rec.autos.simulators

CPR Patch - more details

John Walla

CPR Patch - more details

by John Walla » Wed, 10 Dec 1997 04:00:00



FWIW I've had some very good races with both MTM and the CPR demo on
the Zone. Even using dial-up from home here in the UK it was very
playable.

Cheers!
John

Martin Simar

CPR Patch - more details

by Martin Simar » Wed, 10 Dec 1997 04:00:00


>The patch is intended to address the top issues:-

>1)     Improve AI
>2)     Boosting frame rate
>3)     Game controller null-zone
>4)     Pit lane auto-pilot
>5)     Gamma for track brightness
>6)     Gamma for car brightness
>7)     Rain tires
>8)     Friction levels on the grass
>9)     More precise control over graphics settings
>10)   Assorted bugs

I hope the patch corrects the flickering problem with Riva 128 cards. At
first I thought it was a driver problem but since I tried some other games
(including "F1 Racing Simulation") which use Direct3D without flicker, I now
know the problem is in CART.

Bye!

Martin Simard

VABergfe

CPR Patch - more details

by VABergfe » Wed, 10 Dec 1997 04:00:00

Will there be a new demo with the fixes?  To be honest, I wasn't too impressed
with control or graphics of the original demo (especially when it resided on my
HD at the same time as the F1 Racing Sim demo from UbiSoft), and pretty much
wrote off CPR and decided to wait for F1RS.  BUT, with 25-50% increase in frame
rate and the null zone adjustment it would be worth another look.  So, any
chance of a new demo?

Alan

CART Te

CPR Patch - more details

by CART Te » Wed, 10 Dec 1997 04:00:00

Joe,

You_are_correct - there are some features unique to the 3Dfx hardware that
cannot be accessed using DirectX5 - none that are useful to CPR I might add.

However, the next generation of DirectX, I am assured, will address the
double-pass rendering and the other last couple of items.

The performance issue I stand by - why? Because we just spent 3 weeks and
$$$ on doing the glide port
so we know what we're talking about. This is not secondhand info - we did
the numbers. I have the two versions in front of me right now - we did the
port because WE thought it might buy us some performance. It didn't - end of
story (yeah, right).

So I don't expect for one minute for anyone to support/approve or even
believe this data - but so what? We did the work, we have the data, that's
all I need.

We wouldn't have even done the Glide port if we were so set against it.

My job is to make CPR (and Flight Sim and Monster Truck and X, Y and Z tba)
the best Sims they can be - not to make D3D look good if it isn't.

I don't work in the Systems group - I work in the Games group - I run the
Simulations business for Microsoft - if my products stiff because I made bad
technology choices then it's my profitability - and ultimately my head.

There it is - I have been (painfully) honest on this NG so far and I do not
intend to start changing that now.



>>And before anyone posts their MS *** theory <yawn>, go here and
check
>>out a posting that corresponds exactly with what we found - and these guys
>>have no reason to support MS - they develop a competing product to our own
>>Flight Simulator ...

>>http://www.racesimcentral.net/
14
>>39978.800457020&hitnum=0

>>"The patch is indeed reality. It should be on our web site next week. It
>>does not include native glide support. But its not like that would get you
>>vast speed improvements anyway. We've worked hard to get good efficiency
>>with d3d. The biggest thing you can do for your hardware performance is to
>>make sure your driver supports DrawPrimitive.

>>-James Fleming,
>>lead programmer on fu2"

>Since you posted this, I think it's only fair to point out that Andy
>Hollis, the highly respected sim designer/project leader on LongBow2,
>has stated just as emphatically (if not more so) that the LB2 graphics
>quality and speed can not be done via D3D on any current 3D hardware
>(3dfx or otherwise). Only Glide was fast [and feature rich] enough.

>Not suggesting a ***, just pointing out that not all
>programmer's have reached the conclusions you are posting here. Of
>course, Mr. Hollis was referring to graphical quality as well as speed
>- there are several features of the 3dfx card that can only be
>accessed via Glide. A "feature gap", I might point out, that will only
>grow wider in the future if Microsoft doesn't change their stance on
>the issue of supporting hardware geometry acceleration in D3D.

>I'm not saying I boycott D3D games or anything. But I prefer Glide
>games, as they seem generally more reliable, faster, and of higher
>visual quality.

>Joe

>.

CART Te

CPR Patch - more details

by CART Te » Wed, 10 Dec 1997 04:00:00

Not all D3D games use the same hardware features in the same way.

Diamond's latest drivers for their Riva 128 chipset FIXED the flicker.

I know the other Riva OEMs are working on updated drivers.

It is a driver problem - I assure you.



>>The patch is intended to address the top issues:-

>>1)     Improve AI
>>2)     Boosting frame rate
>>3)     Game controller null-zone
>>4)     Pit lane auto-pilot
>>5)     Gamma for track brightness
>>6)     Gamma for car brightness
>>7)     Rain tires
>>8)     Friction levels on the grass
>>9)     More precise control over graphics settings
>>10)   Assorted bugs

>I hope the patch corrects the flickering problem with Riva 128 cards. At
>first I thought it was a driver problem but since I tried some other games
>(including "F1 Racing Simulation") which use Direct3D without flicker, I
now
>know the problem is in CART.

>Bye!

>Martin Simard

>.

CART Te

CPR Patch - more details

by CART Te » Wed, 10 Dec 1997 04:00:00

It is my intention that we'll release a patch for the full game and a new
demo incorporating the same improvements.

VABergfeld wrote in article

Paul Sander

CPR Patch - more details

by Paul Sander » Wed, 10 Dec 1997 04:00:00

The Zone is not required.  As long as the host can get his IP address to all
clients.

-Paul



>> The patch is intended to address the top issues:-

>> 1)     Improve AI
>> 2)     Boosting frame rate
>> 3)     Game controller null-zone
>> 4)     Pit lane auto-pilot
>> 5)     Gamma for track brightness
>> 6)     Gamma for car brightness
>> 7)     Rain tires
>> 8)     Friction levels on the grass
>> 9)     More precise control over graphics settings
>> 10)   Assorted bugs

>> The good news is that the patch addresses all of these items.

>   Well, if you get all this done, especially the null zone thing, then
>I may just have to go get it. How does this game shape up as an offline
>competition sim. The Zone is good but eliminates the prospects of an
>international comp. Cheers.

Paul Sander

CPR Patch - more details

by Paul Sander » Wed, 10 Dec 1997 04:00:00

It's a driver problem.  When I downloaded the latest Viper driver from
http://www.diamondmm.com, I ran CPR before installing the new driver.  Major
flicker.  I installed the driver, no flicker.

-Paul



>>The patch is intended to address the top issues:-

>>1)     Improve AI
>>2)     Boosting frame rate
>>3)     Game controller null-zone
>>4)     Pit lane auto-pilot
>>5)     Gamma for track brightness
>>6)     Gamma for car brightness
>>7)     Rain tires
>>8)     Friction levels on the grass
>>9)     More precise control over graphics settings
>>10)   Assorted bugs

>I hope the patch corrects the flickering problem with Riva 128 cards. At
>first I thought it was a driver problem but since I tried some other games
>(including "F1 Racing Simulation") which use Direct3D without flicker, I
now
>know the problem is in CART.

>Bye!

>Martin Simard

Scott B. Huste

CPR Patch - more details

by Scott B. Huste » Wed, 10 Dec 1997 04:00:00

Dean,

I for one appreciate having you on this newsgroup. Its nice to know
someone with direct input into what ultimately comes out to the market.

Just a quick question.  Awhile back I mentioned (requested) a FAQ on how
to increase *** and general performance in Windows95 and you thought
it was a good idea and said you would pass it along.  I havent heard
anything since and was wondering where we might stand on that.

Thanks,

Scott B. Husted
http://www.racesimcentral.net/~sbhusted


> Joe,

> You_are_correct - there are some features unique to the 3Dfx hardware that
> cannot be accessed using DirectX5 - none that are useful to CPR I might add.

> However, the next generation of DirectX, I am assured, will address the
> double-pass rendering and the other last couple of items.

> The performance issue I stand by - why? Because we just spent 3 weeks and
> $$$ on doing the glide port
> so we know what we're talking about. This is not secondhand info - we did
> the numbers. I have the two versions in front of me right now - we did the
> port because WE thought it might buy us some performance. It didn't - end of
> story (yeah, right).

> So I don't expect for one minute for anyone to support/approve or even
> believe this data - but so what? We did the work, we have the data, that's
> all I need.

> We wouldn't have even done the Glide port if we were so set against it.

> My job is to make CPR (and Flight Sim and Monster Truck and X, Y and Z tba)
> the best Sims they can be - not to make D3D look good if it isn't.

> I don't work in the Systems group - I work in the Games group - I run the
> Simulations business for Microsoft - if my products stiff because I made bad
> technology choices then it's my profitability - and ultimately my head.

> There it is - I have been (painfully) honest on this NG so far and I do not
> intend to start changing that now.



> >>And before anyone posts their MS *** theory <yawn>, go here and
> check
> >>out a posting that corresponds exactly with what we found - and these guys
> >>have no reason to support MS - they develop a competing product to our own
> >>Flight Simulator ...

> >>http://www.racesimcentral.net/
> 14
> >>39978.800457020&hitnum=0

> >>"The patch is indeed reality. It should be on our web site next week. It
> >>does not include native glide support. But its not like that would get you
> >>vast speed improvements anyway. We've worked hard to get good efficiency
> >>with d3d. The biggest thing you can do for your hardware performance is to
> >>make sure your driver supports DrawPrimitive.

> >>-James Fleming,
> >>lead programmer on fu2"

> >Since you posted this, I think it's only fair to point out that Andy
> >Hollis, the highly respected sim designer/project leader on LongBow2,
> >has stated just as emphatically (if not more so) that the LB2 graphics
> >quality and speed can not be done via D3D on any current 3D hardware
> >(3dfx or otherwise). Only Glide was fast [and feature rich] enough.

> >Not suggesting a ***, just pointing out that not all
> >programmer's have reached the conclusions you are posting here. Of
> >course, Mr. Hollis was referring to graphical quality as well as speed
> >- there are several features of the 3dfx card that can only be
> >accessed via Glide. A "feature gap", I might point out, that will only
> >grow wider in the future if Microsoft doesn't change their stance on
> >the issue of supporting hardware geometry acceleration in D3D.

> >I'm not saying I boycott D3D games or anything. But I prefer Glide
> >games, as they seem generally more reliable, faster, and of higher
> >visual quality.

> >Joe

> >.

Tony Rickar

CPR Patch - more details

by Tony Rickar » Wed, 10 Dec 1997 04:00:00

I was surprised how well it worked in the UK as well, although I guess it
takes some practice - couldn't achieve the same times as stand alone
getting used to the change in frame rate. Still very playable though. (Just
my excuse for losing!)

Tony

XRaceTr

CPR Patch - more details

by XRaceTr » Wed, 10 Dec 1997 04:00:00


> I hope the patch corrects the flickering problem with Riva 128 cards.
> At first I thought it was a driver problem but since I tried some other
> games (including "F1 Racing Simulation") which use Direct3D
> without flicker, I now know the problem is in CART.

Which board are you having problems with? I've got the Riva
128 based Diamond Viper 330 both at home (PCI) and at work
(AGP), and, after installing the new drivers, the demo is flicker
free.

I haven't tried it with any other D3D games that I know of... Is ABC's
Road to Indy a D3D title? That's the only other new title I have tried
recently and I know it doesn't flicker...

- Mike.

====================================================
 Mike Holthouse - Indianapolis, IN
 I'm sick of spam!! To e-mail me, remove the "x"...
====================================================

Jarrod Smit

CPR Patch - more details

by Jarrod Smit » Wed, 10 Dec 1997 04:00:00


> Dan,

> The good news is we have achieved between 25 and 50% (or greater) frame rate
> improvements for ALL users regardless of 3D card type - or no 3D card at
> all.

> We did a lot of benchmarking on Glide (that's what we have been doing for
> the last 3 weeks) but it didn't buy us very much  - you need a really slow
> (yes, s-l-o-w) machine to see any difference at all. So TRi found several
> other ways to achieve much better improvements - for everyone.

> I am sure many of the "3D experts" on this NG will cry foul - but TRi  spent
> weeks on the Glide version and have benchmarked the two versions of the SAME
> game side-by-side. Anyone else who has done this much research and coding in
> this area is welcome to offer their data - but the glib "Glide is much
> faster than D3D" story does not turn out to be the case. I am sure that
> won't satisfy everyone - especially those that love a good MS ***
> theory - but that's the reality.

> And before anyone posts their MS *** theory <yawn>, go here and check
> out a posting that corresponds exactly with what we found - and these guys
> have no reason to support MS - they develop a competing product to our own
> Flight Simulator ...

It's kind of ridiculous for you to catagorically state that the "Glide
is much faster than D3D story" is not the case.  I am not at all
surprised that it isn't the case for CPR or some other applications, but
that's likely for reasons related to the way the 3D engine is designed
and/or the way the physics and AI calculations may dominate what's going
on and not the graphics.  You can't change the fact that D3D is an
additional software layer between the application and the hardware and
therefore eats CPU cycles.  From my recollection of discussions led by
Daryll Strauss (the guy who ported Glide to Linux), Glide is mainly
written in assembly and is a very thin and highly optimized layer
between 3dfx hardware and the application.  It is therefore highly
effecient and doesn't waste many CPU cycles.  It just depends on where
the bottlenecks are in the software.  I'm not surprised that graphics
are not a bottleneck in CPR.  They don't look that great IMO.

Besides that, there is the issue of extra features that can be supported
by a native Glide port versus D3D...I don't know what they are, but
according to Andy Hollis (the producer of Longbow2) they are there.

--

The Scripps Research Institute
http://www.racesimcentral.net/~jsmith

Jo

CPR Patch - more details

by Jo » Wed, 10 Dec 1997 04:00:00


>FWIW I've had some very good races with both MTM and the CPR demo on
>the Zone. Even using dial-up from home here in the UK it was very
>playable.

I never got it working well at all. Horrible stuttering/jumping of
other cars. And I'm less than a hundred miles from Microsoft
headquarters.

Joe

Jo

CPR Patch - more details

by Jo » Wed, 10 Dec 1997 04:00:00


Yup, that's one of the big disadvantages of D3D all right. It seems
even the most exhaustive testing by developers can't eliminate
graphical glitches on at least some cards.

Joe

Jo

CPR Patch - more details

by Jo » Wed, 10 Dec 1997 04:00:00


>So the possibility of GL patch is gone, what a disappointment.  I was hoping
>to see an end to all of the flicker that owners of 3dfx Voodoo Rush cards
>have been getting with the D3d.  I do recognize that CPR is the only D3D
>program that I have with the flicker problem (FS 98, and MotoRacer look
>great).  So if the flicker is still in CPR and GL patch does not happen, I
>can always use another coaster for my coffee table (CPR CD) I was eager to
>have a choice, and to make judgements on my own which mode, GL or D3D  I
>would prefer to run in.

That's a good point, that has been missed on this - Glide games aren't
only faster and easier to program, they're also more reliable. I've
never had a Glide program fail to run on my Monster3d 3dfx setup
(can't say the same for D3D), and never run into the graphical
glitches that are fairly common in D3D games.

Joe


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.