2 00:00:04,713 --> 00:00:09,900 Layering various percussion. Still bossa nova so far.
3 00:00:09,900 --> 00:00:14,912 Adding a danceable rhythm. Better combo than expected.
4 00:00:14,912 --> 00:00:19,767 Cool bass. The basic form of 'Window View' is now complete.
5 00:00:20,800 --> 00:00:23,200 Adding sounds to fill the space.
6 00:00:23,200 --> 00:00:25,600 Adding underwater-like sounds. I love the tremolo effect.
7 00:00:25,600 --> 00:00:28,000 I love music boxes too. Adding one in.
8 00:00:28,000 --> 00:00:30,667 All sounds mixed together.
9 00:00:31,534 --> 00:00:34,200 Upper harmony
10 00:00:34,200 --> 00:00:39,467 Lower harmony
11 00:00:39,467 --> 00:00:41,551 "Cut my hair and face forward."
12 00:00:41,551 --> 00:00:51,300 Window View ft. Kasane Teto SV Production Log
13 00:00:53,867 --> 00:00:55,500 How to Make a 3D Demo
14 00:00:55,500 --> 00:01:03,200 I made the demo using Unreal Engine. A generous software that's free until you earn $1 million in revenue.
15 00:01:03,200 --> 00:01:13,234 It's originally software for making games, but I used the in-game movie feature to create this video.
16 00:01:13,234 --> 00:01:18,134 UTAU Teto and SV Teto models are borrowed from VRoid-distributed models. The models used are listed in the description.
17 00:01:18,134 --> 00:01:23,038 I set up the models to move inside Unreal Engine. Honestly, I'm not an expert. Try searching 'VRoid UE' for more info.
18 00:01:23,038 --> 00:01:32,934 The train is based on a retro steam train asset. I removed unnecessary parts and changed the sizes.
19 00:01:32,934 --> 00:01:42,771 I set up multiple cameras and switched between them.
20 00:01:42,771 --> 00:01:52,667 Screen capturing instead of rendering. Added descriptions to each scene in Premiere Pro. It's done!
21 00:01:52,667 --> 00:01:54,700 Demo vs. Final Comparison
Nowadays it's easy to say that everything is done by AI.
Paranoia is something, indeed, but training your eye, being familiar with art and AI generated pictures and not be blinded by asses are the keys to spot the artifacts and details. The hairband pattern is inconsistent here, the flower buttons are poorly generated and hold some fabric that are already blended. Finally the strands of hair on her neck make no sense. I think Shexyo has some control over what he does, hence making it harder to spot mistakes. Sometimes, there's no effort done in order to hide AI mistakes, like post #11067054 featuring a terrible mess on the top left corner. Advice: Look for the ears, hair, hands and jewelry. Anyway, I was disappointed when I noticed his work was AI generated (at least partially). Anyone can be fooled.
Paranoia is something, indeed, but training your eye, being familiar with art and AI generated pictures and not be blinded by asses are the keys to spot the artifacts and details. The hairband pattern is inconsistent here, the flower buttons are poorly generated and hold some fabric that are already blended. Finally the strands of hair on her neck make no sense. I think Shexyo has some control over what he does, hence making it harder to spot mistakes. Sometimes, there's no effort done in order to hide AI mistakes, like post #11067054 featuring a terrible mess on the top left corner. Advice: Look for the ears, hair, hands and jewelry. Anyway, I was disappointed when I noticed his work was AI generated (at least partially). Anyone can be fooled.
I just came to a realization, the fact Shexyo actually knows how to draw and that being the reason these works aren't as chaotic as many others can be for using AI, it almost feels like what hes doing is generating these pieces and just patching up some minor stuff when you would expect it to be the other way around
We've got to a point in which you might as well call the AI-Assisted artworks as Artist-Assisted AIworks
Paranoia is something, indeed, but training your eye, being familiar with art and AI generated pictures and not be blinded by asses are the keys to spot the artifacts and details. The hairband pattern is inconsistent here, the flower buttons are poorly generated and hold some fabric that are already blended. Finally the strands of hair on her neck make no sense. I think Shexyo has some control over what he does, hence making it harder to spot mistakes. Sometimes, there's no effort done in order to hide AI mistakes, like post #11067054 featuring a terrible mess on the top left corner. Advice: Look for the ears, hair, hands and jewelry. Anyway, I was disappointed when I noticed his work was AI generated (at least partially). Anyone can be fooled.
It could have just been human mistake, your assuming AI error without any proof.
So, if I understand the composition of this, Noa's legs are the ones on the outside, and the ones between them are Sensei's, with the rest of his body hidden both underneath her and beneath the sheets behind her. Did I get it right?
So, if I understand the composition of this, Noa's legs are the ones on the outside, and the ones between them are Sensei's, with the rest of his body hidden both underneath her and beneath the sheets behind her. Did I get it right?
This really isn't something that should be uploaded here. Uploading the cards themselves is already borderline, but this is the equivalent of uploading a scan of magazine page just because it contained one artwotk.
This really isn't something that should be uploaded here. Uploading the cards themselves is already borderline, but this is the equivalent of uploading a scan of magazine page just because it contained one artwotk.
No it isn't. Both of the images are pretty much the only things there. There's nothing wrong with uploading the cards, either. They might not be the most ideal, but they're probably the best ones we can get, and it's not like they're extremely low-res or blurry.
I think it's supposed to be a Ijichi Nijika cosplay considering the artist previous two posts, but I am not sure since he didn't make it clear in the commentary like the others.
Leave a comment