Let's go for some participatory democracy, shall we?

G

voted for the third choice (focused goal).

definately keep a tight focus. don't get ahead of yourselves and MOST IMPORTANT don't try to "out do" hl2.

make an outdoor city map the way you feel it should be done while keeping it simple and focused. more detail and special effects can always be added later. for example, a gui that controlls the weather is a nifty idea, but that should be very last thing done.

good luck to you all...

Well I believe you must consider the real purpose for this, and everyones interpretation of it.

If the purpose is to try and show some weird comprability to HL2 then you have to aim for low end systems, and not use geometry.

If the purpose is to make a real mod type thing that is playable (w/ enemies and so forth) then it must be simpler.

If the purpose is to showcase how pretty D3 can be then I say shoot for the top end b/c anyone with a low end system can still look at the pretty pictures.

My idea would be to make it as pretty as possible while maintaining a 45frames per second on a 6800GT with a pentium 3GHz.

(BTW I have a x800pe and a athon 3200.. so I may be biased

)

What I want to see most is a well-lit environment that looks good and runs well on a mid-range computer. I don't even care if it's outdoors. Seems like it should be a simple thing to me, but no one has done it that I've seen.

I voted 3 but 1 was a close 2nd.

As rich said, it's not about beating engine X it's about proving Doom3 can do more than dark corridors.

I know it can and most mappers here know it can but so far most of the new fan made maps released are using id's textures, models and stylings.

i voted for 3 - the focused one.

if the goal is focused then it will be finished faster.

you can always add in the 'frills' of No. 1 later on if you wish, but atleast you have a base project.

making something for lowend comps is pretty crap...because it would kill all the good stuff about the DOOM engine. =/

Trying to mimick HL2 while also trying to make it as fast as HL2 is kinda pointless. Doom3 has a more advanced graphics engine and of course will run slower on compareable maps.

I voted for no. 3

I voted 1, I don't have a low-end system so I don't care for those losers

.

Duff wrote:

i voted for 3 - the focused one.

if the goal is focused then it will be finished faster.

you can always add in the 'frills' of No. 1 later on if you wish, but atleast you have a base project.

making something for lowend comps is pretty crap...because it would kill all the good stuff about the DOOM engine. =/

I'd go even farther and say its the only way it will finish. If focus is lost I could easily see little things constantly getting added and nothing ever finishing.

I voted for the first option - on Q3W I've seen endless discussion about performance and playability.

All these discussion led to mediocre work where more could have been achieved.

If we do such a project or even many of them, then we should consider to keep two things in mind:

- the capeability of the engine
- the original purpose of the engine

Most of the capeabilites we already know (some undocumented nice features are still waiting for us, but no - it's not that radiosity rumor).
The purpose of the engine was to render limited scenes with an unified lighting model.

Now trying to recreate scenerey the engine wasn't meant to render as it's prime target will always come with a tradeoff. The very last thing I'd like to see in this project are virtual obstacles put up by people who think any virtual given scene should run like butter on a shit machine.

Please - don't go the wrong turn and waste precious time over discussing how many frames are acceptable and what the limit of playability is.

Value artistical freedom over the same restrictions that led to the tradeoffs in Doom3 which humorously led to this exact project and thread!
You're about to make the same mistake twice if you tone the project down now.

I say - give a damn shit about maximum compatibility and low end performance! We are the privileged ones who can work with this technology without having to worry about the sales figures for the Wal Mart and AOHell crowd!

Such projects take time and have a strong geek appeal who, for the most part, are well equipped with the needed tech. I don't mind babyboy sending me a flame email becasue it wont run as good as CS on daddys ancient outlook pc or mommies untouched machine.

Think about it.

BNA! wrote:

I say - give a damn shit about maximum compatibility and low end performance! We are the privileged ones who can work with this technology without having to worry about the sales figures for the Wal Mart and AOHell crowd!

Such projects take time and have a strong geek appeal who, for the most part, are well equipped with the needed tech. I don't mind babyboy sending me a flame email becasue it wont run as good as CS on daddys ancient outlook pc or mommies untouched machine.

Think about it.

Well, that means it's basically gone away from rich's original idea, which I suppose is fine if that's the way it goes - to quote :-

Quote:

This thread is a community project I'm interested in starting in responce to all these "HL2 can do this" posts.

So, have you found something in a game that you don't think Doom 3 can match?

Post a screenshot. Describe it. Ect...

I'll try to prove to you that Doom 3 is just as capable. Anyone interested in lending me a hand is welcome.

I would like everything to be custom made however so that the resulting testmaps, ect... can be released for all to see firsthand running on their system.

Taken literally, that means an open area of medium realistic impact that runs smoothly on todays systems - with leeway, this obviously means todays systems that run Doom3 well.

But sure, if the project goes in the direction of pure eye candy, showcasing what will be playable in 12 months from now on top end hardware, nothing wrong.

Don't forget the title of the topic sets it up for comparison - not by the HL2 fanbois (who gives a shit about them), but by people genuinely interested in seeing whether Doom3 can really "do it too"

The way I see it is that obviously Doom3 is capable of rendering environments as "realistic" - and far more so than the square in city 17 - there's no dought about that - but can it run as effectively ?

That's why I chose 3.

I view it as a balance - balancing performance and realism, which truly would show what the Doom3 engine is capable of.

I'd prefer 3 as well. This is mainly because my machine isn't too high end, although it isn't any crappy CS computer either, heh. But I would like to play the map I've helped with, rather than having my FPS turn to Seconds Per Frame.

What about doing both 1 and 3?

I voted for 3 when I started the poll. However, I see both options possible... option three leads fairly logically towards option 1.

I mean, for starters, we need a textured map that is well-lit and maybe toss some water in for good measure. Once we reach that point, then we've taken care of the most frequent complaint(s) about the Doom3 engine.

This is also a readily-achievable (short-term) goal. It's the kind of thing that this group can put together in a relatively short amount of time and with a reasonable amount of effort (voluntary). It'll be a great milestone and will most likely motivate folks to continue working on the project.

But that's how I see it... I see the "Can Do It Too" part as just the first step in a larger project: harnessing the D3 engine's abilities to a fuller extent. My biggest fear is that we'll spend too much time trying to aim high, yet forget to ever implement or realize any of our goals.

G

I didn't enter a vote above, but here's my thought: create a great outdoors scene right on the outskirts of the city, which should be easy for a quick start, and release that partial test map as an early demonstration while continuing to work on the project; then add the city from there. The city could take a bit longer for the detail you want. I know fps limitations should not cripple the project, but some ballpark target should help. You don't want 6 fps and don't require 60 fps. I would think you want the features to have a decent showing on mid-high systems, maybe 30-45 fps with good to excellent detail.

This next paragraph is just to stimulate some feedback on specific type of target performance for the project, not argument on what is average or what is best. I'm not sure what you guys would consider a mid-high system. Would a lower end AMD 64 3400+ or equivalent Pentium qualify? Maybe with an ATI 9600 or Nvidia equivalent card? I don't think you should require 6800 Ultra or X800XT Platinum or above, because on recent surveys most people don't have those yet; also you probably don't want to require more than 128 or 256 megs of gfx card memory, as most people don't exceed that yet either, and you do want to show this version of the project now rather than a year from now.

Again, my suggestion would be to make a good demo that would run on what you define as a currently mid-high end system, and would not run so well on lower end systems. Also, start with an easier part like the outskirts of the nice cityscape just to get some of it out sooner, rather than waiting for all to gel together. The final version should be great! :idea:

(Edit: clarified gfx card info.)

Setting your goal now is a nice thing to do but just to "keep an eye on them" and not to follow them like a religion.

I'd say we should have a "Minimum requirement" for the final thing to make sure all these low end systems are out of question.

Lets say:
Lowest end DirectX 9.0 card (e.g. Radeon 9600) and a ~2,5Ghz CPU will handle this level with at least 20fps 99% of the time.
Out of the box, without special tweaking on the hardware/software side.

Just keep an eye on these specs and reduce your content enough at the end of the project to make this target a reality.

The least thing we want is a "Naaaah, see Doom 3 can't do it like i said before" reaction!

goliathvt wrote:

What about doing both 1 and 3?

...

But that's how I see it... I see the "Can Do It Too" part as just the first step in a larger project: harnessing the D3 engine's abilities to a fuller extent. My biggest fear is that we'll spend too much time trying to aim high, yet forget to ever implement or realize any of our goals.

G

Yes, that's totally reasonable. What is good there is also experimentation during the building of the initial architecture - such as your cool sun implementation and ideas on implementing weather fx - I like the idea of a podium with a weather control GUI, that would be terrific.

I guess the groundwork will tell us how much further it can be pushed - I honestly hope to see it at least run on my ti4200 - even if it's only at 10fps !

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

whoa182 wrote:

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

Creating a scene that chugs with most people's PCs isn't showing what the D3 engine is capable of. A game engine must not only render lovely images, but render them in a method home PCs can run. to simply create something that looks nice but plays awfully is not demonstrating that the D3 engine can "do it too", just that it can "nearly do it, but falls down short".

If you remember, the aim John Carmack had in mind with the first Doom engine was not graphics or sound or physics or anything like that, it was reliability and speed. Creating something that embraces neither of these things is defeating the object of a powerful engine.

Bauul wrote:

whoa182 wrote:

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

Creating a scene that chugs with most people's PCs isn't showing what the D3 engine is capable of. A game engine must not only render lovely images, but render them in a method home PCs can run. to simply create something that looks nice but plays awfully is not demonstrating that the D3 engine can "do it too", just that it can "nearly do it, but falls down short".

If you remember, the aim John Carmack had in mind with the first Doom engine was not graphics or sound or physics or anything like that, it was reliability and speed. Creating something that embraces neither of these things is defeating the object of a powerful engine.

I disagree. If this map ends up unplayable on my work machine ( GF4 MX400 ) then so be it. It's two generations old and D3 chuggs on it anyway.
If it's not playable on my GF FX5900(@min of 800x600, I play D3 @1152xwhatever normally) at home then I'll be less happy and rate this a faillure.

We are not releasing a game for the mass market but trying to prove what the engine can do on todays hardware, not the hardware that was top-of-the-range when Carmack started on the Engine.

If you want to make a large outdoor map with high res textures that plays great in cards from 2/3 years ago then the answer is simple - use the Q3 engine. The result will not look as good as the D3 engine of course, but what else would you expect?

The reason I say go for HIGH-END PC's only is just as 3Dmark 05 does and the previous versions before that. Show off what Graphics may look like Next Generation or when hardware catches up.

Optimize it has Much as you can but know that within 6-12month this map with be playing Very nice on the new Gen cards. Mabey you could even set your targets at the x800 and 6800 and up cards, Just to show whats around the corner.

When I play a game I would rather have a developer max out the capabilities of the engine to a certain point where it might only be playable on High in a years time, So therefor I would only choose Medium Graphics settings for the moment. But thats me. I like developers to be ' Future proof '

This isnt going to be sold or mass marketed, It can be a benchmark to overcome using the D3 engine on outdoor maps, I dont know. Might give some overclockers something new to get 200fps on lol.

But whatever you go with, I wouldnt expect you guys to aim at low end PC's for such a project. But theres definitly a Lot of ppl that would disagree i guess.

Howabout having a building that you can enter with high-rez textures and all the eye candy a computer can handle blocked by a portal or a seperate load point. Then, to top it off you can have a low-end system's beware sign outside. Or a seperate mod option for the better stuff?

Yeah, I'd love to see that. High-res textures and high-poly models in Doom 3 is something that I've longed for.

An attempt to make something similar to the Unreal 3 engine, and since this area would be meant to "rape" your GPU anyway, the artists could go a little crazy with the polycount of their models, making everything a bit more realistic.

I suggest a bar-scene, similar to this one:
http://www.montedocasal.pt/images/hi-res/bar.jpg

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

Intel17 wrote:

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

I think that demonstrating D3 can do a well lit out door scene without low fps is all part of the project. at the moment, we've basically seamed to aim for a mid range cpu set up, so we have a large out door scene that is perfectly playable on most people's rigs, which I think is by far and away the best idea.

High end systems all the way. I didn't spend 1,500 on a rock-solid workstation so i could work on things designed for some kid using daddy's work computer which is powered by a GeForce 3. Anyway, there is low-end content in the doom 3 community...some of us fortunate enough to have crack-addict computers should make use of those abilities to push doom 3 as far as it can go. Aside from that, Doom 3 is a "heavy" engine, having a well-lit...anything.... will run poorly on a low-end system no matter how you slice it. Some people are just lucky to be able to open the Doom3.exe on thier system! Luckily, that's not me, and it's none of you reading this, so like I said, we oughta crank this to the max because if those of us who can...don't....then no one will.

The reason I say go for HIGH-END PC's only is just as 3Dmark 05 does and the previous versions before that. Show off what Graphics may look like Next Generation or when hardware catches up.

Optimize it has Much as you can but know that within 6-12month this map with be playing Very nice on the new Gen cards. Mabey you could even set your targets at the x800 and 6800 and up cards, Just to show whats around the corner.

When I play a game I would rather have a developer max out the capabilities of the engine to a certain point where it might only be playable on High in a years time, So therefor I would only choose Medium Graphics settings for the moment. But thats me. I like developers to be ' Future proof '

This isnt going to be sold or mass marketed, It can be a benchmark to overcome using the D3 engine on outdoor maps, I dont know. Might give some overclockers something new to get 200fps on lol.

But whatever you go with, I wouldnt expect you guys to aim at low end PC's for such a project. But theres definitly a Lot of ppl that would disagree i guess.

Howabout having a building that you can enter with high-rez textures and all the eye candy a computer can handle blocked by a portal or a seperate load point. Then, to top it off you can have a low-end system's beware sign outside. Or a seperate mod option for the better stuff?

Yeah, I'd love to see that. High-res textures and high-poly models in Doom 3 is something that I've longed for.

An attempt to make something similar to the Unreal 3 engine, and since this area would be meant to "rape" your GPU anyway, the artists could go a little crazy with the polycount of their models, making everything a bit more realistic.

I suggest a bar-scene, similar to this one:
http://www.montedocasal.pt/images/hi-res/bar.jpg

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

Intel17 wrote:

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

I think that demonstrating D3 can do a well lit out door scene without low fps is all part of the project. at the moment, we've basically seamed to aim for a mid range cpu set up, so we have a large out door scene that is perfectly playable on most people's rigs, which I think is by far and away the best idea.

High end systems all the way. I didn't spend 1,500 on a rock-solid workstation so i could work on things designed for some kid using daddy's work computer which is powered by a GeForce 3. Anyway, there is low-end content in the doom 3 community...some of us fortunate enough to have crack-addict computers should make use of those abilities to push doom 3 as far as it can go. Aside from that, Doom 3 is a "heavy" engine, having a well-lit...anything.... will run poorly on a low-end system no matter how you slice it. Some people are just lucky to be able to open the Doom3.exe on thier system! Luckily, that's not me, and it's none of you reading this, so like I said, we oughta crank this to the max because if those of us who can...don't....then no one will.

Let's go for some participatory democracy, shall we?

G

voted for the third choice (focused goal).

definately keep a tight focus. don't get ahead of yourselves and MOST IMPORTANT don't try to "out do" hl2.

make an outdoor city map the way you feel it should be done while keeping it simple and focused. more detail and special effects can always be added later. for example, a gui that controlls the weather is a nifty idea, but that should be very last thing done.

good luck to you all...

Well I believe you must consider the real purpose for this, and everyones interpretation of it.

If the purpose is to try and show some weird comprability to HL2 then you have to aim for low end systems, and not use geometry.

If the purpose is to make a real mod type thing that is playable (w/ enemies and so forth) then it must be simpler.

If the purpose is to showcase how pretty D3 can be then I say shoot for the top end b/c anyone with a low end system can still look at the pretty pictures.

My idea would be to make it as pretty as possible while maintaining a 45frames per second on a 6800GT with a pentium 3GHz.

(BTW I have a x800pe and a athon 3200.. so I may be biased

)

What I want to see most is a well-lit environment that looks good and runs well on a mid-range computer. I don't even care if it's outdoors. Seems like it should be a simple thing to me, but no one has done it that I've seen.

I voted 3 but 1 was a close 2nd.

As rich said, it's not about beating engine X it's about proving Doom3 can do more than dark corridors.

I know it can and most mappers here know it can but so far most of the new fan made maps released are using id's textures, models and stylings.

i voted for 3 - the focused one.

if the goal is focused then it will be finished faster.

you can always add in the 'frills' of No. 1 later on if you wish, but atleast you have a base project.

making something for lowend comps is pretty crap...because it would kill all the good stuff about the DOOM engine. =/

Trying to mimick HL2 while also trying to make it as fast as HL2 is kinda pointless. Doom3 has a more advanced graphics engine and of course will run slower on compareable maps.

I voted for no. 3

I voted 1, I don't have a low-end system so I don't care for those losers

.

Duff wrote:

i voted for 3 - the focused one.

if the goal is focused then it will be finished faster.

you can always add in the 'frills' of No. 1 later on if you wish, but atleast you have a base project.

making something for lowend comps is pretty crap...because it would kill all the good stuff about the DOOM engine. =/

I'd go even farther and say its the only way it will finish. If focus is lost I could easily see little things constantly getting added and nothing ever finishing.

I voted for the first option - on Q3W I've seen endless discussion about performance and playability.

All these discussion led to mediocre work where more could have been achieved.

If we do such a project or even many of them, then we should consider to keep two things in mind:

- the capeability of the engine
- the original purpose of the engine

Most of the capeabilites we already know (some undocumented nice features are still waiting for us, but no - it's not that radiosity rumor).
The purpose of the engine was to render limited scenes with an unified lighting model.

Now trying to recreate scenerey the engine wasn't meant to render as it's prime target will always come with a tradeoff. The very last thing I'd like to see in this project are virtual obstacles put up by people who think any virtual given scene should run like butter on a shit machine.

Please - don't go the wrong turn and waste precious time over discussing how many frames are acceptable and what the limit of playability is.

Value artistical freedom over the same restrictions that led to the tradeoffs in Doom3 which humorously led to this exact project and thread!
You're about to make the same mistake twice if you tone the project down now.

I say - give a damn shit about maximum compatibility and low end performance! We are the privileged ones who can work with this technology without having to worry about the sales figures for the Wal Mart and AOHell crowd!

Such projects take time and have a strong geek appeal who, for the most part, are well equipped with the needed tech. I don't mind babyboy sending me a flame email becasue it wont run as good as CS on daddys ancient outlook pc or mommies untouched machine.

Think about it.

BNA! wrote:

I say - give a damn shit about maximum compatibility and low end performance! We are the privileged ones who can work with this technology without having to worry about the sales figures for the Wal Mart and AOHell crowd!

Such projects take time and have a strong geek appeal who, for the most part, are well equipped with the needed tech. I don't mind babyboy sending me a flame email becasue it wont run as good as CS on daddys ancient outlook pc or mommies untouched machine.

Think about it.

Well, that means it's basically gone away from rich's original idea, which I suppose is fine if that's the way it goes - to quote :-

Quote:

This thread is a community project I'm interested in starting in responce to all these "HL2 can do this" posts.

So, have you found something in a game that you don't think Doom 3 can match?

Post a screenshot. Describe it. Ect...

I'll try to prove to you that Doom 3 is just as capable. Anyone interested in lending me a hand is welcome.

I would like everything to be custom made however so that the resulting testmaps, ect... can be released for all to see firsthand running on their system.

Taken literally, that means an open area of medium realistic impact that runs smoothly on todays systems - with leeway, this obviously means todays systems that run Doom3 well.

But sure, if the project goes in the direction of pure eye candy, showcasing what will be playable in 12 months from now on top end hardware, nothing wrong.

Don't forget the title of the topic sets it up for comparison - not by the HL2 fanbois (who gives a shit about them), but by people genuinely interested in seeing whether Doom3 can really "do it too"

The way I see it is that obviously Doom3 is capable of rendering environments as "realistic" - and far more so than the square in city 17 - there's no dought about that - but can it run as effectively ?

That's why I chose 3.

I view it as a balance - balancing performance and realism, which truly would show what the Doom3 engine is capable of.

I'd prefer 3 as well. This is mainly because my machine isn't too high end, although it isn't any crappy CS computer either, heh. But I would like to play the map I've helped with, rather than having my FPS turn to Seconds Per Frame.

What about doing both 1 and 3?

I voted for 3 when I started the poll. However, I see both options possible... option three leads fairly logically towards option 1.

I mean, for starters, we need a textured map that is well-lit and maybe toss some water in for good measure. Once we reach that point, then we've taken care of the most frequent complaint(s) about the Doom3 engine.

This is also a readily-achievable (short-term) goal. It's the kind of thing that this group can put together in a relatively short amount of time and with a reasonable amount of effort (voluntary). It'll be a great milestone and will most likely motivate folks to continue working on the project.

But that's how I see it... I see the "Can Do It Too" part as just the first step in a larger project: harnessing the D3 engine's abilities to a fuller extent. My biggest fear is that we'll spend too much time trying to aim high, yet forget to ever implement or realize any of our goals.

G

I didn't enter a vote above, but here's my thought: create a great outdoors scene right on the outskirts of the city, which should be easy for a quick start, and release that partial test map as an early demonstration while continuing to work on the project; then add the city from there. The city could take a bit longer for the detail you want. I know fps limitations should not cripple the project, but some ballpark target should help. You don't want 6 fps and don't require 60 fps. I would think you want the features to have a decent showing on mid-high systems, maybe 30-45 fps with good to excellent detail.

This next paragraph is just to stimulate some feedback on specific type of target performance for the project, not argument on what is average or what is best. I'm not sure what you guys would consider a mid-high system. Would a lower end AMD 64 3400+ or equivalent Pentium qualify? Maybe with an ATI 9600 or Nvidia equivalent card? I don't think you should require 6800 Ultra or X800XT Platinum or above, because on recent surveys most people don't have those yet; also you probably don't want to require more than 128 or 256 megs of gfx card memory, as most people don't exceed that yet either, and you do want to show this version of the project now rather than a year from now.

Again, my suggestion would be to make a good demo that would run on what you define as a currently mid-high end system, and would not run so well on lower end systems. Also, start with an easier part like the outskirts of the nice cityscape just to get some of it out sooner, rather than waiting for all to gel together. The final version should be great! :idea:

(Edit: clarified gfx card info.)

Setting your goal now is a nice thing to do but just to "keep an eye on them" and not to follow them like a religion.

I'd say we should have a "Minimum requirement" for the final thing to make sure all these low end systems are out of question.

Lets say:
Lowest end DirectX 9.0 card (e.g. Radeon 9600) and a ~2,5Ghz CPU will handle this level with at least 20fps 99% of the time.
Out of the box, without special tweaking on the hardware/software side.

Just keep an eye on these specs and reduce your content enough at the end of the project to make this target a reality.

The least thing we want is a "Naaaah, see Doom 3 can't do it like i said before" reaction!

goliathvt wrote:

What about doing both 1 and 3?

...

But that's how I see it... I see the "Can Do It Too" part as just the first step in a larger project: harnessing the D3 engine's abilities to a fuller extent. My biggest fear is that we'll spend too much time trying to aim high, yet forget to ever implement or realize any of our goals.

G

Yes, that's totally reasonable. What is good there is also experimentation during the building of the initial architecture - such as your cool sun implementation and ideas on implementing weather fx - I like the idea of a podium with a weather control GUI, that would be terrific.

I guess the groundwork will tell us how much further it can be pushed - I honestly hope to see it at least run on my ti4200 - even if it's only at 10fps !

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

whoa182 wrote:

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

Creating a scene that chugs with most people's PCs isn't showing what the D3 engine is capable of. A game engine must not only render lovely images, but render them in a method home PCs can run. to simply create something that looks nice but plays awfully is not demonstrating that the D3 engine can "do it too", just that it can "nearly do it, but falls down short".

If you remember, the aim John Carmack had in mind with the first Doom engine was not graphics or sound or physics or anything like that, it was reliability and speed. Creating something that embraces neither of these things is defeating the object of a powerful engine.

Bauul wrote:

whoa182 wrote:

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

Creating a scene that chugs with most people's PCs isn't showing what the D3 engine is capable of. A game engine must not only render lovely images, but render them in a method home PCs can run. to simply create something that looks nice but plays awfully is not demonstrating that the D3 engine can "do it too", just that it can "nearly do it, but falls down short".

If you remember, the aim John Carmack had in mind with the first Doom engine was not graphics or sound or physics or anything like that, it was reliability and speed. Creating something that embraces neither of these things is defeating the object of a powerful engine.

I disagree. If this map ends up unplayable on my work machine ( GF4 MX400 ) then so be it. It's two generations old and D3 chuggs on it anyway.
If it's not playable on my GF FX5900(@min of 800x600, I play D3 @1152xwhatever normally) at home then I'll be less happy and rate this a faillure.

We are not releasing a game for the mass market but trying to prove what the engine can do on todays hardware, not the hardware that was top-of-the-range when Carmack started on the Engine.

If you want to make a large outdoor map with high res textures that plays great in cards from 2/3 years ago then the answer is simple - use the Q3 engine. The result will not look as good as the D3 engine of course, but what else would you expect?

The reason I say go for HIGH-END PC's only is just as 3Dmark 05 does and the previous versions before that. Show off what Graphics may look like Next Generation or when hardware catches up.

Optimize it has Much as you can but know that within 6-12month this map with be playing Very nice on the new Gen cards. Mabey you could even set your targets at the x800 and 6800 and up cards, Just to show whats around the corner.

When I play a game I would rather have a developer max out the capabilities of the engine to a certain point where it might only be playable on High in a years time, So therefor I would only choose Medium Graphics settings for the moment. But thats me. I like developers to be ' Future proof '

This isnt going to be sold or mass marketed, It can be a benchmark to overcome using the D3 engine on outdoor maps, I dont know. Might give some overclockers something new to get 200fps on lol.

But whatever you go with, I wouldnt expect you guys to aim at low end PC's for such a project. But theres definitly a Lot of ppl that would disagree i guess.

Howabout having a building that you can enter with high-rez textures and all the eye candy a computer can handle blocked by a portal or a seperate load point. Then, to top it off you can have a low-end system's beware sign outside. Or a seperate mod option for the better stuff?

Yeah, I'd love to see that. High-res textures and high-poly models in Doom 3 is something that I've longed for.

An attempt to make something similar to the Unreal 3 engine, and since this area would be meant to "rape" your GPU anyway, the artists could go a little crazy with the polycount of their models, making everything a bit more realistic.

I suggest a bar-scene, similar to this one:
http://www.montedocasal.pt/images/hi-res/bar.jpg

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

Intel17 wrote:

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

I think that demonstrating D3 can do a well lit out door scene without low fps is all part of the project. at the moment, we've basically seamed to aim for a mid range cpu set up, so we have a large out door scene that is perfectly playable on most people's rigs, which I think is by far and away the best idea.

High end systems all the way. I didn't spend 1,500 on a rock-solid workstation so i could work on things designed for some kid using daddy's work computer which is powered by a GeForce 3. Anyway, there is low-end content in the doom 3 community...some of us fortunate enough to have crack-addict computers should make use of those abilities to push doom 3 as far as it can go. Aside from that, Doom 3 is a "heavy" engine, having a well-lit...anything.... will run poorly on a low-end system no matter how you slice it. Some people are just lucky to be able to open the Doom3.exe on thier system! Luckily, that's not me, and it's none of you reading this, so like I said, we oughta crank this to the max because if those of us who can...don't....then no one will.

Let's go for some participatory democracy, shall we?

G

voted for the third choice (focused goal).

definately keep a tight focus. don't get ahead of yourselves and MOST IMPORTANT don't try to "out do" hl2.

make an outdoor city map the way you feel it should be done while keeping it simple and focused. more detail and special effects can always be added later. for example, a gui that controlls the weather is a nifty idea, but that should be very last thing done.

good luck to you all...

Well I believe you must consider the real purpose for this, and everyones interpretation of it.

If the purpose is to try and show some weird comprability to HL2 then you have to aim for low end systems, and not use geometry.

If the purpose is to make a real mod type thing that is playable (w/ enemies and so forth) then it must be simpler.

If the purpose is to showcase how pretty D3 can be then I say shoot for the top end b/c anyone with a low end system can still look at the pretty pictures.

My idea would be to make it as pretty as possible while maintaining a 45frames per second on a 6800GT with a pentium 3GHz.

(BTW I have a x800pe and a athon 3200.. so I may be biased

)

What I want to see most is a well-lit environment that looks good and runs well on a mid-range computer. I don't even care if it's outdoors. Seems like it should be a simple thing to me, but no one has done it that I've seen.

I voted 3 but 1 was a close 2nd.

As rich said, it's not about beating engine X it's about proving Doom3 can do more than dark corridors.

I know it can and most mappers here know it can but so far most of the new fan made maps released are using id's textures, models and stylings.

i voted for 3 - the focused one.

if the goal is focused then it will be finished faster.

you can always add in the 'frills' of No. 1 later on if you wish, but atleast you have a base project.

making something for lowend comps is pretty crap...because it would kill all the good stuff about the DOOM engine. =/

Trying to mimick HL2 while also trying to make it as fast as HL2 is kinda pointless. Doom3 has a more advanced graphics engine and of course will run slower on compareable maps.

I voted for no. 3

I voted 1, I don't have a low-end system so I don't care for those losers

.

Duff wrote:

i voted for 3 - the focused one.

if the goal is focused then it will be finished faster.

you can always add in the 'frills' of No. 1 later on if you wish, but atleast you have a base project.

making something for lowend comps is pretty crap...because it would kill all the good stuff about the DOOM engine. =/

I'd go even farther and say its the only way it will finish. If focus is lost I could easily see little things constantly getting added and nothing ever finishing.

I voted for the first option - on Q3W I've seen endless discussion about performance and playability.

All these discussion led to mediocre work where more could have been achieved.

If we do such a project or even many of them, then we should consider to keep two things in mind:

- the capeability of the engine
- the original purpose of the engine

Most of the capeabilites we already know (some undocumented nice features are still waiting for us, but no - it's not that radiosity rumor).
The purpose of the engine was to render limited scenes with an unified lighting model.

Now trying to recreate scenerey the engine wasn't meant to render as it's prime target will always come with a tradeoff. The very last thing I'd like to see in this project are virtual obstacles put up by people who think any virtual given scene should run like butter on a shit machine.

Please - don't go the wrong turn and waste precious time over discussing how many frames are acceptable and what the limit of playability is.

Value artistical freedom over the same restrictions that led to the tradeoffs in Doom3 which humorously led to this exact project and thread!
You're about to make the same mistake twice if you tone the project down now.

I say - give a damn shit about maximum compatibility and low end performance! We are the privileged ones who can work with this technology without having to worry about the sales figures for the Wal Mart and AOHell crowd!

Such projects take time and have a strong geek appeal who, for the most part, are well equipped with the needed tech. I don't mind babyboy sending me a flame email becasue it wont run as good as CS on daddys ancient outlook pc or mommies untouched machine.

Think about it.

BNA! wrote:

I say - give a damn shit about maximum compatibility and low end performance! We are the privileged ones who can work with this technology without having to worry about the sales figures for the Wal Mart and AOHell crowd!

Such projects take time and have a strong geek appeal who, for the most part, are well equipped with the needed tech. I don't mind babyboy sending me a flame email becasue it wont run as good as CS on daddys ancient outlook pc or mommies untouched machine.

Think about it.

Well, that means it's basically gone away from rich's original idea, which I suppose is fine if that's the way it goes - to quote :-

Quote:

This thread is a community project I'm interested in starting in responce to all these "HL2 can do this" posts.

So, have you found something in a game that you don't think Doom 3 can match?

Post a screenshot. Describe it. Ect...

I'll try to prove to you that Doom 3 is just as capable. Anyone interested in lending me a hand is welcome.

I would like everything to be custom made however so that the resulting testmaps, ect... can be released for all to see firsthand running on their system.

Taken literally, that means an open area of medium realistic impact that runs smoothly on todays systems - with leeway, this obviously means todays systems that run Doom3 well.

But sure, if the project goes in the direction of pure eye candy, showcasing what will be playable in 12 months from now on top end hardware, nothing wrong.

Don't forget the title of the topic sets it up for comparison - not by the HL2 fanbois (who gives a shit about them), but by people genuinely interested in seeing whether Doom3 can really "do it too"

The way I see it is that obviously Doom3 is capable of rendering environments as "realistic" - and far more so than the square in city 17 - there's no dought about that - but can it run as effectively ?

That's why I chose 3.

I view it as a balance - balancing performance and realism, which truly would show what the Doom3 engine is capable of.

I'd prefer 3 as well. This is mainly because my machine isn't too high end, although it isn't any crappy CS computer either, heh. But I would like to play the map I've helped with, rather than having my FPS turn to Seconds Per Frame.

What about doing both 1 and 3?

I voted for 3 when I started the poll. However, I see both options possible... option three leads fairly logically towards option 1.

I mean, for starters, we need a textured map that is well-lit and maybe toss some water in for good measure. Once we reach that point, then we've taken care of the most frequent complaint(s) about the Doom3 engine.

This is also a readily-achievable (short-term) goal. It's the kind of thing that this group can put together in a relatively short amount of time and with a reasonable amount of effort (voluntary). It'll be a great milestone and will most likely motivate folks to continue working on the project.

But that's how I see it... I see the "Can Do It Too" part as just the first step in a larger project: harnessing the D3 engine's abilities to a fuller extent. My biggest fear is that we'll spend too much time trying to aim high, yet forget to ever implement or realize any of our goals.

G

I didn't enter a vote above, but here's my thought: create a great outdoors scene right on the outskirts of the city, which should be easy for a quick start, and release that partial test map as an early demonstration while continuing to work on the project; then add the city from there. The city could take a bit longer for the detail you want. I know fps limitations should not cripple the project, but some ballpark target should help. You don't want 6 fps and don't require 60 fps. I would think you want the features to have a decent showing on mid-high systems, maybe 30-45 fps with good to excellent detail.

This next paragraph is just to stimulate some feedback on specific type of target performance for the project, not argument on what is average or what is best. I'm not sure what you guys would consider a mid-high system. Would a lower end AMD 64 3400+ or equivalent Pentium qualify? Maybe with an ATI 9600 or Nvidia equivalent card? I don't think you should require 6800 Ultra or X800XT Platinum or above, because on recent surveys most people don't have those yet; also you probably don't want to require more than 128 or 256 megs of gfx card memory, as most people don't exceed that yet either, and you do want to show this version of the project now rather than a year from now.

Again, my suggestion would be to make a good demo that would run on what you define as a currently mid-high end system, and would not run so well on lower end systems. Also, start with an easier part like the outskirts of the nice cityscape just to get some of it out sooner, rather than waiting for all to gel together. The final version should be great! :idea:

(Edit: clarified gfx card info.)

Setting your goal now is a nice thing to do but just to "keep an eye on them" and not to follow them like a religion.

I'd say we should have a "Minimum requirement" for the final thing to make sure all these low end systems are out of question.

Lets say:
Lowest end DirectX 9.0 card (e.g. Radeon 9600) and a ~2,5Ghz CPU will handle this level with at least 20fps 99% of the time.
Out of the box, without special tweaking on the hardware/software side.

Just keep an eye on these specs and reduce your content enough at the end of the project to make this target a reality.

The least thing we want is a "Naaaah, see Doom 3 can't do it like i said before" reaction!

goliathvt wrote:

What about doing both 1 and 3?

...

But that's how I see it... I see the "Can Do It Too" part as just the first step in a larger project: harnessing the D3 engine's abilities to a fuller extent. My biggest fear is that we'll spend too much time trying to aim high, yet forget to ever implement or realize any of our goals.

G

Yes, that's totally reasonable. What is good there is also experimentation during the building of the initial architecture - such as your cool sun implementation and ideas on implementing weather fx - I like the idea of a podium with a weather control GUI, that would be terrific.

I guess the groundwork will tell us how much further it can be pushed - I honestly hope to see it at least run on my ti4200 - even if it's only at 10fps !

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

whoa182 wrote:

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

Creating a scene that chugs with most people's PCs isn't showing what the D3 engine is capable of. A game engine must not only render lovely images, but render them in a method home PCs can run. to simply create something that looks nice but plays awfully is not demonstrating that the D3 engine can "do it too", just that it can "nearly do it, but falls down short".

If you remember, the aim John Carmack had in mind with the first Doom engine was not graphics or sound or physics or anything like that, it was reliability and speed. Creating something that embraces neither of these things is defeating the object of a powerful engine.

Bauul wrote:

whoa182 wrote:

Id say keep this Project on the HIGHER end of PC's capabilities. Even if we only get 15fps with a 9800PRO +, i believe that is acceptable. because we all know computers are getting much faster every year.

TBH I dont care if they people with crap computers cant run it. I think you should SHOW what the engine can do, not be limited because of these gamers have old PC's...

Creating a scene that chugs with most people's PCs isn't showing what the D3 engine is capable of. A game engine must not only render lovely images, but render them in a method home PCs can run. to simply create something that looks nice but plays awfully is not demonstrating that the D3 engine can "do it too", just that it can "nearly do it, but falls down short".

If you remember, the aim John Carmack had in mind with the first Doom engine was not graphics or sound or physics or anything like that, it was reliability and speed. Creating something that embraces neither of these things is defeating the object of a powerful engine.

I disagree. If this map ends up unplayable on my work machine ( GF4 MX400 ) then so be it. It's two generations old and D3 chuggs on it anyway.
If it's not playable on my GF FX5900(@min of 800x600, I play D3 @1152xwhatever normally) at home then I'll be less happy and rate this a faillure.

We are not releasing a game for the mass market but trying to prove what the engine can do on todays hardware, not the hardware that was top-of-the-range when Carmack started on the Engine.

If you want to make a large outdoor map with high res textures that plays great in cards from 2/3 years ago then the answer is simple - use the Q3 engine. The result will not look as good as the D3 engine of course, but what else would you expect?

The reason I say go for HIGH-END PC's only is just as 3Dmark 05 does and the previous versions before that. Show off what Graphics may look like Next Generation or when hardware catches up.

Optimize it has Much as you can but know that within 6-12month this map with be playing Very nice on the new Gen cards. Mabey you could even set your targets at the x800 and 6800 and up cards, Just to show whats around the corner.

When I play a game I would rather have a developer max out the capabilities of the engine to a certain point where it might only be playable on High in a years time, So therefor I would only choose Medium Graphics settings for the moment. But thats me. I like developers to be ' Future proof '

This isnt going to be sold or mass marketed, It can be a benchmark to overcome using the D3 engine on outdoor maps, I dont know. Might give some overclockers something new to get 200fps on lol.

But whatever you go with, I wouldnt expect you guys to aim at low end PC's for such a project. But theres definitly a Lot of ppl that would disagree i guess.

Howabout having a building that you can enter with high-rez textures and all the eye candy a computer can handle blocked by a portal or a seperate load point. Then, to top it off you can have a low-end system's beware sign outside. Or a seperate mod option for the better stuff?

Yeah, I'd love to see that. High-res textures and high-poly models in Doom 3 is something that I've longed for.

An attempt to make something similar to the Unreal 3 engine, and since this area would be meant to "rape" your GPU anyway, the artists could go a little crazy with the polycount of their models, making everything a bit more realistic.

I suggest a bar-scene, similar to this one:
http://www.montedocasal.pt/images/hi-res/bar.jpg

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

Intel17 wrote:

There would be no point in attempting to cater to the lowest common denominator.

D3CDIT should focus on achieving the highest quality visuals it can possibly do. However this would take an extreme amount of time and talent to go through the proper Doom 3 engine content development process, and then making high quality in game assets.

I think that demonstrating D3 can do a well lit out door scene without low fps is all part of the project. at the moment, we've basically seamed to aim for a mid range cpu set up, so we have a large out door scene that is perfectly playable on most people's rigs, which I think is by far and away the best idea.

High end systems all the way. I didn't spend 1,500 on a rock-solid workstation so i could work on things designed for some kid using daddy's work computer which is powered by a GeForce 3. Anyway, there is low-end content in the doom 3 community...some of us fortunate enough to have crack-addict computers should make use of those abilities to push doom 3 as far as it can go. Aside from that, Doom 3 is a "heavy" engine, having a well-lit...anything.... will run poorly on a low-end system no matter how you slice it. Some people are just lucky to be able to open the Doom3.exe on thier system! Luckily, that's not me, and it's none of you reading this, so like I said, we oughta crank this to the max because if those of us who can...don't....then no one will.