Re: GPL Direct3D patch, - quick comparison

Just done a quick test on my own set-up here (set-up described in previous message, but I've got a Voodoo3 2000 card clocked at 175). Set GPL to single race at Monaco with 19 AI cars: Then sat on grid and watched fps in both Glide and in Direct3D. Then raced away from grid and watched fps in both Glide and direct3D. Used exact same GPL settings for both (mirror detail set to car/track, detail slider set half way, skids on, most textures on, 4 sounds, res @ 1000x800, trackside detail to many). Results were as follows: ************************************ NB: Leave car sitting on grid, not moving at all: ************************************ Lowest fps sitting on grid: (Before cars move) Cockpit view: Glide = 29 Direct3D = 28 Arcade view: Glide = 30 Direct3D = 29 ************************************* Lowest fps sitting on grid: (as cars move away) Cockpit View: Glide = 22 Direct3D = 24 Arcade view: Glide = 24 Direct3D = 24 ************************************* Lowest fps sitting on grid: (cars disappear into distance) Cockpit view: Glide = 36 Direct3D = 36 Arcade view: Glide = 36 Direct3D = 36 ************************************* NB: Car drives off with other cars at start: ************************************* Lowest fps sitting on grid: (as cars move away) Cockpit View: Glide = 22 Direct3D = 22 Arcade view: Glide = 24 Direct3D = 22 ************************************* Lowest fps sitting on grid: (cars disappear into distance) Cockpit view: Glide = 36 Direct3D = 36 Arcade view: Glide = 36 Direct3D = 36 ************************************* As you can see not much difference, we are talking only one fps or two at most? The drops to the lowest fps only occur for about a second or two, then quickly rise. (Obviously this is only a *very* small test, on one system at one track, but hey, I've got a life! and I ain't gonna spend three weeks testing every possible scenario!) 8-) I didn't really percieve any visual difference, apart from the Direct3D is much lighter (and realistic) as I've always had fairly dark screens in Glide mode, no matter what I tried to change, and the fact that the mirrors look better. Overall, as there is no obvious fps hit, I'd personally rather use the Direct3D mode, cos I like the lighter screen! 8-) *Peter* - http://www.cix.co.uk/~peterpc/home.html

good test package...you just left out one ***y detail. What speed cpu are you running??? dave henrie

good test package...you just left out one teeny detail. What speed cpu are you running??? dave henrie <[email hidden]> wrote in message news:8j5lb6$imt$[email hidden]... > Just

In article <mav55.30397$[email hidden]>, [email hidden] (Dave Henrie) wrote: > good test package...you just left out one teeny detail. > What speed cpu are you running??? > dave henrie > <[email hidden]> wrote in message > news:8j5lb6$imt$[email hidden]... > > Just I did say.... >>(set-up described in previous message) However, if you missed my previous message, I'm running: Celery 300a @450 on a abit bh6 m/b with 128mb pc100 ram and a V3 2000 card @175 and am using Directx 7a and Win 95b. 8-) *Peter* - http://www.cix.co.uk/~peterpc/home.html

> good test package...you just left out one ***y detail. > What speed cpu are you running??? > dave henrie > > Just I did say.... However, if you missed my previous message, I'm running: 8-) *Peter* - http://www.cix.co.uk/~peterpc/home.html

Uhm, Gamma settings? I find that playing with those causes the exact results in both Direct 3D and Glide, but they need to be set differently for both modes on my Voodoo 3. On the other hand, did you notice that Direct3D handles transparency (railings at Nurburgring, for exampl) pretty awkwardly? The transparency is either on or off, there are no intermediate values (alpha is either full or zero). -Gregor [email hidden] wrote:

Quoted text (10 lines)

> > I didn't really percieve any visual difference, apart from the Direct3D > is much lighter (and realistic) as I've always had fairly dark screens > in Glide mode, no matter what I tried to change, and the fact that the > mirrors look better. Overall, as there is no obvious fps hit, I'd > personally rather use the Direct3D mode, cos I like the lighter screen! > > 8-) > > *Peter* - http://www.cix.co.uk/~peterpc/home.html

Ok I've run a few tests, and here's the results. My system Celery 533a @ 533 (non coppermine) abit BF6 mobo 128 meg PC100 SDRAM Voodoo 3 300 @ 166 4.12.01.0565 beta drivers Sonic storm pro pci sound card All test's at Monaco with a full field of AI, from the back of the grid without moving the numbers mean, FPS before start/Lowest FPS during start/Max FPS after start. All details, effects and sounds at max D3D (high res. mirrors) 640x480 12/11/28 800x600 12/10/29 1024x768 7/4/30 1152x864 2/2/30 1280x1024 1/1/30 D3D (low res. mirrors) 640x480 12/9/27 800x600 12/10/31 1024x768 6/2/30 1152x864 2/1/30 1280x1024 1/1/30 3DFX 512x384 12/10/37 640x480 11/10/36 800x600 10/9/34 960x720 9/8/31 1024x768 9/8/31 1280x1024 9/8/19 1600x1200 untested, my monitor wont handle it Also tested with the settings and screen res.'s I usually use All forward Textures on, rear Textures on track,objects and cars, smoke and skid effects on, detail slider at 1/3, mirrors on cars and track, trackside objects all, max sounds 8 D3D (high res.'s mirrors) 800x600 21/15/31 1024x768 19/13/31 D3D (low res. mirrors) 800x600 21/17/31 1024x768 19/14/32 3DFX 800x600 20/18/36 1024x768 18/17/36 And just for laughs in software mode Max details and sounds etc. 320x200 14/11/22 512x384 10/9/16 640x480 8/7/13 min. details and sounds etc.. (Sega virtual racer mode) 320x200 36/36/36 Wow I can matain max framerate LOL 512x384 28/27/36 640x480 24/22/30 Well after all that it shows that for me at least 3DFX is still the best option. But D3D looks beautiful especially the mirrors, and I still get 36 at all res.'s in training mode. -- Mark Aisthorpe London UK GPLRank +170.03 Remove NO-SPAM to reply <[email hidden]> wrote in message news:8j5lb6$imt$[email hidden]... > Just done a quick test on my own set-up here (set-up described in > previous message, but I've got a Voodoo3 2000 card clocked at 175). > > Set GPL to single race at Monaco with 19 AI cars: > > Then sat on grid and watched fps in both Glide and in Direct3D. > > Then raced away from grid and watched fps in both Glide and direct3D. > > Used exact same GPL settings for both (mirror detail set to car/track, > detail slider set half way, skids on, most textures on, 4 sounds, res @ > 1000x800, trackside detail to many). > > Results were as follows: > > ************************************ > > NB: Leave car sitting on grid, not moving at all: > > ************************************ > > Lowest fps sitting on grid: (Before cars move) > > Cockpit view: Glide = 29 Direct3D = 28 > > Arcade view: Glide = 30 Direct3D = 29 > > ************************************* > > Lowest fps sitting on grid: (as cars move away) > > Cockpit View: Glide = 22 Direct3D = 24 > > Arcade view: Glide = 24 Direct3D = 24 > > ************************************* > > Lowest fps sitting on grid: (cars disappear into distance) > > Cockpit view: Glide = 36 Direct3D = 36 > > Arcade view: Glide = 36 Direct3D = 36 > > ************************************* > > NB: Car drives off with other cars at start: > > ************************************* > > Lowest fps sitting on grid: (as cars move away) > > Cockpit View: Glide = 22 Direct3D = 22 > > Arcade view: Glide = 24 Direct3D = 22 > > ************************************* > > Lowest fps sitting on grid: (cars disappear into distance) > > Cockpit view: Glide = 36 Direct3D = 36 > > Arcade view: Glide = 36 Direct3D = 36 > > ************************************* > > As you can see not much difference, we are talking only one fps or two > at most? The drops to the lowest fps only occur for about a second or > two, then quickly rise. (Obviously this is only a *very* small test, on > one system at one track, but hey, I've got a life! and I ain't gonna > spend three weeks testing every possible scenario!) > > 8-) > > I didn't really percieve any visual difference, apart from the Direct3D > is much lighter (and realistic) as I've always had fairly dark screens > in Glide mode, no matter what I tried to change, and the fact that the > mirrors look better. Overall, as there is no obvious fps hit, I'd > personally rather use the Direct3D mode, cos I like the lighter screen! > > 8-) > > *Peter* - http://www.cix.co.uk/~peterpc/home.html

when you say 3Dfx still is the best, of coarse it should due better with it's proprietary API, but that doesn't mean it's better than Nvidia. you'd have to compare a comparable 3Dfx card to a Nvidia card both running their native API's. however on your system, "...for me at least (Glide) is still the best option" just saying... anyhow I'm really glad to see someone do some side by side tests with low res mirrors on. It did make me curious though that some of the low res tests did worse than the high ones. that leads me to believe that each race start is different, so the results are unbalanced. If you averaged say 25 starts it may be more fair. And finally frame rate really doesn't matter, since the counter seems pretty questionable, I'd go with what looks smoother, Id you can't tell the difference, then why not go with the high res mirrors? Mark Aisthorpe <[email hidden]> wrote in message news:ToT55.4258$[email hidden]...

Quoted text (194 lines)

> Ok I've run a few tests, and here's the results. > > My system > Celery 533a @ 533 (non coppermine) > abit BF6 mobo > 128 meg PC100 SDRAM > Voodoo 3 300 @ 166 4.12.01.0565 beta drivers > Sonic storm pro pci sound card > > All test's at Monaco with a full field of AI, from > the back of the grid without moving > the numbers mean, FPS before start/Lowest FPS > during start/Max FPS after start. > > All details, effects and sounds at max > > D3D (high res. mirrors) > 640x480 12/11/28 > 800x600 12/10/29 > 1024x768 7/4/30 > 1152x864 2/2/30 > 1280x1024 1/1/30 > > D3D (low res. mirrors) > 640x480 12/9/27 > 800x600 12/10/31 > 1024x768 6/2/30 > 1152x864 2/1/30 > 1280x1024 1/1/30 > > 3DFX > 512x384 12/10/37 > 640x480 11/10/36 > 800x600 10/9/34 > 960x720 9/8/31 > 1024x768 9/8/31 > 1280x1024 9/8/19 > 1600x1200 untested, my monitor wont handle it > > Also tested with the settings and screen res.'s I > usually use > All forward Textures on, > rear Textures on track,objects and cars, > smoke and skid effects on, > detail slider at 1/3, > mirrors on cars and track, > trackside objects all, > max sounds 8 > > D3D (high res.'s mirrors) > 800x600 21/15/31 > 1024x768 19/13/31 > > D3D (low res. mirrors) > 800x600 21/17/31 > 1024x768 19/14/32 > > 3DFX > 800x600 20/18/36 > 1024x768 18/17/36 > > And just for laughs in software mode > > Max details and sounds etc. > 320x200 14/11/22 > 512x384 10/9/16 > 640x480 8/7/13 > > min. details and sounds etc.. (Sega virtual racer > mode) > 320x200 36/36/36 Wow I can matain max framerate > LOL > 512x384 28/27/36 > 640x480 24/22/30 > > Well after all that it shows that for me at least > 3DFX is still > the best option. > But D3D looks beautiful especially the mirrors, > and I still get 36 at all res.'s in training mode. > > -- > Mark Aisthorpe > London UK > GPLRank +170.03 > Remove NO-SPAM to reply > > <[email hidden]> wrote in message > news:8j5lb6$imt$[email hidden]... > > Just done a quick test on my own set-up here > (set-up described in > > previous message, but I've got a Voodoo3 2000 > card clocked at 175). > > > > Set GPL to single race at Monaco with 19 AI > cars: > > > > Then sat on grid and watched fps in both Glide > and in Direct3D. > > > > Then raced away from grid and watched fps in > both Glide and direct3D. > > > > Used exact same GPL settings for both (mirror > detail set to car/track, > > detail slider set half way, skids on, most > textures on, 4 sounds, res @ > > 1000x800, trackside detail to many). > > > > Results were as follows: > > > > ************************************ > > > > NB: Leave car sitting on grid, not moving at > all: > > > > ************************************ > > > > Lowest fps sitting on grid: (Before cars move) > > > > Cockpit view: Glide = 29 Direct3D = 28 > > > > Arcade view: Glide = 30 Direct3D = 29 > > > > ************************************* > > > > Lowest fps sitting on grid: (as cars move away) > > > > Cockpit View: Glide = 22 Direct3D = 24 > > > > Arcade view: Glide = 24 Direct3D = 24 > > > > ************************************* > > > > Lowest fps sitting on grid: (cars disappear into > distance) > > > > Cockpit view: Glide = 36 Direct3D = 36 > > > > Arcade view: Glide = 36 Direct3D = 36 > > > > ************************************* > > > > NB: Car drives off with other cars at start: > > > > ************************************* > > > > Lowest fps sitting on grid: (as cars move away) > > > > Cockpit View: Glide = 22 Direct3D = 22 > > > > Arcade view: Glide = 24 Direct3D = 22 > > > > ************************************* > > > > Lowest fps sitting on grid: (cars disappear into > distance) > > > > Cockpit view: Glide = 36 Direct3D = 36 > > > > Arcade view: Glide = 36 Direct3D = 36 > > > > ************************************* > > > > As you can see not much difference, we are > talking only one fps or two > > at most? The drops to the lowest fps only occur > for about a second or > > two, then quickly rise. (Obviously this is only > a *very* small test, on > > one system at one track, but hey, I've got a > life! and I ain't gonna > > spend three weeks testing every possible > scenario!) > > > > 8-) > > > > I didn't really percieve any visual difference, > apart from the Direct3D > > is much lighter (and realistic) as I've always > had fairly dark screens > > in Glide mode, no matter what I tried to change, > and the fact that the > > mirrors look better. Overall, as there is no > obvious fps hit, I'd > > personally rather use the Direct3D mode, cos I > like the lighter screen! > > > > 8-) > > > > *Peter* - > http://www.cix.co.uk/~peterpc/home.html > >

[snip]

Quoted text (8 lines)

> As you can see not much difference, we are talking only one fps > or two at most? The drops to the lowest fps only occur for about > a second or two, then quickly rise. (Obviously this is only a > *very* small test, on one system at one track, but hey, I've got > a life! and I ain't gonna spend three weeks testing every possible > scenario!) > > 8-)

Funny. On my P-II-300, V3-3000 system I lost 5-10 fps nearly everywhere using D3D, not just when the graphics were loaded up. In Glide I get an almost constant 36 fps running at that funny 9??x7?? resolution between 800x600 and 1024x768. D3D must be quite a bit more processor intensive. -Chris- **Jesus saves sinners and redeems them for valuable cash prizes** Sent via Deja.com http://www.deja.com/ Before you buy.

Mark, Something seems wrong in your system. performance to low. I get at Monaco (same conditions as you) about 14/23/36 at 1024x768 in 3DFX I use the GPLEA carset. For fun I checked with a complete 3D-tireset from Tire-Barn (incredible they are) Was about 10/20/36 then My system: C300@450 V3 2000 @ 143 128 MB Sblive Asus MB Jens On Tue, 27 Jun 2000 02:25:19 +0100, "Mark Aisthorpe" <[email hidden]> wrote: >Ok I've run a few tests, and here's the results.<snip> --------------------------------------------- [email hidden] Remove "NOSPAM" before using my email-address JensSchumi homepage at http://members.xoom.com/JensSchu/ Racing simulation, esp. GPL Logitech Wingman FF optimization Classic cars, esp. Morris Minor Radio controlled airplanes

Beta... -Larry Gregor Veble <[email hidden]> wrote: >Uhm, Gamma settings? I find that playing with those causes theexact >results in both Direct 3D and Glide, but they need to be setdifferently >for both modes on my Voodoo 3. > >On the other hand, did you notice that Direct3D handlestransparency >(railings at Nurburgring, for exampl) pretty awkwardly? Thetransparency >is either on or off, there are no intermediate values (alpha iseither

Quoted text (7 lines)

>full or zero). > >-Gregor > >[email hidden] wrote: >> >> I didn't really percieve any visual difference, apart from

the Direct3D >> is much lighter (and realistic) as I've always had fairlydark screens >> in Glide mode, no matter what I tried to change, and the factthat the >> mirrors look better. Overall, as there is no obvious fps hit,I'd >> personally rather use the Direct3D mode, cos I like thelighter screen! >> >> 8-) >> >> *Peter* - http://www.cix.co.uk/~peterpc/home.html > > ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com