The Technology of GTA IV/RDR *Rage Engine*

I still like what I see on my X360/Samsung LCD. Kinda like film grain, makes it look less sterile.
 
hm...
http://diaryofagraphicsprogrammer.blogspot.com/2008/03/light-pre-pass-renderer.html
Wolfgang Engel said:
In June last year I had an idea for a new rendering design. I call it light pre-pass renderer.
The idea is to fill up a Z buffer first and also store normals in a render target. This is like a G-Buffer with normals and Z values ... so compared to a deferred renderer there is no diffuse color, specular color, material index or position data stored in this stage.
Next the light buffer is filled up with light properties. So the idea is to differ between light and material properties. If you look at a simplified light equation for one point light it looks like this:

Color = Ambient + Shadow * Att * (N.L * DiffColor * DiffIntensity * LightColor + R.V^n * SpecColor * SpecIntensity * LightColor)

...

So what you can do is instead of rendering a whole lighting equation for each light into a render target, you render into a 8:8:8:8 render target only the light properties. You have four channels so you can render:

LightColor.r * N.L * Att
LightColor.g * N.L * Att
LightColor.b * N.L * Att
R.V^n * N.L * Att

That means in this setup there is no dedicated specular color ... which is on purpose (you can extend it easily).

....

After all lights are alpha-blended into the light buffer, you switch to forward rendering and reconstruct the lighting equation.

...

This is a direct competitor to the light indexed renderer idea described by Damian Trebilco at Paper .
I have a small example program that compares this approach to a deferred renderer but I have not compared it to Damian's approach. I believe his approach might be more flexible regarding a material system than mine but the Light Pre-Pass renderer does not need to do the indexing. It should even run on a seven year old ATI RADEON 8500 because you only have to do a Z pre-pass and store the normals upfront.

....

<more shtuff>

And asking about scaling (June 28 2007):

http://www.gamedev.net/community/forums/topic.asp?topic_id=453577


It's all so clear ( ;) ). or not. :p

edit: links for my reference when I get home
http://www.gamedev.net/community/forums/topic.asp?topic_id=492723&whichpage=1&#3215336
 
This is very similar to light indexed deferred rendering and to what Drake's fortune does as well (which is even more clever imhio)
 
With the supposed lack of tearing on PS3, I'm wondering if they just left VSync enabled. If that's indeed the case this could serve as nice insight into what people actually prefer - tearing or a touch more juddery fps.

I'll re-enter this thread since I spent a gang of time this morning playing with the PS3 version, since my PC at work died :( I've been playing the 360 one at home semi regularly, and the frame rate on the PS3 definitely seems lower to me, it seems almost stuttery and more prone to pauses. I "sanity checked" it by asking some other people to come in as asked what they thought, and their comments were similar. I haven't noticed tearing on the 360 version, but I'll assume it is there since people claim it. From what I'm seeing on the frame rates, I would also have to assume that the PS3 version is vsync locked and the 360 version isn't. That would explain the tearing people apparently only see on 360, and it would explain the stuttery frame rate I see on the PS3 version.


MazingerDUDE said:
did some more 200% shots

These zoomed in shots are useful if you want to figure out perhaps how something works, or what technique they use. It does not reflect though how you see it on screen though and hence, at least I think, is not a good way to pick a better quality version. I'll sound like a broken record and again suggest people to see both versions back to back first hand, it's the only way.

Mintmaster said:
Doesn't object level sorting give you nearly the same benefit? Or do render state changes often become a big problem that way?

You asked and answered your own question, thats tops in efficiency :) Render states can be a performance issue when dealing with object draw order sorting (for Z benefit), it tends to be game by game basis though. On Baseball, the draw order was incredibly predictable, and the objects in question had similar render states, so we didn't have a z prepass.
 
These zoomed in shots are useful if you want to figure out perhaps how something works, or what technique they use. It does not reflect though how you see it on screen though and hence, at least I think, is not a good way to pick a better quality version. I'll sound like a broken record and again suggest people to see both versions back to back first hand, it's the only way.

Actually I did :smile: I have spent about 10 hours each on both, and I do feel the PS3 version does look better, and I think the reason why is in those shots. I initially thought 360 version looked better, but changed my mind after spending some time on both. Maybe you'd too if you spend just little more time with the PS3 version ;)

BTW, the actual pixel size of any HDTV should easily be bigger than 2 x zoom shots I provided here (I'm playing the game on 46" LCD TV)
 
BTW, the actual pixel size of any HDTV should easily be bigger than 2 x zoom shots I provided here (I'm playing the game on 46" LCD TV)

It depends on both viewing distance (I'm ~10 feet from my 50 inch plasma) and what your tv does to the signal. Most tv's tend to do their own 'processing' to an incoming image, such as rescaling it to fit the native rez of the tv. Or, some tv's have other effects on the final image depending on the tv tech (dlp,lcd,plasma,crt). So what you see in screen shots is not what makes it to your tv.
 
joker454 said:
That would explain the tearing people apparently only see on 360, and it would explain the stuttery frame rate I see on the PS3 version.
Yea I take the recent info as confirmation of the fact - people captured 360 tearing on video/screens, and grandmaster pretty much confirms PS3 is VSynced.
The question this brings up is why - just like the 360 grain, I would think it's intentional. And you and others confirm what I've already seen with my own work before - majority seems to favour VSync-off scenario even in blind tests.
 
My guess is this: R* went for v-lock off on 360 because only the most anal can actually see it. They turned v-lock on for PS3 because the tearing effect would be far more pronounced than the 360 version if it wasn't. The GTA games are inherently fluctuating on frame rate all the time (according to my tests, both versions can run *above* 30fps on average) so it would probably be less noticeable.

They've already compromised on resolution, so why not FPS too?

BTW I'm with joker454 in that the lower frame rate is easily apparent on PS3.
 
Last edited by a moderator:
I've been thinking about joker454's question as to why the preference for the PS3 version among so many and I think I might have an answer. The full screen effect being used could be said to be simulating a kind of atmospheric perspective not usually addressed by real-time 3D graphics. The 360 version may be sharper, but it's to a degree that highlights the artificiality of what you're seeing. The softening effect at work on PS3 actually makes the game seem less obviously rendered.
 
Eurogamer's GTA face-off is now up, complete with the usual galleries (including 1080p), but this time with comparison videos that actually show the difference - one pixel in the vid is one pixel on your display.

The feature has embedded vids at 632x400, but 728x544 versions are also available. All vids were compiled in a totally lossless workflow up until the final h264 encoding (1100kbps at HQ-Insane level using x264 for the larger ones, 900kbps for the smaller ones).

The framerate measurements you've been waiting for are also included :D
 
Last edited by a moderator:
heh, I kinda liked the 360's explosions more because it was brighter or more bright yellow than orange.

edit: in the video, that's what it seemed to me. The screenshots don't show any difference.

Great article.
 
Eurogamer's GTA face-off is now up, complete with the usual galleries (including 1080p), but this time with comparison videos that actually show the difference - one pixel in the vid is one pixel on your display.

The feature has embedded vids at 632x400, but 728x544 versions are also available. All vids were compiled in a totally lossless workflow up until the final h264 encoding (1100kbps at HQ-Insane level using x264 for the larger ones, 900kbps for the smaller ones).

The framerate measurements you've been waiting for are also included :D

So, 20% less pixels and upto 18% less frames pr second for the PS3 version. And a 360 version that actually could look even better if the dithering wasn´t applied on purpose.

That my friends is pretty bad on paper. Is this just the PS3 being "sucky" or the developers not being able to cope with the PS3?
 
Is my first time in my life that I see this , very interesting

Test One: Game Intro
360: 31.990fps
PS3: 26.460fps
See it on EGTV.
Test Two: Clean Getaway
360: 28.624fps
PS3: 23.452fps
See it on EGTV.
Test Three: Final Destination
360: 35.262fps
PS3: 29.041fps
See it on EGTV.
Test Four: Station Face-Off
360: 26.076fps
PS3: 26.081fps
See it on EGTV.
Test Five: Rigged to Blow
360: 26.712fps
PS3: 23.781fps
See it on EGTV.
Test Six: Ivan the Not So Terrible
360: 33.798fps
PS3: 28.313fps
See it on EGTV.
 
So, 20% less pixels and upto 18% less frames pr second for the PS3 version. And a 360 version that actually could look even better if the dithering wasn´t applied on purpose.

That my friends is pretty bad on paper. Is this just the PS3 being "sucky" or the developers not being able to cope with the PS3?

The amount of money they'd supposedly spent, I'd have a hard time believing they didn't have adequate resources to optimize the PS3. But really, I think this speaks more to the 360 being the lead platform. If this had been a PS3 exclusive, I think the PS3 results would have been much better, not that they're bad as is.
 
The amount of money they'd supposedly spent, I'd have a hard time believing they didn't have adequate resources to optimize the PS3. But really, I think this speaks more to the 360 being the lead platform. If this had been a PS3 exclusive, I think the PS3 results would have been much better, not that they're bad as is.

I'm quite sure that if this were an Xbox 360 exclusive, the technological feats would be even greater as well. That goes both ways.
 
Back
Top