"Let's not say it's impossible to complete first."
Aoki Haruhi quickly interjected, "Akanishi-san, you are also a lead programmer and a veteran of Pokeni. If you give up at the very beginning, other colleagues won't be able to do their work in the future."
"Okay, okay, President Aoki," Akanishi Ken replied with a bitter expression.
"We'll solve the problems one by one," Aoki Haruhi said, "Let's start with the large map."
"1000*1000 might indeed be a bit large for a programmer, and slow loading efficiency is understandable, but we can implement chunk loading first."
"You can cut it into 4 chunks, or if that doesn't work, cut it into 16, or 25 chunks. Any number is fine, as long as it meets the demand for smooth loading."
"It's very normal to have stutters or freezes during transitions," Aoki Haruhi said.
"But it would look very awkward during transitions," Akanishi Ken said, "Maybe everything I see in this small chunk is clear, but just a meter away, the map chunk ahead hasn't been loaded yet, so it's all empty."
"Preloading," Aoki Haruhi said, standing up and drawing a diagram on the whiteboard:
He first drew a square, and then drew another square on each of its four sides.
"Look," Aoki Haruhi said, pointing to the center square, "If I move from this area to the area on the east, don't wait until I've arrived to load the square on the right. Instead, you should start loading the square on the right while I'm still moving."
"And then, when I reach the square on the far right, isn't the leftmost one already out of my field of vision? At that point, you can just delete it."
"This means we only load 2 to 3 chunks at a Time, which allows for maximum savings in programming overhead."
Moreover, to achieve this, Aoki Haruhi decided not to include a mount system in the first generation of Monster Hunter.
Players' running speed also needs to have an upper limit.
"However, even with this, there might still be difficulties with the polygon count of monster models," Akanishi Ken continued, looking at the document, "Herbivore Dragons are a type of gregarious monster that often appear in groups, usually around 4 to 6 of them."
"Our current support for model polygon count isn't that high."
"Furthermore, a large scene isn't limited to just one type of monster; many different monsters will be loaded simultaneously. This means that there could be more than 10 monsters appearing in the field of view at the same Time. If this happens, it will cause stuttering, or even lead to system crashes."
Emmmm…
Aoki Haruhi pondered, resting his chin on his hand.
The problems Akanishi Ken mentioned were objectively present, as the current hardware level was limited and had not yet reached the extent of casual extravagance seen in his previous life.
In 2023, resource and memory saving were only necessary on mobile phones; on PC games, this problem hardly existed.
Think about it, if a game can be hundreds of gigabytes in size, who would care about such a small overhead?
Mobile phones are memory-constrained, so there's no other option.
In other words, current game must be developed according to mobile game standards; compress as much as possible, and save as much as possible.
Ah~
Aoki Haruhi couldn't help but sigh, it was still too slow, Time was still too slow, it would be great if it developed faster.
"This problem is actually easy to solve," Aoki Haruhi picked up a marker and wrote three large characters on the whiteboard—
"LOD."
"LOD?"
Akanishi Ken and Koike looked bewildered.
"Level of Detail technology," Aoki Haruhi rephrased in layman's terms.
"What does that mean?"
"Look,"
Aoki Haruhi first drew a very realistic horse on the whiteboard, then next to it, he drew a pixelated horse, and behind that, a very abstract horse that looked like some kind of ghost, probably just a few lines.
"We create models with different levels of precision."
"Within close range, according to the highest priority, load the model with the actual high polygon count."
"If it's far away, with no combat or interaction needs, but still within visible range, you can use a coarser one, with fewer polygons."
"Even further away, if barely visible, you can just load this very, very coarse model, or even not load it at all, just load the terrain."
"In addition, assign priority tags to terrain, materials, and different object models, which will make it easier for the programmer to make judgments."
"For example, terrain is the most important, so during the preloading process, the ground will be loaded first, followed by important environmental elements like large rocks, cliffs, rivers… Just follow this pattern."
"Oh, oh, I understand, I understand." Akanishi Ken and Koike nodded vigorously.
President Aoki is truly terrifying.
He could actually think of such a method.
According to Akanishi Ken's original idea, loading so many things into a single scene would require at least tens of thousands of polygons, and the program consumption would be unbearable.
But as soon as Aoki Haruhi mentioned layered loading and preloading, suddenly, tens of thousands of polygons were reduced to just a few hundred, which was simply terrifying.
This actually has another benefit, which is that it can simulate the real human eye.
When people look at things, within a circular range close by, the clearer it is the closer it is to the focal point, and the blurrier it is the further away it is.
When LOD is implemented, it actually produces a similar effect.
The further away, the coarser the loaded model, the fewer the polygons, and naturally, the blurrier it becomes.
Not only will players not feel uncomfortable due to the reduced polygon count.
On the contrary, it will appear more realistic.
However, just as Aoki Haruhi said, you must control the player's movement speed within a certain range.
Otherwise, if your movement speed exceeds the preloading speed, it's very easy for things to break immersion.
For example, if the floor tile next to you hasn't loaded yet, and you drive a motorcycle over it, you might suddenly fall through the floor.
This is a common bug in many game.
But in reality, it's caused by the preloading not matching up.
…
"Besides that, the programmer needs to do one more thing," Aoki Haruhi said.
"Hmm?" Akanishi Ken looked up in confusion.
"You know about topological operations, right?"
"I… I think so…" Akanishi Ken hesitated a little.
It wasn't that his programming skills weren't good enough; it was mainly President Aoki's overbearing demeanor, just like a teacher lecturing in a classroom, suddenly asking a student a question, which would naturally make you feel a little flustered.
"Then it's easy," Aoki Haruhi said, placing a chalk box on the table, then taking out a piece of chalk and placing it behind the box, with only half of its head exposed.
"In the 3D field, there are actually occlusion relationships between objects."
"For example, this piece of chalk, with half of it hidden behind, from your perspective, only half is visible."
"In this situation, according to general programming logic, regardless of whether the back half is visible or not, would it still be loaded?"
"Hmm," Akanishi Ken nodded, "That's for sure."
"But in reality, we don't need to load the back part, do we? Because for players, the half-piece of chalk hidden behind the box is invalid information, so that part should be clipped."
"Furthermore, as for the chalk box itself, only three of its faces are visible to players; the faces at the back and bottom are also redundant. If these faces are all clipped, the memory overhead will be released instantly."
Nani?!
Akanishi Ken and Koike were both stunned.
A short lesson immediately showed them how powerful President Aoki's programming foundation was.
You don't even know if this falls under the category of programming or mathematics.
"But it's a whole, it can't be cut."
"That's where it comes in. Use a topological algorithm to calculate and clip the invisible parts behind it."
"Wouldn't that require real-Time computation? The overhead caused by this computation doesn't seem small either."
Hmm… Aoki Haruhi stroked his chin; Akanishi Ken seemed to have a point.
He almost forgot about the Time limitations.
Current computing power might truly not be enough to support it.
The polygon loading was reduced, but the real-Time computation increased. It's hard to say which one is more cost-effective.
"Then let's do this," Aoki Haruhi said, "For static object compositions, for example, if these things don't move and their occlusion relationships are fixed, then as I said, after the art production is complete, the programmer will perform the initial clipping and merging work."
"If it's something that changes and moves, and there are such occlusion relationships, then you all need to find a way to optimize it. Even if it's not possible now, it definitely will be in the future."
Uh...
Akanishi Ken frowned heavily.
He felt that the workload had invisibly increased quite a lot.
What was said earlier was about meeting basic needs, but now it has become content for optimization.
As everyone knows, program optimization is the most difficult and Time-consuming.
...
"It's also very difficult to do in terms of AI, isn't it?" Akanishi Ken said uncomfortably.
He could barely find any more questions to challenge President Aoki with; he could only struggle desperately with one last effort.
If this still doesn't work—
We'll have to start working to overcome these terrifying challenges.
Although it's not difficult given that Aoki Haruhi has already provided a clear solution, it's still a lot of work.
Work without thought, and even less a sense of accomplishment.
All he felt was the fatigue of being a tool.
Now Akanishi Ken regretted it a bit; he shouldn't have come to President Aoki so early. He should have discussed it with his subordinates first to see if they could find a solution, and only then, if absolutely necessary, go to President Aoki. That would have been the right way.
Akanishi Ken was once again stunned by President Aoki's terrifying problem-solving ability.
Still too young, still too naive, he underestimated President Aoki's strength.
It makes sense. Since both of these game were proposed by President Aoki himself, he must have had a clear idea and knew that they could be realized before allowing development.
If he hadn't thought of a solution, and it was something that couldn't be realized, President Aoki probably wouldn't even mention it.
"With so many monsters, each having its own behavior patterns, and needing to react differently in various situations, it's not easy to solve."
Akanishi Ken played his last card.
"You just need to make the behavior tree tool, and then hand it over to the planners to design it themselves," Aoki Haruhi said. "As for how intelligent the AI is or to what extent it's done, you don't need to worry about it."
Fortunately, this game was given to Kamiya to do.
If it were Miyazaki Hidetaka—
Can you imagine how terrifying a Female Fire Dragon that can read commands would be?
Miyazaki Hidetaka hasn't yet introduced the concept of reading commands, so everything is still within controllable limits.
The difficulty of the witcher 2 is numerical; it hasn't reached the terrifying level of mechanics.
It won't be like this in a few years.
...
Damn it!
Akanishi Ken was crushed completely. President Aoki's light remark, "A programmer just needs to do his own job," was like unleashing a pair of trump cards, blowing him away.
Lost, completely lost...
Akanishi Ken thought to himself that these two game definitely had to be made!
Indeed, were all efforts in vain?
Is Aoki Haruhi truly unbeatable?
President Aoki is indeed unfathomable!!!
"I have a question."
Art Director Sakata Taiji weakly raised his hand.
"In Mushishi and Monster Hunter, these monsters and characters all require a large number of actions. This workload is a bit too much for the artists; I'm afraid we can't handle it."
Well done, Sakata!
Akanishi Ken cast him an appreciative glance.
This way, these two projects can be dragged into a long race, at least three to five years, and if spread out, the average workload for the programmers will naturally decrease.
Moreover, in the future, with the development of hardware, some work might even be solved by hardware itself.
"Oh, right." Aoki Haruhi only then remembered that Sakata Taiji had raised a very important question.
In previous 3D game, actions were drawn and adjusted by artists, which was actually a rather clumsy method.
We've already entered the Time of 3D, so it's Time to abandon some outdated production methods.
"It's Time to introduce skeletal animation."
"Skeletal animation?!" Sakata Taiji and Akanishi Ken were both shocked.
"Yes."
Aoki Haruhi stood up, "You've all played with some jointed dolls, right?"
"When we move a doll's hand, its internal bones will also rotate with it, then fix the doll's action."
He demonstrated himself, "Anyone who studies art knows that human body movement is achieved through the expansion and contraction of muscles, such as bending the elbow, which uses our biceps.
It's equivalent to our muscles driving the bones to move.
However, when we do 3D actions now, we have to reverse it."
"Reverse it?"
"Yes, reverse it."
"This Time, what we need to do is to let the bones drive the muscles, driving different parts to move."
Aoki Haruhi first drew a person, then drew a skeletal framework inside it.
"We can define the bones for the model. This is essentially creating many connection points on the model, and when these points are connected, they form a skeletal framework."
"After that, as long as the bones move, for example, bending the elbow, then when the skeletal animation plays, the model will be driven.
This can then match the movement effects the artists want."
"The programmer side will provide skeletal animation support, and then the artists will adjust it themselves."
...
Akanishi Ken and Sakata Taiji were both dumbfounded.
Especially Akanishi Ken.
Good heavens, it was clearly a problem that artists could solve, but somehow, it vaguely became the programmer's job?
Sakata Taiji, on the other hand, was thinking:
Nani?!
The entire art production process and method have changed.
It sounds impressive, but he had no idea how it would actually be to implement.
Akanishi Ken covered his face in some pain.
Why am I even sitting here?
Aoki Haruhi indicated that this was just the beginning.
Skeletal animation is standard in 3D game and can save artists a lot of Time. You just need to adjust the actions well, follow a certain pattern, and then once it's in the game, the programmer will handle its execution.
In the future, to achieve better action effects for characters in game, motion capture devices will also be needed.
Once motion capture technology is available in the future, AI adaptation, AI face swapping, and even AI generation will also be needed...
By then, there will be no need for human adjustment; simply input it, and the AI will automatically generate it.
The road is still long, comrades still need to work hard!
...
The programmers and artists discussed each of their concerns in the conference room.
The newly arrived Kamiya Hideki was stunned.
This scene felt familiar—
It was like watching Zhuge Liang debate with the scholars in Romance of the Three Kingdoms.
(No doubt, Romance of the Three Kingdoms was introduced to Japan very early and became extremely popular. Many Japan people particularly love the Three Kingdoms, even meticulously studying every detail within it.
Many classic stories can be recalled at will.
Two years ago, the Romance of the Three Kingdoms TV series was introduced to Japan at a price of 15,000 USD per episode, becoming a hit TV drama. Japan people were particularly enthusiastic about it, and it even once spurred the popularity of related game.)
At this moment, Aoki Haruhi, in Kamiya Hideki's eyes, was like the embodiment of Zhuge Liang.
Seeing the shock in Kamiya Hideki's eyes, Shinji Mikami smiled and patted his hand.
"This is all normal operation; you'll get used to it."
"Nani?!"
This is actually normal operation?
Kamiya Hideki was so shocked. President Aoki's knowledge reserve and problem-solving ability were too strong; it could be described as unfathomable.
Too terrifying.
President Aoki is truly terrifying.
He answers every question.
And once he answers, he can always convince the programmers and artists.
He instantly finds the direction for production and methods to solve problems.
But...
Akanishi Ken and Sakata Taiji didn't need that!
They wished President Aoki couldn't solve it either, so their burden wouldn't be so heavy.
The more they discussed, the more flustered both of them became. How did the workload keep increasing with more discussion?
In the end, there was nothing left to ask.
Sakata Taiji kept sending Akanishi Ken glances: You better shut up!
...
After a five-hour discussion, everyone had nothing left to ask, and the plans had also been reviewed in the meeting.
Walking out of the conference room.
Akanishi Ken's head was buzzing.
Damn it!
All the problems were solved by President Aoki again!!!
Please Support me by becoming my Patreon member and get 20+ chapters.
[email protected]/Ajal69
change @ with a
Thank You to Those who joined my Patreon