimum optimization level of avatar to download (optimization level is kind of useless since most avatars rank as "poor"). As a result, if an avatar wouldn't load because of your safety settings or otherwise, the person would turn into a default robot avatar, but they've recently added "fallback avatars" which are essentially meant to be highly optimized avatars to load when the main avatar cannot.
If it wasn't already clear, VRChat is basically a complete mess when it comes to avatar detail levels is basically the wild west with zero rules enforced about them other than the ground rules about not having NSFW avatars in public worlds, and not having intentionally malicious avatars, such as ones with shaders that distort the views of other people, or fill peoples screens. On the one hand, it's very liberating, but on the other it means a lot of duplication of work because of these VERY, VERY loose suggestions rather than rules.
No.90177
>>90154If you're looking to improve whether it's drawing or modelling one thing that's the best to learn is anatomy.
No.90184
>>90177Yeah, probably. Anatomy is one of those things you can tell is subtly off, but is hard to really pin down without use of various references I suppose. I'm having a slightly difficult time simply because I'm not sure where is best to look. I think I've gotten the broad strokes of the hands down, but now that I'm looking to refine my body, I'm noticing it looks slightly off. I'm having a bit of trouble finding male reference images that aren't absolutely jacked monsters compared to the more slim and average physique I'm trying to go for.
No.90191
>>90190Thank you for the anatomy references. Upon reflection, though, more realistic body types are incredibly uncute. Just phenomenally not nice to look at. Like really, really not what I want at all.
I suppose this might end up being a first attempt for learning, but without anything usable, unfortunately. I did however, save along the way so I'll see about working my earlier body into something more aesthetic and less... uh... masculine.
No.90329
>>90191Oh, be glad you haven't seen the 3D models with all the various skeleton and muscle systems and veins and stuff. Yeah, it's definitely not cute.
It does help a lot to know where the muscles are and how stuff flows, though. Also, to know what is uncute also helps you know what is cute, since you can see it as a lack of the uncute stuff.
The disappointing thing about learning 3D modeling is when you learn that there's no English tutorials from experts or experienced people in making pretty or cute things. It's all monsters and gore and gross stuff because the people work in Hollywood or Western video games and it's all about the violence and monsters.
It might be changing slowly because of stuff like VR Chat, but not at any reasonable speed.
No.90332
>>90329I've progressed a little bit since last time I posted. Still struggling a bit with cuteness vs realism, but I think I'm mostly at a place where I'm fairly happy with how things look. I haven't gotten particularly far so I'm still at a point where I'm at liberty to move stuff around if I don't like it.
Slight uncanny kigurumi feel, but hopefully that fades with shading, texture work, and proper hair instead of being a baldy.
No.90333
If you ever want to put your things into a 3D game engine to do things then give me the "plot" and I'll do it. I'm trained as a gaym programmer and it would take me like a few hours or less to do it.
Typically just pulling or buying assets from
https://sketchfab.com/3d-models/sharp-rock-10-91909721747b46679ee641bb1ffe2ee1and watching Unity graphics tutorials
https://www.youtube.com/watch?v=12gkcdLc77sSkeleton 2d/3d rigging
https://www.youtube.com/watch?v=eXIuizGzY2A No.90334
>>90332Heh, you gave him (her?) pants already?
Well, I got my tablet's cable yesterday and started messing with it. Man, it's hard to get used to, but at the same time it already feels natural.
I should probably get back to messing around with Elden Ring stuff and maybe finally release a mod, but this is fun. Pressure sensitivity doesn't matter much at this early stage of the body, but the positioning and accuracy is really nice.
Kinda weird timing with two people here working on 3D bodies with two vastly different workflows.
>>90333Yeah, I don't know really know what I'll do with my 3D stuff, I'd like to start making money from it after messing with things for a decade. The western porn game market is such a low bar that it's at ground level, but furry and Japanese-style models in VR stuff also sell for a pretty penny.
No.90335
>>90334Make a futa model, make it one of those easy 2D action scrollers. Give me some animations. Split it 30:70
No.90340
Found out about "JXL art" which is what you get when you configure JPEG XL's prediction method by hand and then generate a small file where all the differences from the predicted pixels are set to zero.
https://jxl-art.surma.technology/https://jxl-art.surma.technology/wtf.htmlPic related is one of them converted to PNG. As a .jxl file it was 49 bytes. Here it is as a data URI:
data:image/jxl;base64,/wr6H0GRCAYBAJQASzhDtjn
tAfQ2SC2sBHg4hoawkPkYA
rmGyJcHM7wW6hqivjxIMA=
=
I posted some others at
http://z5lcip4dafatwwa6hvyibi
zpzwycvwp67cjga3hzjhxh
wvuyaqavxnid.onion/tech/1147 / http://l7jqnz3yfe2wtwietafoie
admgqbu7dcmzmey63ktbjt
xal3he4a.b32.i2p/tech/1147
There's some stuff other people have done here, a lot of it more impressive:
https://jpegxl.info/art/ No.90600
Man, feet are hard. I'm not looking forward to hands.
>>90335I'm a long ways off from doing something like that. Yeah, there's premade assets and stuff but you still need to assemble them all in the world.
I just know basic female fantasy body stuff, not environments and animation. Well, I know weighting which is part of animation, but haven't built rigs before, just tinkered with pre-existing ones.
No.90764
>>90600Never sculpted a foot in my life and this took forever. But, I can't get bogged down in one body part since this is is a base body that I'll retopologize later on, so I think I'll stop here for now.
I used a bunch of 2D images as references, which probably made it a lot more complicated. I'm also not a foot guy so I don't know what's attractive, but it seems like rounded shapes were the way to go since that's how it works for every other part of the body that people find feminine and attractive.
Also, like 99% of 2D feet images are from the bottom so the top is a bit of a mystery. I'll have to figure that out later, but it seems pretty rounded and plain so I don't think it'll be much work.
No.90765
>>90764>rounded shapeI like foot fingers that look more like hand fingers more than round ones.
No.90766
>>90764Apparently pointy heels are what they want, that's what I have heard.
No.90769
>>90765>>90766If you have examples could you post them in
>>>/ec/4410 please? Or reply to the ones already there that have what you mean
No.90770
>>90769Hm, long and slender ones that aren't too round at the tip:
>>>/ec/4740>>>/ec/6551>>>/ec/6939>>>/ec/4433and not short and very round ones like:
>>>/ec/6457>>>/ec/7203 No.90784
>>90765>>90766>>90769>>90770Do
NOT listen to any of these
NASTY TROLLS please
>>90764 No.90787
>>90785I have been unfailingly polite.
No.90987
>>90934Fixed up my arms, but it would seem like I need to adjust the proportions of everything to get everything working correct to use as an avatar. Right was before adjusting proportions, notably my real arms are way higher than my avatar's arms. Left is with the torso stretched even more, and the head, neck, and arms raised, which seems to have fixed it, but now the bottom half of the torso is a bit too high and the legs need adjusting. I'm pretty sure I know how to do all that.
One thing I'm really not sure how to do properly however is getting a transparent scanline effect across the monitor face. You can see it there on the eyes on the left, but that's just from applying a texture to the eyes, which does work, but it has problems. Notably, the eye blendshapes deform the eyes into various shapes (e.g. =3 -> X3, @.@, >w<), so that scanline texture gets stretched and squished and doesn't look right. I was thinking maybe I could create a plane and put it just over the face and that would be fine, but it kinda seems like it wouldn't work like that. The scanline texture gets all messed up from the alpha channel somehow. Instead of it just being a line, it's got these sort of scaling artefacts with what looks like a black layer underneath the a blurry scaled version of the gray scanline line, but with regular transparency surrounding it. It's like when you have a layer with transparency with another layer behind it, and then you merge it with the layer below that, and then color select the background and delete it but you're left with this sort of messy edge where the layer that had some transparent colored pixels were.
I just want to have normal looking lines so that I can apply that across the face so every expression has proper looking scanlines... Maybe someone knows how to fix that?
My only other issue is that I think somewhere along the line something I did kinda nuked my expression controller for VRChat, because although my blendshapes and visemes are intact, my actual expression controllers in Unity are completely borked and missing all of the things for how you control expressions. Even the VRChat SDK seems kinda broken. The examples which should normally contain example expression controllers are missing everything related to expressions and attempting to copy in the controller from a separate Unity project did show the animations, but it was missing the hand gesture part. I even went to check that project by itself and now it seems it's gesture are also broken. What the HECK. Anyways, I think the gesture thing being broken is mainly because of the CATS Blender plugin. For whatever reason it decided to combine a bunch of my bones, which has been the source of many of my problems thus far. I tried recreating my finger bones, but to no effect. I think I might need to rename them to the finger bone naming the avatar originally had, but I don't see why that would make everything suddenly appear in Unity, but I'm kinda grasping at straws and just want my expressions to work and I really don't want to have to re-edit the original model from scratch again...
No.90988
>>90987>>I just want to have normal looking lines so that I can apply that across the face so every expression has proper looking scanlines... Maybe someone knows how to fix that?I think I might be the only other 3D guy here and I have no idea.
I'm guessing a "blendshape" is when the vertices move around to different locations to change the shape of the mesh? I think blender calls them shape keys. I don't understand why it would mess up the texture like that, maybe look at the rendering and texture maps. That stuff gives me a headache, though, I've purposely avoided learning any rendering stuff because it's like looking at a spreadsheet.
No.90989
>>90988>I don't understand why it would mess up the texture like thatIt's because the mesh gets deformed. I guess you could think about it like a having a rectangle with 6 vertices and the texture is lines going across it. If the top corner vertices get moved down, the lines will become like an triangle rather than staying horizontal.
I still think that somehow applying a texture to a plane would work I just have no idea how to do it properly because I want it to be semitransparent and having lines by using an alpha channel causes problems. I may have to just have like 20 planes stretched into lines to get it to look correct, but if it works that's what matters.
No.90991
>>90989Well, I'm not sure how you're planning to animate it, or if this is something animated at all, but you could always duplicate sets of eyes with their own textures tailored for each shape, and then your "blendshape" thing grows one set of eyes to full size while the others are infinitely small.
I should probably look into how the 2D face texture animations work on 3D one of these days...
No.91112
>>91028Oh. Yeah, that looks like a basic texture warping since the mesh itself was warping. I thought you meant something more elaborate was happening since you mentioned alphas and stuff. You can have multiple UV sets on a mesh, but I don't know if that's compatible in Unity/VR Chat and I have no idea how to make them.
Anyway, been working on my model while watching anatomy videos and stuff. Anatomy is so hard, and the hands were not fun. I'm not going for realism, but you still need to follow basic rules or it looks wrong and your subconscious picks up on it.
Model from 2 days ago on the right, model right now on the left. The head is a prebuilt head and I haven't done anything with it, it's just there so I can try to figure out how the head is supposed to work. I don't think the 7-8 head height rule for bodies works for anime-style heads. Do I give it any neck detail at all? What about... BLEH!
I don't think nipple-less 3D model boobs are NSFW, but just in case people are sensitive to it I'll put it behind NSFW
No.91148
>>91112Those hips are NSFW.
No.91477
>>91112I am beginning the retopology process, starting with the feet because they're so freaking ANNOYING. GOD I HATE FEET MODELING!!
I can't even tell if this is good or not, so I might have to do it again. Toes don't have a very wide range of motion, though, so it should be fine. I definitely need to flatten the top of these afterwards so the nails can be placed on top.
No.91808
>>91807They're looking pretty good, don't even see anything that would indicate painting the wrong spot accidentally.
No.92105
>>92104...and this is what it looks like now. God damn, I really, really, really, really HATE HATE HATE
HATE HATE HATE feet topology. In hindsight maybe I should have started elsewhere on the body, but at least now I'm back in the hang of getting mental imagery of squares in my short term memory.
I looked at a bunch of 3D models of feet, but they all seem to be done in a relatively rudimentary manner with, understandably, far more focus on the rest of the body. Maybe this is a sign I should make foot fetish art? Wait, for the big bucks I do furry foot fetish 3D art!
No, that won't work because I don't actually know what makes furries or feet attractive. I think the gentle slopes and curves and the spherical shapes seem to be great, it follows the stuff that people find femininely attractive elsewhere on the body. But, the responses earlier show it's not that simple.
If you do want to talk about attractive 2D feet so I can try to understand, try this thread
>>>/ec/4410 while posting examples. It's still a bit of a mystery to me.
Oh yeah, and the toes look pretty basic now, I ended up redoing them in a uniform manner and will give them individual sculpting later on.
No.92154
>>92105I made a post in that thread pointing out the feet that are good.
>>>/ec/8425See also this thread:
>>>/ec/7648 No.92815
>>92799I can definitely see how that's a hobby with all the effort and materials required there. I don't think I'd be very good at detailing things since my hands have always been a little shaky
No.93525
>>93185>>93186Damn, nice work.
No.93607
Got sidetracked by some games and Lala, but back to sculpting.
I'm like 40% happy with the feet and don't want to obsess over them, so I moved on to lower leg. After messing around with the geometry I looked at a bunch of leg diagrams and remembered that there's actually two big calf muscles back there. It's something you never seen in real life in women and very rare in men, but realism is not my goal so I decided to do a danbooru search for "muscular_woman" and "butt" because it seemed like the easiest way to see the back of legs. ("legs" is not a very good tag)
Of the 90 pages I did see about four images with the distinction made, with one of them being a Kemono Friend of all things, so I decided that it can look good. (probably shouldn't have given myself that restriction in hindsight).
The vast majority of the time you see a 3D model of something, it's a straight grid pattern on the limbs, with an occasional breakup on the knees and elbows. That's really not what I want because I want clear lines between muscles so it's easier to morph and would have a better shape at exaggerated proportions.
Bleh. It looks really crappy here, but you can at least see that I'm trying to curve things around the muscle's shape instead of it being a straight line down.
This is giving me a lot more work, but the goal is that I have one body and I will manipulate it to create bodies that are thin, fat, muscular and so on. Skinny and fat are both really really easy, but muscular... Well, at least I'm not going after realism or I'd be in serious trouble.
I decided that it's too hard to draw muscles on a flat leg, so after taking this screenshot I'm going back in zbrush to sculpt something to "trace" with retopology.
No.93608
>>93523What is this exactly? Reminds me of Melty Blood, but I don't actually play it so I don't know what's different or new here.
No.93609
>>93608A base fighting game engine. I'll use it to make an actual game, someday. It uses melty blood assets just for testing. I'm not making a fangame or anything.
No.93611
>>93607>musclesFor those, you can probably get away with using bump mapped textures, ne?
No.93615
>>93611After modding a dozen or so games and messing with their bodies I've learned about the limitations of textures, although it's much easier these days. It can be okay at a distance, but closer up you still need a good shape. Muscle, like pretty much everything else in the body is composed of rounded, curved shapes and it doesn't fit into a grid pattern nicely at all. Even with great textures you still need geometry to move around, and newer engines are also capable of more real-time rendering effects that really benefit from a good mesh. One game technology that crops up now and then is "tessellation" where the engine itself splits up meshes which can then be modified by something like a displacement map. (vid related). These AAA studios have the money and manpower for amazing textures, but they need the geometry there and don't want to create a separate set of detailed meshes, so they just use tessellation.
But anyway, back to the human body.
I guess the best way for anyone to see this themselves would be to boot up something like like Koikatsu with sliders and a mod that lets you bypass the default slider limits. You'll see stuff become very square because no matter how strong the texture might be, the mesh was made in a square grid pattern and the Japanese developer didn't intend for anything other than the breasts to be round and stick out.
No.94196
>>94193Hex editing is a pain, especially since you may not always have templates to mess around with. I've had to use it a lot for my 2hu WoW mod too. It's really useful though, since it lets you mod games that otherwise don't have any dedicated modding tools, within reason. You can also rip sprites from doujin games and stuff with them, it's a good thing to learn provided you have the autism for it.
>it could be informative to finally have access to nexus's modder-only forumWow I'd never heard of that before.
No.94208
MASSIVELY NOT SAFE FOR WORK VIDEO OF 3D COMPUTER-GENERATED MONSTER HUNTER GIRLS WITH PENISESheh
hehehehe
HEHEHEHEHE
EHEHEHEHEHE
HAAHAHAAHAHAHAI don't know why this video is so choppy when it normally records fine (and you can see 55+ FPS in the upper right of the screen...)
Days of frustration, but the relief and euphoria of seeing it work is truly an amazing feeling.
The neck armor is not displaying, which hides the gap between the custom body and neck, but I think I've already fixed it for Minoto, and for player armor that area is hidden. (it's a naked neck for testing purposes)
I'm pretty happy with how this turned out. Unfortunately with the way the physics modding works I don't think I can keep the physics on the cloth on her outfit, which is a huge shame.
I'm getting better at Substance Painter, both manually painting and also using the procedural stuff that uses baked maps to influence things. The diffuse (color) texture seems to look good both with dark and light skin, which is a relief. Seems like it clashes with the less colorful default head texture, though.
I'm really happy with most of the body. The legs look great and I'm surprised at how much I like the armpit creases. I had no idea how it'd look in-game. Unfortunately, I can't NOT make it deform weird because of the way the mesh is aligned there, and it's something I need to learn from. Or maybe I didn't weigh it correctly, but I don't want to do that again. It's supposed to be a fantasy athletic body with muscles, but I need to get better at doing that area. I've looked at anatomy images and stuff so I know how to do it better in the future.
The roughness (shininess) texture seems like it needs adjustment, because the armpit and other creases are supposed to be noticeably shinier than the rest, for sweat collecting there. Also, I really need to get better at making butts. I can make them decent from behind (heh), but at an angle like at the end of the video it looks bad. I see no value in realism, but it looks too abnormal even in a fantasy way. Well, more stuff to learn from as I travel upward for the new body I've been procrastinating about:
>>93607I'm starting to gain confidence that my 3D is at a level that people would pay for, but I need to create animations from scratch still, albeit people have very low standards (myself included) when it comes to pornographic animation quality. Plus, people that make dickgirl stuff often just have other people's models which they then attach someone else's penis model to, which looks pretty bad generally. There's a serious lack of passion. I think I'd have more fun and motivation making an H game than the usual 30 second blender animation fare.
Ah, yeah, I also need to start sculpting heads, but people like significantly less detail there so I don't think it's as difficult as it could be. I've saved many reference images and models to learn from at least. I'm looking forward to trying to find a good style there as the face and hair is perhaps the only place you can be creative on a 3D body.
Alright then, time to do more perverted stuff!
No.94209
>>94196>Wow I'd never heard of that before.Yeah, you need to hit a certain threshold of downloads or endorsements. It was significantly harder to get in a decade ago, but these days far more people are into modding. It's like comparing youtube views today to 15 years ago.
No.97145
>>93607Time to get back to work after so much procrastination.
I was spending way too much time in Blender making the topology and using my imagination. I'm going back to sculpting, this time with lots of large definition so I know exactly where to place geometry. This will be so much easier and I really should have done this before.
I'm not going to bother making it look perfect, I just need the muscles to be in the general area correctly. and I'm not sure how much of this detail I'll try to keep, especially in the quads. I need to look at more anatomy guides, but strangely no one covers Amazonian fantasy muscle, just real life men. And there's no professional artist-minded sets of those steroid women to use as a guide, either, so I had to dig through a bunch of stuff which is honestly quite hideous.
While I'm doing this I'm trying to keep in mind where fat collects, too, so the body I'm making could also morph into those impossibly curvy and soft fantasy bodies, which again have no anatomy guides because they don't really exist.
BLEH!
No.97760
I've been trying to write a story about a relationship between a mega NEET and magical imouto
and it's gonna be awesome
but writing is so hard.
and it's true what they say about writing it's the last bastion for the failed artist after they realize that drawing is too hard.
nooooo!!!!
No.97911
>>97760ACTUALLY the last bastion is becoming a critic.
anyway I believe in you
No.98754
>>97145Late I know but have you tried to find resources related to pro wrestling media?
No.98764
>>98754It's not late, this is a slow process because I'm bad at focusing on it.
If you mean real life examples I've been using bodybuilders and fitness women and it works decent enough. The ideal are proper poses in black and white with good lighting and that's what people pay for (or pirate) but it just isn't there for female musculature.
The biggest problem I'm running in to is finding the right balance for a "passive" fantasy muscle appearance without making one of those 2% bodyfat extreme bodybuilders that look alien. Abs are simple and people have grown to expect to see them in their active state even when someone is sleeping, but for other parts like limbs and back it's a lot more complicated with the shape varying wildly with activity. I've noticed that the quads in the legs are especially a nightmare as they have a bunch of overlaying muscles that can really alter the shape of the leg in basic movement.
In 2D you can draw the proper muscles flexing per pose, but no one, not even studios with billions of dollars, animate muscles actually flexing and stretching in 3D (and rarely in 2D). I'm also trying to make this a body can turn into a regular "anime" appearance of low detail, which is easy, but also a plump body, which is actually more difficult than you'd think.
I think I'm going to try experimenting with morphs/shapes that trigger on animation, but it will be a lot of work and I doubt it'd work with physics.
Pic is just a 3D butt without any color, but I'll spoiler to protect the virgin /qa/ eyes anyway. No.107600
>>107595It was a hectic bunch of months wasn't it
No.110909
I've decided to make my desktop less boring. It was a cluttered mess of dozens of icons, but there is a better way.
I'm going to organize things into folders like "temporary stuff", "AI related", "Stuff to post" and so on. I've made the icons the largest possible size, 256x256, and I'm going to make custom ones using stable diffusion merge. My monitor is 1440p but is very wide so I've got lots of space. I plan to make the icons a lot more interesting than pic related, but it's an example.
My basic process is this:
1. Generate the images, 512x512 (minimum size that doesn't affect SD's quality massively)
2. Do Basic edits like thickening lines, doing some perspective stuff and making background transparent. ("simple background" tag is very nice for this)
3. Paste and scale the image into a 256x256 size and save as PNG.
4. Import into gimp that can save as .ico
5. Change the desired shortcut icon!
I also found a cool trick while searching to change the icon of drives:
>If you want to assign an icon to a hard drive, place the .ico file in its root directory along with an Autorun.inf file created in Notepad.
(inside the notepad image is:
[autorun
icon="iconname.ico"
So I'll be changing my drive icons, too.
No.111607
made a website to "release" some of my programming projects
https://chino-chan.gitlab.io/programs.htmlin retrospective, making my own desktop youtube client was a good decision with most of other clients targeting android only and youtube clamping down on invidious and ad blockers
No.111608
>>111607So you where the Anonymous who made these programs, good work!
No.114061
>>111607development has sped up significantly since then and is going well so far!
implemented some features i wanted to do for a long time and fixed lots of bugs
my programs have also gained some users other than me
>>111608thanks
No.114063
Ah, I wasn't want to bump this since my progress has been slow, but I have been doing my 3D modeling stuff very VERY slowly. There's all this excitement about AI stuff, including 3D, but none of it has any relation to topology. There's been previews of AI attempts to create static meshes and textures, but that's the fun stuff (and it's not good enough for anything other than background stuff anyway.
If AI could do re-toplogy, which is what I've been procrastinating about for a YEAR I'd be in business. Alas, no, like real life the LLM stuff is just trying to remove the fun stuff and use humans for dull manual labor.
I can't overstate how mind-numbingly dull this whole process is. Just looking at this image fills me with dread. I spent 8 hours doing it a few days ago and I don't even know why. I moved up the thigh, but then I went back and redid some spots and naisdonaisdioasdas GOD I HATE TOPOLOGY! I WANT TO CREATE, NOT DO 3D MATH!
>>114061Very cool! You could post them in the software thread, too!
>>94952 No.114073
>>114063>You could post them in the software thread, too!thought about that before but i felt that thread was for somewhat noteworthy programs so posting my own stuffs there feels like promoting something more than its worth
if someone likes my programs enough he can post there
No.115920
>>115918>>115919You're killing her!!
No.115929
>>115919i feel like dust should spread out more but that looks really cool
No.115931
>>115920but she's already dust!
>>115929right, i made it a bit more spread out
No.116030
I'm working on a fangame for Touhou based off Rune Factory/Stardew like gameplay and primarily inspired by EraToho TW games: for context, EraToho TW is in effect a sandbox dating sim with a timestop gimmick (hence The World for TW) and the goal of romancing the girls and have sex with them. Sex won't be the focus for this project but I won't close the door on it since adding eroge scenes might be something I do when the game is done. Until then it exists as part of romance but it's not explicitly shown.
For engine I started with Gamemaker Studio 2 but chose Godot because I'm not afraid of programming and Godot's node based system and UI tools drew me in, but I'm holding off on coding until I have a solid idea of direction and what I actually want. Unlike Rune Factory, Harvest Moon, and Stardew Valley, this is not going to be a farming focused game rather it's intended to be a sim/rpg of sorts where the focus is shifted to the characters (and the dating of) and world. Whether this is a bad decision I don't know since I haven't talked about it at all outside of close friends. The idea was originally born of spite of EraToho TW's eng development which is cancer to wade through and suffers from too many cooks in the kitchen, the engine is also pushed to its absolute limit which makes content creation for it a nightmare plus all the Japanese autism bits about touching parts claimed by others.
The ultimate goal is for the player to settle into Gensokyo, socialize and romance the characters, get stronger, play a part in the growing incident, and start a family later on if you play your cards right. I'll be borrowing ideas from TW and hammer them out as I need to fit this goal. The nitty gritty I have so far is in terms of the core of the game and crafting with alchemy and spellcraft. The rest of what I got are scattered ideas I'm working on sorting through and tying together/discarding. As a sandbox, the core loop is simple: Take care of any daily tasks at home then spend the rest of the day as you see fit. Whether that be socializing with characters and giving them gifts, exploring the world, taking part in events, getting into fights, training, crafting, etc.. In short it's the Stardew Valley approach but give the player more direction on things to do, but I know some people want to b-line towards their favorite character and that's okay. It's designed to start out slow and limit where you can go as to not overwhelm you with choice. Over time you'll become stronger and with things like flight and knowledge of shortcuts and locations you can access more of the world and its characters.
I probably made a mistake in jumping into crafting systems with just outlines for everything else, but anyway: Crafting should be thematic to the setting and not just a "click recipe gain product" style of crafting. Alchemy is based off of Feng Shui and balancing the five elements of reagents and their yin/yang disposition to get a result. Spellcrafting (making spellcards, danmaku, and researching other sorceries) is based off the Eight Taoist Trigrams and ordering them to get different meanings and elements. Just those two alone, on top of making it "fun" as well as complex and thematic is a daunting task. There will be farming, but as I said it's not the focus as this isn't a farming sim, and is to act as a zen activity, supplemental source of income, and feed into crafting by growing things you'd otherwise have to buy or forage. Getting stronger and learning magic is linked to characters as an incentive to meet, talk, cultivate a relationship, and learn from them. If you want to use Ofuda and talismans then you should learn from the shrine maidens Reimu and/or Sanae, if you want to contract a shikigami then speak to an expert like Ran, and so on. Crafting feeds into combat, and combat feeds into progressing the game and character subplots.
I know this is incredibly ambitious, but I still want to have a go at it. Right now I'm doing my research and playing related games to get a feel for how it could play, and add to and organize my design document whenever I have time to write.
No.116031
>>116030Sounds really cool! Apart from the obvious things you need to create, I think creating a lot of daily lines for everyone is something you need to plan for. The first in-game day in these games when you first notice a repeated line really changes the atmosphere forever.
Well, at least in Rune Factory games they tried to write a lot of them, but for the Harvest Moon games the townspeople didn't have many lines to begin with and the world felt a lot more barren as a result. I guess that's the difference between a game like Rune Factory that was heavily focused on the people and Harvest Moon in which you spent most of the time farming.
No.116032
>>116031You make a good point. Poor or too little dialogue can make characters, and the world, one dimensional and little more than set-pieces or glorified vending machines. Interaction leads to depth; being able to splash in the water and watch the ripples is more engaging than a still puddle.
Something I never liked about these games is that you talk to NPCs once per day and receive random dialogue. As someone else put it: "Walking up to a gyrating mass of pixels and getting out some fatuous dialogue as a friendship meter arbitrarily fills up is not going to make me interested in them as a character. In fact it does the opposite because I can't help but see them for the computer code they are when it's just a matter of talking to them mechanically and reading the same line until inexorably we become friends." While this is more of a symptom of the medium, I want to introduce player interaction with the dialogue and this is how you increase their affection/friendship because just speaking to someone one-sidedly doesn't make you friends. Even something like a binary choice and an extra line could add a lot as your character has a voice and isn't just assumed to agree with everything with no input. A little scene in RF 4 comes to mind, where Bado tries to scam you for amnesia meds and you can choose how to react to that. It's a little touch but it's an example that trivial interactions may lead to depth—or at least an impression of—for a character.
This is another monumental task just in the writing department since there are *a lot* of 2hus and just one unique line a day for a dozen or so characters can number in the thousands. AI can come in real handy here, not for generating their words verbatim because that's scummy, but for giving me topics to work from since at some point I'm going to run out of things to say and talk about.
No.116033
>>115973feels more like a tripping balls simulator
No.116034
>>116031Harvest Moon depends on the game. Back to Nature has pretty lively people for example but A Wonderful Life doesnt, though I wonder if its on purpose because you're kind of an outsider
No.116043
>>116041.NET executables are run through the
mono filename.exe command, but I don't see any EXE file in that folder? You can use the
file filename command to describe what kind of files are those. If they're Linux executables (ELF) you can run them like this if you're in the same folder:
./filename (chmod is used to give you the permission to do so, not to run them).
No.118050
>>118027Love seeing a new program from you!
However it immediately closes on me on both Win-10 and Linux Mint.
The output only says:
playlist: playlist.txt, scale: 1, font: /home/anonymous/Folder/ikamp/source.otf, ui: ui.txt
window size 300x300 No.118051
>>118050it's similar to mpv that it immediately closes if there is no file to play
either you provide filename as commandline parameter or you add file names to playlist.txt
the easiest way to use it is to associate ikamp as the program to open music files and open from file manager
No.118059
>>118051Thank you for your answer (and for the cute artwork)!
The instructions you gave all worked well on Linux.
On Windows 10:
Opening mp3-files with ikamp through association in the file manager and as a parameter both worked.
The playlist (playlist.txt) did not work for me on Win-10 though (I used full paths to the files and made sure to use utf-8 encoding).
Using the option 'playlist[playlist.txt
]' does not play mp3s listed in playlist.txt but plays files I put in the same directory as the ikamp.exe as a test
(since the ikamp reference on the download-site says "if playlist is not supplied, the files in the same directory as the file argument are added to the list.").
When trying the full path to playlist.txt as an argument, nothing is played and the program closes.
But that might be an issue with my Windows-system or me not understanding of course. As said, everything works nicely on Linux.
Am going to play with the ui-options now.
No.118060
>>118059>The playlist (playlist.txt) did not work for me on Win-10 though (I used full paths to the files and made sure to use utf-8 encoding).for now the text file must use unix (lf) line ending, not windows (crlf)
>Am going to play with the ui-options now.hopefully you can have some fun with it
by the way i just updated it to support borderless window so you can do something like pic related
No.118082
I've been working on a very basic markdown to HTML program in Haskell, mostly to teach myself the language. I don't know if I'll actually end up finishing it, but I've learned a lot from it already.
No.118083
I'm working on a new DE suite based upon Openbox and some other applications. I'm going to hopefully use it as the default DE for a Linux distro I've been working on for some time. It'll run on one tty while my current WM runs on another tty. The idea is being able to switch back and forth between them with the same set of applications running on each.
I've been playing around with s6 and seatd lately. Been writing a lot of init scripts/services for s6 to hopefully replace my current OpenRC init. For the most part everything is working as intended but s6 doesn't really have the best documentation and the author isn't a fan of hand holding so I've been having to spend many hours reading the source code to understand how s6 functions. I've made a little management application for it similar to 66. It serves as a frontend for s6 init and the package manager. Right now it only has CLI but once I finish my Openbox extensions I'm going to write a GUI for it using Qt.
I've also been forking a lot of GTK stuff lately to keep GTK2 support alive in applications that have recently removed it. GTK2 integrates better with Qt than GTK3+. That way I can keep themes consistent between applications.
The long term goal is to move a lot of old Windows machines over to my distro while retaining support for the Windows-only applications they run. I'm building this for people that can't play sysadmin role on their own machines. Using a combination of wine and VMs to support all that old Windows stuff. I have a lot of old hardware that needs a Windows 3.11 box to function. Once of which will require doing some reverse engineering because it has an old serial port proprietary dongle for DRM. If I can figure that out I'll have a good pay day because the company that produced it went out of business decades ago and there is a good bounty for whomever manages to figure out how to run the application outside of the Windows 3.11+486 ecosystem.
>>118060Good work anon. I'm going to use this as my main music player for the next few days. I haven't compiled it yet or read the docs but I'm wondering can it use mpd server or does the music have to be on local machine/NAS?
No.118085
>>118083very nice work
>I've also been forking a lot of GTK stuff lately to keep GTK2 support alive in applications that have recently removed it.saw the writing on the wall for a long time, that's the reason i made my own gui toolkit from scratch for my programs
>I haven't compiled it yet or read the docs but I'm wondering can it use mpd server or does the music have to be on local machine/NAS?there is no source code yet, but the provided binary should just work
it can play files from remote servers if they're mounted on the local filesystem, but yes it has no network capability otherwise
No.118088
>>118085Please let me know if and when you decide to release source code. I'd be willing to help out.
>saw the writing on the wall for a long timeSame. It's a shame what they did to GTK. I don't really like Qt either but at least they haven't taken away basic functions like GTK has been doing.
For my personal stuff I've moved away from Qt and GTK. But I need to support a lot of older applications so I'm stuck with them both. I hate what Gnome has become.
No.118091
>>118088the platform abstraction layers are mostly ready for release now, including audio output and display/input system
there are several other modules i made for this which i want to release as independent libraries so lots of api cleanups are needed before they're ready
>For my personal stuff I've moved away from Qt and GTK.what are you moving to?
No.118092
>>118091For personal stuff I've mostly moved on to terminal/CLI + my own dwm config. I only do GUI stuff for other people now or abstract it away from my base system as much as possible. Qt whenever I do need GUI if not I use GTK2. If you compile from source most GTK applications still work with GTK2-only. For stuff that's dropped support for it I use older versions. But most of that stuff has already transitioned to Qt.
No.118378
>>114063I'm commenting on this again to give myself extra motivation to power through it. I spent like 20 hours the past few days (with waning attention spans causing me to do other stuff) mapping out the legs with
quads with a polygon flow that follows the muscles and major leg parts. Not a single tri!
This was supposed to be kind of low poly but that was thrown out the window. I guess I'm going to have a modeling/posing/rendering body and then I'll have to make a game-ready one.
I've moved onto the 'fun' areas of the body if you know what I mean so I'll need a NSFW warning on the next update which won't be until I've put a lot more time into it.
Man, this is frustrating, but satisfying to see the finished geometry. Doing muscles feels 50x more complicated than doing the simple tube shape that people are used to with this stuff.
No.118858
I've been doing a bunch of design stuff for an animated project that may or may not end up actually happening.
In the past, I've made the mistake of not designing my characters upfront, instead just kind of winging it in the storyboard and pulling everything together afterward. That approach leads to crummy designs, because I'm trying to contrive a character from a storyboard scribble, and usually weird unnatural drawings, because the storyboards don't account for the finalized design and its limitations.
This time, I'm doing it the other way around, and it feels frustrating; not because it's wrong, but because I'm not used to it, and I'm a creature of habit. Actually designing a character, as opposed to polishing doodles to make them somewhat consistent, feels foreign to me. All of my designs end up with weird proportions and clumsy combinations of shapes.
As an aside, I find it frustrating how few resources there are for more stylized character design, at least ones that are more in depth than "this is what a square looks like". Drawing people that look like real people is a skill that takes a long time to build, but there's obvious steps to get there and plenty of reference. What isn't obvious, at least to me, is then taking that knowledge and applying it to simplified, graphic characters.
No.119212
I've been trying to get used to a new drawing tablet. It's the first time I've ever had one of those tablets with a screen and it's definitely an adjustment.
No.119213
i don't think there's much point to new tablets. As long as it works and was made last decade it does everything current decade tablets do
No.119214
>>119213the only reason I replaced my old tablet was because the pressure sensitivity got all wonky and I couldn't find a way to fix it. I'd been using it since highschool, and if it didn't break I'd have happily used it into the next century.
No.119215
>>119213There's possibly accuracy improvements from better sensors and such. There's also quality of life stuff like hotkeys on the tablet or on the pen itself, but none of it this stuff is really necessary. Also if you use one with a screen then all the various monitor improvements find their way to tablets. If you're doing things like rotating camera around for 3D modeling and stuff you really do want a good refresh rate.
Oh yeah, I remember reading about tablets trying to become as thin as possible between the drawing space and the screen, so there's accuracy gained there.
No.119217
>>119215but you know, all the art you see on pixiv and boorus might be pro grade tablets, but I doubt they all use display screens to draw
No.119226
Tablets without screen are ergonomically superior to display tablets because you can look at the eye level and draw at the hand level. Display tablets have a display but suck in every other way possible.
No.119227
>>119226This.
I tried drawing on a laptop with integrated tablet and after drawing I found the perspective was distorted because my eyes weren't perpendicular to the screen.
It seems the only sane way to do that is to hang the whole tablet vertically and draw like painting on a real canvas.
No.120678
>>118378I'm still steadily working on the body and I did not realize the butt would be such a pain in the butt. I'm really glad I did this now because it would be far more difficult if I wasn't able to use this as a starting point for the upper body.
The butt has too many complications for what I'm after:
Green - High polygon density for, uhh, you know.
Magenta - Ideal polygon flow for the butt
Aqua - The actual biological contour of the butt
Yellow - The flow of polygons towards the side (also part of the butt shape)
Purple - Shapes to preserve in lower back
Orange - General polygon flow for the belly,
VERY important for obvious reasons
I wracked my brain and did a bunch of experiments and I guess I'm going with the grey row of polygons you see in the middle with a diagonal kind of flow which seems very counterintuitive to how I was doing things, but I think it will align with most of my goals. The extra polygon density will flow from green area into the belly which will also need a high density of polygons due to its variable size.
Also, after deciding this is going to be a medium-high polygon body, I redid the feet topology. This was a major headache in a way similar to the butt. I wasn't able to do everything (the pink and yellow topology is less than ideal) but I really got sick of messing with it. I know foot reviewer doesn't like how this looks, but it needs to be blocky and spread out so I can actually see what I'm doing.
Retopologizing is the absence of joy. I hate this so much. Blogging about it relieves some stress.
No.120679
>>120678kkkkkkkkkkkkkkkkimo
No.120680
How difficult is 3d modeling to get into?
No.120681
>>120680Pretty easy, I guess? Look up the famous Blender doughnut tutorials and follow along and I think you'll be able to decide if you're into it or not. I learned the basics before that stuff was around so I don't really have anything to link you. There's all sorts of different specializations to go into, one of the more obvious one being soft modeling versus hard. (Organic versus mechanical basically. People devoted to modeling cars or buildings will have a different skillset from someone modeling people)
The main issue is that unlike drawing there's a lot of uncreative and tedious parts of it, unless you work in a team and have offloaded that to someone else.
And no, AI is only seeking to automate the parts that people enjoy, like sculpting and texturing. (and it's not very good at it)
No.120683
>>120681I downloaded and installed Blender already, it was much easier than I thought it would be.
I guess I should give that doughnut tutorial a try.
No.120684
I think it's fair to say that it's theoretically interesting and technically hard
No.120685
Do you need to know how to draw?(because I don't).
No.120687
>>120685Nope. Just give it a try if you're interested. A lot of the stuff people learn for 2D transfers over (anatomy, movement, shading, etc) but a lot of it is irrelevant, like perspective or line work. No one should learn to draw specifically for 3D, but it's certainly an advantage if you already know it.
No.120688
It's mechanically tedious but with enough grind you're guaranteed certain results. You can buy models on sketchfab and focus on animating if that's preferable. Lots of options, but 3d modelling is a long process without the automation toolkit people build up through learning what's boring
No.120689
>>120688What do you mean by automation toolkit?
No.120691
>>120689Hard to explain if you don't try it. But in university we did some scripting such as increasing the poly counts of low res models. With scripts and adons you can create a circle out of a triangle. Things like that
No.120841
I tried a tutorial to make an anime girls head in Blender, it was going alright but with some flaws, those flaws then manifest themselves when I tried creating a subdivision and smoothing the surface as when I tried doing that it made a mess. Vertices are annoying.
But then I got distracted and tired out Unreal because I wanted to have an idea of that program as well, so I would know what I should be working on. So I tried out the program and I manged to make the projectiles explode when they hit things and make a sound when they did it and then I watched a few videos about Unreal to learn more about it. It's also quite complicated, though it seems that a lot of the complication can be worked around by getting assets on the unreal market place, they don't just sell character models and such but they sell back end things too. Things like inventories and health systems would be a pain but you could buy Blueprints that deal with that, though I think I could implement a basic inventory system and a basic health system.
I think that with the little that I know about it now, if I tried to I could make a game without having to buy assets. It would basically be a shooter where you shoot non moving girls(I might see if I can get them to move) and their clothes come off when you do, though I would need the character models to do that which would mean going back to Blender. I should probably make a gun or two as well, but I might make the environment in Unreal first.
No.121235
someone that isn't me should make a online multiplayer ecard game
No.121236
I meant the ecard game from kaiji
No.121250
>>120841 Again.
It's coming along easier than I thought it would. I made a material for the Landscape and was playing around with that, I made it this green Hue. The Unreal Marketplace has lots of free assets including a few environment ones, I got the tress from one of those and then edited the material to turn them greener as well, I also got grass from another asset pack(and turned the material greener too). I'll do more work on the landscape later, this is just a rough sketch.
Oh and the buildings there were a free asset as well.
Another asset I found was an animal pack which included this Fox and some animations so I made a player movement model and then used the Fox for it and then Used the Animations for that, he can walk, run and jump, though the Jump feature is a jumping attack where he bites, but that's just what the animation was. It was easier than I thought but I don't know why I thought it would be hard.
There was an AI asset pack on the Marketplace as well so I played around with that for a bit, I got the AI to target my Fox and shoot at him, though he shoots over the fox and does not track the fox after he starts shooting.
Bleh....
I'll deal with that some other time.
I should probably add a human player character so I get a better perspective on the game and not one from so low down, but all of the free Human models look ugly.
No.121378
Cool! Yeah, reference images are generally a profession of their own: concept artist. Otherwise you generally use real life people for references, with scans from all sorts of angles and muscle study and anatomy and so on. You're supposed to arrange images in a way that makes sense to you. Generally the thought is before you can make a stylized human you should know a real human. I'm not sure if the Japanese ever do that, though, and maybe that's why their stuff is so different.
I should mess around with that Unreal stuff sometime, but I told myself not to do it until I'm done with my model which I keep procrastinating about...
No.121397
>>121378>concept artistI suppose their is that as well, somebody somewhere else said that all 3d artists use a 2d artist's image to make their own 3d work, but I don't know how true that is.
I don't really mean references like that though, I mean character sheets. This is the one I am using for example(spoilered because it's lewd). I could only find two that had what I wanted.
I guess studying human anatomy would help. Particularly sense I am still not sure what art style I want my game to be in. I like anime art style but a more realistic style would fit the game better, I would ideally want something like Dead or Alive but that's way in the future when I get better.
No.121462
>>121457they look better than the other 3D feet
No.121539
>>121538I meant subdivision modifier not subsurface modifier.
No.121540
I showed it to my mum and she said the breasts look weird.
Sigh...
No.121545
now that my vision is not obscured by all those black dots i no longer think the feet look better than the other feet. toes have a messed up teardrop shape and there are big toe gaps. the relative sizes of the toes are alright i think....
No.121552
>>121545I agree, I'm going to go back and thicken the toes, I might try to angle them down more as well at them movement they are inline with the foot so don;t contact the ground, but I'm not sure anybody would notice that, though maybe they would.
This is the image I used for reference for the toes and general shape from the bottom.
No.121557
>>121556those low poly toes are kinda cute
No.121559
I am currently working on an anonymous git instance that I will integrate with a torrent tracker and imageboard (and in time other stuff).
Users will be randomly assigned username from preexisting list of names. Users must be invited by existing user. Users must log-in to access everything behind what I'm calling web 1.0 wall (fake webpage). Once logged in all interaction with other users will be totally anonymous of pseudonymous (using assigned name). The hope is this will prevent people from bringing existing identity into the community and the fake web 1.0 wall will prevent discovery of the community by the wider internet.
I'm also trying to finish the IPFS. But the code is pretty bad so I haven't been able to yank out their crytpocoin scam yet without breaking access to existing files on the network. I've also been unable to get deduplication on Windows clients/server working correctly. It seems to work on *nix but not that well.
No.121562
>>121557You wouldn't lick a girl's polygons.
No.121634
>>121633aren't her hands too short
the wrists are supposed to be at the same level as crotch if i remember correctly
can't be assed standing up to check...
No.121645
>>121634it's his first thingimajig. Wouldn't be concerned with anything more than completing a project
No.121648
framebuffer would mean like.. he's rendering a window on top of the window
No.121653
>>121647if you want to do what you said in x11, there is the xshape extension exactly for that
in my video there are no windows at all and all programs (including the linux console) render to an area of the screen directly, so any rendering overwrites whatever is on the screen previously even for contents from other programs
No.121655
>>121633Looks like she's tensing her shoulders quite a bit
You also have the classic problem with the breasts; the nipples should not be pointing straight ahead, but away from the torso in an angle.
Otherwise, great work! You got farther than I ever did
No.121662
>>121634>>121655The two are related, getting the shoulder right is tricky. as you can see here her right arm which has not been rotated is fine.
I did try to move them apart a bit but because I started with two half spheres for breasts the nipples were just the apex of that and so moving the apex to be inline with my reference image would have distorted it a lot. But yes, it certainly is something that needs to be done in the future.
No.121672
>>121540You're basically running into the thing I did when I started out (and probably what most people in this niche do) and that's using 2D images as a direct reference instead of learning anatomy. 2D generally looks horrendous when it's not 2D because it doesn't make sense in 3D space.
Tamamo's breasts here
>>121397 don't really form a coherent shape if you compare the front and side views.
Anyway, the thing with breasts is that they start from the shoulders. Above the pink line in this image this girl looks exactly the same as someone that's completely flat, and that's the main problem. Even if you ignore the anatomy issue, for purposes of animation they should start way higher than they do there because that's where their movement needs to start.
In the left part of this image is the side view of the breasts. The top of breasts are a straight line when seen from the side and the perkiness determines the angle of that straight line. From the front, breasts form two 6's and the ligaments extend up and to the shoulders and armpits.
Anatomy is fun!
No.122599
>>121642>>121653Any reason why you'd want to remove windows?
No.122690
>>122599it's not really about "removing" because in the framebuffer backend there is no window system in the first place
but because of the abstraction used by my gui system, the program code doesn't know about this fact and still thinks it's in a window even though there isn't any
No.122982
>>122961Where can I get Chino-anon's software?
No.122992
>>111607 >>118027Great work! Cute set of tools
ikachan is a great little booru browser. While it's nice that it lists tags alphabetically, as far as I could see it unfortunately omits the tag categories like Artist, Character, and Copyright (the first and most useful (imo) things I look for when browsing Danbooru, for example). Just a small thing, though, all the info is still there. That aside, it's quite a functional and convenient tool~
I got
ikatube to browse, but somehow couldn't get it to play. It's probably because I didn't use
ikafb+pwm-xft. I've ZERO confidence that I can use ikafb and
not bork my system lol
ikamp is great! This is something I'll be keeping for sure. Played with the config a bit and it looks nice. The new borderless option looks good. I likely don't have whatever compositor (probably
ikafb+pwm-xft) is needed to draw border around the window, but nothing wrong with rectangles! haha
Thanks for these, Chino-chan ^-^
No.122995
>>122982Linked in the replies of the comment I made just before seeing yours. For ease, it's
https://chino-chan.gitlab.io/programs No.123012
just released a new version of ikamp with spectrogram support!
>>122961for ikatube you can use the same theme file included with the ikachan distribution (ikachan_colortheme_bo
cchi.json), you can use that as a template
i actually have worked on a gui theme editor for my programs, so you can use it once it is released, just need some polish
>>122992>as far as I could see it unfortunately omits the tag categories like Artist, Character, and Copyrightthis is unfortunately a limitation of some other boorus ikachan supports, like gelbooru, so they're not always available
>I got ikatube to browse, but somehow couldn't get it to play.do you have mpv in your path? it should work with that
>I likely don't have whatever compositor (probably ikafb+pwm-xft) is needed to draw border around the windoware you using the native wayland version? window border is not supported on gnome in this case
No.123014
>>122992>>123011>ikafb+pwm-xftikafb is a library that my programs use to draw ui and is already bundled in, and pwm-xft is an old window manager that i used previously
so they should'nt have any effect on the operation of the ika* programs
No.123020
>>123012Wonderful. Looks great, thanks for that ikamp update! I got the spectrogram working on my end and did some other small tweaks in the UI file to make it look pretty without being
too cluttered. Took a while, but I finally understand how to use it properly now.
>>123011 >>123014I actually figured out what my issue with ikatube was. It was in the .json file where I overlooked the default program was
mpv! My $PATH was all correct, but ikatube was looking for a program that just wasn't there. I've used VLC exclusively for
everything for the past 15+ years. I didn't even consider that it'd default to another program that I've never used haha
I usually use FreeTube on desktop and NewPipe on Android since they're privacy-focused YT front-ends, but this one works fine now. Nice to have some more fun tools~
No.123021
>>123020 >>123012 >>123014I also forgot to mention that I'm on GNOME yep. I got a new laptop recently and can't be arsed to do any ricing lol. Pop!_OS does exactly what I need for daily tasks and it's been good for many years ^^
Cheers!
No.126375
>>73454I have been meaning to make a database of Scottish computer science exam questions tagged by topic to make my tutoring job easier, but I haven't found the energy.
No.126390
>>126387>all the site contentdo you mean displaying contents from all booru sites at once or something else?
No.126393
>>126392ikachan already displays all contents on gelbooru by default
No.128872
>>128867You have very good taste in aesthetics!
No.128991
>>128987I forgot to mention that one of the main things I'm hoping to eventually do is provide a seamless client to server interface. Basically, you can send off projects to a powerful encoding server that would spit out finished encodes in various formats. In other words; batch processing.
Right now the situation is you have to track down a bunch of stuff manually. The existing repos are a bit garbage. No one has properly collected all that stuff together and documented.
The core of this project is basically an IDE for video editing/scripting. In time I'll provide a GUI for common filters and an easy way for other people to do the same through python scripting. When I started this was basically a text editor with syntax highlighting + mpv to display the results in real time. Then I moved on to providing a GUI for ffmpeg and the codecs it supports.
So instead of loading a .VOB in one application. Splitting out chunks of it. Saving that as its own file. Importing it into my script. Editing it in another application. Saving the script. Opening that in ffmpeg and having it spit out an h.264/265/whatever encode. Which I open in yet another application for muxing. While doing a similar process separately for each audio track. Then another round of editing for subtitles in yet another application. Then muxing it all together. Then testing in mpv. I do most/all of it from one application.
As I said before ultimately my goal is to provide an all-in-one solution to mastering video and audio. With the recent changes in the Linux landscape like pipewire this is finally becoming possible without many hours of configuration to each individual system. Linux finally got something akin to Video for Windows in the last couple of years. So it's ready to take on the role Windows has been required to use for decades now. An application like this is badly needed for both sides of video before and after the editing step. Meaning it would be useful for making clean video to work with from traditional sources (IVTC/color correction/deblocking/lots of other common errors) and the final step before distro (encoding into different formats for distro/creating a true lossless master to store the video long term).
There are a lot of great applications for editing video already. But no one has really bothered to make an application for producing good sources to use within such applications. They all pale in comparison to the filtering you can do with the *synth scripting. I've noticed most "professionals" don't even bother with cleaning up video proper. Since the tools they use for editing are so badly lacking proper filtering.
I think if I work at this hard enough over the next year or so I might be able to turn it into a full time job. I plan to release it for free. But I feel like people would happily donate for something like this. I see so many people requesting fixes and features for applications like MeGUI, Staxrip, Vdub and things like DGIndex all of the time. Most of those haven't been touched in years now.
I'm kind of amazed that we're halfway through the 2020s and most of the serious video work is still being done in 20-35+ year old software. When I said "XP era" software before I was wrong. A lot of this stuff is Windows 98-era software. A lot of it is still 32-bit only!
It's a pipe dream. But if I had the money I'd love to make something like a modern Amiga system. Where my own Linux distro came pre-installed with a suite of video/audio editing software. It'd have some custom hardware to do stuff like rip raw data from Laserdiscs and a proper capture card to get at analog sources. It'd be powerful enough to do all the processing locally. But if the end user wanted they could also purchase a powerful server from us that they could offload the processing work to (or rent a server from us). It would be very useful for studios that process a lot of video and re-master old video for release on modern formats like BluRay or the web. If I could offer that I think I could turn it into a thriving business and compete with the existing players in the market. I'm not kidding when I say no one is offering this and hasn't since the late 80s-early 90s. Professional studios only very recently got away from analog formats and went all-digital. So the market is ripe for someone to swoop in and offering an all-in-one solution to these problems.
I plan to spend the next year or two working on this full time. I'm hopeful that my choice to use Qt won't come back to bite me in the ass. One of the reasons I picked it was the fact that I could build cross-platform GUI using python. Since python is what is used for modern *synth scripting. Similarly, I'm building it on Gentoo right now since python is used by portage. I've been building it on Gentoo and testing it on Arch, Fedora and Windows 10. I want to throw an Apple device into the mix but I can't afford one right now. I'm hoping one comes up on the used market for cheap soon.
Sorry to ramble.
No.133337
>>133294>>133295>>133296>>133298>>133301>>133302>>133303>>133304>>133305>>133308>>133331I have access to AnimeBytes and I downloaded the raw BD, but for the torrent
https://nyaa.si/view/507347 it's still stuck at the metadata download so I can't actually consolidate the files. Surely my client's connectable... private trackers I'm on report a green light
No.133483
wow... im finally seeding
https://nyaa.si/view/507347 to someone... sorry my upload speed is so slow
No.133534
>>121642didn't know ikaanon posted here
No.133735
>>133727looks good but if your really going to one day take it seriously, study anatomy and don't fall into the trap of "drawing everyday = progress"
study and observe the artists that you like and incorporate it in to your art
No.133736
>>133727better than I could ever imagine my first day turning out
looks like you really like 90s anime from it though
No.133737
>>133735>>133736Thank you. I don't want to treat it as something that I have to spend hours practising instead of just drawing for enjoyment - before I purchased my tablet I was drawing on paper and a friend linked me a website with drawing exercises that quickly sucked the fun out of it. I do intend to do some, although I'd rather learn by trying to emulate artists I like, and there are plenty of those.
For this picture I was just following some guides I found on Pixiv, although I do like his art style.
https://www.pixiv.net/en/artworks/56572916https://www.pixiv.net/en/artworks/56546373Getting used to the tablet is another factor and I've already damaged my pen nib while trying to draw clouds
No.133756
>>133727I'm using my tablet that sat for too long as well; still can't get into a regular habit of daily usage yet. I find that if I stop for any reason it's so hard to get back into the rythm. Been working on Pippa in my own style off and on for two weeks now.
I had a face I liked more and all her hair decorations done but somehow I hit something I don't quite understand that deleted or covered my page and I accidentally closed it without saving which was a huge demotivator.
Lets goooo self improvement!
No.133757
>>133756pippa the self-ripper
No.133967
>>133737if you open up the pen there are nibs inside
No.135228
>>111607just released my youtube client as a mpv plugin!
now i can search and play youtube videos in the player