Gears of War 4 DX12 Performance Review @ [H]

[Shintai]

Yep, and its the developers that makes them obsolete, not the IHVs. Either by not targeting the uarch at all or by using features that doesn't exist in the older. If I was a non GCN 1.3 owner I would worry about tessellation. Kepler does fine against GCN 1.0 in Civ6 for example.

 

[FlipCody]

Running this game at 1440p with my GTX 970 and ultra settings as described on this article and I'm getting 48.4 fps with the in-game benchmark. I have been playing the game with these settings from the beginning (I'm in Act 5) and it has played great (usually at 45 - 55 fps). I have been really impressed how well this game is optimized and it looks great. I bought the version of game that includes PC and xbox one version (for $60) and have found the difference with ultra settings at 1440p on PC versus the xbox one painful and have played 95% of game on PC. I keep looking for reason to upgrade to GTX 1060 from GTX 970 and I have not found compelling reason yet...

 

[heatlesssun]

I keep looking for reason to upgrade to GTX 1060 from GTX 970 and I have not found compelling reason yet...

I don't have the benchmarks committed to memory but the 970 to 1060 isn't that big of a difference overall is it?

 

[razor1]

the 1060 should give you a bit more performance but if we are talking about the 3 gb 1060 you might see some dips depending on what settings you are using, otherwise they should be fairly close, edge to the 1060 of course.

 

[noko]

I hate to bring this up from a certain point of view - did Async compute on AMD have areas where it was greater then 7% improvement and other areas with less? The average was 4% with the highest of high settings and 7% with Ultimate if I remember right. That to me may mean that some sections or conditions you would see greater then that. Also where you may indeed need the added performance. If one section game gets a 10%+ boost etc. then this would be more important from a play ability standpoint.

Second point is, seems like the more shaders you have the more effective or chance that async compute will add performance. Anyone test out the Fiji line of cards yet to see if async compute adds more then 4% or 7%? I have not bought the game, not sure if I will but looks like a good DX 12 game for once.

 

[Shintai]

Running this game at 1440p with my GTX 970 and ultra settings as described on this article and I'm getting 48.4 fps with the in-game benchmark. I have been playing the game with these settings from the beginning (I'm in Act 5) and it has played great (usually at 45 - 55 fps). I have been really impressed how well this game is optimized and it looks great. I bought the version of game that includes PC and xbox one version (for $60) and have found the difference with ultra settings at 1440p on PC versus the xbox one painful and have played 95% of game on PC. I keep looking for reason to upgrade to GTX 1060 from GTX 970 and I have not found compelling reason yet...

A GTX 1060 is more or less a GTX980. You need 1070 or better to make a real difference.

 

[misterbobby]

There would be no logic at all in upgrading from a 970 to a 1060. A 1060 is less than 20% faster overall than the old 970. That would be a ridiculously useless upgrade after all this time. And the 3 GB version would be even less sensible than that.

 

[britjh22]

Now the real question, if DX11 was possible in GOW4 would it still be faster than DX12. Yes, it performs well in DX12, but it would be really interesting to see if it fails the same way most other "DX12" titles do. It's a shame that DX11 isn't an option.

 

[Armenius]

Now the real question, if DX11 was possible in GOW4 would it still be faster than DX12. Yes, it performs well in DX12, but it would be really interesting to see if it fails the same way most other "DX12" titles do. It's a shame that DX11 isn't an option.

Who cares? This is one of the first true DX12 games and it performs phenomenally well on most hardware configurations. Most other titles fail at DX12 because they were initially built for DX11.

 

[CSI_PC]

Who cares? This is one of the first true DX12 games and it performs phenomenally well on most hardware configurations. Most other titles fail at DX12 because they were initially built for DX11.

And even more surprising is how well it works with the UWP platform that has been a nightmare for developers in various ways.
As you say this is a well developed game.
Cheers

 

[heatlesssun]

And even more surprising is how well it works with the UWP platform that has been a nightmare for developers in various ways.
As you say this is a well developed game.
Cheers

The problems with UWA games were mostly due to immaturity of the platform and most of the issues have been resolved at this point from my understanding.

 

[CSI_PC]

The problems with UWA games were mostly due to immaturity of the platform and most of the issues have been resolved at this point from my understanding.

For sure but it goes beyond that, Gears of War is the only game to launch without problems, relative to the issues usually experienced due to UWP.
And Forza Horizon 3 launched reasonably close to it and had problems that can be tied back to UWP, with the experience the developers also had with early workings on DX12.
Cheers

 

[heatlesssun]

For sure but it goes beyond that, Gears of War is the only game to launch without problems, relative to the issues usually experienced due to UWP.
And Forza Horizon 3 launched reasonably close to it and had problems that can be tied back to UWP, with the experience the developers also had with early workings on DX12.
Cheers

The major issue I know about for Forza Horizon 3 is framerate consistency especially at higher settings. FH3 is a pretty demanding title especially at higher settings. Solid performance is a problem with AAA across the board especially at release so I don't know specially how much UWA played into FH3.

 

[CSI_PC]

The major issue I know about for Forza Horizon 3 is framerate consistency especially at higher settings. FH3 is a pretty demanding title especially at higher settings. Solid performance is a problem with AAA across the board especially at release so I don't know specially how much UWA played into FH3.

It plays a fair bit into it as it also impacted the engine, along with the headaches of Vsync/etc-window-full screen/triple buffering/etc influenced by its mode, however it is fair to say it is not ideal conversion but that is hardly an excuse considering they have been working with DX12 now for awhile.
A lot of its issues comes back to UWP in some shape or form whether directly and indirectly.
Gears of War is just as demanding graphically btw and runs smoothly (relative to other AAA UWP games) at launch at 60fps.

But yes moving forward more games should be better, point is that Gears of War that came out just after FH3 is a great game from both a DX12 and UWP context, and so far the only AAA one to date to do both well.
Cheers

 

[heatlesssun]

It plays a fair bit into it as it also impacted the engine, along with the headaches of Vsync/etc-window-full screen/triple buffering/etc influenced by its mode, however it is fair to say it is not ideal conversion but that is hardly an excuse considering they have been working with DX12 now for awhile.
A lot of its issues comes back to UWP in some shape or form whether directly and indirectly.
Gears of War is just as demanding graphically btw and runs smoothly (relative to other AAA UWP games) at launch at 60fps.

But yes moving forward more games should be better, point is that Gears of War that came out just after FH3 is a great game from both a DX12 and UWP context, and so far the only AAA one to date to do both well.
Cheers

Actually Forza Horizon 3 is much better rated on the Windows Store than Gears of War 4, 4.4 vs 3.2 out of 5 stars with around 6,000 more ratings left that for GoW 4. Windows Store ratings have been pretty brutal for other UWA games with far fewer ratings than for FH 3 so it looks to have done very well compared to all other UWA games including GoW 4.

 

[britjh22]

Who cares? This is one of the first true DX12 games and it performs phenomenally well on most hardware configurations. Most other titles fail at DX12 because they were initially built for DX11.

I care, we are all assuming that the reason it runs well is because it was designed for it, without DX11 performance it's just an assumption. For me DX12 is the only selling point for W10, and so far it hasn't added enough value to make the switch and deal with the windows store, so it would be cool to see if we are reaching that tipping point.

 

[CSI_PC]

Actually Forza Horizon 3 is much better rated on the Windows Store than Gears of War 4, 4.4 vs 3.2 out of 5 stars with around 6,000 more ratings left that for GoW 4. Windows Store ratings have been pretty brutal for other UWA games with far fewer ratings than for FH 3 so it looks to have done very well compared to all other UWA games including GoW 4.

Those rating do not necessarily have anything to do with how well a game is implemented technically in terms of DX12 and UWP.
I think I would go by in-depth analysis by a few of the better tech sites out there than general reviews
And I am talking about sites that understand VSYNC/flip model/how to accurately analyse frames/impact of window-borderless/etc.
Those good tech sites (only 3-4 of them out there) all came back with the same conclusion, Gears of War 4 is the best DX12 with UWP to date, and FH3 had some serious technical issues that needed to be analysed to be seen.

Cheers

 

[heatlesssun]

Those rating do not necessarily have anything to do with how well a game is implemented technically in terms of DX12 and UWP.
I think I would go by in-depth analysis by a few of the better tech sites out there than general reviews
And I am talking about sites that understand VSYNC/flip model/how to accurately analyse frames/impact of window-borderless/etc.
Those good tech sites (only 3-4 of them out there) all came back with the same conclusion, Gears of War 4 is the best DX12 with UWP to date, and FH3 had some serious technical issues that needed to be analysed to be seen.

Cheers

Forza Horizon 3 current ratings on the Windows Store

5* - 5185
4* - 1524
3* - 404
2* - 204
1* - 389

And pretty much all of the gaming sites gave the PC version A class ratings. That's just not the sign of a game that ever had serious technical issues. Not saying there weren't any but nothing serious like any number of games that release these days.

 

[CSI_PC]

Forza Horizon 3 current ratings on the Windows Store

5* - 5185
4* - 1524
3* - 404
2* - 204
1* - 389

And pretty much all of the gaming sites gave the PC version A class ratings. That's just not the sign of a game that ever had serious technical issues. Not saying there weren't any but nothing serious like any number of games that release these days.

Yep those excellent tech sites who know how to use actual frametime analysis tools (not talking general fps here) with environment parameters and also show how poor FH3 was in comparison relative to Gears of War 4 are all wrong

 

[heatlesssun]

Yep those excellent tech sites who know how to use actual frametime analysis tools (not talking general fps here) with environment parameters and also show how poor FH3 was in comparison relative to Gears of War 4 are all wrong

If FH3 was as problematic as you're claiming there's no way it would have been as well reviewed and rated because those problems would have shown up in game play without the need to do low level testing.

 

[Derangel]

I care, we are all assuming that the reason it runs well is because it was designed for it, without DX11 performance it's just an assumption. For me DX12 is the only selling point for W10, and so far it hasn't added enough value to make the switch and deal with the windows store, so it would be cool to see if we are reaching that tipping point.

I think it's pretty safe to assume that a game that is developed around DX12 is going to use it better than one developed around DX11 and later ported to DX12. The same thing happened when the industry started moving away from DX9.

 

[CSI_PC]

If FH3 was as problematic as you're claiming there's no way it would have been as well reviewed and rated because those problems would have shown up in game play without the need to do low level testing.

You misunderstand the context of what many of us are talking about in terms of its technical design including environment.
Also if you feel that strongly why respond to me and not even Kyle/Brent who did the review and this whole thread is part of?

I did not notice you responding to his review when he states:

Gears of War 4 is a extremely fun and well optimized game. It is a Windows 10 exclusive on the PC, but makes use of DX12 and as of right now is the best performing DX12 game we have experienced. This game will run near to the highest possible in-game settings across all GPUs.

His perspective aligns with those that did the technical tests I go on about who showed clearly the weaknesses of FH3 and also themselves mention Gears of War 4 is the best DX12 game they have experienced, which is also UWP.
Anyway the technical capturing tools used well cannot be mislead, the tools show what happens in these games from a technical perspective and the factors I mentioned earlier, and align with the various good reviewers.

Cheers

 

[heatlesssun]

You misunderstand the context of what many of us are talking about in terms of its technical design including environment.
Also if you feel that strongly why respond to me and not even Kyle/Brent who did the review and this whole thread is part of?

I did not notice you responding to his review when he states:

His perspective aligns with those that did the technical tests I go on about who showed clearly the weaknesses of FH3 and also themselves mention Gears of War 4 is the best DX12 game they have experienced, which is also UWP.
Anyway the technical capturing tools used well cannot be mislead, the tools show what happens in these games from a technical perspective and the factors I mentioned earlier, and align with the various good reviewers.

Cheers

All I am saying is that if Forza Horizon 3 had serious technical issues at launch that straightforward gameplay reviews would have picked up on that. I don't think [H] did this type of review for FH3. The only common issue that I've been aware of with FH3 was stuttering that was generally alleviated by adjusting settings. I'm not arguing that GoW 4 might be better optimized and that was never my point. I have both games, I think FH3 is graphically more appealing and one of the best racing games to come to the PC and it performs very well and did at launch though it's received some patch that probably has improved some performance issues.

That's not to take anything away from GoW4 or [H]'s technical analysis. But FH3 is great as was solid at launch. It's a great DX 12 UWA title.

 

[CSI_PC]

All I am saying is that if Forza Horizon 3 had serious technical issues at launch that straightforward gameplay reviews would have picked up on that. I don't think [H] did this type of review for FH3. The only common issue that I've been aware of with FH3 was stuttering that was generally alleviated by adjusting settings. I'm not arguing that GoW 4 might be better optimized and that was never my point. I have both games, I think FH3 is graphically more appealing and one of the best racing games to come to the PC and it performs very well and did at launch though it's received some patch that probably has improved some performance issues.

That's not to take anything away from GoW4 or [H]'s technical analysis. But FH3 is great as was solid at launch. It's a great DX 12 UWA title.

We are going full circle and I guess either disagree or more likely have different PoV.
Adjusting setting is not resolving the issue but just a way of hiding or more likely masking them perceptually, some of which come back to UWP and also their experience with it.
Gears of War 4 has none of those issues nor other DX12 quirks a few recent games exhibit, it is the best game to date not just for DX12 but also UWP, that was my point and FH3 was a good example of one with multiple issues at launch.
To be fair maybe the technical reviews analysing frames should look at this again since the Windows 10 Anniversary Update.

Cheers

 

[heatlesssun]

We are going full circle and I guess either disagree or more likely have different PoV.
Adjusting setting is not resolving the issue but just a way of hiding or more likely masking them perceptually, some of which come back to UWP and also their experience with it.
Gears of War 4 has none of those issues nor other DX12 quirks a few recent games exhibit, it is the best game to date not just for DX12 but also UWP, that was my point and FH3 was a good example of one with multiple issues at launch.
To be fair maybe the technical reviews analysing frames should look at this again since the Windows 10 Anniversary Update.

Cheers

You keep referring to FH3 having multiple issues at launch and that's not what I or many people experienced. Some stuttering at higher settings for some that could be mitigated with different settings. That's not at all uncommon with PC titles, including plain old Win32 games. In any case FH3 looks to have been a much bigger deal than GoW4 at far as UWA games go, a lot more people seem to have bought it and like it.

I'm not trying to take anything away from GoW4, but FH3 is seriously great and there are plenty on this forum that agree.

 

[CSI_PC]

You keep referring to FH3 having multiple issues at launch and that's not what I or many people experienced. Some stuttering at higher settings for some that could be mitigated with different settings. That's not at all uncommon with PC titles, including plain old Win32 games. In any case FH3 looks to have been a much bigger deal than GoW4 at far as UWA games go, a lot more people seem to have bought it and like it.

I'm not trying to take anything away from GoW4, but FH3 is seriously great and there are plenty on this forum that agree.

Again as I keep saying I am approaching this from a technical PoV and system environment.
PCGamesHardware had this to say who analysed in depth the frame behaviour:

With Microsoft's Windows and Crossplay Initiative, PC gamers will also benefit from the high-gloss funcizer Forza Horizon 3. In principle, this is a reason for joy, because gameplay and scope correspond exactly to what friends of the well-groomed fun course desire. Regrettably, Forza Horizon 3, however, is suffering from the same problems as previous Windows store games: the Universal Windows Platform, or UWP for short, has a lot to do with image output. We recommend the following video, which shows a part of the problem in word and moving picture impressively

Not linking the video as it is in German.
But their article if you translate it: http://www.pcgameshardware.de/Forza.../Specials/Benchmarks-Test-DirextX-12-1208835/
Also Computerbase.de

A good game, but a technical accident on the PC
Forza Horizon 3 content a very good game become, makes the PC currently but just not fun. Microsoft's Universal Windows Platform slows down all tested graphics cards massively, only the GeForce GTX 1080 creates the test sequence without changing between 30 and 60 FPS - and even in Full HD only with reduced details.


The verdict is harsh: Microsoft would have to Quantum Breakeven Forza Horizon 3, not to publish in this state for the PC. Currently interested in the desire for high and above all constant FPS in racing games from the purchase for the PC is therefore discouraged. Short-term improvement is not in sight, because an update of the game is not enough. Instead, a Windows update of UWP is necessary. And this is apparently planned for the year 2017.

https://www.computerbase.de/2016-09/forza-horizon-3-benchmark/3/

Both of these analysed this from a frame and system environment perspective.
Cheers

 

[heatlesssun]

Again as I keep saying I am approaching this from a technical PoV and system environment.
PCGamesHardware had this to say who analysed in depth the frame behaviour:
Not linking the video as it is in German.
But their article if you translate it: http://www.pcgameshardware.de/Forza.../Specials/Benchmarks-Test-DirextX-12-1208835/
Also Computerbase.de

https://www.computerbase.de/2016-09/forza-horizon-3-benchmark/3/

Both of these analysed this from a frame and system environment perspective.
Cheers

And this just doesn't make any sense having played FH3 for quite a few hours since early release in October. I have settings on Ultra now at 5760x1080 and while it doesn't maintain 60 FPS it stays about 45. I've limited refresh to 60 FPS and it stays very smooth at these settings. It wasn't all that bad for me at launch either but I don't notice any stuttering now either.

FH3 is beautiful, it's the best looking game for the PC I think this year, better than Doom or even GoW4 and it performs well on a variety of hardware but it's very taxing on higher settings.

I get that you're arguing benchmarks, I'm talking about real world game play and that's far more important. That's sort of the whole idea behind [H] and gaming isn't it? Real world performance and enjoyment? Again, there's no way a game is this well received AT LAUNCH and is anywhere near as problematic as you're suggesting. We've seen PLENTY of PC games at launch that get decimated in reviews because of technical issues. FH3 was not one of them fortunately.

Right now, FH3 would be my Game of the Year on the PC. It's a fantastic arcade style racer with that's packed with content.

 

[JustReason]

And this just doesn't make any sense having played FH3 for quite a few hours since early release in October. I have settings on Ultra now at 5760x1080 and while it doesn't maintain 60 FPS it stays about 45. I've limited refresh to 60 FPS and it stays very smooth at these settings. It wasn't all that bad for me at launch either but I don't notice any stuttering now either.

FH3 is beautiful, it's the best looking game for the PC I think this year, better than Doom or even GoW4 and it performs well on a variety of hardware but it's very taxing on higher settings.

I get that you're arguing benchmarks, I'm talking about real world game play and that's far more important. That's sort of the whole idea behind [H] and gaming isn't it? Real world performance and enjoyment? Again, there's no way a game is this well received AT LAUNCH and is anywhere near as problematic as you're suggesting. We've seen PLENTY of PC games at launch that get decimated in reviews because of technical issues. FH3 was not one of them fortunately.

Right now, FH3 would be my Game of the Year on the PC. It's a fantastic arcade style racer with that's packed with content.

It isn't worth the time, trust me. I get what you are saying and agree. He makes it sound as if this is the Batman release all over again. I liked the way FH3 ran and I didn't have issues running it either. I guess that is the good we get with [H] as they show playable settings which more closely mirror what an individual playing the game would see. Most of these other site throw a bunch of settings at it and scream fail if it doesn't play well maxed. The question they should be answering is if it plays well at certain setting and is the decrease in settings affecting the gameplay experience in any varying degree.

Although I do believe DOOM should be GotY simply because what they were able to do performance wise for all hardware. Not my kind of game but I bought it in support of their effort and I must say it is really nice to see a game done right.

 

[razor1]

It isn't worth the time, trust me. I get what you are saying and agree. He makes it sound as if this is the Batman release all over again. I liked the way FH3 ran and I didn't have issues running it either. I guess that is the good we get with [H] as they show playable settings which more closely mirror what an individual playing the game would see. Most of these other site throw a bunch of settings at it and scream fail if it doesn't play well maxed. The question they should be answering is if it plays well at certain setting and is the decrease in settings affecting the gameplay experience in any varying degree.

Although I do believe DOOM should be GotY simply because what they were able to do performance wise for all hardware. Not my kind of game but I bought it in support of their effort and I must say it is really nice to see a game done right.

Simply because it was able to perform on all hardware, or did it remind you of the first doom? I would pick the later, the game experience is a lot more important to me then what the hell its running on or how fast the FPS's are.

Great to have a person sit here and dictate why he thinks a game should get GOTY because it shows up on a certain hardware vendors radar. WTF are you talking about again?

 

[CSI_PC]

And this just doesn't make any sense having played FH3 for quite a few hours since early release in October. I have settings on Ultra now at 5760x1080 and while it doesn't maintain 60 FPS it stays about 45. I've limited refresh to 60 FPS and it stays very smooth at these settings. It wasn't all that bad for me at launch either but I don't notice any stuttering now either.

FH3 is beautiful, it's the best looking game for the PC I think this year, better than Doom or even GoW4 and it performs well on a variety of hardware but it's very taxing on higher settings.

I get that you're arguing benchmarks, I'm talking about real world game play and that's far more important. That's sort of the whole idea behind [H] and gaming isn't it? Real world performance and enjoyment? Again, there's no way a game is this well received AT LAUNCH and is anywhere near as problematic as you're suggesting. We've seen PLENTY of PC games at launch that get decimated in reviews because of technical issues. FH3 was not one of them fortunately.

Right now, FH3 would be my Game of the Year on the PC. It's a fantastic arcade style racer with that's packed with content.

I am not arguing benchmarks, I am basing this upon tools that analyse frames and behaviour, not quite the same thing.
As I mentioned if you change settings you are not resolving the integral issues with the game-UWP but just masking them, hence why those tools are useful.
Remember that it was FRAPs/FCAT as tools that helped to identify frame behaviour issues in the past, and became a defacto tool to use and not just rely upon perception or that some settings work but not others without identifying why.

Cheers

 

[heatlesssun]

I am not arguing benchmarks, I am basing this upon tools that analyse frames and behaviour, not quite the same thing.
As I mentioned if you change settings you are not resolving the integral issues with the game-UWP but just masking them, hence why those tools are useful.
Remember that it was FRAPs/FCAT as tools that helped to identify frame behaviour issues in the past, and became a defacto tool to use and not just rely upon perception or that some settings work but not others without identifying why.

Cheers

Having to adjust settings to optimize performance isn't anything unique to UWA games. And the implication in the article you linked earlier said that the issues were are the core of UWA and would only be fixed via an update to Windows 10 in 2017. And clearly that's wrong otherwise wouldn't GoW4 been affected as well?

Again, it all comes down to gameplay, FH3 is awesome in that regard and if you have the hardware to run it runs very well at higher end settings. You're not going to crush 4k at 60 FPS without a Titan XP at max settings though.

 

[heatlesssun]

It isn't worth the time, trust me. I get what you are saying and agree. He makes it sound as if this is the Batman release all over again. I liked the way FH3 ran and I didn't have issues running it either. I guess that is the good we get with [H] as they show playable settings which more closely mirror what an individual playing the game would see. Most of these other site throw a bunch of settings at it and scream fail if it doesn't play well maxed. The question they should be answering is if it plays well at certain setting and is the decrease in settings affecting the gameplay experience in any varying degree.

Although I do believe DOOM should be GotY simply because what they were able to do performance wise for all hardware. Not my kind of game but I bought it in support of their effort and I must say it is really nice to see a game done right.

Doom is getting a lot of praise for it's solid performance across the board and that's well deserved. But it's mostly a graphical upgrade of an old game, nothing much new. And that's fine and I appreciated Doom for that. FH3 is a much different game obviously, I think the open world driving is more demanding that a corridor shooter and FH3 I think is visually stunning, beyond Doom. In any case, both are great in their own right. FH3 isn't like that Batman release as you point out, it just needs more hardware than Doom to crank up and perform well.

 

[CSI_PC]

Having to adjust settings to optimize performance isn't anything unique to UWA games. And the implication in the article you linked earlier said that the issues were are the core of UWA and would only be fixed via an update to Windows 10 in 2017. And clearly that's wrong otherwise wouldn't GoW4 been affected as well?

Again, it all comes down to gameplay, FH3 is awesome in that regard and if you have the hardware to run it runs very well at higher end settings. You're not going to crush 4k at 60 FPS without a Titan XP at max settings though.

No but understanding why is.
And that is how it is known what the issue is with FH3.
This was one of the primary reason FCAT/FRAPs exist, to work out technically how the game is behaving, so while you and others are happy to adjust settings to avoid or mask so you do not perceive the actual issue it does not technically reflect how well the game is designed.
I mean that is the point we have been discussing for last few days where you raised FH3 as being better or at least equal but I am on about from technical analysis/tools, technically from a development and engine perspective Gears of War 4 is better than FH3 in context of DX12 and UWP development, as explained in the 2 articles I linked - used those because it is a source separate to what I am saying.
And you can check the history of why FRAPs/FCAT came about, to understand unusual frame behaviour that is not fully quantifiable.

Just to also add, even though most will not perceive why, Gears of War 4 also has a particular quirk relating to UWP and dynamic resolutions, but overall it is still the best designed DX12+UWP game out there.

Cheers

 

[CSI_PC]

..... In any case, both are great in their own right. FH3 isn't like that Batman release as you point out, it just needs more hardware than Doom to crank up and perform well.

Just want to say JustReason is being deliberately petty towards me due to another thread where he is not happy with the discussion the two of us are having and even started being insulting (Roy Taylor thread).
Kinda annoyed he is making it out like I am saying FH3 is like Batman, and that is effing bullshit.
What I have highlighted is nothing like that, and also used two technical reviewers known for well articulated measurement/analysis of frame behaviour beyond normal reviews.
This has been part of the gaming industry for years now, as I explained with FCAT/FRAPs.

Cheers

 

[heatlesssun]

No but understanding why is.
And that is how it is known what the issue is with FH3.
This was one of the primary reason FCAT/FRAPs exist, to work out technically how the game is behaving, so while you and others are happy to adjust settings to avoid or mask so you do not perceive the actual issue it does not technically reflect how well the game is designed.
I mean that is the point we have been discussing for last few days where you raised FH3 as being better or at least equal but I am on about from technical analysis/tools, technically from a development and engine perspective Gears of War 4 is better than FH3 in context of DX12 and UWP development, as explained in the 2 articles I linked - used those because it is a source separate to what I am saying.
And you can check the history of why FRAPs/FCAT came about, to understand unusual frame behaviour that is not fully quantifiable.

Just to also add, even though most will not perceive why, Gears of War 4 also has a particular quirk relating to UWP and dynamic resolutions, but overall it is still the best designed DX12+UWP game out there.

Cheers

I did read over those articles. The basic issue with FH3 like any high speed racer is consistent framerates. Cranking FH3 up and trying to hit 60 above 1080P does require good hardware. I think the early issues with FH3 had to do with frame pacing to deal with significant drops below 60 and the issues with UWA and vsynch. I just think the issue was very minor at release and looks to have been fixed now without the need for a new version of Windows.

 

[CSI_PC]

I did read over those articles. The basic issue with FH3 like any high speed racer is consistent framerates. Cranking FH3 up and trying to hit 60 above 1080P does require good hardware. I think the early issues with FH3 had to do with frame pacing to deal with significant drops below 60 and the issues with UWA and vsynch. I just think the issue was very minor at release and looks to have been fixed now without the need for a new version of Windows.

These are not minor issues, sorry but this is a technical and fundamental design consideration and performance behaviour.
This is not just me saying that but also those articles!
It would be like saying Scott Wasson (now employed by AMD) was wasting his time developing tools exactly for this type of thing because we just turn off some options or not use MSAA, cap are resolutions and fps,etc.

It could not be fixed without improvements to UWP/Windows 10.
Hence why I did say to be fair it should be revisited since the Windows 10 Anniversary update as that may had helped.
Why do you think most reviewers are saying that Gears of War 4 is the best DX12/UWP game to date in terms of performance and associated behaviour, rather than say it is just behind FH3 in its current state.

 

[heatlesssun]

These are not minor issues, sorry but this is a technical and fundamental design consideration and performance behaviour.
This is not just me saying that but also those articles!
It would be like saying Scott Wasson (now employed by AMD) was wasting his time developing tools exactly for this type of thing because we just turn off some options or not use MSAA, cap are resolutions and fps,etc.

It could not be fixed without improvements to UWP/Windows 10.
Hence why I did say to be fair it should be revisited since the Windows 10 Anniversary update as that may had helped.
Why do you think most reviewers are saying that Gears of War 4 is the best DX12/UWP game to date in terms of performance and associated behaviour, rather than say it is just behind FH3 in its current state.

Both Gears of War 4 and Forza Horizon 3 require the Anniversary update. All I am saying is that both run fine on my rig at around the highest settings at 5760x1080. Just understand from my perspective since I have and play both and they run well that whatever issues you're talking about they are at most not real problems. I've seen real problems with PC games where UWA had nothing to do with it.

 

[JustReason]

Both Gears of War 4 and Forza Horizon 3 require the Anniversary update. All I am saying is that both run fine on my rig at around the highest settings at 5760x1080. Just understand from my perspective since I have and play both and they run well that whatever issues you're talking about they are at most not real problems. I've seen real problems with PC games where UWA had nothing to do with it.

That is a big problem with forums and the internet, personal experience against review sites. For years I was always being told that having an all AMD rig meant playing a terrible Skyrim. Yet I had 100 mods, mostly texture upgrades, 2xSSAAeq on 7700@1150 x2 and a [email protected] running 75fps (capped) with dips into the 50s in the problematic cities. Of course I had to do some major tweaking to the .cfg files, fix some of the issues that is, but point is I was having an outstanding gameplay experience where the rest of the web said I shouldn't. Some people just believe <almost fanatically, everything a review site tells them. The part they never consider is that it is nearly next to impossible to get the EXACT same results as they do for any number of reasons be it background processes running to just personal preference on settings. I never run DoF or Radial Blur, hate em cant stand em. So when a site benches a gam the y tend to have those on so it is likely to a great degree that I will have higher framerates.

Honestly this is why I tend to read the consumer reviews, some will be very biased but there are a few that are honest, and take what they say over a review site.

 

[CSI_PC]

Both Gears of War 4 and Forza Horizon 3 require the Anniversary update. All I am saying is that both run fine on my rig at around the highest settings at 5760x1080. Just understand from my perspective since I have and play both and they run well that whatever issues you're talking about they are at most not real problems. I've seen real problems with PC games where UWA had nothing to do with it.

Bah yeah sorry I meant the update to WDDM 2.1, think this was sometime in November.
Anyway I think we can both agree that those articles I linked are going beyond reviewing and are using tools to analyse the frame behaviour of the engine and its interraction with DX12-UWP and various settings including post processing.
Even DigitalFoundry had WTF moments (they mention it in their online video) with FH3 once they started drilling down in comparison between Titan XP-1060-480-390-970/resolutions/settings/framerates/etc but do not have the necessary tools and possibly experience to prove why (unlike the 2 articles I linked).

BTW I am not saying FH3 is a bad game, I am saying that Gears of War 4 is the best game to date with regards technically how implemented with DX12+UWP, I think that keeps getting lost in this debate.

Cheers

 

[CSI_PC]

Forza Horizon 3 current ratings on the Windows Store

5* - 5185
4* - 1524
3* - 404
2* - 204
1* - 389

And pretty much all of the gaming sites gave the PC version A class ratings. That's just not the sign of a game that ever had serious technical issues. Not saying there weren't any but nothing serious like any number of games that release these days.

Coming back to customer reviews can you link which page those stats were from as trying to understand whether that is just PC or combined PC+XBox/region and which digital version.

Thanks