Quantcast

Page 207 of 216 FirstFirst ... 107157197203204205206207208209210211 ... LastLast
Results 3,091 to 3,105 of 3226

Thread: PS2 vs Dreamcast Graphics

  1. #3091
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Feb 2019
    Posts
    410
    Rep Power
    5

    Default

    I just posted the image of silent hill 3 to decorate the post since it is one of the most beautiful games of the sixth generation.

  2. #3092
    Master of Shinobi
    Join Date
    Sep 2012
    Posts
    1,282
    Rep Power
    38

    Default

    Quote Originally Posted by Cloofoofoo View Post
    Heres a few "myth" ps2 games people mention. Grand prix challenge(2002) by Melbourne house( guys who did le mans on the dc) and F1 racing championship(2001) by video systems. Grand prix challange especially is usually touted 12 mpps - 20 mpps well its no where near, its running very similar to f1 season 2000 two years prior. I guess like test drive lemans the Melbourne house guys like to hype things lol. F1 racing championship was terrible looking on the dreamcast and while better on the ps2 it isnt all that much better.

    Grand prix challange - 2002
    86,550 tris per frame 60fps = 5.1 mpps
    Is it possible that in scenes where they show more/all the cars, it reaches a significantly higher number?

  3. #3093
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    Quote Originally Posted by zyrobs View Post
    Is it possible that in scenes where they show more/all the cars, it reaches a significantly higher number?
    Significantly no. In Grand Prix challenge it peaks out when the cars are together in the beginning. Now the polygon count might be higher in other tracks but it isn't a big jump. As a matter of fact Grand Prix challenge runs less consistent that f1 season 2000 by ea. I would say that one out performs it despite being 2 years older. As that one really hovers like 70k to 90k per frame at 60 fps while Grand Prix challenge just stays at 75k tris per frame unless you get more than 2 or cars pretty close to you then it jumps to around 88k Tris per frame.

    F1 racing championship runs up to 24k per frame in certain condition but likes to dip to 15k to 19k per frame at 60 fps.

    Also remember these games seem to use lod so the farthest should be pretty low and will not impact polygon count considerably unless u can get them ALL in your face.l, which is unrealistic as they are spaced out to prevent that. So even if you see all cars doesn't mean it will impact the scene count as much as you think.

  4. #3094
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    So I was testing horror ps2 games: fatal frame 3 , clocktower 3 , silent hill 2. Silent hill 2 flips the hell out right at the beginning so I wasnt able to test it much. But all these titles show the same behavior. Basically I extracted the scene and found that everything has backface culling and they cull everything right outside the frame. Basically very little polygons is being rendered because most of it is culled unless its in the frame and facing the camera. This kinda makes sense for silent hill 2 and fatal frame 3 because they seem to use alot of fullscreen filter (fatal frame even keeps a dense render to texture polygon quad in front of the enemys to make the cool ghostly ripple effect, didnt realize such a cool effect would be so simple) as this would save fillrate but not so much for clocktower 3 as its very basic. Clock tower 3 would probably work on the dc as - is but the other two would require some reworking as you would need to removed the framebuffer effects they depend alot for their looks.

    Also tested chaos legion. The enemies are numerous on screen(more than 20) but they are very low poly and there is basically no draw distance( just enough to be room distance). You could probably run something like this on the dreamcast if you minus the depth of field @ 30 fps.

    Silent hill 2:
    intro cutscene
    6,691 tris per frame x 60 fps = 401k per second


    Wandering before the game flips out
    9,508 tris per frame x 60 fps = 570k per second



    Clock tower 3
    In a room
    2,942 tris per frame x 60 fps = 176k per second


    Wandering around
    3,707 tris per frame x 60 fps = 222K per second

    '

    Fatal frame 3
    Right at the intro
    7,389 tris per frame x 60 fps = 443K per second


    Vs an enemy
    5,489 tris per frame x 60 fps = 329k per second



    Chaos legion(ps2 version)
    Right at the beginning
    15,140 tris per frame x 60 fps = 908k per second


    First mob of 20+ creatures
    21,546 tris per frame x 60 fps = 1.2 mpps


    Second mob with a leader beast + summon
    16,985 tris per frame x 60 fps = 1.01 mpps
    Last edited by Cloofoofoo; 05-16-2020 at 04:15 PM.

  5. #3095
    Raging in the Streets Yharnamresident's Avatar
    Join Date
    May 2013
    Location
    British Columbia
    Posts
    4,080
    Rep Power
    65

    Default

    Quote Originally Posted by Cloofoofoo View Post
    I wouldn't say they were garbage, the world grix 1 and 2 played great and back in the day go decent scores for graphics. Some of the crappy Ubisoft ones for Europe even had online play. So while graphics weren't their merit they did have other positives.

    I might try chaos legion , maybe silent Hill 2 next.

    I think it's more surprising that some Wii games if you cut out the multi texture and frame buffer effects and lowered the polygon count just tiny bit they might run on the DC.
    The problem with the Dreamcast F1 World Grand Prix games is they were pieced together. Like the first game uses N64 and original Dreamcast assets, then the sequel also uses assets from the PC version. So the sequel uses N64, original Dreamcast and PC assets. How about just make the god damn Dreamcast port from the ground up.
    Certified F-Zero GX fanboy

  6. #3096
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    Quote Originally Posted by Yharnamresident View Post
    The problem with the Dreamcast F1 World Grand Prix games is they were pieced together. Like the first game uses N64 and original Dreamcast assets, then the sequel also uses assets from the PC version. So the sequel uses N64, original Dreamcast and PC assets. How about just make the god damn Dreamcast port from the ground up.
    Can't disagree with that. It's obvious they felt it wasn't worth the effort. Considering they did from the ground up work for the ps2 version.(video systems basically worked on all these f1 games)

  7. #3097
    Raging in the Streets Yharnamresident's Avatar
    Join Date
    May 2013
    Location
    British Columbia
    Posts
    4,080
    Rep Power
    65

    Default

    I will give you guys a preview of a post that has to be made.



    Certified F-Zero GX fanboy

  8. #3098
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    Quote Originally Posted by Yharnamresident View Post
    I will give you guys a preview of a post that has to be made.



    I mentioned this one before. I said it it hovers 16k to 24k per frame. Updated UI and more light sources and blur

  9. #3099
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Feb 2019
    Posts
    410
    Rep Power
    5

    Default

    based on this pics do you still believe textures are better in dreamcast version? to my eyes on ps2 is much better, low budget game by ps2 standards, but a typical dreamcast game.

  10. #3100
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    Since R Racing evolution is a very pretty racing with its real time reflections on the cars and specular maps on the road in some areas. Doesnt work well with the ps2 emulator so I decided to extract using the gamecube emulator. Was suprising what I found, if you take away the fancy effects its pretty much running at a dreamcast-esque polygon count and it doesnt seem to be doing sim physics, more arcade like. I guess if you remove the reflections and replace it with static environment maps and simplify the lighting you got a dc version, the cars off the distance use lods. I guess looks are deceiving and how art direction is everything. Like I was saying the common game seems to be with in grasp.

    R Racing Evolution -2003(GameCube version):

    Monaco start line - 20,063 tris per frame x 60 fps = 1.2 mpps



    Monaco middle of the race - 18,849 tris per frame x 60 fps = 1.1 mpps



    Motegi track zoomed out - 33,292 tris per frame x 60 fps = 1.97 mpps



    Motegi middle of the race - 22,106 tris per frame x 60 fps = 1.3 mpps


  11. #3101
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    Quote Originally Posted by Yharnamresident View Post
    The problem with the Dreamcast F1 World Grand Prix games is they were pieced together. Like the first game uses N64 and original Dreamcast assets, then the sequel also uses assets from the PC version. So the sequel uses N64, original Dreamcast and PC assets. How about just make the god damn Dreamcast port from the ground up.
    You know I went back to check those ubisoft made ones and found that monoco grandprix, racing simulation 2 , f1 racing championship are all actually the same exact game released over and over in Europe. The extremely low polygon count comes from the fact the game heavily uses backface culling on everything and barely draws anything out of the frame so even if you pile up all the cars it wont matter. They probably did this to save fillrate because the grass and road actually use bumpmapping. Yup bumpmapping on the grass and road but its hard to see because the mipmap use destroys it as its very up close and vicious. kinda realized it when i played it on the tv then went to check on the emulator to confirm. One of the few games besides rayman 2 and tomb raider the last revelation to use bumpmapping heavily. In this game its like rayman , a tileable texture that makes more grain/detail.

    F1 racing championship (dreamcast version)
    Bumpmapping on the road and grass:


    The bumpmap texture itself:


  12. #3102
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    One more for the gamecube. The little I could get of star wars rogue squadron 2 for the gamecube. The little I captured seemed to match jak 3 polygon count wise but I guess this game has more effects like bump mapping. still below 10 mpps though.

    Star Wars rogue squadron 2 - gamecube - 2001

    Mission select - 76,823 tris per frame x 60 fps = 4.6 mpps



    Hanger/ship select - 136,665 tris per frame x 60 fps = 8.1 mpps


  13. #3103
    The Future Is Yesterday Hedgehog-in-TrainingRaging in the Streets Leynos's Avatar
    Join Date
    Feb 2018
    Location
    My Chair
    Posts
    4,835
    Rep Power
    59

    Default

    A launch game doing 8MPPS is nuts for 2001.

    Life!? ... What console is that on?



    [PSN] Segata-S //[Switch] FC-SW 3892 5228 2895 //[XBL]Dogi99


  14. #3104
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Feb 2019
    Posts
    410
    Rep Power
    5

    Default

    ps2 airblade 8,4M november 2001, ps2 F1 2000 6M june 2000

    now that we are closer to the truth because we know now that Dolphin doesn't measure polygons accurately, the claim that the gamecube made the game with the highest polygonal count of the gen, can no longer be claimed, the highest possible count in Rogue Squadron 2 can only be gauged in the Hangar, during gameplay it's impossible all effects and slowdowns, should stay within the limit of 6M.

    see here; RE4 is credited with making 7.9M at the highest polygonal count

    https://www.sega-16.com/forum/showth...ardware/page28

  15. #3105
    Hedgehog-in-Training Hedgehog-in-TrainingRoad Rasher
    Join Date
    Aug 2018
    Posts
    389
    Rep Power
    8

    Default

    Quote Originally Posted by SegaAMD View Post
    ps2 airblade 8,4M november 2001, ps2 F1 2000 6M june 2000

    now that we are closer to the truth because we know now that Dolphin doesn't measure polygons accurately, the claim that the gamecube made the game with the highest polygonal count of the gen, can no longer be claimed, the highest possible count in Rogue Squadron 2 can only be gauged in the Hangar, during gameplay it's impossible all effects and slowdowns, should stay within the limit of 6M.

    see here; RE4 is credited with making 7.9M at the highest polygonal count

    https://www.sega-16.com/forum/showth...ardware/page28
    Ugh what are you deaf? Air blade doesn't do 8mpps, it just doesn't even get close. I've told you that a bunch of times. It caps out at over 6 mpps. To be 8mpps it would have to reach jak 3 type of geometry. Also since he's using the dolphin counter resident evil 4 is obviously not doing 8 mpps. Regardless it seems ps2 is deadlocked with GameCube at peak polygon count except the GC can obviously push far more effects and use bumpmapping while doing so while ps2 can't. That's not a put down to the ps2 it's just the GC had better hardware feature set.

    Edit: so if we compare timeframe 2001 gc was doing 8mpps vs ps2 7 mpps(first Jak game fight?).
    Last edited by Cloofoofoo; 05-19-2020 at 08:12 PM.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •