🧵 Untitled Thread
Anonymous at Sat, 14 May 2022 15:24:58 UTC No. 897139
You experts and professionals in the field, how many polys does this beauty have?
Anonymous at Sat, 14 May 2022 15:47:25 UTC No. 897145
At least 10
Anonymous at Sat, 14 May 2022 16:41:06 UTC No. 897153
Do you really expect me, in a Saturday afternoon that I'm passing drinking good whiskey, in probably another continent from you, to count the polygons of your fucking jpeg picture? Why? What, exactly, is the purpose behind that question?
No. Wait. I do not care. I absolutely give zero fucks.
Anonymous at Sat, 14 May 2022 17:04:25 UTC No. 897155
>>897139
between 40k - 60k
Anonymous at Sat, 14 May 2022 17:18:43 UTC No. 897156
>>897139
Middle is the most attractive.
Anonymous at Sat, 14 May 2022 18:35:19 UTC No. 897169
>>897153
>drinking good whiskey
Bullshit. I bet you drink some supermarket swill.
Anonymous at Sat, 14 May 2022 19:14:52 UTC No. 897175
>>897174
>malt whiskey
My god, I can smell the poor taste across the ocean.
Anonymous at Sat, 14 May 2022 21:39:17 UTC No. 897200
>>897153
Yes, i will explain.
When this movie came out it was practically the most absurd 3D at that time, it would be interesting to find out if the current games come close to that amount of triangles... but good drinking anyway.
>>897155
Seems like a believable amount according to this wireframe.
Anonymous at Sat, 14 May 2022 22:48:10 UTC No. 897204
Anon do you understand the concept of subdivision
Anonymous at Sun, 15 May 2022 01:02:39 UTC No. 897222
I like big butts and I cannot lie. No literally I cannot.
Anonymous at Sun, 15 May 2022 11:13:56 UTC No. 897285
>>897174
>malt
Nig
Anonymous at Sun, 15 May 2022 15:38:35 UTC No. 897338
>>897227
Yea looks close to the base Genesis 2 low res 16k poly.
Anonymous at Sun, 15 May 2022 18:06:53 UTC No. 897367
>>897227
no
the base body is close but look at ops pic, hands and face have a higher density.
the head is obviously one level higher subdiv.
Anonymous at Sun, 15 May 2022 18:42:01 UTC No. 897369
>Every one of her 60,000 hairs was separately animated and rendered, at a render farm consisting of 960 Pentium III-933MHz workstations that took 1.5 hours to render every frame. Her model is estimated to be made up of around 400,000 polygons, as are the other main characters of the movie
Anonymous at Sun, 15 May 2022 20:53:18 UTC No. 897380
>>897139
As you can see the Aki model isn't that heavy and in the sub 100K range like most modern game characters. But it's built with subdivision in mind so when you see it on screen it is the number of polys in the basemodel squared to the subdivision level. Whatever is needed for her curvature to look perfectly smooth at the distance from camera so prob between 2-4 subdivisions depending on the shot.
So Aki is maybe 40-90K polygons that is actually hand edited by the artist, and a couple of hundred thousand to low millions is what you're looking at on the subdivided surface in the rendered frames of the released film.
Anonymous at Sun, 15 May 2022 21:15:46 UTC No. 897381
>>897369
This is bullshit there already was hair sim back then
Anonymous at Mon, 16 May 2022 01:58:56 UTC No. 897435
>>897367
you over estimat the polycount.
~25k polys is accurate, definitely under 30k
The comparison model is under 20k polys.
OPs model shows more polys in the neck and hands and probably the feet, but not in the face. The face only looks more dense because the the hair from the back of the head showing through the front of the head.
add ~100polys each to the neck, each hand and feet on the comparison model and you have a very similar density distribution.
Anonymous at Mon, 16 May 2022 02:03:35 UTC No. 897436
>>897367
you over estimate the polycount.
~25k polys is accurate, definitely under 30k
The comparison model is under 20k polys.
OPs model shows more polys in the neck and hands and probably the feet, but not in the face. The face only looks more dense because the the hair from the back of the head showing through the front of the head.
add ~1000 polys each to the neck, each hand and feet on the comparison model and you have a very similar density distribution.
Anonymous at Mon, 16 May 2022 02:52:12 UTC No. 897441
>>897381
they literally chose to do this anon
Anonymous at Mon, 16 May 2022 03:41:33 UTC No. 897448
>>897227
>>897367
>body is close
Look at OP boobs - round to pixel precision.
Look at yours - like 8 sides.
I'm not even /3/ but are you guys retarded?
Didn't those old programs use nurbs so there are no polys until surfaces are divided up to render one tri per pixel or whatever?
Anonymous at Mon, 16 May 2022 05:56:05 UTC No. 897454
>>897448
Sub-D and NURMS had displaced NURBS quite some time before The Spirits Within.
NURBS was pretty archaic and something few where dabbling in by the turn of the Millennium.
CV-surfaces are very hard to work with compared sub-d and something few people was dabbling in past the early-mid 90's.
Anonymous at Mon, 16 May 2022 06:21:20 UTC No. 897460
>>897436
20K sound plausible, my fem-model looks like it has a very similar poly density and it's about 20K.
It's a game model not meant for sub-d but it can be used to demo what sub-D is all about for people ITT that perhaps are new to all this.
Anonymous at Mon, 16 May 2022 09:02:56 UTC No. 897491
>>897153
No one knows or cares who you are, faggot
>dood drink lmao
Shut up
Anonymous at Mon, 16 May 2022 11:07:29 UTC No. 897501
>>897227
How did you get quad version of game models?
Anonymous at Tue, 17 May 2022 01:28:50 UTC No. 897663
>>897501
He didn't see that thigh? how all the wrong triangles are iso-lined?
Anon prob ran some 're-quadify' function on it or did a very sloppy cleanup manually.
Anonymous at Tue, 17 May 2022 03:48:38 UTC No. 897675
>>897663
>>897501
this is right. I just ran the quadrangulate operation in maya for the sake of getting a more comparable model.
Anonymous at Tue, 17 May 2022 04:52:09 UTC No. 897679
>>897448
>I'm not even /3/
you should stop there and asks questions next time instead of calling others retards. it really just makes you sound like one.
OPs image is a poly model with NURBS hair; Built in 2000, so likely in Power Animator, and that approach would be the norm at the time. I started 3d in 1998, so I know quite well how it worked back then. There were not a lot of options.
The underside of the breasts are rounder in OPs image because of subdivisions in the bikini top and perhaps subdivisions in the rendered models. Otherwise, you can visually count the number of polygons that make up the breast area and in entire model for that matter, to see that they're very similar.
Also note, that the comparison mesh has denser hair including the bun, and bows and lace in the top of her outfit that account for additional polygons if you want to consider the differences. All that in a model under 20k polys. Adding an extra 5k on top of that is generous.
Anonymous at Tue, 17 May 2022 07:39:14 UTC No. 897720
>>897369
>>897381
>>897441
>Every one of her 60,000 hairs was separately animated and rendered
>Her model is estimated to be made up of around 400,000 polygons
These points and numbers are nonsense. Tech writers have no idea what they're writing about, especially back then. They may get a few talking points and specs from someone and write complete horseshit articles to impress and fill magazines. I had very close experience with this.
you can see the wireframe in OPs pic. there's no need to imagine what it looks likes to understand what is a realistic poly count is, or how the hair was made.
the hairs,or eye lashes, brows, possibly the clothing were nurbs based surfaces, so of course they cant give a polycount on the mdels, so they give some made up *estimate* of 400k polygons to make it sound impressive.
Being partially made of nurbs means that the tellelation and polycount varies depending on the rendering resolution and distance from camera any character is at any time.
Anonymous at Tue, 17 May 2022 07:45:32 UTC No. 897722
>>897369
>>897381
>>897441
>Every one of her 60,000 hairs was separately animated and rendered
>Her model is estimated to be made up of around 400,000 polygons
these points and numbers are nonsense. tech writers have no idea what they're writing about, especially back then. they may get a few talking points and specs from someone and write complete horseshit articles to impress and fill magazines. I had very close experience with this.
you can see the wireframe in OPs pic. there's no need to imagine what it looks likes to understand what is a realistic poly count is, or how the hair was made.
the hairs,or eye lashes, brows, possibly the clothing were nurbs based surfaces, so of course they cant give a polycount on the mdels, so they give some made up *estimate* of 400k polygons to make it sound impressive.
being partially made of nurbs means that the tesselation and polycount varies depending on the rendering resolution and distance from camera any character is at any time.
Anonymous at Tue, 17 May 2022 07:50:13 UTC No. 897725
>>897369
>>897381
>>897441
>Every one of her 60,000 hairs was separately animated and rendered
>Her model is estimated to be made up of around 400,000 polygons
these points and numbers are marketing nonsense given to impress the audience.
you can see the wireframe in OPs pic. there's no need to imagine what it looks likes to understand what is a realistic poly count is, or how the hair was made.
the hairs,or eye lashes, brows, possibly the clothing were nurbs based surfaces, so of course they cant give a polycount on the mdels, so they give some made up *estimate* of 400k polygons to make it sound impressive.
being partially made of nurbs means that the tesselation and polycount varies depending on the rendering resolution and distance from camera any character is at any time.
Anonymous at Tue, 17 May 2022 13:43:35 UTC No. 897786
>>897675
>>897663
>>897501
Worth noting some games, especially older ones, do sometime encode quads in the model file and split them at load time.
Anonymous at Tue, 17 May 2022 22:00:13 UTC No. 897865
>>897794
Spoopy
Anonymous at Wed, 18 May 2022 03:37:14 UTC No. 897917
>>897679
>you should stop there
And yet I was clearly right that there are more polygons used in the shaded images in OP than in the 'mesh', (look at the shoulder of the arm on our left - hard corners - shaded, no corners) whereas you are left embarrassing yourself talking about visually counting the polygons in the mesh
Anonymous at Wed, 18 May 2022 12:55:32 UTC No. 898000
>>897917
>I was clearly right
no, the only smooth here is your brain. this is a discussion talking about polygon count in the model, not how you look at breasts.
Anonymous at Wed, 18 May 2022 12:57:52 UTC No. 898001
>>897917
>I was clearly right
no, the only smooth thing here is your brain. this is a discussion talking about estimated polygon count in the model, not how you focus on breasts.
Anonymous at Thu, 19 May 2022 06:16:12 UTC No. 898217
>>897794
Oh, it's you
Anonymous at Thu, 19 May 2022 22:34:37 UTC No. 898396
>>898001
so what happened to the shoulder between the 'model' in wireframe and the actual model in grey?
Anonymous at Thu, 19 May 2022 23:18:25 UTC No. 898400
>>897794
Fucking kek. This is not your true skill. Stop holding back and post your artstation.
Anonymous at Fri, 20 May 2022 01:22:20 UTC No. 898405
>>898396
wireframe is the base model
smooth-shaded model is typically just the base model rended with smooth shading, but because the aliasing is different it will look different from the wireframe. however i can see a lot of the same linear edges in the top of the shoulder and the side of the hip, and because it makes absolutely no sense to subdivide the model just to take a screenshot of an unaliased shaded model, im going to say it probably wasnt subdivided.
rendered model however clearly shows additional subdivision.
now, OP showed the wireframe and asked for a poly count estimate, and everyone is clearly making comparisons to the wireframe and not the rendered image.
Anonymous at Sat, 21 May 2022 19:01:58 UTC No. 898816
Bamp
Anonymous at Sat, 21 May 2022 22:36:25 UTC No. 898845
>>897145
He's not wrong
Anonymous at Sat, 21 May 2022 22:41:20 UTC No. 898847
>>897153
You sir, are a blatant liar. You spelt whiskey with an 'e', revealing that you are in fact drinking dirty dish water.
Anonymous at Thu, 26 May 2022 16:31:31 UTC No. 899657
>>897156
T. Dural lover
Anonymous at Fri, 27 May 2022 22:29:48 UTC No. 899930
>>899141
60k at least, but they used nurbs
Anonymous at Tue, 31 May 2022 03:49:28 UTC No. 900568
>>899121
5000
Anonymous at Sun, 5 Jun 2022 07:02:59 UTC No. 901508
>>897227
Triss, what a miss!
Anonymous at Fri, 17 Jun 2022 17:08:26 UTC No. 903978
>>897139
all those Tris, jesus christ and im struggling to get a job with the ability to fully quad topologise and animate a bitch wtf
Anonymous at Tue, 21 Jun 2022 08:35:52 UTC No. 904544
>>897139
the visual difference between 200K verts and 100K-60K verts is hard for most people to notice and depending on the art style/how textures are applied.
I think around 60K is right.
Anonymous at Tue, 21 Jun 2022 20:45:39 UTC No. 904628
>>899141
Looks like nina willams
Anonymous at Sat, 25 Jun 2022 22:07:16 UTC No. 905262
>>904628
Kek... Very funny
Anonymous at Sun, 26 Jun 2022 12:02:21 UTC No. 905391
>>897139
From wiki:
Each character's base body model was built from more than 100,000 polygons,[34] plus more than 300,000 for clothing alone.[23] Aki's character model bears 60,000 hairs, each of which were separately and fully animated and rendered.[30]
Anonymous at Mon, 4 Jul 2022 01:59:54 UTC No. 906608
Boomp
Anonymous at Fri, 15 Jul 2022 18:54:00 UTC No. 908398
>>897139
atleast two subdivisions levels an really low. it has really good bump maps.
Anonymous at Fri, 15 Jul 2022 20:13:17 UTC No. 908403
>>897153
Nobody gives a shit about you getting drunk off of $20 store brand whiskey on a weekend afternoon.