Image Quality and Framebuffer Speculations for WIP/alpha/beta/E3 games *Read the first post*

Looks legit to me!
Are you sure about that?
q2qwr2nuu5m.jpg
 
Your argument is 'it looks better than the previous version of the game so it must be fake'?
The environment (or most of it) is probably real. But the character isn't.
I don't even think they rendered this in-engine. From the looks of it they added the character and his flashlight beam in post.
 
The environment (or most of it) is probably real. But the character isn't.
I don't even think they rendered this in-engine. From the looks of it they added the character and his flashlight beam in post.
This isn't the thread for that. In terms of IQ and pixel counting, the presence of aliasing on the neon sign and the lower resolution in the light ray suggests the screenshot is taken from in-engine, and hence is a legitimate representation of the game's rendering resolution and AA amount, although JPEGing makes that hard to discern acurately. But looks native 720p to me with some form of AA.
 
Not sure if this will help with an analysis of Tekken Tag Tournament 2, particularly the Prologue version just released on PS3:

http://yaplog.jp/fightorflight/archive/13

The same person provided a link with a Namco presentation about the arcade version, and looking at the slides, the resolution switches often dropping as low as 720x720:

http://gigazine.net/news/20110907_tekken_cedec2011/

hmmmm, they already have soul calibur's engine established, why go through the pains of making a new engine that clips away at pixels and reduces everything?

The hair alpha maps and impressive facial animations of soul calibur 4 were some of the best work I've seen in any fighting game.

The engine used for this tekken game is making a justifiable sacrifice for 60 fps...but at what cost? The usual HDR lighting, crisp textures and normal maps, hi-rez meshes, vast landscapings, and insane particles effects are the reasons of making a morphological frame buffer setup.....not flying confetti whenever someone takes a hit.

In this generation of fighting games there has been a couple of good examples of rendering, like "fight night round 4" which brought about some of the best facial deformation physics yet used in any fighting game without the digression of pixels.

That's my two cents.:smile2:
 
Last edited by a moderator:
Don't think its using the Soul calibur engine but rather the Tekken 6 engine, which makes sense.

Do remember that Tekken 6 had am amazing per object motion blur and still remains the only fighting game to use object based motion blur, and when not using it the resolution went up to 1366*768 (for xbox) and applied with 2*MSAA for PS3 at 576p. And it also had a level with dynamic Time of Day.
 
It also has drop dead gorgeous character models and the facial animations go far above Soul Calibur Levels (heck, they already did in Tekken 5) Game looks phenomenal in motion, and in the end that's what counts for a fighting game.
 
As the slides show, it starts off at full 720p but with *sparks* and two characters, it goes down to 1024x720 and gradually to 720x720 when you have 4 character models on screen with full effects. To keep the frame rate consistent and ultimately the controls responsive of course, and Harada had mentioned before that motion blur did take them a while to implement with T6.

What I'm curious about is the res of the console Prologue version (which incidentally was done before the arcade version was out). The res seems to be consistent and closer to the 1024x576 of T6.
 
Don't think its using the Soul calibur engine but rather the Tekken 6 engine, which makes sense.

Do remember that Tekken 6 had am amazing per object motion blur and still remains the only fighting game to use object based motion blur, and when not using it the resolution went up to 1366*768 (for xbox) and applied with 2*MSAA for PS3 at 576p. And it also had a level with dynamic Time of Day.

It also has drop dead gorgeous character models and the facial animations go far above Soul Calibur Levels (heck, they already did in Tekken 5) Game looks phenomenal in motion, and in the end that's what counts for a fighting game.

respect.

Don't want to go off topic but i will say tekken 6's characters do have a couple of bells and whistles of it's own to commend as well as the lighting.

I was viewing some of these shots that were exhibited and i was dissecting a couple of them. my regards to people stating weather or not tekken has the best of anything particular.

According to the statements in this demonstration the pixels are 1024x720 for 2 people on screen dropping about 100,000 pixels per character that enters.

http://gigazine.jp/img/2011/09/07/tekken_cedec2011/P1960699_m.jpg
http://gigazine.jp/img/2011/09/07/tekken_cedec2011/P1960708_m.jpg

and the effects too can also result for the 100,000 pixel deficit even when only two characters are being rendered.
http://gigazine.jp/img/2011/09/07/tekken_cedec2011/P1960727_m.jpg
maybe even more i think.
 
Laa-Yosh said:
Anyway, thanks for the insight! It's always nice to get more info about these issues... would not have thought that even 1-2% can matter.
One of the biggest lies they teach in software-engineering education is that optimization is a process of removing peaks out of profiling graphs.
More often then not - "flat" graphs are when your work really starts.
 
Back
Top