Current Generation Games Analysis Technical Discussion [2020-2021] [XBSX|S, PS5, PC]

[Shortbread]

-The main character model appeared without some details at the beginning of the game and this problem has been fixed.
- The quality of lighting on PS4 has been decreased.
- Some reflections have been fixed / improved in all versions.
- Average resolution has increased on all platforms.
- Textures have improved their quality, while shadows are kept low quality.

- Performance has not undergone notable changes. PS4 performs better, but it is still insufficient.
- The PS5 version has slightly worse performance than the 1.04 version. The same has happened with SeriesX.
- In version 1.05 there is a problem with the interface on PS4 / Pro. It takes too long to load some items.
- Pop-in is still unacceptable, considerably higher on PS4 / Pro.
- The number of NPCs corresponding to each version remains the same, even on PS5.

 

[Deleted member 11852]

Yes, you are correct that updates are not required to play Xbox One games on Series consoles.

Thought as much. So this is perhaps something unique Microsoft have done for Cyberpunk.

 

[Jay]

Thought as much. So this is perhaps something unique Microsoft have done for Cyberpunk.

Doubt it.

I think it's more likely that CP actually does a check during startup and mandates the patch.
No different to how games check for updates when they start, apart from this time it checks the system profile against the current version.
Games have access to smart delivery api.

 

[Shortbread]

This video benchmarks the new Cyberpunk 2077 1.05 patch and the hex edits (with 1.04), but we also talk about the memory pool allocation misconceptions spread by the community.

A lot of people insisted that the memory_pool_allocation CSV actually did something and insisted we retest everything for this change, but fortunately, CDPR saved us all the time and said that this file wasn't connected to anything. As usual, you must exercise control and care when benchmarking components, otherwise it's easy to be misled by bad data. In the case of the CPU Hex Edit, there are benefits in some cases, but not to the degree most of the internet seemed to expect -- it doesn't universally benefit Ryzen, just some of the lower-end Ryzen CPUs on CPU-bound settings. The new 1.05 patch also adds this SMT balancing to CPUs at 6 cores and fewer. Either way, these benchmarks update the numbers for the 3300X and 5600X. The 5900X will be unchanged, as are the Intel CPUs.

TIMESTAMPS
00:00 - SERIOUS (imaginary) GAINS!
03:03 - New Cyberpunk 2077 Patch 1.05 Notes
06:20 - Hex Edit Benchmarks at 1440p/Medium (AMD R5 5600X, R3 3300X)
07:20 - When an Improvement Can Happen
08:00 - AMD R3 3300X Improvements at 1440p/Medium
08:20 - CPU Load Balancing in Cyberpunk 2077 (R5 5600X)
10:31 - Frametime Improvement in AMD R3 3300X at 1080p Low
11:58 - Thread Activity on R3 3300X
12:55 - 1080p Low CPU Benchmarks and Bigger Differences
14:30 - Serious FPS Consistency Issues in Cyberpunk
16:38 - Cyberpunk Update 1.05 CPU Benchmark Updates
19:39 - Conclusion

 

[BRiT]

Thought as much. So this is perhaps something unique Microsoft have done for Cyberpunk.

Doubt it.

I think it's more likely that CP actually does a check during startup and mandates the patch.
No different to how games check for updates when they start, apart from this time it checks the system profile against the current version.
Games have access to smart delivery api.

Its hard to know for sure because of the incomplete reporting on what the exact message was and how it looked. Was it an OS Message or was it a Game Message.

As far as we know, the devs were aware of a huge patch they were working on when the game went gold and perhaps it was also Xdk Gen9 Aware, so upon startup on nextgen console they could have simply displayed a message to install the patch. They may not have even needed to use any SmartDelivery Apis. Did it behave the same on base Xbox One consoles too? Or what about the One X console?

If we wanted to dig further, we'd have to have access to physical media of the game and do our own controlled experiment and report the findings. With a focus on various states of console and network connectivity (wired cable), like offline mode and not connected, offline mode and connected, online mode and not connected, and online mode and connected.

 

[Deleted member 11852]

Its hard to know for sure because of the incomplete reporting on what the exact message was and how it looked. Was it an OS Message or was it a Game Message.

It's in the video - start at 1:50.

 

[Shortbread]

Important: The minimum resolution of SeriesX is 1800p, not 1080p as it says in the video.

- Better reflections in Series X thanks to SSR
- The resolutions oscillate the same as in the original game, however it maintains higher numbers because this DLC does not have such open environments.
- The Oldgen versions suffer some drop in framerate in some specific load areas. Nothing serious.
- This DLC also makes use of the improved lighting and shading system in Series S|X.
- Thanks to the more linear scenarios, a more visually spectacular result is achieved with respect to the base game.
- The Coalition is, without a doubt, one of the best developers when it comes to taking advantage of the Unreal Engine.

 

[BRiT]

It's in the video - start at 1:50.

That kind of behavior seems odd, like they started to install from physical media while being in online mode and declined the patch when prompted what to do, so now the system knows there's a required patch. If I had CP2077 Physical Media I'd do a more controlled examination of the issue.

I think devs can mark patches as being optional or required for play. There were a couple times a game had a patch and the system asked me if I wanted to download the patch or continue and play offline. This was back in the One X days.

 

[Shortbread]

- Pop-in on ps4 is still quite aggressive, but has had a slight improvement on PS4 Pro.
- The quality of some shadows on PS5 has improved slightly, but is still below Xbox.
- Liger performance improvement on PS4. There appear to be no changes to PS4 Pro or PS5.
- I have not suffered crashes or freezes during recording.
- The menu suffered a bug in 1.05, it has already been fixed.
- The number of npcs has seemed to increase slightly in some areas on PS5, but still far from Series S | X.
- I have noticed an improvement in vehicle collisions.

 

[Shortbread]

- The tearing has been slightly reduced on Xbox One X, on Xbox One it is still high.
- For the first time, I have not had any crashes while recording.
- I have noticed an improvement in vehicle collisions.
- The framerate has undergone slight (very slight) changes on Xbox One.
- Freezes on Xbox One seem to have disappeared.
- Series X has reduced its NPCS in some areas of the game, probably a bug.
- The framerate of Series X suffered certain problems in patch 1.05, now it seems to have the same stability as 1.04 again.

 

[Shortbread]

Destiny 2 frame rate test showing the framerate/fps on PS5 and Xbox Series X|S. This test is using the 60hz mode which is the only mode available outside of PvP. Subscribe for more gaming tech analysis: http://bit.ly/VGTechSubscribe

Timestamps:
0:00 - Xbox Series X
4:58 - Xbox Series S
9:55 - PS5

The versions tested were 1.003.000 on PS5 and 0.1.24550.0 on Xbox Series X|S. Default video options were used for this video. In the scenes tested, the game still seems to usually run at 60fps when using the higher fov settings http://bit.ly/3pm3VqY

PS5 and Xbox Series X use a dynamic resolution with the highest resolution found being 3840x2160 and the lowest resolution found being approximately 2346x2160 on both consoles. PS5 and Xbox Series X often render at a resolution of 3840x2160, but in the scenes tested, the resolution on PS5 was below 3840x2160 more often than Xbox Series X was.

Xbox Series S uses a dynamic resolution with the highest resolution found being 1920x1080 and the lowest resolution found being approximately 1536x1080. Drops in resolution below 1920x1080 on Xbox Series S seem to be uncommon.

Performance can vary somewhat and the scenes here aren't exactly like-for-like which should be kept in mind when comparing the performance of the consoles here.

All three consoles use lower resolution buffers for certain effects.

The real-time cutscenes are letterboxed and run at 30fps http://bit.ly/2WN8mis

Stats: http://bit.ly/3hg66tv
Frames Pixel Counted: http://bit.ly/2WKvzlg

Watch in 4K and 60fps for the best experience.

FYI: Some commented that the 3.0.1 patch was available for both next-generation systems, so I'm not a 100% sure if VG Tech patched for the latest version or correct version. And some are mentioning that the graphics in the video look no better than the prior editions.

Yeah, it seems the 1.03 is the standard PS4 version and possibly the 0.1.2 patch is the standard XBO version.

Those who are already playing on PS5 got the 1.03 update but it was only 1.8gb, they were downloading the standard patch for the PS4 version and not the full PS5 upgrade.

 

[VGA]

FYI: Some commented that the 3.0.1 patch was available for both next-generation systems, so I'm not a 100% sure if VG Tech patched for the latest version or correct version. And some are mentioning that the graphics in the video look no better than the prior editions.

Yeah, it seems the 1.03 is the standard PS4 version and possibly the 0.1.2 patch is the standard XBO version.

1.003.000 is the latest version on PS5 as of right now. Note that this number is taken from the PS5 dashboard and not in game.

 

[BillSpencer]

FYI: Some commented that the 3.0.1 patch was available for both next-generation systems, so I'm not a 100% sure if VG Tech patched for the latest version or correct version. And some are mentioning that the graphics in the video look no better than the prior editions.

Yeah, it seems the 1.03 is the standard PS4 version and possibly the 0.1.2 patch is the standard XBO version.

Looks like Xbox is showing some muscle.
I remember back in the day Bungie producing technically identical PS4 and Xbox One versions for Destiny 1, this was because Xbox was lead platform

 

[Shortbread]

With another year over, we come to that time when I allocate some semblance of order to the proceedings.

We were not short of contenders and quality in 2020 but I pick my Top 7 and here I attempt to demonstrate why.

 

[Shortbread]

- I have restarted the game every time I changed the display mode to avoid possible errors.
- Higher resolution of textures on PS5.
- Now vegetation is only affected in S Series depending on the display mode.
- Rain lighting appears to be different on PS5.

- Resolution is virtually identical on PS5 / Series X.
- The reflections on some standard models are different between PS5 and Xbox.
- Less anisotropic filtering in Series S.
- Series S has less NPCS. Neither version has npcs in FPS mode.

- Reflections on some vehicles are different in each version.
- The FPS has been impossible for me to measure since I have not made this video in my usual workplace.
- The ambient occlusion in PS5 seems somewhat more noticeable in some points such as vegetation.

 

[pjbliverpool]

I assume that video vastly undersells the PlayStation games (except Sackboy which looked great) because based on that, they really, really didn't deserve those positions.

 

[PSman1700]

I assume that video vastly undersells the PlayStation games (except Sackboy which looked great) because based on that, they really, really didn't deserve those positions.

Its NXgamer, thanks DF for existing. Carry on.

 

[Shortbread]

I assume that video vastly undersells the PlayStation games (except Sackboy which looked great) because based on that, they really, really didn't deserve those positions.

I view one's personal choices and/or opinions like any others... I don't take them personal.

That being said, my top 10 choices of the year based strictly on visuals would be...
1. Microsoft Flight Simulator 2020 (PC)
2. Demon Souls Remake (PS5)
3. Cyberpunk 2077 (PC)
4. Ghost of Tsushima (PS4)
5. Call of Duty: Black Ops Cold War (PC)
6. Watch Dogs: Legion (PC)
7. Spider-Man Miles Morales (PS5)
8. Doom Eternal (PC)
9. The Last of Us Part II (PS4)
10. Cloudpunk (PC)

 

[pjbliverpool]

I view one's personal choices and/or opinions like any others... I don't take them personally.

That being said, my top 10 choices of the year based strictly on visuals would be...
1. Microsoft Flight Simulator 2020 (PC)
2. Demon Souls Remake (PS5)
3. Cyberpunk 2077 (PC)
4. Ghost of Tsushima (PS4)
5. Call of Duty: Black Ops Cold War (PC)
6. Watch Dogs: Legion (PC)
7. Spider-Man Miles Morales (PS5)
8. Doom Eternal (PC)
9. The Last of Us Part II (PS4)
10. Cloudpunk (PC)

I know a lot of people rate Demons Souls very highly and I can only assume there's something major about it that doesn't come across in the online videos. Obviously I've not played it so I'm not in a great position to comment.

 

[Shortbread]

I know a lot of people rate Demons Souls very highly and I can only assume there's something major about it that doesn't come across in the online videos. Obviously I've not played it so I'm not in a great position to comment.

This can be said about most games. Some gamers believe CP 2077 NPCs geometry is no better than GTA V... or that SM: Miles Morales RT is low quality... or that Flight Simulator 2020 realistic looks diminishes when traveling (flying) closer to the ground surface. So yes, subjective and personal choices...