#game optimization

[thesirensims]

Sims 3 Reset: Bittersweet CC Cuts, Bigger Stories 🧜🏽‍♀️💎

So recently, I have been going through a lot with my Game. I have been going through a lot of growing pains trying to fix my CC folder and one thing that always happens to me is that first I start out small-ish, not very small, but small enough and the game is working even though it takes forever to load and all that stuff. But I keep thinking to myself that it's worth it as long as everything loads. Or I keep thinking that it's worth it once I get all the things (CC) I want.

Those of you who watch my YouTube content, I am definitely a create-a-sim girlie. I absolutely love creating sims. I'll stay in there for hours. If it was possible that I knew the game would never crash and I could add all the create-a-sim CC I wanted, I would be in there for hours creating modernized versions of cartoon characters, creating beautiful sims, creating main characters for stories. I always have a story in my head. But recently I realized that I'd be doing too much and I think we all go through that sometimes in our CC journey. We pick up everything we find, whether it is that you're a builder. So you pick up every furniture piece of CC or a new wall or a new pattern. Or if you're a create-a-sim girlie like me and every time you see a hair that be looking good that was originally a sims 4 hair, you're like, oh my god. Hell, yeah. I got to stop myself.

So I'm just going to bring you guys through what I plan to do. Honestly, I am going to be focusing on my stories more. Honestly, I am such a story girlie, and I've been losing that. I've been spending so much time trying to work with aesthetics and I've made so many create-a-sim videos, a good like 10 to 20 that haven't even been posted. I feel like I can take a take a little break, for right now. So, I will just be putting in what is already needed. For example, if you've seen a let's play that I've done and you like the way a certain sim looks that CC will be in my game. When it comes to makeup and stuff like that, I'll just leave all of that in there. But when it comes to hairs, I'm going to have to be a little bit more careful. Also, when it comes to clothes, I'll just try to only put in what's necessary.

Now when it comes to accessories, that's a little bit more difficult. Because accessories always be loading slow, but you kind of need the accessories. Or maybe you don't. Please let me know below. These sims don't need necklaces and nails and glasses. Even just saying that hurt me. But yeah, this is what I plan to do.

Also as it comes into the building because I do have a lot of build by CC and as some of you know watching some of my videos, I've been getting into my building. Well, not really building more like interior designing and you can see the progression of how much better I got. But honestly, I think at this point I'm just going to be using other creators houses that they post here on this platform and giving them credit in the videos. That'll have to be it for now. That way I get the small, curated amount of CC that they placed in their houses. And if anybody's like, oh my god, where'd you get that? I'll just be like I downloaded this from this creator. Just get that because even though I do like doing it myself, my game is not stable enough where I can do every little creative thing I want in there right now. My storytelling is being sacrificed because I'm spending so much time with all the aesthetics.

So please give me any tips you have for better game optimization, and the way that you curated your Sims 3 saves folder. I'm not going to delete all my CC. It's in a database, but I'm not going to be putting all of it in my game anymore. I'm slowly finding default skins, default eyes, and maybe even a little bit of default makeup. It's a bittersweet, but it has to be done. Thank you, guys, for taking a moment to read this.

[elderscrollsconceptart]

whats your computer specs? just wondering since how the fuck are you getting that good quality pictures of the npcs ingame sob

I have a RTX 4090 gpu, Ryzen 7 7800X3D cpu with a light overclock, and 64gb of RAM. Game is installed on an SSD.

I play in 4k, at ultra settings with the Quality DLSS setting on. I get 50-60 FPS outdoors (thick forest can get as low as 40fps) and get solid 60 indoors invariably.

The game needs several stability patches as it doesn't seem to be detecting the VRAM in my GPU. I'm "powering thru it" thanks to my PC being very strong, but the game definitely has issues detecting ppls hardware from what I've read on other forums.

I'd imagine with patches and NVIDIA driver updates things will play alot smoother in a few weeks/months time.

[askagamedev]

Regarding your latest answer about optimizing games. Can you elaborate more on how the optimization gets to happen? Like ofcourse they do want to make their npcs look as good as they planned to so how do they optimize without reducing quality in general?

Optimization tends to be tailored to the specific tasks at hand. Programmers on large projects often build or use out-of-the-box tools called profilers that measure performance in specific areas of the code. Based on the profiler reports, the engineers can see where the majority of processing time is being spent and then consider looking into optimizing the heaviest offenders.

If we're optimizing the visuals specifically, then it's a question of what saves the most processing time. The best optimization is often not to render anything at all. If the hardware doesn't have to render and animate a thing, it saves a lot of processing time. We do this by not rendering models that the player can't see (e.g. stuff off-camera, stuff behind walls, stuff too far off in the distance, etc.), and also by not rendering the parts of the model that the player can't see (e.g. if you're looking at the front of a building, you can't see the back of the building at the same time without a mirror or something behind it).

That's only one example of optimizations we make. We load cheaper models and/or textures for things that are far away (LODs), we can reduce the frequency of animation updates for characters in the background or on the peripheral edges of the camera that players aren't focusing on, and so on. There are as many ways to save processing time as there are ways to spend it. If you have the time and interest, I suggest looking at the #game optimization tag on this blog, since I've written at length about various means of optimization in past posts.

[Join us on Discord] and/or [Support us on Patreon]

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Long questions: Ask a Game Dev on Tumblr

Frequent Questions: The FAQ

[simlicious]

Just curious: Your Downloads and Mods folders are huge! How long does the game load for you?

hi @cas-sims!

I have not played Sims 3 in a regular world for years, mainly because I also had trouble with heavy lag and it ruined the fun 😥 If I actually want to play, I need to do a lot of optimization beforehand! My test save that I use for testing and making screenshots for my CC is on a CAS world which has only 4 lots on it and loads pretty quickly. The drawback is that there is no scenery at all for outdoor pics and it does look very bland. It's just a flat green grass area with a road on it, that's it. It takes just under 5 minutes to load the game and the save (I actually timed it). My game runs on a PCIE4 M.2 SSD, I bet this definitely helps with the loading times! I also have a good chunk of CC as merged packages. I also use the HQ mod for my screenshots, and because of it, my game is lagging very badly, especially when applying CASt patterns to objects (which makes actually decorating/making the previews for my patterns a game of patience). I do have the smooth patch, but I think my game has a lot of optimization potential still.

stead of having 3x the amount of CC installed in one game.

I know from some gameplay simmers that they often get rid of store stuff they do not use and disable some EP's and SP's they rarely use (like ITF and Island Paradise, which are known to be especially laggy). I'm sure you know this, but in case people read this who are new to TS3, the AnyGameStarter tool at MTS has this functionality and can create profiles so you can have several "separate" games with different mods and CC. Gameplay simmers tend to pay attention to how CC performs in-game/they look at polycounts beforehand, and also reduce their amount of mods. They also tend to play in worlds that have optimized routing. Also, merging CC is a must. It might also be a good idea to make 2 or 3 different profiles with AnyGameStarter and play with different CC in all of them and if you get bored, you can switch over to another profile and play a different save with different CC. My game is definitely overdue for a refresh, and I plan on using the Sims 3 Performance Guide on Steam to get it into a more playable state. It is one of the best performance guides I have come across, so maybe it will help your game too?

[askagamedev]

In a talk I saw that Spider-man (2019) had redundant assets on disk for shorter seek times while loading. How common is this, and how much data was generally redundant? Presumably the install sizes will be more efficient if not smaller thanks to the next console generations' loading times?

It’s pretty common to do optimizations like this. We often have to trade disk space for performance. This is because there’s a fairly significant amount of overhead involved with loading a file - you have to find the file (via its address) on disk before you can read it in addition to actually reading and parsing the file. If you read one file, you have to spend time to find one file on the disc. If you read a thousand files, you need to spend a thousand times the time to find a single file on the disc. Those load times add up, so we look to try to reduce that as much as possible. As an example, the Call of Duty games have long preprocessed all of their level data into a huge contiguous single chunk of data that can be directly copied into system memory. That data chunk contains all of the level data, including things like animations, models, rigs, textures, etc. that are duplicated across multiple level files. This means they only have to load one file from disc for a level instead of hundreds.

We’ve been doing tricks like this ever since we had discs. One of the oldest tricks in the book back during the Playstation 1 days was to pad our disc with dummy data because of how CDs work. The disc burners would write from the inside of the disc outward but the outside of the disc spins faster than the inside due to rotational physics. This meant that the closer our data was written to the outer edge of the disc, the faster the laser seek time would be (since that part of the disc spins faster). Thus, if we had any unused space left on the disc, we’d fill it up with dummy unused data in order to push the actual game data further out from the disc center, ensuring it had a faster read time. EA Sports actually got in trouble for this and had to issue a recall for Tiger Woods ‘99 because the data they used to pad the disc was the Jesus vs Santa precursor video to Southpark.

[Join us on Discord] and/or [Support us on Patreon]

The FANTa Project is being rebooted. [What is the FANTa project?]

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Long questions: Ask a Game Dev on Tumblr

Frequent Questions: The FAQ

[askagamedev]

I am well aware that I might simply be in a minority here when it comes to caring about framerate, but why do so many console games focus more on graphics than framerate? Do the general consumer not care that much about framerate? I hardly ever see any AAA games on console with 60FPS. And why don't consoles have graphic options like on PC for people who may want higher FPS rather than high graphics? Thank you for any answers. Love your blog!

Frame rate is hard to improve past a certain point, just because of the realities of math and physics. The faster you want something to happen (like drawing a video game frame on a screen), the less time you have to do it. Rendering at 10 frames per second means you have (1 second / 10 frames = 100 milliseconds) of real time to do all of the necessary calculations, game logic, and graphical setup for that frame. If your hardware can’t get it done within that 100ms, then you won’t have the right data to draw on the screen. If it takes 125ms instead, you’ll have to run at 8 frames per second because 1000ms / 8 = 125ms. Increasing the frame rate from 30 to 60 means that you’re going from (1000ms/30 = ~33.33ms) to (1000ms/60 = ~16.667ms). The shorter that amount of time you have between frame updates, the less work your processors and hardware can do.

The other thing you have to realize is that what developers care about isn’t maximizing frame rate wherever possible, but maintaining a consistent frame rate. It’s easy to hit 120 fps when you’re looking at a wall with no other moving models on screen. It’s a lot harder to maintain that frame rate when there are 40 other players all performing complex animations and activating self-shadowing light-emitting particle effects at the same time on screen. That’s primarily because swings in frame rate will affect gameplay - your reflexes and muscle memory are keyed to specific speeds and reaction times that you’ve gotten used to. Sudden drops or changes in frame rate throw player reactions completely off, especially in key high-pressure moments. That is an absolutely terrible gameplay experience, so we devs try to avoid it at all costs - even if it means that we lock the game at 30 fps when it can run at 60 frames 95% of the time. That change in reaction time (usually when gameplay is at its most hectic) is too important to give up. Most of the time, we just don’t have the time or opportunity to optimize those situations down to maintain 60 fps to fit within that 16.6 millisecond frame window, not without sacrificing the visual quality further than the art director and team leadership are comfortable with.

The FANTa Project is currently on hiatus while I am crunching at work too busy.

[What is the FANTa project?] [Git the FANTa Project]

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Long questions: Ask a Game Dev on Tumblr

[askagamedev]

What does a patch note like "improved performance and stability" usually mean in more concrete terms?

Stability in games basically means just one thing - how often the program crashes. “Improved stability” means “we stopped (some) crashes from happening”.

“Performance” is a broad term that generally means the speed at which the game runs. “Improved performance” generally means that the engineers have improved the speed at which some elements of the game run. This could be under certain specific circumstances like when there are lots of enemies on the screen at once all performing visually intensive attacks, it could be improvements to the UI system handling menu navigation, or it could be the speed at which the game saves and loads. Internally, we tend to call improving performance “optimization”. Most of optimization is figuring out when we don’t need to do something and then not doing it under those circumstances. For further reading about optimization, there’s actually a [#game optimization tag] on this blog.

The FANTa Project is currently on hiatus while I am crunching at work too busy.

[What is the FANTa project?] [Git the FANTa Project]

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Long questions: Ask a Game Dev on Tumblr

[askagamedev]

A question on hitboxes. Why are hitboxes a seperate structure from the character models? Since the game has to render the character and know the character's model at the exact location at any time, why isn't that data used as the hitbox?

Because the model has a huge number of polygons and we have to play at 60 frames per second. Here, let me visually demonstrate what I mean.

Here’s Ryu hitting Zangief with a Shoryuken:

Now… let’s say that we’ve got just one big hit box around Zangief and a hurt box around Ryu’s fist here. The hurt box around Ryu’s fist intersects Zangief’s hit box, ergo Zangief gets hit.

How many collision checks do you need to make here to see if Zangief gets hit? Just one - does Ryu’s hurt box overlap with Zangief’s one hit box? If yes, Zangief gets hit.

But let’s say that we decide we want a more accurate representation of Zangief’s body. Like… say, instead of just one big box, let’s put a bunch of smaller boxes on Gief’s body. Like so:

The more small boxes we add, the closer an approximation it becomes to Gief’s actual body. Right? Now here’s a question - how many collision checks do I need to make in order to tell if Ryu hit Gief this frame?

If you said “One for every green box”, you get a cookie. Now… how many green boxes are there in this new representation of Gief? I count 11 here - that’s ten more than before. On any given attacking frame, in the worst-case scenario, I will need to check whether any of the 11 boxes have been hit by any of the hit boxes of the attacking move. 

Did you see what I added there? I said “Any of the hit boxes of the attacking move” - there might be more than one. For example, look at a move like this:

Yun is attacking with his leading (left) palm, but he also extends his right palm at the same time. Shouldn’t that also have a hurt box? If so, that’s another set of checks against all of the hit boxes that need to happen this frame. 

Total number of collision checks I need to make = (# of hit boxes) x (# of hurt boxes)

Are you starting to see why we don’t use the full model for hit detection? There are often thousands or even tens of thousands of polygons that go into a fully rendered model. When we take a swing, we would have to check the polygons we mark as damaging against every single one of those that can be hit in order to be sure whether we’ve hit it. With thousands of polygons potentially colliding, that can easily mean millions of collision checks every frame. We gain some accuracy, but it slows gameplay to a crawl and the added gain in accuracy isn’t even particularly fun. Overwatch, as an example, is extremely lenient with their collision detection, and not particularly worse for wear because of it:

Overall, that tradeoff really isn’t worth the performance cost. We always have to prioritize performance and frame rate over missing by inches, and that’s why we use hit boxes that only generally approximate the shapes of our characters. It’s much faster to handle these calculations this way.

PS. If you’re interested in the actual math behind collision detection (which is a serious issue in practically all video games), I highly recommend the book [Real Time Collision Detection] by Christer Ericson. Fair warning - it’s VERY math-heavy, especially with regards to 3D math. If you haven’t enjoyed college-level linear algebra at the very minimum, it may just seem like gibberish to you. It’s also a technical book and priced to match. You have been warned.

Got a burning question you want answered?

Short questions: Ask a Game Dev on Twitter

Long questions: Ask a Game Dev on Tumblr

Frequent questions: The FAQ