Digital Foundry Article Technical Discussion Archive [2016 - 2017]

[Jay]

My take from article is they have delivered what they said they would in regards to native 4k.
I don't know how anyone can claim otherwise if a game isn't native 4k but has much higher settings applied.

I don't see the initial base resolution being the driver if something is native or not in the future on 1X. I think it will be all about if native looks a lot better, and if the engine already has an alternative rendering method in place.

I can see just as much chance 1X uses CBR even if the game is 1080p on the XO, if the engine is capable and if it looks better for it I.e. ramped up settings and effects. Smaller studios will go for the easy wins dependent on what the engine provides they are using.

 

[BRiT]

Remember, before the griping begins, the data reported is months old and the games are not optimized for X.B.O.X. yet.

 

[AlNom]

There's a PR messaging thread, btw.

 

[Silent_Buddha]

Microsoft's Xbox One X Benchmarks Revealed: 4K vs 900p/1080p + Back-Compat!:

http://www.eurogamer.net/articles/digitalfoundry-2017-microsofts-xbox-one-x-benchmarks-revealed

Important to keep in mind for people that might just quickly skim though it.

To get these PIX metrics, all of the titles underwent the most basic of ports onto Project Scorpio development hardware - and this presents challenges in terms of getting data representative of final games. Firstly, an early iteration of the operating system was in place, while Scorpio-specific hardware features are entirely unused. Furthermore, ESRAM from the base Xbox One is mapped directly to GDDR5 in the new console with no tweaking - meaning that there will be plenty of read/writes between memory areas that are entirely wasteful and would not happen in a shipping title. On the flip-side though, these are entirely GPU-based benchmarks - meaning that memory contention issues between CPU and GPU aren't factored in. However, this applies for both base Xbox One and X performance.

Basically a look at how very quick and dirty ports on an unfinished/unoptimized OS look. Any level of optimization or additional work put into an XBO-X version should show better results. IE - making use of Scorpio features or refactoring the ESRAM useage for a game. As well, improvements in the OS before launch might improve "dumb" ports as well.

Still fascinating insight into things.

Regards,
SB

 

[Jay]

 

[iroboto]

I wonder if all these recent materials by DF should be compiled and stickier as primer knowledge for discussion of next gen, which would undoubtedly be 4K.

Come to think about it, there's always a bit of good information every once in a while in our tech threads, would have been helpful to mine those values out too so that people don't have to read 50 page threads for developer notes.

Anyway just thinking out loud. This was a neutral piece I thought it explained things well.

 

[AlNom]

8K-16K texture atlases and shadowmaps ought to confuse people.

 

[Cyan]

PlayStation Now PS4 game performance comparison, screengrabs and video.

http://www.eurogamer.net/articles/digitalfoundry-2017-playstation-nows-ps4-game-performance-analysed

things shall evolve but man do I dislike streaming.

 

[Silent_Buddha]

PlayStation Now PS4 game performance comparison, screengrabs and video.

http://www.eurogamer.net/articles/digitalfoundry-2017-playstation-nows-ps4-game-performance-analysed

things shall evolve but man do I dislike streaming.

It always surprises me how much lag people can adjust to. I guess using a controller ameliorates that somewhat as you can't get remotely close to the twitch controls of KB/M (for analog controls, digital controls like button presses or the D-pad will be the same), so it's going to be far less noticeable when you do a quick "right-left-right" analog movements (much slower with a controller) or need to quickly move your aiming reticle onto someone at the edges of your screen or behind you.

But that said, it's still hugely noticeable with controllers when using digital (on/off) style controls in games like platformers. Combine that with the image quality losses (sometimes massive) and I just don't see the value. And that's even before you talk about pricing.

Still, considering there is no other way to access back catalog games on PS4, it's not like users have a choice if they want to play older titles on a current gen PlayStation.

Regards,
SB

 

[zed]

I havent played any online games since quake3 arena but 200msec should be playable, IIRC any time I had under 300msec latency I was happy.
The image degradation is pretty crap though, its certainly an excellent option for hotels or something

 

[Silent_Buddha]

I havent played any online games since quake3 arena but 200msec should be playable, IIRC any time I had under 300msec latency I was happy.
The image degradation is pretty crap though, its certainly an excellent option for hotels or something

I always see that brought up, but in the case of online games back on modems where 200-300 ms lag was considered normal, that's completely different.

On the one hand you have lag between input and output. IE - if you ever hook up a PC to a TV without extremely low input-display lag, you'll immediately notice massive lag between moving your mouse and the mouse pointer moving on the TV. I see this a LOT with TV's that people consider good for console gaming, but when you hook up a PC, the input-output lag is immediately noticeable (both in "game" mode or PC mode). BTW - some modern TVs do much better with this. Also this isn't something that CRT TVs suffered from.

On the other hand there's network lag where what you shoot at isn't really where you shot at and then various methods for accounting for that and attempting to make it fair to people at each end of the connection.

So in the case of Quake 3 Arena, the input-output lag was very low and consistent with local single player games. Had the controls suffered from 200-300 ms lag while the network connection was a steady 10 ms, everyone would have considered the game completely unplayable. But fortunately, it didn't suffer from that. Instead it just suffered from 200-300 ms network lag (for modem users, ADSL users had 100-200 ms usually, while cable users generally had 30-50 ms) and network compensation code, and very low input-display lag which was much easier to compensate for.

Regards,
SB

 

[Cyan]

I havent played any online games since quake3 arena but 200msec should be playable, IIRC any time I had under 300msec latency I was happy.
The image degradation is pretty crap though, its certainly an excellent option for hotels or something

200ms is playable I can tell. That's from the wonderful days of Call of Juarez online on the X360. I think I never cried so much -with laughter- while playing a game. Most of the players were American and very few European, except a guy from the UK and very few others, plus I was on a 1MB ADSL connection.

So, from the slow connection and the overseas distance, the player with the best connection being the host, 200ms was pretty normal, rarely below that. The game was still very playable, but on close fighting both my brother and me lost quite a few highnoon duels, so to speak.

Maybe there was some kind of compensation too, as Silent has mentioned already, but anyways, it wasn't unplayable, just disadvantageous.

 

[Allandor]

200ms is playable I can tell. That's from the wonderful days of Call of Juarez online on the X360. I think I never cried so much -with laughter- while playing a game. Most of the players were American and very few European, except a guy from the UK and very few others, plus I was on a 1MB ADSL connection.

So, from the slow connection and the overseas distance, the player with the best connection being the host, 200ms was pretty normal, rarely below that. The game was still very playable, but on close fighting both my brother and me lost quite a few highnoon duels, so to speak.

Maybe there was some kind of compensation too, as Silent has mentioned already, but anyways, it wasn't unplayable, just disadvantageous.

Well online games are another thing here. You have immediate controller reaction (well a few frames later) to what you do, but the enemies might get a bit jumpy (or you ). But that can be compensated because at least the rest of the world reacts like it should. But if you move your controller and the picture is not moving things are getting really really odd.
I remember times with Unreal when forcing tripple buffer mouse controls got so laggy it was not possible for me to easily select something from the menu because the mouse seems to drift. And that on a fast CRT monitor.
Now another 4 frames is much worse than just tripple buffering. And that on a TV that itself has a "nice" lag.

 

[Scott_Arm]

Yep. What we call network lag and what we call input lag are very different things, even though they're measured in the same units.

 

[Tkumpathenurple]

There's a point they make with which I particularly agree: PSNow needs to provide access to your entire library. It's one thing to provide access to a selection of PS3 and PS4 games, it's quite another to provide access to that selection and everything you've ever bought digitally.

 

[Clukos]

 

[Karamazov]

it's a nice showcase, but i was more impressed to see my childhood anime coming to life in dragon ball fighter Z.

 

[AlNom]

http://www.eurogamer.net/articles/d...itch-minecraft-patch-1080p-confirmed-and-more

 

[HokutoNoKen]

Input lag tested: COD vs Battlefield - which is the most responsive FPS?

http://www.eurogamer.net/articles/digitalfoundry-2017-console-fps-input-lag-tested

-----

-----

 

[function]

Whelp, consoles are now squarely in PC territory - new systems can give increased performance that will offer real world advantages in competitive games.

In a way this is both good and bad - potentially bad for balance and fairness in games without a stable frame rate cap; good because games can continue to push boundaries while offering the option of high and stable levels of performance.

No doubt this situation will occur on MS platforms too. It's the new reality, I guess.