Proof Death Stranding 20% better then DC edition

[Derik-For-Real]

Here is proof, Death Stranding DC has 20% performance loss compared to base game.

The new game fails to deliver the same or better average fps then base game which is concerning.

I did performance logging through amd adrenalin software, nd shared on reddit the data for the game, and a nice guy was interested to tested both game nd send his data to me to compare. Both games were play tested in same region nd play style to perform accurate comparison.

We have very similar results check it out !

My results; https://imgur.com/gallery/srjvCvp

His results; https://imgur.com/gallery/4LuBu0w


my build;
Aorus xtreme x570
AMD 5800x OC 4.6 ghz
CPU cooler, NZXT z73 360m aio
32gb 4x 3600mhz
NVME 2tb samsung evo plus 970
Sapphire nitro plus SE 6800xt (stock)
Cooler Master PSU 1200 watt
Case, Cooler Mastrer c700m


His build;
Intel I5 10600k (stock)
16 gb G skill DDR4
Radeon RX6600 (Hellhound) (stock)
WD Blue M.2 SSD

At this point Kojima Production & 505 needs to be called out for breaking the game performance in DC, the game plays nd looks the same as base, only difference is a few items nd mission, no graphic upgrade, we paid monney and expect to receive similar or better performance then base.

This is really bad, you know it nd I know it, fix it ASAP kindly ! I am shocked that this is even happening, Kojima games usually are solid no complaints, first time complaining.

[Shrimps Prawnson]

Hello everybody, I'm the second benchmark. I honestly thought this was going to be just a problem for OP so I decided to do a comparison just to prove him wrong. Well that bit me in the ass. He was completely right. The DC doesn't utilize the hardware nearly as well, the frame rates are consistently lower and it's a bit disappointing that they are literally 20% lower than the 2 year old base game.

[cyäegha]

i'll admit that this doesn't come as a massive surprise; as i recall, there were benchmarks around DC's release that corroborate similar performance drops, to say nothing of long-standing engine-level issues, such as the unexplained frame rate cuts and jacked-up frame pacing caused by something as pithy as plugging in a controller
anecdotal though it might seem, and not backed up by empirical data of my own, it also feels a ♥♥♥♥♥♥♥ worse than the OG release did, with ghosting up the wazoo
and in saying that, the OG release was no saint

it's not the only major thing kojipro borked in transition: the world-state system that controls road accreditation, among other things, isn't working either
in the original release it would cycle quite frequently, usually upon changing regions and like clockwork upon changing chapters
after 120 hours of DC, i'm still seeing the exact same names on the same segments of road, sometimes from players that haven't touched DC since release

[Surihix]

There is also another longstanding issue which was present even on the OG release and is still there on this version as well.

The game would lock to half of the refresh rate that it was running on during cutscenes. so if you are playing at 60fps, it would lock to 30 during cutscenes.

if you pause and unpause the game, it goes back to 60. its pretty jarring to see this drop occur and I have no idea why no one at Koji Pro noticed this issue earlier when QA testing.

[Derik-For-Real]

Originally posted by Surihix:

There is also another longstanding issue which was present even on the OG release and is still there on this version as well.

The game would lock to half of the refresh rate that it was running on during cutscenes. so if you are playing at 60fps, it would lock to 30 during cutscenes.

if you pause and unpause the game, it goes back to 60. its pretty jarring to see this drop occur and I have no idea why no one at Koji Pro noticed this issue earlier when QA testing.

yes this is damn true, totally agree !

This bug is also shocking, so in Death Stranding when on bike nd your doing a long wheelie, and you want to go right or left, the directional input gets suddenly reversed, left becomes right nd right becomes left, this bug is fixed in Death Stranding DC, but did not deserve a fix in Death Stranding Base. I mean that is pretty bad, nd unethical, you dont need to force people like that to get dlc update.

[Waiwai]

Image is blur on mobile...

Anyway, does that mean both tests got results of CPU UTIL stayed almost equal to base, while GPU UTIL went lower than base?

[Shrimps Prawnson]

Originally posted by Waiwai:

Image is blur on mobile...

Anyway, does that mean both tests got results of CPU UTIL stayed almost equal to base, while GPU UTIL went lower than base?

My less powerful RX6600 ran at higher clocks and utilization on the base version, his RX6800 ran at lower clocks and utilization on the base version. (compared to DC)

Both experienced a 20% drop in FPS on the DC version compared to the base game.

OP was getting 101 fps on base game vs 85 on DC (4k)
I was getting 107 fps on base game vs 89 on DC (1080p)

[FTC]

Yeah, my rig exceeds the recommended and I am struggling to maintain 50FPS on LOW SETTINGS.

AMD Ryzen 5 (6 core, 3.6 GHz unboosted)
Geforce GTX 1650 Super
16 GB Ram.

Yeah, it's low-mid budget PC, but I have ran every single game besides this with no problems.

[Waiwai]

Originally posted by Shrimps Prawnson:

My less powerful RX6600 ran at higher clocks and utilization on the base version, his RX6800 ran at lower clocks and utilization on the base version. (compared to DC)

Both experienced a 20% drop in FPS on the DC version compared to the base game.

OP was getting 101 fps on base game vs 85 on DC (4k)
I was getting 107 fps on base game vs 89 on DC (1080p)

I am on PC now. Image is clearer now.

About the test, the result seems clear, GPU UTIL hadn't reached 100%, meaning GPU work is dragged down by something, and only one question left.

Did both tests run in offline and no structure environments?

Because after I did the max 120 FPS measurement in DS:DC, it looks like it is something related to how many structures surrounding you, rather than just merely performance.

[Derik-For-Real]

Originally posted by Waiwai:

Originally posted by Shrimps Prawnson:

My less powerful RX6600 ran at higher clocks and utilization on the base version, his RX6800 ran at lower clocks and utilization on the base version. (compared to DC)

Both experienced a 20% drop in FPS on the DC version compared to the base game.

OP was getting 101 fps on base game vs 85 on DC (4k)
I was getting 107 fps on base game vs 89 on DC (1080p)

I am on PC now. Image is clearer now.

About the test, the result seems clear, GPU UTIL hadn't reached 100%, meaning GPU work is dragged down by something, and only one question left.

Did both tests run in offline and no structure environments?

Because after I did the max 120 FPS measurement in DS:DC, it looks like it is something related to how many structures surrounding you, rather than just merely performance.

No, I dont think it has any thing to do with structures or gpu utilisation at some degree. For example my gpu utilisation is different then the other guy, but we loste about 20 % average fps both which is allot, nd translates to poor optimization in game engine nd hardware.

About your structure claim, how did you arive at this conclusion ? I mean, why do you think structure cause fps loss, how you test this. Lets assume structures generally causes overal fps dips in both games, but does it change the fact that one of the games runs better then the other, I dont think so. The play test we did for both games was exactly the same, to ensure both game run nd played in same scenario to have a clear understanding of the outcome.
This

Also If you want to help, provide real data with performance logging for both games, We used amd adrenalin software. You can use same layout I made.

[Chorizo_Joe]

Yeah kinda bummed I progressed so much in DC before I noticed the performance drop. Now I have to choose to either lose 15 hours of progress in DC and go back to standard edition or just deal with the crappier performance for the remainder of the story. Guess it's on me though, I should have read the forums before forking over 10 bucks for a downgrade.

[Waiwai]

DC vs base version comparsion, location-specific only comparison

Results from Nvidia performance overlay

DC version:
https://steamcommunity.com/sharedfiles/filedetails/?id=2810836774

Base version:
https://steamcommunity.com/sharedfiles/filedetails/?id=2810836858

Hardware:
AMD Ryzen 7 5800X
Gigabyte AORUS B550 motherboard PRO AC
Gigabyte AORUS Nvidia GeForce RTX 2070S 8G
2 x 16 GB DDR4 RAM 3200MHz
M.2 NvMe, OS disk and DS:DC installation
SATA3 SSD, base installation

OS:
Windows 11 Pro 21H2 (22000.675)

In-game settings:
Very high graphics settings
DLSS:Quality
Motion blur off
Depth of field off
Vsync on
Max FPS 120
Shared network, all shared

DLSS library:
DC 2.3.7 (installed in current DC version)
Base 2.1.19 --> 2.3.7 (copied from DC version)

=====
* Last updated, DC: 15th May 2022
** Last updated, base: 21th May 2022
*** Only 7 seven locations had measured in the base version

East of Timefall Farm, the Torii (the screenshots):
119 - 120 FPS, 35% CPU, 75% GPU
(Base: 119 - 120 FPS, 35% CPU, 75% GPU)

Weather Station:
88 - 90 FPS, 36% CPU, 61% GPU
(Base: 115 - 119 FPS, 44% GPU, 80% GPU)

Distribution center of Lake Knot:
77 - 85 FPS, 44% CPU, 61% GPU
(Base: 98 - 102 FPS, 48% CPU, 71% GPU)

Crafter:
83 - 85 FPS, 40% CPU, 62% GPU
(Base: 106 - 111 FPS, 49% CPU, 75% GPU)

Incienerator, central region:
91 - 94 FPS, 44% CPU, 62% GPU
(Base: 119 - 120 FPS, 41% CPU, 81% GPU)

West coast, Chiral Relay:
106 - 120 FPS, 39% CPU, 74% GPU
(Base: 119 - 120 FPS, 32% CPU, 81% GPU)

Captial Knot:
106 - 112 FPS, 39% CPU, 72% GPU
(Base: 114 - 120 FPS, 39% CPU, 81% GPU)

Cutscreens:
60 FPS, 20% CPU, 40% GPU

Racetrack:
90 - 95 FPS, 42% CPU, 60% GPU

Timefall Farm:
75 - 85 FPS, 46% CPU, 66% GPU

Cosplayer:
85 - 88 FPS, 34% CPU, 55% GPU

Mama's Lab:
81 - 92 FPS, 44% CPU, 58% GPU

South Knot:
88 - 95 FPS, 42% CPU, 60% GPU

Film Director:
79 - 86 FPS, 40% CPU , 63% GPU

Middle Knot:
85 - 93 FPS, 41% CPU, 65% GPU

Middle Knot Crater coast:
82 - 86 FPS, 45% CPU, 63% GPU

Roadside factory ruin:
75 - 80 FPS, 42% CPU, 60% GPU

Elder:
82 - 86 FPS, 43% CPU, 63% GPU

Lake Knot:
86 - 93 FPS, 42% CPU, 62% GPU

Distribution center of Mountain Knot:
92 - 95 FPS, 41% CPU, 63% GPU

Mountain Knot:
85 - 90 FPS, 48% CPU, 63% GPU

Heartman's Lab:
86 - 89 FPS, 39% CPU, 64% GPU

West coast mountain:
95 - 101 FPS, 41% CPU, 70% GPU

Private Room (Central region):
114 - 120 FPS, 34% CPU, 70% GPU

Edge Knot:
100 - 105 FPS, 40% CPU, 65% GPU

Private Room (Eastern region):
119 - 120 FPS, 28% CPU, 69% GPU

Graveyard on ruined factory:
77 - 82 FPS, 40% CPU, 65% GPU

Distribution center of Captial Knot:
94 - 100 FPS, 40% CPU, 68% GPU

Port Knot:
98 - 104 FPS, 40% CPU, 70% GPU

Map/cargo/info menu:
119 - 120 FPS, 29% CPU, 59% GPU

[Waiwai]

The reason I said about structures was because there is heavily different between places with lesser structures and place with many structures.

The DS:DC definitely has something on its own, otherwise it won't have this huge different in FPS from location to location, and it is not something requiring you to compare with the base version to see the difference.

I don't consider base version and DC version comparison is valid and makes a lot of sense, before you can get more variables under controlled.

Rather than asking why DC and base have huge difference, better than getting the FPS difference solved in DC version first.

Check the location of screenshots, the Torii top in games, eyes toward South Knot. Both got 120 FPS and same load, if there is really 20% performance difference, it should reflect on here too.

Your average tests have too many variables. I cannot consider them to be valid tests.

Check the GPU utilisation, when FPS goes down, the GPU utilisation also goes down as it should be. This means something caused bottleneck to GPU, causing GPU cannot work at maximum.

I never got 80% GPU utilisation in DC version, but I got it easily to over 80% GPU utilisation in the base version, therefore I usually get better FPS on the base version.

The base version is netiher better nor worse than DC version. They perform same, when you eliminate many uncontrollable variables in the comparison.

That is why I said about structure (enviroment). If you ignore structure number difference, test result will always vary. Not to mention many players moved from the base to DC now, making the shared network has large difference.

[VisciousFishes]

Clean save on the DC version or a transferred save from the OG cut? I hear (not 100% sure if this is right) that Clean saves have less issues with FPS than those with saves transferred from the OG release.

I too experienced drops between DC and OG versions. But I transferred my save. I haven't bothered trying a new save yet, but I haven't completed the OG once yet, so I am not really minded to start a new save yet again to test the theory.

[Waiwai]

Originally posted by jakethebusker:

Clean save on the DC version or a transferred save from the OG cut? I hear (not 100% sure if this is right) that Clean saves have less issues with FPS than those with saves transferred from the OG release.

I too experienced drops between DC and OG versions. But I transferred my save. I haven't bothered trying a new save yet, but I haven't completed the OG once yet, so I am not really minded to start a new save yet again to test the theory.

My DC save is a brand new start, not transferred from the original.

[VisciousFishes]

Originally posted by Waiwai:

Originally posted by jakethebusker:

Clean save on the DC version or a transferred save from the OG cut? I hear (not 100% sure if this is right) that Clean saves have less issues with FPS than those with saves transferred from the OG release.

I too experienced drops between DC and OG versions. But I transferred my save. I haven't bothered trying a new save yet, but I haven't completed the OG once yet, so I am not really minded to start a new save yet again to test the theory.

My DC save is a brand new start, not transferred from the original.

Okay that's interesting. I will have a look at a new save and do a comparison.