ID buffer and DR FP16

What's with the dithered shadow there?

Maybe devs should combine A2C dither with CBR and then the pixels will magically align.

 

We wil have a full Digitalfoundry comparison later on final version code and Xbox One X version will be better because the Xbox One X is more powerful than PS4 Pro and it is reflected by the 499 dollars price. it is as simple as 1+1 =2 or than the PS2 was less powerful than the og Xbox. I am a PS4 only guy and I read the spec of Xbox One X I knew it, now it is time for all fanboy to accept it... Is it the end of the world? No you can buy one if it is so important. I will play the game I wait on PS4 pro and be happy when I will have a new 4k TV and it was the same thing for people buying xbox One and not a PS4 three years ago when they were Halo, Gears of War and Forza fan. If you have only one console and your most anticipated game is The Last of US 2 you need a PS4 or a PS4 Pro...

If you want the best console version of multiplatform games it is time to buy an Xbox One X...

 

This is a technical discussion about the impact of the ID buffer and double-rate FP16 features.

 

We have no idea of the usefulness for Rise of the Tomb Raiser of the the PS4 Pro functionnality without technical documentation. I don't even understand why the image are into this topic. This would be a better place to compare the image into a digitalfoundry topic... Nixxes never talk about double rate FP16 for example...

 

@Shifty Geezer Apparently they are using their own reconstruction technique called "Sparse rendering" which would explain the lack of CB artefacts and the blurrier 4K image. Maybe they couldn't do the same CB used on Pro because of the lack of ID buffer on XBX ?

Previously some thought we would compare XBX game running natively versus CB games on Pro. Now we know the XBX can't properly run this game at native 4K and it's best to run it using their own sparse rendering solution, a reconstruction mode, all modes on X with missing effects (Depth of field and some reflections) compared to the Pro version.

And with all this power advantage against Pro (GPU but more importantly better bandwidth and more memory), we are currently comparing both very similar 4K images of the same game. Currently the main advantage of XBX is higher res textures on foreground objects (Difference of textures is actually aggravated by the lack of DOF on XBX game) and less artefacts (thanks to sparse rendering) but at the cost of a generally softer image (probably due to their own solution) so apparent lower resolution, at least in some scenes.

Based on this game alone, I wouldn't advise buying XBX just to play the best multiplatform games. Seems to me it's a tie for this game. DF seem to praise the sharper textures but much of the clarity difference comes form the missing effects on the XBX game. Also the 60fps has the same performance problems on XBX, except that the drops are mitigated by the fact that vsync is disabled (so plenty of screen tearing) versus triple buffering on Pro.

 

You say missing effects on XBO-X but it could just as well be missing or incorrect rendering features on PS4-P. Nixxes could clarify things obviously, but it's not likely that they would comment officially on it.

Regards,
SB

 

There is no DOF on Xbox One X

 

In native 4k the game is not running at locked 30 fps on Xbox One X, it is normal to cut some effect but it is strange to not implement the effect on the 4k sparse rendering mode...

Wait and see it is not the final version of the game.

 

Sure, and on the other hand the checkerboard rendering used on the PS4 doesn't cover all pixels, meaning that there is far less pixel coverage than on the XBO-X.

It isn't a case of XBO-X is the only one at a deficit WRT rendering or that PS4-P is the only one at a deficit WRT rendering the scene. As well, it's been noted that the XBO-X version is still a work in progress. Any missing features may or may not get implemented before the final version is released to consumers.

Regards,
SB

 

Like I said wait the final Digitalfoundry face off...

After I don't expect multiplateform title to be the best use for ID buffer or dual rate fp16.

This is many work for a tiny market out of first party I don't think all this features will be pushed very far...

 

It's a game using reconstruction which is what the ID buffer is for. Makes sense to have a look at the two games and compare them. Without technical documentation we are guessing a bit, but it's also fairly obvious some of what's going on, and in this case the ID buffer seems to be going unused (results are no different to other CB rendered games on PS4). Had PS4Pro shown far better results, we could have speculated on the positive influence of said ID buffer.

 

Unlikely. The results look akin to other CB games rendered on PS4, without ID buffer.

Wrong thread. Not to mention advising purchasing decisions on one single multiplat is a really dumb idea.

 

ID buffer can have other use than reconstruction...

 

I think sebbbi 's earlier post in this thread sort of sums it up fairly well when it come to whether developers will implement or not.

This is pretty similar imo to the tiled resources scenario.

https://forum.beyond3d.com/posts/1994345/

 

"Checkerboard" is a "sparse" technique if one thinks about it, so it's probably just a catch-all term.

They could try an even lower res DOF for extra cinematic focus framing of cut-scenes if they really need performance, but otherwise I don't see how it affects general gameplay, so it's a weird thing to happen.

 

I'm pretty sure that Pixel Shader Invocation Control has been in the PS4 since the beginning.

 

Is there any info about whether they used id buffer and/or fp16, or about their cb resolve, or post processing? I haven't seen anything to assert the usefulness of either ID buffer or fp16.

The frostbite presentation was showing a much better cb resolve than what we are seeing here. Some launch titles had exactly this artifacts, others didn't.

Maybe TR mid-gen upgrade was done by two completely different team, with different budgets, and different delivery schedule. Or different amount of "help".

You need a frame grab and zooming into a tiny square to show something different. Doesn't qualify as obvious even considering the sub-par resolve. If you show both images on two monitors at a normal distance, most people will notice the colors and gamma differences long before they start squinting at the leaves, maybe they'll notice a subjective sharpness/roughness/smoothness/softness. Might be different in motion though, specially depending on how the artifacts either average out making it look better, or otherwise pop randomly like aliasing.

 

Don't forget that it's a just a demo... so the unstable framerate at native 4k might change in the final version. Same thing for the DOF effect.

Fair enough. I cancel what i said.

 

Maybe I missed the boat here but this is my understanding of the comparison of Tomb Raider modes... HDR and Dolby Atmos is present in all modes and currently exclusive to Xbox but that may change.
4K mode: X1X is native 4k with higher detailed textures vs PS4Pro using CBR 4k but more consistent frame-rate compared to the X1 demo.
Enriched mode: X1X is using CBR 4k with higher detailed textures vs PS4Pro running at 1080p both at locked 30fps. This mode moves settings to higher level LOD, TressFX, ambient occlusion, reflections, and volumetric lighting and texture filtering.
High frame-rate mode: X1X and PS4Pro both running at 1080p with X1X hitting 60fps more consistently. Not sure of other visual differences.

 

I'm sure Xbox One X has ID buffer feature in the gpu states in DF Scorpio reveal. Who says it doesn't?