Work In Progress / Ignite (Working Title)

Author
Message
MXS
Valued Member
15
Years of Service
User Offline
Joined: 10th Jan 2009
Location: Cybertron
Posted: 11th May 2020 20:04
this is the most interesting game made by game guru. you have set the bar high with the game play.
more than what meets the eye.Welcome to SciFi Summer

Windows 7 home premium 64bit gtx770 sc acx 2gb gpu boost 2.0

AmenMoses
GameGuru Master
8
Years of Service
User Offline
Joined: 20th Feb 2016
Location: Portsmouth, England
Posted: 11th May 2020 21:37
Of all the projects I am involved with this is probably the most fun one.

The scripting challenges are sufficiently interesting and challenging and I find the whole concept to be fresh and game worthy.

I really hope this makes it to a fully fledged game release!
Been there, done that, got all the T-Shirts!
PM
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 12th May 2020 17:40
Thanks Amen, that means a lot coming from you - I really appreciate all your help on the project

AE
synchromesh
Forum Support
10
Years of Service
User Offline
Joined: 24th Jan 2014
Location:
Posted: 18th May 2020 00:36
I cant fault it .. just amazing.
Its a game I want to buy and play.
The only person ever to get all his work done by "Friday" was Robinson Crusoe..
PM
darimc
GameGuru TGC Backer
17
Years of Service
User Offline
Joined: 19th Jan 2007
Location: Nova Scotia, Canada
Posted: 21st May 2020 13:38
Just here to echo what everyone else said -- the game looks fantastic! I think you have a really creative and original idea here, and it's clear you put a lot of care into your projects. I'm definitely interested to see where this goes.
Signatures are overrated.
Wolf
Forum Support
16
Years of Service
User Offline
Joined: 8th Nov 2007
Location: Luxemburg
Posted: 16th Jun 2020 23:09
Jupp! Really inspiring to see what you came up with so far.
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 10th Dec 2020 23:02 Edited at: 11th Dec 2020 00:07
UPDATE
Well, well, well, long time no see. 2020 has been a busy and creative year for me; I wrote two short films, started a YouTube series, released assets onto the store, and found my workload at my day-job triple the moment we went into lockdown. Notice a pattern? None of that had anything to do with Ignite. I guess I was just burnt out (no pun intended) and couldn't face it.

One reason I lost interest in Ignite over the summer was my frustration with an early three-level build that refused to work as intended. I mean, if I couldn't get three simple levels to run correctly, what chance did I stand of getting six or seven levels done? I was also at a crossroads regarding whether I should port the game to Game Guru Max, in hindsight I'm glad I didn't.

Instead I spent a lot of time thinking about texture optimisation. It could be that one of the reasons Ignite wasn't running correctly was because I had made little effort to optimise it. I figured I'd have to eventually, I just didn't expect to start hitting memory issues after only three short levels. Still, whether or not lack of optimisation is the problem, it's still a good habit to get into.

What follows is a long and rambling post about my journey into the world of texture optimisation.

Texture optimisation
Optimisation is all about saving resources to improve performance. When it comes to textures, there are two basic attributes you can tweak to save memory; resolution (size), and compression (format). Let's talk resolution first.

Traditionally, assets bought made by third parties are never optimised by their creators - I'm guilty of this too! Why? Because we don't know what the user will want to use that asset for. It makes sense to provide textures in the highest quality and highest resolution you can to give the user more freedom. But most people won't notice the difference between a 4096 x 4096 (4K) texture and a 2048 x 2048 (2K) texture in-game, not unless they are either getting really close to an asset, or it's really big, or they are running a big high-resolution monitor. To test this out, let's look at this switch I made for Ignite ages ago. I originally mastered it with a 2K texture.



As you can see, the difference between 2K and 1K is almost negligible, and even at 512 x 512, the texture still holds up at a distance. Only at 256 x 256 does noticeable pixilation occur, but at this point your texture is 1/64th the size of the 2K original.

Every time you half the length of a side of your texture (e.g. 4096/2 = 2048, or 2048/2 = 1024), you divide its area by 4. This means a 1024 texture is a quarter the size and therefore takes up a quarter of the memory of a 2048 texture. A 512 x 512 texture takes up a quarter of that, so 1/16th the memory of a 2048 texture, often it's even less than that. Here's how those four examples seen above measure up (we're still working in PNG for now):

2048 x 2048 = 6.24MB
1024 x 1024 = 1.46MB
512 x 512 = 343Kb
256 x 256 = 79.7Kb

Now obviously I didn't just want to go around converting every texture to 256 x 256, it would look awful! But I spent a week going through my third party and self-made assets and, one-by-one, reducing the texture resolution down until it became unacceptable. Again, it's all about the amount of screen space the asset occupies. Here's a big wooden table, it takes up a lot of the screen because the player can get really close to it.



In the end, I still went with a 2K texture here because the player can see it up close so well. But that's still a quarter of the size it used to be. I always tried to go one step further than I thought I could, just to see if the texture quality fell apart.



Equally, there were other entities that are tiny in-game, yet had enormous textures. I'd made this combination lock to have a 2K texture because I felt it was important the player can read the numbers. But it turns out, you can still read them perfectly well at 256 x 256, because the entity is so small in-game, and there's enough contrast in the texture itself.



Of course, many assets for Game Guru now take advantage of PBR; they'll have not one texture map but usually three or four; _color, _normal, _gloss, _ao, _metalness, _illumination, or any combination of those. The saving in memory only multiplies with additional textures. 5 x 1024 textures make take up 10MB of memory, whereas 5 x 512 textures would take around 2.5MB (usually even less because _ao, _metalness, and _gloss are greyscale). Game Guru Classic is a little behind the times when it comes to texture stacking; in other engines, the _ao, _gloss, and _metalness maps, which are all greyscale, can be combined into a single texture, as in Max's use of '_surface' textures. Classic can't do that, but that does give you the freedom to resize the individual maps as you choose, potentially further saving memory.



The other tool at your disposal is format. PNG textures are popular because they offer less compression than JPG, and allow for transparency, AKA alpha channels. However, every PNG texture you use must be converted to DDS by your graphics card in order to display. This conversion happens behind the scenes in Game Guru and puts a little extra load on your graphics card. Some people will tell you Game Guru is limited by its memory cap, not its graphics, therefore the smaller filesize of a PNG is better than a bigger filesize of a DDS. This is rubbish, DDS comes in a multitude of flavours and can be compressed in a number of ways to make it both better for your GPU, and your RAM.

I don't know if this is just Game Guru, but another unintended up-shot of DDS textures over PNG is that they appear crisper in-game. Maybe this has something to do with your GPU not needing to convert them.



A good program for converting textures is Paint.net. But it comes with a huge list of compression types to choose from, so how do you know which is best?

Some people swear by BC1 compression; it's been around the longest and is the most supported. However, I find the compression to be too harsh, and it also only features a 1-bit alpha channel, meaning your alpha is either on or off, opaque or transparent, there's no ability to do semi-transparent materials. Yes you save filesize, but at what cost?! The compression on BC1 is also quite noticeable. In order to save filesize, BC1 groups similar colours together into colour blocks, which often results in a subtle but noticeable shift in colours, and strange mottled patterns throughout the texture. These artefacts that are particularly noticeable in low contrast, low saturation areas like on metal but you'll see these purple and green blocks pop up on lots of DDS textures, including in stock Game Guru assets.



To save time, here's the compression formats I use now, having done lots of experimentation:

_color = BC7 (little compression, 8-bit alpha channel allows for semi-transparent materials)
_normal = BC3 (a little compression, but less than BC1. You could go BC1, but the green and purple colour-blocking you notice in the above example can mess with the subtleties of your normal map)
_ao = BC4 (a greyscale format, discards all colour data, saving you about 50% filesize)
_metalness = BC4 (see above)
_gloss = BC4 (see above)
_illumination/_emissive = BC7 if low compression is a priority, BC1 if solid colours of illumination.

All of this combines to save some serious memory. I had a set of 4K PNG textures weighing in a 27.1MB. After resizing them to 2K, and saving them as DDS with the above compression, I got that down to 12MB, 44% of what it was originally. Every megabyte saved is a little less strain on the already inefficient memory management of Game Guru.

I've enjoyed doing these experiments, laborious though it has been. Now that winter approaches, I'm gradually beginning to pick up Ignite again, but from now on I'll be going through every asset to ensure it both looks great, and is optimised to ensure smooth, and hopefully stable, gameplay.

My fingers hurt now

AE
Wolf
Forum Support
16
Years of Service
User Offline
Joined: 8th Nov 2007
Location: Luxemburg
Posted: 11th Dec 2020 04:34
Impressive and conclusive write up, well done.

I have nothing more to add except kudos!
PCS
7
Years of Service
User Offline
Joined: 7th Jul 2016
Playing:
Posted: 12th Dec 2020 18:31
This work is outstanding , keep up the awesome work
Windows 7 Professional 64-bit
Intel(R) Pentium(R) CPU G3260 @ 3.30GHz (2 CPUs), ~3.3GHz RAM 16GB NVIDIA GeForce GT 730
DirectX Version: DirectX 11
GubbyBlips
5
Years of Service
User Offline
Joined: 14th Jan 2019
Location:
Posted: 19th Dec 2020 19:36
Take notice! Great info, wish this was more acknowledged.
PM
Nevin
8
Years of Service
User Offline
Joined: 19th May 2015
Location: U.S.A
Posted: 21st Dec 2020 09:23 Edited at: 21st Dec 2020 13:55
That was a good read. It's given me some good ideas

I am vary good at not getting burnt out when working on projects. What I do is:

1. Set a release date. This is a big one. When we get close to that date our minds go in overdrive trying to finish up. This is good b/c your body wants to hit that goal. You can push the date back if needed.

2. Make list. As you finish a task, check boxes. The mind loves to finish task.

3. Make time to work on your project. But make sure its scheduled. 1 hour a day, or what I do is only work on projects in the morning.

4. when working on your project no fackbook, youtube, etc ... Music is fine ( no videos of music, this gets me )

5. Even though its more of a hobby. Act like its your job.

6. Don't add everything. Don't add every possible feature you can think of. But save them and think of what's needed. The rest can be added when your ahead, or after release.

7. long walks, a hot shower, or shooting some hoops (basketball) helps me think of ideas.

8. play other games. games, even ones that are not the same as the one your trying to make can give you inspiration. do this and number 7 if your stuck, out of energy to work on project, or feel you need some downtime from the project


9. focus on one project. if your making a game then dont make another game at the same.


10. dont procrastinate. if you have the energy and time dont waste the day on tik tok

I hope this helps
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 6th Feb 2021 10:57
In my previous update, I wrote at length about my journey into the world of texture optimisation. Turns out, I was only at the beginning of that journey and much of what I wrote in that post was only half-true. Since then I've done a lot more research and tests, and I've come to some different conclusions now, particularly over what format to use when.

I've decided to share my newfound knowledge with the community in a comprehensive three-part tutorial. Check it out if you're interested in getting better performance out of your games:
https://forum.game-guru.com/thread/222504

AE
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 11th Mar 2021 23:59 Edited at: 12th Mar 2021 00:02
This was a test post. I'll do a proper update soon, but since you're reading this, maybe you'll appreciate this model I made recently; a smashed up wooden walkway, which is now a small jumping puzzle for the player in level 1 of Ignite.



AE

Attachments

Login to view attachments
Wolf
Forum Support
16
Years of Service
User Offline
Joined: 8th Nov 2007
Location: Luxemburg
Posted: 24th May 2021 13:48 Edited at: 24th May 2021 13:49
Hello there!

I recently played the version you've sent and here are some of my impressions:

Overall I liked the idea, with a big asterisk attached... I will go into that later on but you already know what is coming.

First of all, the really good stuff: The intro cinematic sets the mood, is well made and goes on for just the right amount of time. Very well done and very clean presentation. A trend that will go through the rest of the game.
The HUD models of the torch and candle are brilliant! The fire effects are convincing and seeing the damn candle burn down in real time really surprised me. Lately I see more and more people do things in game guru I did not expect to be possible. This game here is one of them.
Now the gameplay is very stripped back and simplistic which I enjoyed a lot. Modern games tend to drop a ton of features that tend to vomit hud elements all over your screen so seeing something so minimalistic where you take things one step at a time was decent. Especially since you relied on the most basic of plattforming and switch puzzles but still managed to make it engaging.
Sound design is decent too, especially the disembodied whispering that grows louder when the candle burns out.

However, there are some things that didn't quite hit the mark for me.
The dungeon you traverse is not bad but it looks very clean and cartoonish. There are some really neat areas and light effects but its too sterile to have any atmosphere to me. The game tries to have a certain creepy vibe to it but the whole castle was too friendly, orderly and brightly lit to really sell me on that. I think some different textures and a few more broken clutter objects would help. This game revolves around light and darkness at its core and that is sadly, due to technical limitations its weakest part. Especially since I also have a flashlight

Also it feels generally more like a level than a place. I never knew what part of a castle I was supposed to be in. this is not necessarily a bad thing if you take more of a "functionality over immersion" approach but I figured I'd point it out. Other things I noticed are weird collision where walking up to a wall pushes me away from it. I've never seen this before in a GG game so I don't know how this happened. Player movement speed and jump height feel sluggish as well, I'd maybe tweak it a little. Depends on how you like it of course. Then there is the fact that falling into the water didn't do anything to my candle You can see that in my playthrough video

Now on the big asterisk *



That being said, there is a lot of creative work in this and I'm looking forward to how it progresses.





-Wolf
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 24th May 2021 19:29
Hello everyone. Just to put Wolf's post in context, I recently released a short three-level standalone of Ignite to a select few Game Guru developers to gauge their opinion of it. I had intended to do the same thing a year ago but my 2020 build was riddled with bugs and instability problems, largely because I'd spent no time optimised it. At all. Flash forward twelve months, and now I have a fairly solid build of the opening three levels which, although not without its bugs, has garnered some insightful feedback which I'll bear in mind going forward.

Now, Wolf, to address your feedback directly

I'm pleased you enjoyed this vertical slice of the game and that some of the details like the fire on the torch and the candle burning down impressed you, it certainly took a lot of hard work from AmenMoses to get those working! It was a deliberate choice to pair things back in these opening levels as I didn't want to overload the player with information. Like you, I've played a number of modern titles that pile these mechanics on far too heavily in the game's opening chapters, whereas I prefer to introduce things gradually and with a purpose. I noticed your recording cut out after level 2, was that because of a crash or simply because you stopped recording? I'm just curious because level 3 introduces some more gameplay elements and another type of light to keep things interesting.

One of the reasons my 2020 build failed miserably was a lack of optimisation. I've been on a journey of discovery these last twelve months and part of that has involved recreating almost every asset from scratch, but this time with optimised textures. I've also rebuilt all three levels from scratch. In an effort to not stress out my graphics card too much, I deliberately kept the decoration and architecture sparse in this build, but I do intend to add more later on once I know how much VRAM leeway I have. This may not start to happen until I have more levels in the game and I know roughly what my VRAM requirements are. You make a good point about the levels feeling like levels rather than places; I'm aware of it, it bugs me too a little, but I'm finding it hard to balance the need to control the flow of information to the player, the pacing of the level, and the overall functionality of the space. I'm hoping later levels, which will be less tutorialised, will not suffer from this quite so badly.

It's worth noting the lighting effects themselves have turned out rather differently in standalone than they do in test game. The lightmapping is less consistent in standalone and, as I'm sure you noticed, the dynamic light that is supposed to be paired with the flaming torch doesn't appear to work at all in this build, I know not why. I've also never seen the water bug before; water is supposed to kill you! Definitely some kinks to iron out.

Lastly, to address the big asterisk - it's the elephant in the room at this point - yes, I see what you're saying and I do agree that Game Guru is far from ideal for making this sort of game. Unfortunately, I'm a stubborn so-and-so, and I feel compelled to finish it to some degree before I stand any chance of turning my back on Game Guru. Even if the Game Guru version of Ignite functions as little more than a test bed for a Unreal/Unity/[insert whatever other engine here] version, it still has the potential to teach me an awful lot which will come in handy if I do decide to remake it elsewhere. Hell, I've remade it four times in Game Guru already and each time I've learned something new! You yourself still toil away for months on FPS Creator games despite that engine's own shortcomings, so I'm sure you understand the value of "the journey" as opposed to just the destination

Anyway, I'm super happy you played my game and it didn't make you wince! Still lots to be ironed out, polished up, reconfigured, but it's nice to finally have something to show for 2+ years of work. Thanks for playing!

AE
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 24th Jun 2021 19:01
VIDEO UPDATE
It's been quite while since my last update, so I felt I should do another developer vlog to summarise my learnings and progress (or lack thereof) in a more engaging way. Hope you like it!



Whilst my progress on Ignite has been a little sporadic this past year, one thing that hasn't changed is my gratitude for your support. My thanks go out to everyone in this community for their continued help and support; your comments really motivate me to keep working on this!

AE
GubbyBlips
5
Years of Service
User Offline
Joined: 14th Jan 2019
Location:
Posted: 27th Jun 2021 05:58
Hi there AE. Nice podcast- again.
Lots of work, so very much work, frustration and some success?!
What I think would be interesting is if you could get a demo version of Ignite finished at whatever level of functionality is possible in Game Guru Classic. Perhaps just the 3 levels you mention. And then if MAX would EVER get released and usable..., see if any of the project would port over to it.
That would be a good experiment/ demonstration for if MAX is gonna support any portion of Grand-pa GG. Of course that sounds like a lot more work, whether you have the interest, time and ambition? I just think it would be pretty interesting, and looking forward to playing the game at any time.
PM
Bisella
GameGuru TGC Backer
9
Years of Service
User Offline
Joined: 30th Apr 2014
Location: Verona - Italy
Posted: 27th Jun 2021 07:13
a magnificent work
The software I use? ... Autocad, sketchup, Spazio3D, pconplanner, keyshot, palettecad, Kerkythea, paint.net, enscape, AxixGameFactory, PD particles,leadwerks, s2game engine.......in the next future blender
PM
Avenging Eagle
18
Years of Service
User Offline
Joined: 2nd Oct 2005
Location: UK
Posted: 27th Jun 2021 19:34
GubbyBlips wrote: "What I think would be interesting is if you could get a demo version of Ignite finished at whatever level of functionality is possible in Game Guru Classic. Perhaps just the 3 levels you mention. And then if MAX would EVER get released and usable..., see if any of the project would port over to it."


I do think there would be merit in seeing a direct comparison between a MAX version of Ignite and a Classic version. One thing I'm particularly excited for in MAX is dynamic lights with dynamic shadows, I think that could really bring the game to life. But I think it would be foolish of me to begin a MAX version of Ignite whilst the engine is still deep in development. Maybe once v1 is released, I can look into it

Bisella wrote: "a magnificent work"

Molte grazie, Bisella!

AE
science boy
15
Years of Service
User Offline
Joined: 3rd Oct 2008
Location: Up the creek
Posted: 13th Aug 2021 19:44
Avenging if you want my dungeon kit not the new but previous which is tstill i believe good enough drop me a om and i will send it to you. If you want of course im not dissing what you have at all just offering a possible option for you. Dont have to mention me or anything as long as you dont sell the kit im happy or give it away. Same to you wolf
Up to you guys.
Anyway babbling away here
Im 50 now aint a kid travelled the world been in terrorist bombing shot at near kidnapped. Have an audi a house a fiancee a cat and ex dj and promoter and now home Manager. Have a degree and lots of quals and this is actually all true

Login to post a reply

Server time is: 2024-04-18 23:01:49
Your offset time is: 2024-04-18 23:01:49