Jump to content

  • Log In with Google      Sign In   
  • Create Account


Ambient occlusion simulation for AAA projects


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • This topic is locked This topic is locked
91 replies to this topic

#61 dummynull   Members   -  Reputation: 90

Like
-1Likes
Like

Posted 11 August 2011 - 01:55 AM

So, who set -1 to my previous post? Does the truth hard so much? Or you just blind/daltonic winapi gui programmer? I hate everyone who makes "next-gen" graphics of green, yellow and brown colors with crasy amount of bloom and hdr(oh yeah, hdr for games is the separate topic).

Posted Image

Uploaded with ImageShack.us

Sponsor:

#62 Hodgman   Moderators   -  Reputation: 27838

Like
2Likes
Like

Posted 11 August 2011 - 02:01 AM

If this is not interesting for AAA developers, they are idiots.

Images by themselves aren't interesting. Also a good thing you're using a new account here with that kind of attitude.


Also, as I said earlier, developer's aren't interested in buying secret ideas.
The reason for buying middleware are:
1) it's quicker / cheaper than implementing the idea yourself.
2) you don't have the expertise to implement the idea yourself.
3) to reduce risk by using a proven implementation.
4) to reduce risk by gaining contracted support and guarantees.

For a AAA game developer, then reason #2 is invalid -- they will have the expertise to build it themselves.
Reason #1 is only valid if you're offering an implementation of an idea that they already want; if they don't know what your idea is, then this reason is also invalid.
Reason's #3 and #4 aren't valid for you if targeting AAA developers either, as you don't have a console implementation or an existing middleware company with a track record.


Perhaps you should re-think your business plan, and target a demographic other than AAA developers? Or consider a different way to profit from your work, such as using it to get a job for a AAA developer?


So, who set -1 to my previous post?

I set -1 to your previous post, because I thought you were very rude to David, who was simply asking for an explanation of the pictures. He's a professional who knows what AO is, but without any text to go with those pictures, they were fairly meaningless.


#63 dummynull   Members   -  Reputation: 90

Like
-1Likes
Like

Posted 11 August 2011 - 02:42 AM

Hangman
Ok, understood. There are no ways, nobody interesting. Last years i was unsure about correct choice to be graphic programmer, now it's clear that better to change specialisation, all game graphics is on the edge without future, dx11, 12, 13 won't help here. Well, i'll ask producer and manager today what about of making same kind visual fx as other modern games, if it's enough for most successful companies.
Good luck everyone, and good work.

#64 Hodgman   Moderators   -  Reputation: 27838

Like
7Likes
Like

Posted 11 August 2011 - 02:52 AM

Has anyone really been far even as decided to use even go want to do look more like?? You’ve got to be kidding me! I’ve been further even more decided to use even go need to do look more as anyone can. Can you really be far even as decided half as much to use go wish for that? My guess is that when one really been far even as decided once to use even go want, it is then that he has really been far even as decided to use even go want to do look more like. It’s just common sense.

#65 bwhiting   Members   -  Reputation: 651

Like
1Likes
Like

Posted 11 August 2011 - 03:08 AM

the bloom comic combined with the latest epilogue from hodgman just made my day

#66 fanaticlatic   Members   -  Reputation: 458

Like
0Likes
Like

Posted 11 August 2011 - 03:14 AM



It is true what everything above has been said.
I did tried the same thing with my shadow mapping algorithm (CSSM) and got the same response...
I've pitched idea to Crytek and instead of showing interest for the algorithm, Cevat Yerly replied with the yob offer (witch I ended up declining because working on technology that is more inferior than my home projects for mire 4.5K euros a month would probably drove me insane - it would be a waste of my time).

The thing is I want to share my algorithms with the rest but it is getting so darn hard to public them (my old mentor is helping me rewrite them properly).
Basically for the PHD you only need three mayor papers so it would be a shame if I just release them without profiting in ECTS points.


you declined to work at the company thatmade the Crysis series :o ?
im pretty shure they have very sophisticated thechnology.
i would have accepted.


I do admit they have done some amazing work with the shaders. It is most impressive, but everything else is just average (somethings even below average)...
Engine design today is just awful (not just Crytek, but every mayor player including Carmak and the rest). Engines are very limited and most of the development goes into graphic efects which in my opinion shouldn't even be part of the engine.

Consider this graphic engine design and compare it with how things are generally done and then make conclusion if thing that Crytek is doing is advance or not:
  • graphic engine without limitations - fully generic graphic engine (doesn't try to understand the data just how to send it to the graphic card) - can be written in less than 2000 lines of code and there you go, you have the holy grail of engine design and no one will ever be able to write better (DX11) graphic engine. You can send it voxel data, triangle meshes, sparse voxel octrees, + anything else that will be invented and you newer have to change a single line of code because the engined doesn't care about the data - it just cares about the rules for calculating appropriate rendering technique for the supplied geometry data, material data and render mode (lighting, depth, color, z-prepass..) - all this + the shaders can be put in game data (not engine). NO ONE DOES THINS LIKE THIS - everyone is harcoding effects/passes into their engines which is too much work and more and more limited. I hate when they do tech demo of some new graphic engine and then spend 15 minutes showing graphical effects. The moment they do that, I know that their graphic engine is crap. Instead, show my how flexible it is - will it be able to run something that will come up in 3 years from now without me having to "heavily modify it" (I hate when they say that) or buy new version of the engine.
  • Camera Space Shadow Mapping was developed as a single line of code implementation for shadow that will never fail you and will work for every case (even big pointlights). Basically it is just one function that generates 3 shader parameters. If at one point in time we decide to user sparse voxel ocrees or something else - no problem - in our shaders we just ignore those 3 parameters. With other engines multiple shadow algorithms are implemented and they take on most of the graphic engine code that will eventually be obsolete. The moment you implement some other shadow algorithm (cascaded, cube, volumes...) you've made assumption about how things will be render and that means that your graphic engine is not generic any more - it is not all-powerful and it will soon be full of old useless code....
Don't get me started on physic engine limitation, network limitations, spatial limitations, logical limitations... It will newer be able to run a game with land-able dynamic planets big spaceships and some decent multilayer (I expect nothing more than peer-to-peer MMO in completely dynamic world with out of scale objects). Basically it is designed for small things - small static levels where you can only shoot at things and that is all you will ever be able to do...


I may be finding problems where there are none hear but you approach sounds very slow for triple A development and probably a reason why the type of modular engine design you mention hasn't been adopted by the mainstream game dev community as yet. If you can get around that your definitely onto a winner.

I think the approach your suggesting will certainly have a lot of current potential in regards of iphone and other handheld markets where there is a proliferation of different hardware requirements, but unless you can ensure the speed of the engine + scripts is comparable to todays engines you may struggle to get traction in the console and pc space.

Hopefully you'll post a link to your phd once completed and good luck.



"I have more fingers in more pies than a leper at a bakery!"

#67 dummynull   Members   -  Reputation: 90

Like
-4Likes
Like

Posted 11 August 2011 - 03:28 AM

Ha-ha-ha, -2 for my post with comics are showing the truth. Rest in peace, graphics programmers, you are absolutely incompetent, all you can do is just to multiply matrix and vectors. You are dead already, the appendage of computers, nothing more.
Graphics of Crysis 2, Unreal engine dx11 - all deep throat sucks. God bless japanese developers.

#68 Digitalfragment   Members   -  Reputation: 750

Like
3Likes
Like

Posted 11 August 2011 - 04:18 AM

Ha-ha-ha, -2 for my post with comics are showing the truth. Rest in peace, graphics programmers, you are absolutely incompetent, all you can do is just to multiply matrix and vectors. You are dead already, the appendage of computers, nothing more.
Graphics of Crysis 2, Unreal engine dx11 - all deep throat sucks. God bless japanese developers.


I can multiply quaternions too. Can't forget the quaternions.
Best thing about flame baiting and insulting people is that it takes no skill. Incompetent people can do it just as well as anyone.
Not sure how incompetence comes from being honest and clear about how the industry works.

>Not really interested in the AO at all after seeing those as far as quality is concerned. For any interest, it would have to be proven to be faster than most other AO implementions and have zero impact on artist workflow/pipeline. Also, you're not likely to get an interest from other developers by your attidude
Why should i proove here something? I told that already, it's faster, it's not ssao.


I'm not asking you to prove anything here. I'm saying that if you seriously expect any AAA developer to take interest, that's what it would take. Saying its faster because its not SSAO means nothing. I can say the world is flat, but it doesn't mean it really is. You've created a technique that runs on PC hardware, great. However:
- Performance can be drastically different to the point where its unusable on consoles because of differences in hardware architecture.
- Non screenspace tricks don't scale nicely with scene complexity. Does your trick rely on lots of preprocessing? Then its useless for anyone that wants dynamic scenes.
- Does it add a crazy amount of overhead memory or CPU/GPU time per vertex? Then its a hard sell, as thats why we all moved to SSAO instead of topological AO solutions.
- Is it just a signed distance field inside of a volume texture? That's nothing new, and its a crazy assed amount of overhead by the size of the object, which is worse than per vertex.


In the end, no one cares about smart asses and snake oil salesmen. Yes, your screenshots are nice to look at. Granted. But those alone aren't going to be enough for anyone to buy in.

#69 bwhiting   Members   -  Reputation: 651

Like
1Likes
Like

Posted 11 August 2011 - 04:54 AM

how bout a demo (one you dont even have to share)

a demo that implements a couple of version of ambient occlusion i.e. a few SSAO methods, maybe one on the CPU, and any other methods you have come across.
then compare them with yours in a test scene at a high resolution

then just simply post the framerates achieved.

something like (full HD res, complex scene x-polys, y-objects). :
v1 - 48fps
v2 - 80fps
v3 - 51fps
mysecretversion - 113fps

that coupled with a few screen shots should be enough to at least get some attention.
talk about the benefits of your method... maybe even more attention.

but there is no guarantee that anyone would buy it, some will wonder how you did it..some might try and replicate it, others may just admire it. who knows!

maybe its better to publish your method with your name all over it and call it after yourself, that could prove better in a couple of ways:
1. everyone gets to benefit from your work! games look better and it might inspire someone to do something else with it.
2. you might make more money though better jobs/clients/whatever gained because your name will be attached to it!

just a thought

#70 dummynull   Members   -  Reputation: 90

Like
0Likes
Like

Posted 11 August 2011 - 05:42 AM

Digitalfragment
>- Performance can be drastically different to the point where its unusable on consoles because of differences in hardware architecture.
Sorry, if i don't have devkit this doesn't mean i don't know about them.

>- Non screenspace tricks don't scale nicely with scene complexity.
Partially, depends from games/engines fx already implemented. Many games with depth rendering (deffered or separate prepass) will work fine at no cost for non screenspace calculations.

>-Does your trick rely on lots of preprocessing? Then its useless for anyone that wants dynamic scenes.
No, even animated characters(actors) working fine. Haven't i said that already?

>- Does it add a crazy amount of overhead memory or CPU/GPU time per vertex? Then its a hard sell, as thats why we all moved to SSAO instead of topological AO solutions.
No.

>- Is it just a signed distance field inside of a volume texture? That's nothing new, and its a crazy assed amount of overhead by the size of the object, which is worse than per vertex.
No. It's not and no overhead like that. The size i used allow to lit good buildings up to 50*50*50 meters of size and small objects like cars, peoples. Objects like guns are too small, so better to combine with other methods.


bwhiting
>how bout a demo (one you dont even have to share)
Job is the obstacle, can't learn different engine api fast enough and i'd like to sale this as fast as possible, not advertise several months, better to spend that time for other work, not abstract profit. Don't know, may be athoring software like Virtools let me do this faster, i'll check out.

#71 kauna   Crossbones+   -  Reputation: 2165

Like
0Likes
Like

Posted 11 August 2011 - 06:20 AM

No. It's not and no overhead like that. The size i used allow to lit good buildings up to 50*50*50 meters of size and small objects like cars, peoples. Objects like guns are too small, so better to combine with other methods.


What is the memory foot print for building of 50x50x50 meters? How long is the preprocess time?
Practically, you are saying that SSAO or a variant is needed with your AO algorithm to handle small scale objects such as weapons?

Best regards!






#72 samoth   Crossbones+   -  Reputation: 4521

Like
1Likes
Like

Posted 11 August 2011 - 06:40 AM

Look, you're not going to get anywhere here like this.

If your claims are true (i.e. you have a working implementation which has been secretly implemented in a game with 10k+ players), the most reasonable chance that you have is to get a verifiable proof for this claim from this game's creator (this could be something like "Uses the awesome secret technique X by Dummynull" on their blog, or they could give you a kind of written statement). Whatever it is, it should be verifiable.

Then send this evidence, a demo or a copy of the game / screenshots / detailled high quality video on a CDROM, and a description of how your algorithm works to a handful of major studios (whichever you deem "good enough" for your standards). Do not mention a NDA (that's just ridiculous from a solicitant's position), and try hard to avoid calling them idiots or saying that either their product or a notable competitor's product sucks. Arrogance does not sell. If you cannot risk (do not want to trust) them to look at your stuff, stop here and find a different occupation. Seriously. You'll not get anywhere.

Be sure to use proper language and enable your text program's spellchecker and have someone else read over your cover letter (twice and thrice!), and use all best practices (proper envelope, good paper) as if applying for a job.
If the executive assistant's assistant's assistant who is opening the envelopes isn't satisfied with the overall impression during the first two seconds, your technique may be as awesome as it could possibly be... it won't make it to the desk of anyone in the development department, but go right into that round filing under the table.

Your chance of getting an answer will still only be somewhere around 1% to 5% (depending on how stunningly awesome your technique is), but in every other case, your chance is exactly zero.

Also, you might want to edit your posts or ask a moderator to delete this thread because in case someone does look at your stuff and then googles a bit, it would be quite unfavourable if they came across a thread full of statements such as "they're all iditots" which aren't precisely the kind of thing you expect to hear from someone you do business with.

#73 dummynull   Members   -  Reputation: 90

Like
0Likes
Like

Posted 11 August 2011 - 07:17 AM

kauna
>What is the memory foot print for building of 50x50x50 meters? How long is the preprocess time?
On the screens i showed earlier it was 2 megabytes and it's computed at distance about 500 meters around (more memory means greater detail of course). Additionally memory for textures required for bilateral filtering and post process texture, the same as ssao requirements. But for some situations it's better to make several computations, closer to camera, n*2mbytes.
Preprocess doesn't exist, everything is realtime.

>Practically, you are saying that SSAO or a variant is needed with your AO algorithm to handle small scale objects such as weapons?
Yes, in general. It possible to make small size computations also (several size scale could be computed faster, like ssao with more smaller range sampling), but for better quality i should spend linear computation time for small objects, so it's about twice slower than for big and middle size objects (except filtering, it's constant, as in ssao, could be done in same pass with ssao filtering). Also i did good enough ssao (require normals, slower) as complex solution, but it doesn't matter and could be done by anyone after some reseach. Here is an examples of complex ssao i tried to mix (without now AO):


Uploaded with ImageShack.us

SCREENSHOTS REMOVED
Uploaded with ImageShack.us
These are even better with new AO together.

#74 way2lazy2care   Members   -  Reputation: 782

Like
0Likes
Like

Posted 11 August 2011 - 07:58 AM

Has anyone really been far even as decided to use even go want to do look more like?? You’ve got to be kidding me! I’ve been further even more decided to use even go need to do look more as anyone can. Can you really be far even as decided half as much to use go wish for that? My guess is that when one really been far even as decided once to use even go want, it is then that he has really been far even as decided to use even go want to do look more like. It’s just common sense.


http://horror.break.com/exploding-head-montage-head-trauma/exploding-head-montage-head-trauma

Awesome.

#75 MJP   Moderators   -  Reputation: 10243

Like
1Likes
Like

Posted 11 August 2011 - 11:28 AM

I think it's now obvious that you were setting your sights too low with AAA developers. Clearly, you should be going for at least AAAA developers.

#76 dummynull   Members   -  Reputation: 90

Like
0Likes
Like

Posted 11 August 2011 - 12:22 PM

Guys, i'm bored. When i came here to ask, expected something like this, but not such negative to me without any trust. How dare are you to tell me that i'm lier, bad and rude if you guys just trolling me? What actually do you want from me by saying all of that? I'm not the competitor, i'm not advertising, i wasn't angry. If someone making things better than others, is that bad? You are redicilous if it's like that. Proove me then that you can do better, otherwise shut up or just help me.

#77 gsamour   Members   -  Reputation: 140

Like
1Likes
Like

Posted 11 August 2011 - 12:33 PM

Guys, i'm bored. When i came here to ask, expected something like this, but not such negative to me without any trust. How dare are you to tell me that i'm lier, bad and rude if you guys just trolling me? What actually do you want from me by saying all of that? I'm not the competitor, i'm not advertising, i wasn't angry. If someone making things better than others, is that bad? You are redicilous if it's like that. Proove me then that you can do better, otherwise shut up or just help me.



Judging by your attitude, you're looking like the troll to everyone else. I wish I could help you, but I don't have money to give you. Even if I had the money, I have a choice of many algorithms to choose from. Many of them are freely available, so why would I pick yours? If you say it's because of quality, then I'd say this... if I was making a simulation where I absolutely needed your AO, then I'd talk to you (but how can I tell if I need your AO if I don't know your history and I've only seen a few images?). But if I'm making a game where graphics is just one part of a much larger equation, then I would think twice before asking for your top secret algorithm.


#78 dummynull   Members   -  Reputation: 90

Like
0Likes
Like

Posted 11 August 2011 - 02:34 PM

gsamour
I don't know what you or anyone else here calling good quality of ao (realtime, any method, gi or other complex allowed) and what kind of objects or scene is correct to compare. Show me the screenshot how ao must looks like in final game and if this technique exist (dynamic objects at least). What it changes if i'll do photorealistic demo and developers ignore that the same way as now? Just show me what is right.

#79 gsamour   Members   -  Reputation: 140

Like
0Likes
Like

Posted 11 August 2011 - 02:46 PM

gsamour
I don't know what you or anyone else here calling good quality of ao (realtime, any method, gi or other complex allowed) and what kind of objects or scene is correct to compare. Show me the screenshot how ao must looks like in final game and if this technique exist (dynamic objects at least). What it changes if i'll do photorealistic demo and developers ignore that the same way as now? Just show me what is right.


There is no "right" and "wrong". The answer is "it depends". For example, why would a AAA developer want photorealistic results on a cartoony game?

#80 mind in a box   Members   -  Reputation: 552

Like
1Likes
Like

Posted 11 August 2011 - 03:12 PM

What it changes if i'll do photorealistic demo and developers ignore that the same way as now?


If your technique really works as you described and at the same quality you have shown, and even at a decent speed, then you created something awesome, indeed.

Can't you simply publish a paper and send this to some of the companies?
Wouldn't it be nice to have a few projects listened at the bottom where it has been used already?
Maybe some cool indie-game catches it up and those AAA-Devs get so see it in action just in time?

I've been monitoring this thread since your very first post. But I do not get your point.

Are you afraid that someone makes your technique "better" than you got it?
Think of the Crysis' SSAO. Everybody knows that this was the first in-game implementation, even though there are so many more now.

What it changes if i'll do photorealistic demo and developers ignore that the same way as now?

Who said that all developers ignore you? They're just angry because you're saying pretty much this to them:
"I have that awesome tech, but I won't tell you little guys how it works :)"

That makes the interested people kind of angry, don't you think?




Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS