#solowork

Its been awhile since i did a proper update, so i think its about time i do some actual work while i have time and not only write, but show how the project is going.

i wont lie, with the change in scenery, work has taken a hit. I normally tend to get most my work done between 10 pm and 4 am, but such time doesn’t work well with a partner that keeps busy during the day. and the lack of such amazing creative people working next to me, motivation took a hit.

Not sure where to start, so lets start with the more interesting. 

Particles. Since I didn’t have anything in mind when i started, but knew i would be looking into collision detection at some point, i decided to start there. 

The game was not to be a realistic looking game so i went back to the beginning themes and what the game was about. I had got much inspiration from cartoons like bugs bunny and road runner, and the way they dealt with characters hurting themselves or setting off bombs have this amusing feel to it. this was something i wanted. 

Finding tutorials on YouTube, i set to work.           

Although i have yet to put any explosions in the game currently, the cartoon style got me and i would try to base all other effects along the same line. 

After playing around creating a few different yet similar types, i wish to see them in action.

For this i needed to start work on the collision detection. The idea was -  depending on how hard you hit a surface and where, the type of particles, noise and at what volume would change.

Rather then having everything tied to the character who would collide with everything, i decided to have the ground calculate where it was hit (the white flooring/walls would now detect everything). The thought was that if Stan hit and slid down a wall, the impact dust/smoke would spawn at the point of impact rather then on Stan who might  have moved on from that spot. The result i think worked out quite well.

Once dealing with the actual point of collision and how i would spawn the effects in, i set to work on detecting how hard Stan would hit a surface. This actually wasn’t an issue and was more of a copy and paste with a little modification since the wizard ( https://oiral.tumblr.com/) helped with, and implemented such a thing back near the start of the project (since future mechanics i want would need it). 

Once a few of the effects were done, i tied sound to it all. The sound of Stan hitting a wall at high speed then falling from a height to go splat was very amusing (more then it should have been. So much so i sat back watching and listening to him hit objects for far too long before getting back to work).

One of the criticisms i had when testing was the need to feel more like a stunt man. With that in mind one of a few ideas that came to me and hope to do was having stunts to have more impact and weight. a sense of actual danger. Having Stan hit the ground or wall at high speed would have An ambulance race across the screen when having to repeat a failed attempt.

Although amusing at first, because of the trial and error nature of the game, this might get annoying fast, so i still need to tone it down and call it every other time rather then every time. plus need a siren sound effect.. 

Talking about sound, Greg reached out after reading my last post and looped the track, so guess that takes care of that job (anything i don’t have to do is awesome!). 

 The import of some world assets was great and really brings the game to life. Stan has yet to be done yet but he is coming.

(no lighting or post have been done)

With trying to get the feel for being a stunt man, i tried to think how someone like that gets started. In my head they probably have always done crazy things like jump from the roofs of builds etc. (no i haven't researched that haha just a guess).

i had done a tutorial that would walk you through placement of items and the gameplay, but it wasn’t well implemented or well received. Oddly, even though i had the play button flashing to indicate to push it, people completely glossed over that fact (could be the play testers, but il take it on myself that maybe it way not clear).

To date the above gif is all i have done in re-inventing the tutorial. By removing all buttons but the play button, this would see no confusion in what needs to be pushed. This was to be one of a set of small  tutorials that would walk the players though the play style of the game. In a brain storming meeting with my partner, this didnt seem the best way to go about it, so this too will be replace with something a little more better suited and the above will become part of the earlier levels as to how Stan becomes a stuntman. 

       So i was going to leave the assets all 3D since they are all very low poly, but due to performance testing, i plan on changing all assets into 2D sprites. Because i want it playable on most phone and all are different, getting the best  performance is of the most importance. 

So as you may see, turning the world into sprites made a huge difference. The aim was to get Batches and SetPass calls down as far as i could. 

Running two tests for each, i set one at monitor screen size and the other at phone screen size just to see if that helped at all too. From what i took away was models into 4K sprites were the way to go (will drop the size even more too), dropped both Batches and SetPass calls by roughly 3/4, and this was without any optimization in any other way then switching out assets or Occlusion culling (as i did on the top left, which would see even better results. Occlusion culling = not seen, not calculated). The screen size saw most notably a spike in FPS and less strain on the CPU (not fully sure if this would truly transfer over to the phone yet since phones although small, actually are high res, but interesting to know anyways). 

On the subject of performance, something it seems i will need to do is object pooling.. because the game is trial and error, it has people creating and destroying items constantly.. which is a major hit to the game performance.. this means a complete over haul of what systems are there currently. Not as hard as i thought it would be, but will take some time before i can get to it..

For those that are unsure what that meant, the way i understood it when learning about it was - the current system works like this - 

        I create items, and if i don’t want or need them, i throw the items in the bin, then when i need another item, i create another.  As you might notice this is not really the best system, as the rubbish collection will become a problem, and its a bit of work to always create a new item from scratch.

         Object pooling is more like - I create a new item,  and if i don’t want or need them, i place them in a store room. Then when i need another item, i go pick that item up off the shelf.   

The UI placement is getting an over haul for the next play test too, to see if it works better then the last. Also UI placeholders replaced with actual assets. Although these are still not final as they may not fit the game once it all comes together. 

Test of the new layout and look. ignore the font.. 

i also did a small intro screen for when booting up Stan, needs quite a bit of polish still, but as a first pass i think i can live with it for now.