I'll tell you what the setup is for my game in the levels:

Bloom = off
No lightrays
Real Time Shadows
And I replaced ALL animated trees with static ones in a more recent build of the first level,
Vegeation ~40
All shaders set to high except terrain shader

The amount of assets is really not that high, I did so much changes to the first level of Hunted yet and all I got was maybe 10 fps...

In the end I think we all should'nt be talking this deep about saving fps here... and doing that there to gain some fps...
I really think it should be solve by TCG.

Just my opinion.. And I want to thank you guys for your useful posts and Teabone I am not hurt No worries Much rather I am thankful for everyone who's telling me what's wrong. That's the only way to get better and to learn.

- Duchenkuke

Thanks, Teabone. Yeah, I know the frustration well. It makes me crazy too.

See, although you are correct, Lee needs to optimize the engine more, this is not really the attitude to have as a game developer.

It is ultimately your responsibility to optimize your game to get as much performance out of it as you can, even squeezing another one or two FPS makes a world of difference, and it's down to you to make sure your assets work, all the engine optimization in the world isn't going to be much help if you don't optimize your assets in the game itself.

All major publishers know this, and they spend a lot of time optimizing things outside of the engine, a little tweak to a model here, a tiny change to a script there, a new texture over there, and so on.

Most of these little tips you've been given are used by most major devs, even those using multi million dollar engines that are optimized out that wazoo, they still use smoke and mirrors tricks to get that tiny bit of extra performance.

So never give up on trying to get that one extra FPS, don't get into that "good enough" rut.

I agree to what Bleidos says when it comes to "never stop optimizing" as we do that, too.

However, you have to set yourself a certain goal for calling it "release-worthy". That's something different than "good enough". On Father's Island we said: "As soon as it runs on the medium rig above 20 fps no matter the scene, we put it on Steam".

This didn't mean we would stop afterwards. Now the same scene runs at 35 or more.

But we had to make a certain goal to achieve to roll the game out to the customers. Otherwise you end in another trap many indiedevs fall into - they optimize and polish and re-iterate so long, that after some time they don't like the whole picture anymore. Perhaps because of the fact that they had re-visited every inch of ther game hundreds or thousands times. Over and over again. There is a huge graveyeard of "almost-finished" games out there because of that.

And I know that Hutned is still being optimized, so all good here.

Yes, it certainly is!

I am going to release a patch soon. This patch is going to get the game 10+30 more fps. Had to put down some stuff but performance is more important now..

Don't worry belidos, I know it's my responsibility and I will do everything I can - I will prove it soon.

As I am currently testing all this FRAMERATE goodness, here are some tipps from an advanced mental patient...I mean user...yes! :

* Turn OFF the bloom and all other native post processing. Witness up to 100 or more additional frames. Add exterior post processing until Lee will add shaders that don't flat out murder the performance.

* Camera distance down to about 20 or whatever does not occlude the visible terrain. (Planning scenes with fog that still have a low camera distance is a good idea for extra frames. Learn from the glorious PSX age!! )

* In your script bank are zone scripts that deactivate terrain and water. Use this whenever indoors or one of the 2 can not be seen. You can test the impact this has by pressing F11 and then 1 and 2!

* Transparent objects can cause more slowdowns than animated ones. However, both hits are not that significant. Look at how forests are modelled and designed in games from 2005 to 2010. Learn from that. Note that most current gen games also don't spray trees everywhere...especially not the cross plattform ones. (I know that some do but I'm trying to get a point across here!)

* DEACTIVATE THE BLOOM ALREADY

* Pro Tip from Belidos: If you spread out level parts around the map you avoid polygon clusterhumps. You need to clever design these with teleporting and what not.

* Don't cheat on your wife!

* Quality settings don't make that big of a difference in look whenever you know how to unify your games look with texture work and know a thing or 2 about level design other than drag n' drop. Design your game to look good no matter what the player turns off.

* If your mindset is "But I can play the latest COD on max settings, I demand everything I do in GG, no matter how many trees I place, to run at 60 fps min!" please click here

* Pro Tip: remove faces from models the player can not see. Optimize the size of your meshes and textures to a minimum. Decrease unnecessary resolution of textures...increase necessary ones. (Terrain is necessary, that wheel on the heap of garbage is unnecessary). Add LOD's to anything above 10k polies.

* Test game currently running at up to 60 fps on a terrible rig. Beta will be sent out soon to show how it will run on normal to high end rigs.



-Wolf

haha good post wolf

I already did most of the things you mentioned. I think I figured it out now.
Until the patch gets released, some time will pass. I have to change some levels entirely... The amount of bushes I've placed.... AARGH and that was just months ago... I learned so much in recent time haha

I agree, I have been spending an enormous amount of time on my game optimising it as much as I can. I've gone from 16 fps - 24 fps range to a far better 40 - 90 range. Mostly, it runs at 60 but struggles in one area a little still at 40 (far better than 16 though) and looks much more to my liking, I couldn't live with less than 30 fps and even that, when I got it there, was too slow in the end. As an example of time and effort, I can safely say I put in a normal weeks work into this weekend alone, as I was finally seeing paydirt Knocked me about a little yesterday, dog tired, but I can't work any other way when I get the bit between my teeth. Now though, I have a really smooth running game with more in it than when it was struggling to hit 25. Still a way to go though as far as actually finishing it though. Ertlov is quite correct with the testing and testing until you are sick of it. I do that with every game I make really.

The level I mentioned, I have been working on that for a couple of months on and off. It's gone through at least 20 revisions, a couple of those being complete tear down most the objects and start over moments. It's a lot of work, and when you are as picky as I am even harder. I have almost given up a couple of times, thinking my idea is just too ambitious, the level too big. But in the end tenacity and hard work has paid off quite well, and although I am still unsure if I will finish it, it's a lot closer than it was just last week.

Actually, here's another, lol. Oh and Duckenluke, I'd be proud of your game, yes it needs optimising and your on the case by the sounds, but otherwise it is very well presented and put together. I think most posts here are aimed at other games that have been released in the main. I will say this though the second video, in honesty seems fair comment. He complains about loading and framerate which is understandable as both are pretty awful in places. GG is in itself slow at loading stuff, the sheer amount of scenery in this is a long wait for GG to load.

Do you know what art is?
Art are the "Al Capone" Cuba cars still working JUST with the creative minds of the people.

What I mean with this?, so, make a decent game with GG nowaday is a challenge, get a decent game like Ertlov did, is like a miracle, there are a lot of pieces in the puzzle, and all them are very important for the final result.
The fact you care about each piece is the key, about getting the expected final result.

Make a decent game with the tools we have, is not easy; there are many workaraunds around the work, but finally you end becoming specialist in the art of "thinking out the box".

Most of the pcs we are using now, some day becomes older, malfunctioning, impossible to get work, just for trash; well I cay say you than many af them are working now like a charm in Africa, how they do it? "creative minds of the people" is the key.

3com

Thx guys for the empowering words As mentioned, I am on it - (fixing framerate) the aim is at least 30 fps. Some parts have much more frames already. It's mainly the outdoor levels with the insane amount of trees that's causing the bad performance...

Oh I also went through all of the Game-Guru stock trees and made non-animated versions of them. I believe anything that has animations within the model, even if not used, still take up more unnecessary memory -- correct me if im wrong.

But yeah I went through the foliage folder and converted all the animated items there to single frame static meshes. While its nice to have animated trees... some of the best outdoor games out there dont have them. It all comes down to optimizing in the end. I've saved up as much as 15 FPS by not having animated trees.

I would think you still have all the extraneous data in the model (bones, vertex assignments, animation frames, etc.) that it would certainly add up to more memory usage. Thanks for your hard work!

I just published Version 1.7

Has major performance improvements - If there is anyone who has time and the drive to test it, i'd be glad

and...

YAY I finally did it This update was very important.

I'll check it out tonight.

Although to be honest, i wasn't having an issue with it in the first place, i tried it again last night and was getting 30 to 40 in the outdoor scenes, and 40 to 50 indoors, pretty reasonable considering i don't have a beast of a machine.

cool ! Now you should have even more ^^
Thanks for testing

Hmn, yeah, the FPS is absolutely terrible :p

hahahaaa nothing to add here xD