"Yes, but how many polygons?" An artist blog entry with interesting numbers

Cloofoofoo said:
And so was metal gear solid 2( being developed since 98 and look at the difference I mean). We cant always be too wishful. Code Veronica seems to hammer the Dc in it's weakest areas , multiple light sources and if the person who reversed the format is right alot of animated bones. So basically hits the cpu hard limiting the polygon density.

The person who reversed it already mentioned it's actually very easy to negatively affect performance since the game is already at heavy load.
Click to expand...
Why are the models double?
 
Cloofoofoo said:
And so was metal gear solid 2( being developed since 98 and look at the difference I mean). We cant always be too wishful. Code Veronica seems to hammer the Dc in it's weakest areas , multiple light sources and if the person who reversed the format is right alot of animated bones. So basically hits the cpu hard limiting the polygon density.

The person who reversed it already mentioned it's actually very easy to negatively affect performance since the game is already at heavy load.
Click to expand...
I know CV pushes heavily DC hardware, but probably with more modern development tools they could make an smartes approach, using more efficiently DC´s resources, and probably achieving better looking visuals, pushinsg even less the machine. I mean, the 2 models in 1 doesn´t seem to be much efficient or agile. May be sometime in the future community can achieve similar looking and animated character models, using just one model each and liberating more space to implement another stuff in due to make the game look even better.
 
xaeroxcore said:
I know CV pushes heavily DC hardware, but probably with more modern development tools they could make an smartes approach, using more efficiently DC´s resources, and probably achieving better looking visuals, pushinsg even less the machine. I mean, the 2 models in 1 doesn´t seem to be much efficient or agile. May be sometime in the future community can achieve similar looking and animated character models, using just one model each and liberating more space to implement another stuff in due to make the game look even better.
Click to expand...
Agreed, if they had more time, and ofcourse better tools, it would definitely get better graphically. It's already been proven that ninja 2 (which was never used) was getting better performance out of Dreamcast. Imagine what could have been 😍
 
xaeroxcore said:
I know CV pushes heavily DC hardware, but probably with more modern development tools they could make an smartes approach, using more efficiently DC´s resources, and probably achieving better looking visuals, pushinsg even less the machine. I mean, the 2 models in 1 doesn´t seem to be much efficient or agile. May be sometime in the future community can achieve similar looking and animated character models, using just one model each and liberating more space to implement another stuff in due to make the game look even better.
Click to expand...

You'd think so huh. Well according to the same guy they did just that when they did code Veronica gun survivor on Naomi. They simplified it a bit ( apparently missing model envelope data, no multi model approach.) And still required the zombie models to be lowered in poly count to display of them.

It's not two models being rendered but two models kept in memory for other things like animation related things.

Code Veronica already does take a smarter artistic approach. Polygon count wise it isn't that different to what's on scene to blue stinger and seven mansions( stage models are half code Veronica but character and enemy models are 2x to 3x denser) , what makes it look superior is the smarter approach to modeling and texturing.
 
Sonicfan said:
Agreed, if they had more time, and ofcourse better tools, it would definitely get better graphically. It's already been proven that ninja 2 (which was never used) was getting better performance out of Dreamcast. Imagine what could have been 😍
Click to expand...

Ninja 2 isn't a magic bullet. It's faster than ninja 1 but it's just meant to be an abstraction to be handled far easier than kamui 2. Also kamui 2 is faster than ninja2 since it's more low level and was already in use by the time DC does both Japan and state side. Just something to keep in mind. Ninja2 would have just made it easier to access that performance and features than they already had with kamui 2.
 
Nesh said:
Why are the models double?
Click to expand...

He means models stored in memory. Seems many duplicates were packed for linking multiple animations or even morph targets .

that is correct, not all of it renders at same time but gets processed and loaded into memory to call it when its needed, cv dont flush it memory when you swap rooms and everything from the previous rooms remains in memory in case the next room can need such data, also the rdx files and the sound files dont get acceses directly from disc as you play, they get uncompressed/copied into memory and read from there, ive made many experiments to chekout this and i can also supply some "unlocked" rooms of the game wich can show how the console suffer exesive slow downs when enabling even just minor hidden stuff in the room (not being rendered by default),
Click to expand...
 
Cloofoofoo said:
You'd think so huh. Well according to the same guy they did just that when they did code Veronica gun survivor on Naomi. They simplified it a bit ( apparently missing model envelope data, no multi model approach.) And still required the zombie models to be lowered in poly count to display of them.

It's not two models being rendered but two models kept in memory for other things like animation related things.

Code Veronica already does take a smarter artistic approach. Polygon count wise it isn't that different to what's on scene to blue stinger and seven mansions( stage models are half code Veronica but character and enemy models are 2x to 3x denser) , what makes it look superior is the smarter approach to modeling and texturing.
Click to expand...

So CV takes full advantage from DC, it could be, but is not the graphical pinnacle on the system...I mean, Shenmue 1 and specially 2 are more complex games and look more advanced...Also Headhunter, so there is no reason to think that CV could not be hypthetically improved with more modern develpment tools.

Heck, there are even people porting RE 2 and Tomb Raider 2 to Saturn



So i wonder how much time it takes to see something similar finally on DC...
 
Cloofoofoo said:
He means models stored in memory. Seems many duplicates were packed for linking multiple animations or even morph targets .
Click to expand...
The quote you have below though makes more sense which keeps on memory linked rooms and it's objects. I am confused with the morph targets and multiple animations equalling character duplicates stored in memory at all times though for each room you are in. I understand it if there is a special cut scene that can be generated in the room that requires extra animations and models. But most rooms dont have those.
 
xaeroxcore said:
So CV takes full advantage from DC, it could be, but is not the graphical pinnacle on the system...I mean, Shenmue 1 and specially 2 are more complex games and look more advanced...Also Headhunter, so there is no reason to think that CV could not be hypthetically improved with more modern develpment tools.

Heck, there are even people porting RE 2 and Tomb Raider 2 to Saturn



So i wonder how much time it takes to see something similar finally on DC...
Click to expand...
Head hunter is a pretty poor example since it's still about better art style rather than polygon count . The developer has mentioned multiple times after game logic plus lighting plus culling head hunter does around 10,000 to 15,000 at 30 fps( the whole 50k isnt whats rendered but full number before culling like not rendering unseen rooms or stuff offscreen ). Pretty much we code Veronica/ seventh mansion / blue stinger sit at.
Shenmue does polygons at the cost of lighting , you can clearly see it's lighting isn't as complex as code Veronica.

You might be waiting for a long while , current homebrew solutions do like 5,000 to 11,000 at 30 fps in game situations, which well under what the official kit did.
 
Cloofoofoo said:
Head hunter is a pretty poor example since it's still about better art style rather than polygon count . The developer has mentioned multiple times after game logic plus lighting plus culling head hunter does around 10,000 to 15,000 at 30 fps( the whole 50k isnt whats rendered but full number before culling like not rendering unseen rooms or stuff offscreen ). Pretty much we code Veronica/ seventh mansion / blue stinger sit at.
Shenmue does polygons at the cost of lighting , you can clearly see it's lighting isn't as complex as code Veronica.

You might be waiting for a long while , current homebrew solutions do like 5,000 to 11,000 at 30 fps in game situations, which well under what the official kit did.
Click to expand...

But this is exactly what i meant with smarter approach. Headhunter probably do less things on the pure technical side, but manages to look better (or at least more modern) than CV and also offers a semi open world with driving freely stage on the bike. Yes CV looked stunning at the time of release, but was quickly looking outdated vs Shenmue (despite CV being superior in some areas as lighting, that you mentioned), and even more by 2001 vs PS2 MGS2, Devil May Cry, Silent Hill 2 and and, to a lesser extent, on DC itself with Shenmue 2 and Headhunter....

You put Shenmue 2 or Headhunter vs the average PS2 post 2002 third person title and they both does not look as dated in front on them as CV looks...

Code Veronica looks great in some aspects, as lighting and main character facial expressions (and even Claire, Wesker and Chris playable models) but in other areas always felt like a 32 bit game with "128 bits" graphics. You can see it on its tanky controls (RE 3 was more advanced in that area), circle shadows, Steve´s and zombies blocky hands...I´m sure they would find a way to fix all that stuff or at least hide it better, with more modern tools and wider experience with 6th gen systems....

Also i wonder, how above are on graphical level RE Outbreak, Devil May Cry or Silent Hill 2 vs CV. May be with an smarter approach CV would have look closer to the visuals shown by those titles.

Hope in a few years we have the answer via homebrew solution.
 
xaeroxcore said:
But this is exactly what i meant with smarter approach. Headhunter probably do less things on the pure technical side, but manages to look better (or at least more modern) than CV and also offers a semi open world with driving freely stage on the bike. Yes CV looked stunning at the time of release, but was quickly looking outdated vs Shenmue (despite CV being superior in some areas as lighting, that you mentioned), and even more by 2001 vs PS2 MGS2, Devil May Cry, Silent Hill 2 and and, to a lesser extent, on DC itself with Shenmue 2 and Headhunter....

You put Shenmue 2 or Headhunter vs the average PS2 post 2002 third person title and they both does not look as dated in front on them as CV looks...

Code Veronica looks great in some aspects, as lighting and main character facial expressions (and even Claire, Wesker and Chris playable models) but in other areas always felt like a 32 bit game with "128 bits" graphics. You can see it on its tanky controls (RE 3 was more advanced in that area), circle shadows, Steve´s and zombies blocky hands...I´m sure they would find a way to fix all that stuff or at least hide it better, with more modern tools and wider experience with 6th gen systems....

Also i wonder, how above are on graphical level RE Outbreak, Devil May Cry or Silent Hill 2 vs CV. May be with an smarter approach CV would have look closer to the visuals shown by those titles.

Hope in a few years we have the answer via homebrew solution.
Click to expand...
Totally agree with everything you say here! CV looks wonderful, but was definitely outdone by the likes of shenmue 2 and headhunter! Those are bigger games with better textures, and models (outside of hh not having facial animation especially for npcs) given time and better tools, DC would definitely compete well with later 6th Gen games! Shenmue 2 and hh personally had better lighting than cv to me, atleast artistically. Especially S2 at night scenes, etc! As far as the likes of dmc/sh2/mgs2, I truly believe those could be done, with some consessions ofcourse; be smart about the reduction in polys and fillrate effects, you have an excellent port!
 
Cloofoofoo said:
The developer has mentioned multiple times after game logic plus lighting plus culling head hunter does around 10,000 to 15,000 at 30 fps( the whole 50k isnt whats rendered but full number before culling like not rendering unseen rooms or stuff offscreen ).
Click to expand...
Yeah I was tech art lead for Headhunter and sometimes it was even less htan that. Also regarding culling I would say most game of the time had some form of LOD and culling system. I mean there is reasons why levels often was built a bit mazy like, not just for the gameplay but to be able to block stuff. For DC/PVR that had deffered rendering it was not such and issue with overdraw but on other consoles like PS1, Saturn culling was needed just not for processing but draw reasons.

Also about shen Mue not having as advanced lighting as Code Veronica, the lighting in headhunter is super simple (I designed the system) its just maximum two directional (paralell lighths) and just one color (so not possible to have two colored lighting on objects) per object.
 
StefanHQ said:
Yeah I was tech art lead for Headhunter and sometimes it was even less than that. Also regarding culling I would say most game of the time had some form of LOD and culling system. I mean there is reasons why levels often was built a bit mazy like, not just for the gameplay but to be able to block stuff. For DC/PVR that had deffered rendering it was not such and issue with overdraw but on other consoles like PS1, Saturn culling was needed just not for processing but draw reasons.

Also about shen Mue not having as advanced lighting as Code Veronica, the lighting in headhunter is super simple (I designed the system) its just maximum two directional (paralell lighths) and just one color (so not possible to have two colored lighting on objects) per object.
Click to expand...

Amazing bro!! First of all, i´d like to thank for your work in one of the Dreamcast games i love the most. It gave me many great gaming moments 21 years ago!!

Headhunter really impressed me at the time for the quality of the stages specially. At the beginning, remember the Stern Mansion and being shocked cuz to me it looked prerendered quaility, but in real time. Also the Grey Wolf boss fight...The rain effect, lightings and reflexions of stage objects still look unreal to this day....You really felt back in the day that with Headhunter, Rez, Shenmue 2, Test Drive Le Mans, Virtua Tennis 2, Ikaruga and DOA2, DC wasnt that far on the visual level from PS2.

Now you proof us you can make a fantastic looking game with less than 15k polys on screen at once....Question is: Do you think DC could have been pushed further? Graphical level of the likes of PSP´s GTA LCS, Need For Speed Most Wanted or Burnout Dominator, could have been reached by DC? or even that is out of league?
 
You must log in or register to reply here.

Similar threads

A
Full transcript of John Carmack's QuakeCon 2005 Keynote
3 4 5
Replies
98
Views
32K
nemisis
N
Back
Top