how long to build a game and system advice request

Bill76262

Newbie
Nov 11, 2018
31
42
I'm debating on jumping into game development as a hobby and I'd like some advice from people that have made games.

I've learned a little of RenPy and I believe I can figure out the rest. Likewise I've downloaded DAZ studio and I've watched some tutorials and played a little there.

What shocked me was looking at the change-log of some awesome games and the developer, rightfully, boasts of the number of renders and it's 1,000 just for the update. Initially I thought that meant 1,000 unique screen images but later learned individual aspects of what is shown on the screen can be rendered individually.

The advice I'm looking for.

Excluding render time how long does it take to create a single screen image? Just looking at DAZ, writing the script and working with Ren'Py aside. In theory I'll get faster with more experience and using ready made downloads is far easier than making everything from scratch. But if it takes 2 hours to create the average screen image and if the average game has 1,000 images (I genuinely have no idea) then I can't see myself doing that. My goal would be to create screen images of the quality we see in Big Brother by Dark Silver.
https://f95zone.to/threads/big-brother-v0-13-0-007-dark-silver.1519/

How many screen are in a mid-size game? I know a lot of images repeat, what I'm looking for is how many unique images are in a mid-size game?

Lastly, I'm comfortable buying a new computer to aid in this but I don't know what to get. Some articles debate rendering through a GPU instead of the CPU and discussions of cores and blocks of letters that don't spell words like PCIe. I don't know what any of that means and that part I don't want to learn. If there is a game developer computer on Amazon or something I could buy then awesome. If that doesn't exist then suggestions on what specifically to get would be very helpful. I have a budget of $2,500. I'm a middle-class, white American. If I can throw money at this problem I want to do that.
 

Droid Productions

[Love of Magic & Morningstar]
Donor
Game Developer
Dec 30, 2017
7,174
18,354
For PC, if you're going DAZ you're most likely going it iRay path, a raytracer made by NVidia. As such your big leg up is going to be a good NVidia GPU; I'd recommend the (yes it's expensive, yes it's worth it). If you're more cost constrained, a used 1080 will likely be the best bang for the buck. Do not get an ATI card, since NVidia iRay for some strange reason doesn't work that well on it ;) Apart from that, normal gaming specs apply. Get a decent CPU (though for rendering, 95% of the load is on the GPU), 32GB of Ram, a 512GB SSD for anything that's access critical, and a 3TB HDD for all the stuff (including DAZ assets) that isn't.

When devs list out 1,000 images, that very often include frames of animation. These are of course much cheaper to do than 1,000 hand posed images, especially if you use You're still going to spend hours finding the right assets, morphing your characters and assembling the environment. Once that's all done, some animations go quick; other take longer as you re-render to improve lighting, fix small bugs, etc. For Love of Magic, I've got 16 human characters, each with somewhere between 5 and 30 full-body animations, and 5-15 animating facial expressions. That took about 6 months of dedicated work.
 

lancelotdulak

Active Member
Nov 7, 2018
556
557
Youve decided that X is going to happen and you want people to tell you how long X will take. X isnt how it works.

The 'programming' for a vn is trivial.

Daz and graphical development. How good do you want the images to be? I just TOTALLY rebuilt a character who was done over the last few days. For the renders with just her a rooom and maybe 1 character.. 2 hours minimum (im all about Quality). Doing the poses after everything is done... an hour? Changing the poses slightly for another scene.. 20 minutes? Rendering EACH of those? 1 to 4 hours.. per scene. Developing the original character.. days minimum. Redoing them.. a day. Redoing them again later.. a day. Just finding the right assets.. luck. Im STILL stuck on building the island scenes i still cant get out of my mind.

On the other hand.. Theres a game called Corruption.. it's fantastic.. the graphics are shit. Big Brother has midlevel graphics.. brilliant game the creator is an idiot to have abandoned.

If you want to make a passable game.. that isnt like most of these who take stock characters in stock rooms and make a game noone cares about. Id say 10 hours to develop each character. An hour per scene to pose them. Another hour as you discover mistakes and fix them. For 200 scenes for an alpha. Yes you can cheat and turn their head slightly and render and call it a scene but we're being truthful here. 200 scenes for an alpha. You decide on render time... for garbage quality but gets the point across.. 10 minutes per scene. For passable renders.. an hour. For extremely high quality? 2 to 6 hours per scene (you dont actually have to be there or watch). So 200 x 4 hours say (really good renders) = 800 hours plus another hundred to develop the characters and scenes. And this is AFTER youve learned daz and tossed all the garbae you were so proud of before. BTW this is someone who is very good with Daz and spent a long time learning it......


Im not trying to disuade you because the process is FUN!!! I havent played my guitar in forever because i have to turn on amps find pics etc.. when i can click on daz and make something cool. In other words.... you cant preplan this. It's art. It's going to take the time it takes.
 

lancelotdulak

Active Member
Nov 7, 2018
556
557
Note for your computer: If you search for a Dev computer youre going to see costs from 10k low end to 40k

For a realistic system. Cpu isnt that big a deal as long as you have a ryzen 7 or better or TOP of he line of the previous generation (im on a 1700x. My previous system.. a top of the line amd consumer chip works fine.. ). Motherboard. B350 or 450 minimum. You can go much higher and get a Threadripper with pcie4.. with a Lot of gpu slots and the bandwidth to handle them. Thats $$$. Id suggest a ryzen mb with 2 pcie slots. One to connect your monitors to. One solely for rendering. A gtx 1060 will work. Everything above that will make your life more and more pleasant. Buy used. Buy the best 10 series cards you can. Unless you have a lot of money then buy 20s . Ram. MINIMUM 24gb . id suggest 48 or 64. It matters w large scenes. Your boot drive should be an ssd. A 200+gb drive will be cheapish and will handle your os. you can put daz assets on a hard drive. You need multiple terrabytes. 2 minimum and you'll be deleting stuff. If you have money? Stack some 2 /4 gb ssd's. Put another fast ssd in for windows to use. (it will ask). M2 drives if you can afford it.

BTW you can also build amazing games on a last gen system with 8gb ram and a gtx 1060. IT will be less fun but you can do it.
 
  • Like
Reactions: Bill76262

anne O'nymous

I'm not grumpy, I'm just coded that way.
Modder
Donor
Respected User
Jun 10, 2017
10,978
16,236
What shocked me was looking at the change-log of some awesome games and the developer, rightfully, boasts of the number of renders and it's 1,000 just for the update.
It totally depend of the game, and the amount of time between each updates. The last game I remember that had this much renders in his last update was A Wife and Mother, and it took nearly four months for the said last update ; which still lead to something near to an average of 9 renders/days, but include few animations if my memory is correct.


Excluding render time how long does it take to create a single screen image?
It totally depend of the complexity of the scene and the level of quality you want to reach.
There's games where the dev clearly just added the models, choose a pose and clothes, and rendered the scene as it ; with the tons of default you can expect from this. For them, assume that a scene take less than 10 minutes to be built, including the reflection time. And there's games that rely on the right angle of camera, the right lighting, and where a single scene can need a few hours to be built ; potentially including intermediary rendering, to validate that the preview was effectively accurate and to track more efficiently the possible defaults.

This said, the notion of "average time needed to build a scene" isn't something that really exist.
Lets say that you want to build a scene that happen in the MC's bedroom. The first time you'll need near to a full day to built it, because you'll have to build the said bedroom.
There's already made rooms, but if you want your game to not look like all the other, even by using one of those rooms, you'll want to tweak it, to make it look unique by removing this, adding that. You'll also need many tries to find the better angles for the camera, the better places for the lights, and all.
But once you've done that, you save the scene for the bedroom, and the next time you'll built a scene happening here, you'll just have to make few adjustment, and place the characters themselves. It will significantly reduce the building time.


My goal would be to create screen images of the quality we see in Big Brother by Dark Silver.
So average quality. It totally depend of yourself (some works faster than other) and your actual experience, but I would say that each scene would need between an average of 30 minutes and 1 hour to be built.
But keep in mind that, like for the room, what I said above still apply. The further you'll go in your game, the more you'll have saved scenes. Therefore, the more you'll be able to load one and just need to adjust the position of a character, and few little details like this ; which can make you pass from 30 minutes of building to just 10.


How many screen are in a mid-size game?
It totally depend. Some can do marvelous things with 100 unique renders by updates, while others can go further than 1000/updates. And, obviously, it also depend of the time between two updates and the length of the said update.


I have a budget of $2,500.
If it was just a question of change, with this you can buy a decent computer for rendering. Decent enough for it to need around 15 minutes to render an averagely complex scene. But it's not just a question of change, and price seem to be higher in the US, so I can be wrong.
But even with a lower budget, the rendering is rarely the blocking point. There's way to batch render the scenes, so you can built them on your free time, then let your computer render them while you're at works or sleeping. In this case, the blocking point will always be the time needed to build the scene.
 

mickydoo

Fudged it again.
Game Developer
Jan 5, 2018
2,446
3,557
If I have everything I need, the most I can do in a day is 10 renders, that's setting things up and render time. So at 30 days a month, that's 300 renders. That does not include coding time, fuck ups and asset chasing, or setting up completely new scenes. I aim for 250 renders in 4 - 6 weeks. By yourself on a not bad pc, that's the most anyone can do. Keep in mind I have been at it for 12 months or so now, and although I keep learning and getting better, the speed at what I can achieve it has plateaued a fair bit. Also keep in mind, I have days now and again when I get fuck all done.
 

Saki_Sliz

Well-Known Member
May 3, 2018
1,403
1,011
Been making games for over 10 years, but I don't make the typical Visual Novel Games you see on this site. I've done mostly academic work and youtube indie, and the number 1 thing I hear from various people who hope to instruct new game makers is that they should plan for their first project to be a 1-month long endeavor. I think that is worthless advice, I think the first thing people should do is experiments, trying to see how to implement each feature that interests them. What ever you do, don't try to plan out and make a masterpiece, no one ever manages to finish before they loose motivation. I myself prefer to experiment with making game engines in recent years so I haven't been making any game with my newer art techniques (daz). But then again trying to do very basic project forces you to ask questions that are fundamental to making games, ie how to track player status. doing a simple exploration game is usually the best because you don't really plan, but rather you just keep adding and experimenting as you do things.