Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MrMacRight posted a YT review of the game for iPhone and noted all the crashing. I was more curious about the smeared vaseline look (that basically is the calling car of upscaling these days, well ignoring DLSS). I still can't figure out why MetalFX in these instances looks so poor for them supposedly using iPhone hardware for upscaling. I wonder if @leman or @Homy have any ideas as to why.

Digital Foundry said the internal res of RE 4 on iPhone is roughly 300p which they suspected being upscaled to 720p so that could be the case with DS too. Here is a first impression of DS with different settings at 1440p on M1 Max.

 
  • Like
Reactions: Irishman
Digital Foundry said the internal res of RE 4 on iPhone is roughly 300p which they suspected being upscaled to 720p so that could be the case with DS too. Here is a first impression of DS with different settings at 1440p on M1 Max.

I guess I am concerned that developers are using MetalFX as a poor crutch to get games working on macOS/iOS and not even at native resolution. It would be one thing if these games ran using MetalFX to scale to the panels native resolution (because then they can use higher quality textures) but in a lot of cases that isn't true (more so on iOS than macOS).
 
  • Like
Reactions: Homy and Irishman
This is sad. Embracer cancels the next Deus Ex game that's been in development for two years which makes Eidos Montreal to lay off 97 people. Was hoping it would come to Mac.

Clearly they were scared of more dystopian predictions coming true.
 
  • Like
  • Haha
Reactions: goro123 and Homy
I guess I am concerned that developers are using MetalFX as a poor crutch to get games working on macOS/iOS and not even at native resolution. It would be one thing if these games ran using MetalFX to scale to the panels native resolution (because then they can use higher quality textures) but in a lot of cases that isn't true (more so on iOS than macOS).

Yeah but, Death Stranding runs at 60 fps on M2, I believe.
 
No clue, it was from the same video from MrMacRight, comparing the iPad version with the MacOS version.
Ah, I watched that video, but missed that part (or at least I vaguely remember him showing a Mac in terms of showing the difference in image quality)
 
  • Like
Reactions: Irishman
Ah, I watched that video, but missed that part (or at least I vaguely remember him showing a Mac in terms of showing the difference in image quality)

Found another one from MacProTips here:
Comparing an M1 Pro and an M3 Max. The latter was tested at MetalFX OFF, 1440p, and Max graphics! Just buttery smooth!!

Getting between 80 and 120 fps!
 
  • Like
Reactions: bousozoku
Found another one from MacProTips here:
Comparing an M1 Pro and an M3 Max. The latter was tested at MetalFX OFF, 1440p, and Max graphics! Just buttery smooth!!

Getting between 80 and 120 fps!
Yeah it is like the MetalFX implementation is broken or something because the M3 Max seemed to get the same framerate with it on or off.
 
  • Like
Reactions: Irishman
Yeah it is like the MetalFX implementation is broken or something because the M3 Max seemed to get the same framerate with it on or off.

I didn’t see that anywhere in the video. Where did you see it specifically stated? You may be inferring that part.
 
I didn’t see that anywhere in the video. Where did you see it specifically stated? You may be inferring that part.
Earlier they showed DS on the M3M with MetalFX on, and it averaged around 80-100 frames. Then they showed it with it off (or at least claimed it was off) and it showed the same 80-100 frames (but in a different part of the game).

So either it wasn't actually off, that part of the game shown was tougher than the parts shown earlier or MetalFX doesn't work if the game is CPU bound (as in the game isn't GPU bound and thus isn't getting any more performance with MetalFX on).
 
  • Sad
Reactions: Irishman
Earlier they showed DS on the M3M with MetalFX on, and it averaged around 80-100 frames. Then they showed it with it off (or at least claimed it was off) and it showed the same 80-100 frames (but in a different part of the game).

So either it wasn't actually off, that part of the game shown was tougher than the parts shown earlier or MetalFX doesn't work if the game is CPU bound (as in the game isn't GPU bound and thus isn't getting any more performance with MetalFX on).

Yes, something's not right but hard to compare when it's not the same sequence in the game. In both cases the frame rate goes down to around 70 fps at most but the quality settings aren't the same. DS has four levels: Low, Medium, Default and Very High. In the last part the settings are at max according to the video. In the first part Model detail and Memory for Streaming are at Default, not Very High. AA is also Off instead of TAA.

That means that either MetalFX was also ON in the last part or the first part was more demanding since it does have the same frame rate with MetalFX with lower settings. Another detail is that MetalFX is on Ultra so perhaps it doesn’t matter much at Ultra. Someone made a comparison between different MetalFX settings and the difference between Off and Ultra is 10 fps at 4K.

So according to Notebookcheck the performance of M3 Max 40c without MetalFX is like RTX 4060 Ti 16GB, 3070, 3080 Ti laptop or 2080 Ti at 4K Max.
 
Last edited:
  • Like
Reactions: Irishman
Yes, something's not right but hard to compare when it's not the same sequence in the game. In both cases the frame rate goes down to around 70 fps at most but the quality settings aren't the same. DS has four levels: Low, Medium, Default and Very High. In the last part the settings are at max according to the video. In the first part Model detail and Memory for Streaming are at Default, not Very High. AA is also Off instead of TAA.

That means that either MetalFX was also ON in the last part or the first part was more demanding since it does have the same frame rate with MetalFX with lower settings. Another detail is that MetalFX is on Ultra so perhaps it doesn’t matter much at Ultra. Someone made a comparison between different MetalFX settings and the difference between Off and Ultra is 10 fps at 4K.

So according to Notebookcheck the performance of M3 Max 40c without MetalFX is like RTX 4060 Ti 16GB, 3070, 3080 Ti laptop or 2080 Ti at 4K Max.

Sounds more like notebookcheck AND Reddit got it wrong.

Thankfully.
 
A rerelease of a remake less than a year after the initial release.

Sometimes I wonder about this industry.

The Gold Edition would include the main game and all DLCs with a lower price than if you would buy them separatly so I see no harm in such a release. It also means their done with the game and not going to release more DLCs so why wait with a Gold Edition?
 
Sounds more like notebookcheck AND Reddit got it wrong.

Thankfully.

Not sure what you mean. Who did get what wrong? In Tsai’s video the he gets 50 fps at least at 4K Max without MetalFX with M3 Max 40c.

In the same area MacProTips gets 70 fps at least at 1440p with Model Detail and Memory for Streaming at Default (High), AA Off and MetalFX Ultra with Temporal Upscaling on M3 Max 30c.

In the second area MacProTips gets 70 fps at least at 1440p Max without MetalFX on M3 Max 30c.

In the ”start area” the Reddit user gets 70 fps 4K Max without MetalFX on M3 Max 40c.

Notebookcheck has no numbers for DS on Mac so I don’t know how they could be wrong. We know neither which part of the game they benchmarked. I made that comparison based on the Mac numbers. If we use Tsai’s 50 fps it’s comparable to anything between RTX 2070 desktop and 3060, 3070 Ti/3080 laptop and 2080/3060 Ti.
 
  • Like
Reactions: Irishman
Not sure what you mean. Who did get what wrong? In Tsai’s video the he gets 50 fps at least at 4K Max without MetalFX with M3 Max 40c.

In the same area MacProTips gets 70 fps at least at 1440p with Model Detail and Memory for Streaming at Default (High), AA Off and MetalFX Ultra with Temporal Upscaling on M3 Max 30c.

In the second area MacProTips gets 70 fps at least at 1440p Max without MetalFX on M3 Max 30c.

In the ”start area” the Reddit user gets 70 fps 4K Max without MetalFX on M3 Max 40c.

Notebookcheck has no numbers for DS on Mac so I don’t know how they could be wrong. We know neither which part of the game they benchmarked. I made that comparison based on the Mac numbers. If we use Tsai’s 50 fps it’s comparable to anything between RTX 2070 desktop and 3060, 3070 Ti/3080 laptop and 2080/3060 Ti.

Well, maybe it’s more a matter of poor sources, and a 10 fps difference at 4k with MetalFX on and off is also not reflected in any other sources.
 
Well, maybe it’s more a matter of poor sources, and a 10 fps difference at 4k with MetalFX on and off is also not reflected in any other sources.

Still not sure about your point. Are you saying it should run faster, slower or what? The sources are not poor. They’re testing different things under different circumstances. Tsai does test MetalFX On/Off and the difference is about 7 fps so it’s comparable to the result on Reddit. Anyway I think we can conclude that DS runs fine on most Macs.
 
  • Haha
Reactions: Irishman
Still not sure about your point. Are you saying it should run faster, slower or what? The sources are not poor. They’re testing different things under different circumstances. Tsai does test MetalFX On/Off and the difference is about 7 fps so it’s comparable to the result on Reddit. Anyway I think we can conclude that DS runs fine on most Macs.
On higher end Mac’s this game seems CPU limited. Wonder when Apple will add frame generation, maybe this WWDC?
 
On higher end Mac’s this game seems CPU limited. Wonder when Apple will add frame generation, maybe this WWDC?

Maybe but it could also be because they're using MetalFX Ultra in the videos and the 10 fps difference is normal at the highest quality at 4K. In the first video I posted last Wednesday you can see there are four different settings: Performance, Balanced, Quality and Ultra. The difference between Ultra and Balanced is about 20 fps there. Here is another comparison between MetalFX Temporal and Spatial with the same settings.

I wish there was a demo like Resident Evil 4 so we could test it ourselves.
 
  • Sad
Reactions: Irishman
Resident EVil 4 Gold Edition is coming Feb 9. It's like the Deluxe Edtion with Separate Ways included (the base game, Separate Ways story DLC, and previous DLC Packs). Don't think it's coming to MAS since Village doesn't have Gold Edition there.

 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.