"The Official DOOM 3 [H]ardware Guide" at [H]

Aw come on, you know what I meant.

That Z-fill graph doesn't look too promising though ... :?

 

A demo in DOOM 3 does not reprsent actual gameplay. This has been verified with Robert Duffy.

 

Can you expand on that at all? In what sense?

Cheers.

 

imagine that.

 

Why is there a max of 60fps ? I thought only the physics engine was capped at 60fps, or did they bench with vsynq+triple buffering?

 

Just a quick question for anybody who has Doom3 (Kyle, Dave).

I have a laptop with a stupid widescreen (1280*800) Is it possible in Doom3 to set the resolution to widescreen format?

Its quite a powerfull machine, Athlon64 3400+, mobility Radeon 9700 128MB, 512MB RAM. I can probably play in 1024*768 medium quality according to the hardware review, but if I need to play at 800*600 then Ill only be using a tiny bit of my screen (I dont use scale to monitor with 4:3, cos everything looks stupid then).

So, is setting something like 800*500 a possibility?

Ali

 

To join Dave in his request for clarification.

Does Robert mean timedemo'ing a demo don't give us the same average performance stats as per when we were recording the demo? Or does he mean demo playback don't give us the same performance stats as per when we were recording the demo? The difference, in case you got confused , is timedemo and demo-playback.

Assuming the former (timedemo'ing a demo don't give us the same average performance stats as per actual game play), so, say, I play the game as normal without recording a demo. I use FRAPS at the same time, ensuring it starts logging where I want it to start logging, and ending when I want it to. At the end, FRAPS gives an average FPS stat, say 60fps.

I repeat all of the above, except that I record a demo with FRAPS also running. After that, I playback the demo (not timedemo it) while using FRAPS to log (from start of demo playback to end). Demo ends, FRAPS gives average fps result. Would it be 60fps, Kyle?

 

I know this has been explained, but i still don't see one single good reason why ID would cap the maximum framerate.

 

Id wants people to focus on IQ instead of framerate. 60 fps is perfectly smooth and Doom 3 is a rather slow game as well.

 

Yeah i get that, but why would they cap it? I mean, the IQ options would still be the same, because it's a rather "closed options" game. Why cap the framerate? It doesn't really affect IQ, so... I still don't get it.

 

To make sure people don't favor framerate over effects. People with really shitty cards can of course turn off some effects but those with powerful card might as well turn everything on since they're not gonna get more than 60 fps anyway.

 

Oh... Well, if that's the official reason, it sucks. How about in 2 years time when cards will be able to run the game at full details at a much higher framerate? Still capped at 60...?

 

london-boy:

> Oh... Well, if that's the official reason, it sucks. How about in 2 years
> time when cards will be able to run the game at full details at a much
> higher framerate? Still capped at 60...?

Someone will probably find a way to remove the cap but id would probably prefer that you turn on ultra quality and increase AA and filtering.

 

fps cap

Didn't it have something to do with timing issues for things like jumping? I seem to remember reading something about that.

 

Seen how there's not much to increase (far too few options in there, although i'm sure countless patches will come out to fix that issue), i hardly believe that.
I mean, i could understand capping the minimum framarate, making sure that the right settings were chosen to make sure the game never goes under, say, 20fps... But capping the MAXIMUM framarate?? Meh...

 

Here

http://www2.hardocp.com/article.html?art=NjQ0LDI=

"Unlike a timed demo, this real-world gameplay stresses the CPU by using the AI, sound, and physics engines, providing a record of real gameplay system performance"

Seems time demos are purely for testing video performance, and since this guide was meant to inform the user about how well their system would handle Doom3, a timedemo would not be sufficient to provide that information ... that's what I think anyway.

 

Er, re-read what I said. I stressed on demo playback using FRAPS. Not benchmarking/timedemo'ing a demo. The question is if the playback of a recorded demo mirrors the gameplay the demo is recorded/based on.

UT2003 demo benchmarking works fine. UT2003 demo playback is screwed, because the normal demo playback code is screwed. I'm not about to suggest DOOM3's demo playback is screwed, so I'm asking for clarifications. I could ask John but I'll be understanding for once and won't bug him for what (to him) should be a rather unimportant email at this point in the life of DOOM3.

 

WRT the article I am slightly confused. I am now debating whether to get a 6800 regular, or GT. Last time I got a subpar card the 9500pro I upgraded to a 9800 pro within 7 months and that seems a silly thing to do again...

However I am thinking of building another system around march next spring so I could survive...

Any of you "insiders" have a clue about when GT's will become widely available? They still seem to be out of stock so many places that the price is artificially inflated...perhaps they will mysteriously become ubiqiutous when half-life 2 is launched and demand falls

 

I'm guessing that Doom 3 records physics/AI information so is not CPU bound when it otherwise would be. As a result, timedemos are not indicative of actual gameplay performance.