DAYTONA arcade from SEGA
Jonas Elfström
What hardware is there inside the arcademachine that runs
Daytona from SEGA?
It's the best texturemapping I've seen... Looks like 640x512x15 to me
and blizzing fast too.
It also seems to be "real" 3d and not DOOM 3d but that might be an
illusion.
I've also seen Ridge Racer (or what's it called) on Playstation but
I wasn't impressed... (the spec. for that machine promises alot more)
Does anyone know if there is a development kit for PC for Playstation
(I've heard rumours)?
--
Jonas Elfstroem !The more you suffer !ECS...@klecs1.ericsson.se
ERICSSON MOBILE !The more it shows you! ->voice: +46 19 584455<-
COMMUNICATIONS !really care Right?! /*amigafreak, R.E.M. */
This is my point of view not Ericsson's /*Offspring, SOM, DM */
Robert Glidden
Lockheed's first downmarket move from military simulation. They have
announced they are working on PC chip set.
Kind of makes you think accelerated 3D on PCs could be a very big deal.
Jonas Elfström (ECS...@klecs1.ericsson.se) wrote:
: Just curious...
Steve Woodcock
: Just curious...
:
: What hardware is there inside the arcademachine that runs
: Daytona from SEGA?
:
Hello Jonas:
I work for Lockheed-Martin in the game development group. We build
the boards used in several Sega games (Daytona, VF, VF2).
Daytona uses the Model 2 board. It's not *exactly* the same as
the Real-3D chipset (as a previous poster had said), but it's in the
same family. VF1 used the Model 1 board, while VF2 uses the Model 2.
The biggest difference between the two is, as you say, the texturing
capabilities. The Model 2 is also used in Desert Tank, the first game
done by our in-house Lockheed-Martin development team.
The texture mapping is all done in the hardware using an onboard
coprocessor specifically designed to push polygons around. The top speed
of the board is somewhere around 250,000-300,000 textured polygons per second.
Unfortunately I can't really go into tons of detail about the board
for proprietary reasons. Everything I've mentioned above can be found in
the press, particularly in Edge and Next Generation magazines.
: It's the best texturemapping I've seen... Looks like 640x512x15 to me
: and blizzing fast too.
: It also seems to be "real" 3d and not DOOM 3d but that might be an
: illusion.
:
It's real 3D. It's a pain in the butt sometimes, but it's real.
: I've also seen Ridge Racer (or what's it called) on Playstation but
: I wasn't impressed... (the spec. for that machine promises alot more)
: Does anyone know if there is a development kit for PC for Playstation
: (I've heard rumours)?
:
Yes, there is a development kit available. Actually, I believe there
are two: one's a 'standalone' type of kit, and the other is for the PC.
I believe the cost of a Playstation PC kit is somewhere around $18,000;
a similar kit for the Sega Saturn runs around $15,000. NOT cheap.
My personal opinion of Ridge Racer is that the database/graphics
work is nice, but the game suffers from a lack of variety (i.e., a single
race track) and some polygon priority problems. My hunch is that they
probably rushed it out the door for the Playstation release, but that's
just a guess on my part.
==============================================================================
Steven Woodcock _
Senior Software Engineer, Gameware _____C .._.
Lockheed Martin Information Systems Group ____/ \___/
Phone: 407-826-6986 <____/\_---\_\ "Ferretman"
E-mail: wood...@gate.net
swoo...@oldcolo.com
wood...@escmail.orl.mmc.com
Disclaimer: My opinions in NO way reflect the opinions of the
Lockheed Martin Information Systems Group, although
(like Rush Limbaugh) they should. ;)
Motto:
"...Men will awake presently and be Men again, and colour and laughter and
splendid living will return to a grey civilization. But that will only come
true because a few Men will believe in it, and fight for it, and fight in its
name against everything that sneers and snarls at that ideal..."
-- Leslie Charteris
THE LAST HERO
Graeme Love
ECS...@klecs1.ericsson.se "Jonas Elfstrm" writes:
> What hardware is there inside the arcademachine that runs
> Daytona from SEGA?
All recent SEGA arcade machines use their Model N series of boards. The
Daytona board is the Model 2 board. As far as I know, it's basically a lot
of pre-packaged chips which SEGA have bought.
> I've also seen Ridge Racer (or what's it called) on Playstation but
> I wasn't impressed... (the spec. for that machine promises alot more)
Only if you believe the hype that is surrounding the machine. I work for
a company which is working on Playstation projects and the specs which I've
seen advertised for the machine are a tad (!!?) embellished :) It's still
a powerful machine though, more so than the Saturn. I've only seen screen
shots of Daytona on the Saturn, but from what I've seen it's no exactly an
arcade perfect conversion (i.e. 320x224, mostly 4-bit textures, track update
a bit shoddy)
> Does anyone know if there is a development kit for PC for Playstation
> (I've heard rumours)?
The current development system for the Playstation is a set of boards which
fit in your PC. However, even if you could get them from Sony, the cost
around 20,000 British pounds for the full kit (and Sony claim that they're
not making a profit on them either).
See Ya
Graeme
--
Graeme Love
gl...@cadp35.demon.co.uk
Peter.Mart...@p6.f36.n237.z2.fidonet.org
CHRS: LATIN-1 2
In a message of 12 May 95 Jonas Elfström wrote to hfik bhndmbbm:
JE> Just curious...
That's ok.
JE> What hardware is there inside the arcademachine that runs Daytona
from
JE> SEGA?
I don't remeber precisly, but there are two 32 bit processors for the data
processesing, and there's one for the CD-rom drive. BTW i think it is
RISC processors.
JE> It's the best texturemapping I've seen... Looks like 640x512x15 to me
JE> and blizzing fast too.
JE> It also seems to be "real" 3d and not DOOM 3d but that might be an
JE> illusion.
It's cool isn't it :-)
JE> I've also seen Ridge Racer (or what's it called) on Playstation but
I
JE> wasn't impressed... (the spec. for that machine promises alot more)
They had to halve the framerate, and there are still slowdowns :-(. Even
on Parodius there are slowdowns :-(
JE> Does anyone know if there is a development kit for PC for Playstation
JE> (I've heard rumours)?
hhhhmmmm. [finding EDGE from the desk drawer]... here it is... 12000
pounds! It is the Psy-Q thingy as far as i remeber..... Yep, and there's
also one for the Sega Saturn, but i haven't got the price of it here.
I hope this was helpfull to you.
Peter Martin Hartwig
--- Spot 1.3a Unregistered
* Origin: -+- Spot -+- (2:237/36.6)
Peter.Mart...@p6.f36.n237.z2.fidonet.org
CHRS: LATIN-1 2
In a message of 13 May 95 Graeme Love wrote to hdgl bhnebdjk:
GL> All recent SEGA arcade machines use their Model N series of boards.
The
GL> Daytona board is the Model 2 board. As far as I know, it's basically a
GL> lot of pre-packaged chips which SEGA have bought.
Sega Rally runs on a Model 2, but i don't think Daytona is running on it.
I think it is older..
Richard Walter
> MSGID: 2:237/36.6 2fb7e459
> CHRS: LATIN-1 2
>In a message of 13 May 95 Graeme Love wrote to hdgl bhnebdjk:
> GL> All recent SEGA arcade machines use their Model N series of boards.
>The
> GL> Daytona board is the Model 2 board. As far as I know, it's basically a
> GL> lot of pre-packaged chips which SEGA have bought.
>Sega Rally runs on a Model 2, but i don't think Daytona is running on it.
>I think it is older..
Nope, it's the Model 2 board also, the Model 2 has been around longer than
you think!
- Rich
Bernie Roehl
Steve Woodcock <wood...@gate.net> wrote:
> I work for Lockheed-Martin in the game development group. We build
>the boards used in several Sega games (Daytona, VF, VF2).
> My personal opinion of Ridge Racer is that the database/graphics
>work is nice, but the game suffers from a lack of variety (i.e., a single
>race track) and some polygon priority problems.
So the chipset doesn't do Z-buffering?
--
Bernie Roehl
University of Waterloo Dept of Electrical and Computer Engineering
Mail: bro...@sunee.uwaterloo.ca Voice: (519) 888-4567 x 2607 work
URL: http://sunee.uwaterloo.ca/~broehl
Graeme Love
bro...@coulomb.uwaterloo.ca "Bernie Roehl" writes:
> In article <3p1a6i$19...@news.gate.net>,
> >race track) and some polygon priority problems.
> ^^^^^^^^^^^^^^^^
> So the chipset doesn't do Z-buffering?
>
Oh no, if you use the library's method of rendering then each polygon is
depth sorted using it's average Z value. This causes the priority problems.
You can see this in Toshinden as well, as bits of the characters swap depth
order. A bit crap, but not easy to get around.
--
Graeme Love
gl...@cadp35.demon.co.uk
Paul Hsieh
: In article <D8pA9...@watserv3.uwaterloo.ca>
: bro...@coulomb.uwaterloo.ca "Bernie Roehl" writes:
: > In article <3p1a6i$19...@news.gate.net>,
: > >race track) and some polygon priority problems.
: > ^^^^^^^^^^^^^^^^
: > So the chipset doesn't do Z-buffering?
: >
: Oh no, if you use the library's method of rendering then each polygon is
: depth sorted using it's average Z value. This causes the priority problems.
What the heck?? AVERAGE Z-value?? Geez ... . Is the VRAM really *THAT*
expensive? Use a real Z-buffer for gosh sakes ...
: You can see this in Toshinden as well, as bits of the characters swap depth
: order. A bit crap, but not easy to get around.
Are we not software developers? Have our brains turned to putty? There *IS*
and easy work around: Subdivide the "problem" polygons into smallers ones
so that the sorting works itself out better. Do I win a prize?
Paul Hsieh
Steve Woodcock
: Graeme Love (gl...@cadp35.demon.co.uk) wrote:
: : In article <D8pA9...@watserv3.uwaterloo.ca>
: : bro...@coulomb.uwaterloo.ca "Bernie Roehl" writes:
: : > In article <3p1a6i$19...@news.gate.net>,
: : > >race track) and some polygon priority problems.
: : > ^^^^^^^^^^^^^^^^
: : > So the chipset doesn't do Z-buffering?
: : >
: : Oh no, if you use the library's method of rendering then each polygon is
: : depth sorted using it's average Z value. This causes the priority problems.
: What the heck?? AVERAGE Z-value?? Geez ... . Is the VRAM really *THAT*
: expensive? Use a real Z-buffer for gosh sakes ...
: : You can see this in Toshinden as well, as bits of the characters swap depth
: : order. A bit crap, but not easy to get around.
Toshiden is a Sony Playstation game, is it not? Daytona is both a
Sega game and in the arcades....not at all the same hardware.
: Are we not software developers? Have our brains turned to putty? There *IS*
: and easy work around: Subdivide the "problem" polygons into smallers ones
: so that the sorting works itself out better. Do I win a prize?
Every polygon costs a certain amount of CPU power to process. If you
split up too many 'problem' polygons into smaller ones, you end up trading
realism for detail elsewhere in the sim.
Steven
==============================================================================
Steven Woodcock _
Senior Software Engineer, Gameware _____C .._.
Lockheed Martin Information Systems Group ____/ \___/
Phone: 407-826-6986 <____/\_---\_\ "Ferretman"
E-mail: wood...@gate.net (Home)
swoo...@oldcolo.com (Alternate Home)
wood...@escmail.orl.mmc.com (Work)
Scott Le Grand
> Paul Hsieh (phsieh@jedi) wrote:
> : Graeme Love (gl...@cadp35.demon.co.uk) wrote:
> : : In article <D8pA9...@watserv3.uwaterloo.ca>
> : : bro...@coulomb.uwaterloo.ca "Bernie Roehl" writes:
> : : > In article <3p1a6i$19...@news.gate.net>,
> : : > >race track) and some polygon priority problems.
> : : > ^^^^^^^^^^^^^^^^
> : : > So the chipset doesn't do Z-buffering?
> : : >
> : : Oh no, if you use the library's method of rendering then each polygon is
> : : depth sorted using it's average Z value. This causes the priority problems.
>
> : What the heck?? AVERAGE Z-value?? Geez ... . Is the VRAM really *THAT*
> : expensive? Use a real Z-buffer for gosh sakes ...
Z-buffers are expensive and consume bandwidth during rendering. The Atari jaguar
is the only platform I know of with true hardware support for Z-buffering. Even
so, not too many people use it because of the increased rendering time... The
lack of rendering errors does look tres' cool though and there are some nifty
FX you can pull off with one object being eaten by another as well that are
tougher to do on a machine without this feature...
So why do Z-buffers cost more? Well, if you just blindly spew pixels, you
can machine gun write them with a blitter or a REPSTO kind of instruction
for solid fill instructions, use a series of stores for gouraud shading, or
if you're texture mapping. read a bunch pixels, flip the bus direction,
write a bunch of pixels, flip the bus, repeat. In z-buffering, you must now do
2 way transfers for all kinds of polygons, and incur extra read and write
overhead for texture-mapped polygons (not to mention all those annoying
z-comparisons).
However, the results look great and if you do use it, be sure to flaunt
it or it's a real waste...
> : : You can see this in Toshinden as well, as bits of the characters swap depth
> : : order. A bit crap, but not easy to get around.
>
>
> Toshiden is a Sony Playstation game, is it not? Daytona is both a
> Sega game and in the arcades....not at all the same hardware.
But they all get the same errors from the same vintage 1977 polygon rendering
algorithms on steroids :-)... I wonder if U64 will rely on z-buffering since
all SGI hardware now must support it in hardware or software for openGL...
> : Are we not software developers? Have our brains turned to putty? There *IS*
> : and easy work around: Subdivide the "problem" polygons into smallers ones
> : so that the sorting works itself out better. Do I win a prize?
>
> Every polygon costs a certain amount of CPU power to process. If you
> split up too many 'problem' polygons into smaller ones, you end up trading
> realism for detail elsewhere in the sim.
Ya know, I have no problem with the polygon priority errors between objects
in the SEGA arcade machines. After all, just about every platform except for
the jaguar consistently has them, and the jaguar avoids it by using a hardware
Z-buffer at the cost of rendering speed. However, and this is a BIG however, it
really sucks when it happens within objects... C'mon, hasn't anyone ever heard of
binary space partition trees? I mean sheesh... I notice this all the time
in Wing Warz with the helicopter landing gear and it's really silly...
Foley and Van Dam maintain that the best way to render polygons is to
use a 1 scanline Z-buffer kind of algorithm. One of these days I have to
try this out... Does anyone have any real benchmarks on this?
Fortunately, only technerds ever notice this stuff or half the videogames
business would collapse...
Scott@4Play
Graeme Love
> binary space partition trees? I mean sheesh... I notice this all the time
Yup, the best way to stop these sorting problems is the good old BSP tree.
> Fortunately, only technerds ever notice this stuff or half the videogames
> business would collapse...
>
> Scott@4Play
I agree, even with the flickery polygons in Toshinden it's still a good game ( and that's what
counts).
Good job I didn't mention the Playstation's texture warping problem then ;-)
--
Graeme Love
gl...@cadp35.demon.co.uk
Jeff Lander
polygon hardware is a 2D sprite engine on Saturn and Sony.
It does not interpolate Z in the hardware so a Z buffer engine
is impossible in hardware.
You could do it in software but then the powerful Saturn becomes
a 33Mhz AT. Pretty damm slow....
That is why we all use average-min-max Z sorts on the Sega and
Sony. My kingdom for Z buffering, Perspective correct, texture
mapper in hardware.
Jeff Lander
Dan Gosselin
:You miss a major point on Z-Buffering on the Saturn. The
Then you should get into (3DO) M2 programming, since the M2 will have
Z-buffering in hardware, besides 7-10 times the power of the PlayStation.
That system will rule... I can't wait.
-Dan
Jim Boulton
writes:
>
> You miss a major point on Z-Buffering on the Saturn. The
The Jaguar has hardware z-buffering, but unfortunately it's only
fot 2 meg of memory which is VERY limiting... :(
--Jim
Graeme Love
>
> Then you should get into (3DO) M2 programming, since the M2 will have
> Z-buffering in hardware, besides 7-10 times the power of the PlayStation.
> That system will rule... I can't wait.
>
> -Dan
The power of the system does not automatically guarantee that the games will be any good. Given that
the Playstation has not yet been fully released world-wide, it's already being talked down in the
media which helped hype it so much. This constant churning out of new hardware by the console
companies in search of the highest specs is not good for the developers, and consequently, not good
for the consumer. It takes time for developers to learn how to get the best from a machine, and they
are just not being afforded that time these days in the constant pressure from publishers to get
their products on the market.
Just my two cents (pennies, deutchmarks, phennigs, etc.) about the state of the industry that I work
in. Just don't tell my publishers ;-)
--
Graeme Love
gl...@cadp35.demon.co.uk
David Nagy
: > Then you should get into (3DO) M2 programming, since the M2 will have
: > Z-buffering in hardware, besides 7-10 times the power of the PlayStation.
: > That system will rule... I can't wait.
: The power of the system does not automatically guarantee that the games
: will be any good.
The earlier poster wasn't talking about the availability of "good
games". He merely mentioned that he wished that he had access to
hardware z-buffering, hardware perpective correction, MIP mapping,
etc...
[...]
: media which helped hype it so much. This constant churning out of new
: hardware by the console companies in search of the highest specs is not
: good for the developers, and consequently, not good for the consumer.
I don't want to get into a silly system advocacy discussion here, but the
real beauty of the M2 is that it's O/S, development enviroment, tools, etc.
are very similar to the the original 3DO's. I agree that re-inventing the
wheel every couple of years is not conducive to good software.
: It takes time for developers to learn how to get the best from a
: machine, and they are just not being afforded that time these days in
: the constant pressure from publishers to get their products on the market.
It must be a pain. :)
: Just my two cents (pennies, deutchmarks, phennigs, etc.) about the state
: of the industry that I work in. Just don't tell my publishers ;-)
And I'm just some guy on the outside looking in...
Dave Nagy
Robert Kedward
>From: bro...@coulomb.uwaterloo.ca (Bernie Roehl)
>Subject: Re: DAYTONA arcade from SEGA
>Date: Wed, 17 May 1995 02:15:07 GMT
>In article <3p1a6i$19...@news.gate.net>,
>Steve Woodcock <wood...@gate.net> wrote:
>> I work for Lockheed-Martin in the game development group. We build
>>the boards used in several Sega games (Daytona, VF, VF2).
>>...
>> My personal opinion of Ridge Racer is that the database/graphics
>>work is nice, but the game suffers from a lack of variety (i.e., a single
> ^^^^^^^^^^^^^^^^
>So the chipset doesn't do Z-buffering?
>
> Bernie Roehl
> University of Waterloo Dept of Electrical and Computer Engineering
> Mail: bro...@sunee.uwaterloo.ca Voice: (519) 888-4567 x 2607 work
> URL: http://sunee.uwaterloo.ca/~broehl
I don't think that it matters that much. I play Daytona, Ridge Racer and
Sega Rally ALOT. I enjoy all of the games but I dont find the occasional
glitch here and there off-putting. Admitedly (sp?) Sega Rally has alot more
problems (ie seeing the sky through the road, cars flickering badly when
close to the screen.) the latter most likely because the hardware decides it
has drawn enough polygons for the frame and not drawing any more to keep the
frame rate high. But as I say these problems aren't too severe and do not
detract from the gameplay.
Rob.K.
Steve Woodcock
: I don't think that it matters that much. I play Daytona, Ridge Racer and
: Sega Rally ALOT. I enjoy all of the games but I dont find the occasional
: glitch here and there off-putting. Admitedly (sp?) Sega Rally has alot more
: problems (ie seeing the sky through the road, cars flickering badly when
: close to the screen.) the latter most likely because the hardware decides it
: has drawn enough polygons for the frame and not drawing any more to keep the
: frame rate high. But as I say these problems aren't too severe and do not
: detract from the gameplay.
That is usually the most common reason why polygon glitches happen;
there's usually logic to try to avoid pushing more polygons than the
graphics engine can actually manage, but sometimes the tests can fail
and something gets dropped.
Other times, polygon problems are caused by accidently getting in
a position to view the wrong side of a single-sided polygon, and/or from
having twisted vertices on a polygon that aren't apparent under most
circumstances. Most of the tools on the market do let you check these
kinds of things, but it seems like one or two always slip through.
Steve
==============================================================================
Steven Woodcock _
Senior Software Engineer, Gameware _____C .._.
Lockheed Martin Information Systems Group ____/ \___/
Phone: 407-826-6986 <____/\_---\_\ "Ferretman"
E-mail: wood...@gate.net
swoo...@oldcolo.com
wood...@escmail.orl.mmc.com
David Wu
Considering that the PSX has only 2 megs of RAM, and 1 meg of video Ram, I
doubt that they'd waste 600k for a full screen Z-buffer...