Digital Foundry Article Technical Discussion [2021]

Status
Not open for further replies.
it does not align everytime that's why it's weird and should be taken with caution, because if as some say here in photo mode only GPU power talks, why is that in the corridor scene the difference is negligible ?
Yes that scene is especially strange. You could assume (from the PC benchmarks) that the RT pipeline limits here. But than we would see a greater divide between the two.
Than we can just think of a strange situations, where one thing leads to the next which makes the whole RT calculation ineffective and the system is just able to squeeze a bit more out of it.

There is, reflecting glass, that reflects the windows from behind (and the wall), than the glass is partially transparent with an alpha-texture (I guess) behind that (for the grid), where you look through and see at least another light-source (red light). Everything comes together in this scene:
- multiple reflections over a large area
- multiple light sources (of different color)
- shadows from multiple light-sources (and reflections of those)
- possible alpha texturing

And like the PC version + 6800 XT shows us, this is just to much for AMDs RT solution here. Even at half RT res the much smaller chip in PS5 & xsx can just get on par.
 
Clearly, hitman 3 isnt an outlier. Hence the discussion being mainly Control now. The difference is 44% for the ones playing and using the game as a benchmark.
again, stop with this 44% difference as its not a case, 1) xsx has drops to 40ties when ps5 keep 60 2) using df miami scene and comparing to 5700xt xsx has ~1.12x (had to chek it) advantage over ps5
 
Honestly, both platforms are performing quite well for such a demanding game. PS5 has bit of an edge in actual gameplay (minus the stuttering as well), and XBSX has bit of an edge in the uncapped RT photomode demo run-through. In general, neither system are really outperforming each other in Control.
 
Your claiming agendas and whatever. The first post in this thread clearly shows this isnt welcome. Attacking people escalates things. Keep it to yourself, not here.
Seems you need to read the OP and re-watch the DF video...

Clearly, hitman 3 isnt an outlier. Hence the discussion being mainly Control now. The difference is 44% for the ones playing and using the game as a benchmark.
...and understand the definition of outliner.

We spent most of the pre-launch with you slagging off the PS5 '9.2TF' BS, and now a couple of crumbs seem to imply there's a bigger gap than the actual reality. Yes, it's called confirmation bias.

I on the other hand will openly accept data which falls within the expected 'ball park' (let's say 15-25%) but I (feel rightly) question anything that falls outside that. DF sampled 25 areas and threw away 4 due to (like the Hitman 3 game) there not being good enough data to confirm anything. They state that the average is (unsurprisingly) within my aforementioned guidelines of 'good enough to accept as truth'.

You however are more interested in the bigger gaps to prove your pre launch bias. If I were to just talk about the incredibly close performance deltas (of which there are significantly more) and try to use that to prove the PS5 is as powerful as the XSX I'd expect to be called out too.

You are even putting words into Alexs mouth suggesting that Control with RT on XSX could be capped at 50FPS - he never said that (happy to be proven wrong), he talked about the potential of lowering the resolution to 1080p and unlocking to 60FPS, but the photo mode is just showing what the GPU is capable of - nothing more, the problem is there are other parts at play and something is holding the XSX back in gameplay mode.
 
Last edited:
  • Like
Reactions: snc
Seems you need to read the OP and re-watch the DF video...


...and understand the definition of outliner.

We spent most of the pre-launch with you slagging off the PS5 '9.2TF' BS, and now a couple of crumbs seem to imply there's a bigger gap than the actual reality. Yes, it's called confirmation bias.

I on the other hand will openly accept data which falls within the expected 'ball park' (let's say 15-25% but (I feel rightly) question anything that falls outside that. DF sampled 25 areas and threw away 4 due to (like the Hitman 3 game) there not being good enough data to confirm anything. They state that the average is (unsurprisingly) within my aforementioned guidelines of 'good enough to accept as truth'.

You however are more interested in the bigger gaps to prove your pre launch bias. If I were to just talk about the incredibly close performance deltas (or which there are significantly more) and try to use that to prove the PS5 is as powerful as the XSX I'd expect to be called out too.

You are even putting words into Alexs mouth suggesting that Control with RT on XSX could be capped at 50FPS - he never said that (happy to be proven wrong), he talked about the potential of lowering the resolution to 1080p and unlocking to 60FPS, but the photo mode is just showing what the GPU is capable of - nothing more, the problem is there are other parts at play and something is holding the XSX back in gameplay mode.

Who is this aimed at? I have never mentioned none of these things.
 
Seems you need to read the OP and re-watch the DF video...


...and understand the definition of outliner.

We spent most of the pre-launch with you slagging off the PS5 '9.2TF' BS, and now a couple of crumbs seem to imply there's a bigger gap than the actual reality. Yes, it's called confirmation bias.

I on the other hand will openly accept data which falls within the expected 'ball park' (let's say 15-25%) but I (feel rightly) question anything that falls outside that. DF sampled 25 areas and threw away 4 due to (like the Hitman 3 game) there not being good enough data to confirm anything. They state that the average is (unsurprisingly) within my aforementioned guidelines of 'good enough to accept as truth'.

You however are more interested in the bigger gaps to prove your pre launch bias. If I were to just talk about the incredibly close performance deltas (or which there are significantly more) and try to use that to prove the PS5 is as powerful as the XSX I'd expect to be called out too.

You are even putting words into Alexs mouth suggesting that Control with RT on XSX could be capped at 50FPS - he never said that (happy to be proven wrong), he talked about the potential of lowering the resolution to 1080p and unlocking to 60FPS, but the photo mode is just showing what the GPU is capable of - nothing more, the problem is there are other parts at play and something is holding the XSX back in gameplay mode.
But control works better during gameplay on ps5 it drops frames on series x just the same plus has stutters it literally drops to 20s during raytracing mode on series x, I think this photo mode benchmark doesn't prove anything. Alex said it could run 1080p 60 with raytracing on series x just by looking at photo mode but that's a weird conclusion since they perform similar during gameplay and the ps5 even smoother,

so judging by gameplay it's the ps5 that can smoothly play 1080 60 with raytracing.


I mean here's series x dropping to 28 during gameplay it's just similar to ps5.
 

Attachments

  • IMG_1076.PNG
    IMG_1076.PNG
    832.3 KB · Views: 24
But control works better during gameplay on ps5 it drops frames on series x just the same plus has stutters it literally drops to 20s during raytracing mode on series x, I think this photo mode benchmark doesn't prove anything. Alex said it could run 1080p 60 with raytracing on series x just by looking at photo mode but that's a weird conclusion since they perform similar during gameplay and the ps5 even smoother,

so judging by gameplay it's the ps5 that can smoothly play 1080 60 with raytracing.


I mean here's series x dropping to 28 during gameplay it's just similar to ps5.

Well yes (and no) - I'm actually agreeing that Control plays better (or at least as good when stutter is fixed) on PS5 in all modes - it's a wash.

I believe Alex has drawn the conclusion that because of the photo mode showing such a good performance gain (well, to the expected level of gap) that 'if dropped to 1080p XSX could run RT capped at 60'.

I can see why he has concluded this from the photo mode, but in the same breath we have the 2 consoles in gameplay neck and neck. So my conclusion (and I'm no expert - Alex knows a lot more than me!) is that something in the gameplay element of the game is dragging the XSX back...as such if both were dropped to 1080p 60 with RT I'd personally expect a similar performance.

From memory (in this tread) I believe Alex said the texture streaming is off in photo mode and others have said the CPU is not really utilised at all in photo mode...I would ass-u-me if those were being utilised the graphs would level off to a similar figure (from my ass) of 30-50?. We've seen examples where the photo mode varies wildly (from 32 to 60+) so you can imagine such a varying FPS would be jarring and hence the 30FPS cap.

I guess the main question is, has the PS5 got something that in game boosts performance (eg IO/cache scrubbers) or is the XSX being held back due to immature dev kits (or whatever the technical term is)...I'd guess the latter, this photo mode (for me) cements the XSX potential and the paper specs show it shouldn't be any more bottle-necked than PS5 (unless I missed something).
 
Well yes (and no) - I'm actually agreeing that Control plays better (or at least as good when stutter is fixed) on PS5 in all modes - it's a wash.

I believe Alex has drawn the conclusion that because of the photo mode showing such a good performance gain (well, to the expected level of gap) that 'if dropped to 1080p XSX could run RT capped at 60'.

I can see why he has concluded this from the photo mode, but in the same breath we have the 2 consoles in gameplay neck and neck. So my conclusion (and I'm no expert - Alex knows a lot more than me!) is that something in the gameplay element of the game is dragging the XSX back...as such if both were dropped to 1080p 60 with RT I'd personally expect a similar performance.

From memory (in this tread) I believe Alex said the texture streaming is off in photo mode and others have said the CPU is not really utilised at all in photo mode...I would ass-u-me if those were being utilised the graphs would level off to a similar figure (from my ass) of 30-50?. We've seen examples where the photo mode varies wildly (from 32 to 60+) so you can imagine such a varying FPS would be jarring and hence the 30FPS cap.

I guess the main question is, has the PS5 got something that in game boosts performance (eg IO/cache scrubbers) or is the XSX being held back due to immature dev kits (or whatever the technical term is)...I'd guess the latter, this photo mode (for me) cements the XSX potential and the paper specs show it shouldn't be any more bottle-necked than PS5 (unless I missed something).
If they've are both necklaces and neck during gameplay then they can both play at 1080 60 with raytracing similarly. The photo mode doesn't make any sense as a measure of performance what Alex should have said is the series x would lock 1080p 60 during photo mode instead.

Cause both consoles drop to 20s in raytracing mode, If the unlocked frame rate was tested during gameplay then it would have made sense but until then it doesnt. It's like looking at a map screen and measuring performance then claiming console x is stronger than b. Photo modes have always behaved different to gameplay since ever some racing games run 30fps during photo mode and 60fps during gameplay and nobody has ever used photo mode as a benchmark.
 
Alex said it could run 1080p 60 with raytracing on series x just by looking at photo mode but that's a weird conclusion since they perform similar during gameplay and the ps5 even smoother,
We can't draw such a conclusion, because the game has a capped framerate. That's precicely why Alex used the Photomode, because it takes the CPU out of the picture and shows the GPU power.

Who knows this game could be running at a higher average if the FPS was unlocked on Xbox than PS5 with the caveat of having heavier drops.
 
Last edited:
But control works better during gameplay on ps5 it drops frames on series x just the same plus has stutters it literally drops to 20s during raytracing mode on series x, I think this photo mode benchmark doesn't prove anything. Alex said it could run 1080p 60 with raytracing on series x just by looking at photo mode but that's a weird conclusion since they perform similar during gameplay and the ps5 even smoother,

so judging by gameplay it's the ps5 that can smoothly play 1080 60 with raytracing.


I mean here's series x dropping to 28 during gameplay it's just similar to ps5.

If you look at the frame time graph on the photo you attached, you can see that it's going from a steady 33ms, to several frames at 100 ms, back to a steady 33ms. That's not normal frame to frame variance (and the game is triple buffered no less), so it's almost certainly hitching caused by some other element of the game / development environment and not the GPU's ability to render regular, in-gameplay frames.

This kind of hitching will be present in the engine on Xbox whatever your target frame rate is. It's just an issue that the Xbox versions of the game have at this time, for some reason. Even PC seems to fare better in this regard.

It doesn't tell you how the renderer will perform above 30 fps at either 1440p or 1080p in RT mode. For the other 99.999% of frames that aren't stuck at 100ms for some obscure reason, the best way we have of inferring RT mode performance is to push as many RT frames as we can with the frame cap off. If we do that, it's clear that the XSX is a bit better at maintaining higher frame rates, in the general case, in RT quality mode.

In gameplay, I would expect some of the heavier full screen alpha effects to hit the PS5 a little less hard, so there's probably that to take into account, but then again at a lower 1080p XSX would probably be pretty okay in that regard.

But those sharp, sudden drops to 100ms followed a full recovery are probably going to be there whatever frame rate you target on XSX / XSS. There's no reason to let 3 or 4 frames every few tens of thousands of frames make you chose a particular, lower frame rate especially when doing so won't get rid of the problem anyway.
 
If you look at the frame time graph on the photo you attached, you can see that it's going from a steady 33ms, to several frames at 100 ms, back to a steady 33ms. That's not normal frame to frame variance (and the game is triple buffered no less), so it's almost certainly hitching caused by some other element of the game / development environment and not the GPU's ability to render regular, in-gameplay frames.

This kind of hitching will be present in the engine on Xbox whatever your target frame rate is. It's just an issue that the Xbox versions of the game have at this time, for some reason. Even PC seems to fare better in this regard.

It doesn't tell you how the renderer will perform above 30 fps at either 1440p or 1080p in RT mode. For the other 99.999% of frames that aren't stuck at 100ms for some obscure reason, the best way we have of inferring RT mode performance is to push as many RT frames as we can with the frame cap off. If we do that, it's clear that the XSX is a bit better at maintaining higher frame rates, in the general case, in RT quality mode.

In gameplay, I would expect some of the heavier full screen alpha effects to hit the PS5 a little less hard, so there's probably that to take into account, but then again at a lower 1080p XSX would probably be pretty okay in that regard.

But those sharp, sudden drops to 100ms followed a full recovery are probably going to be there whatever frame rate you target on XSX / XSS. There's no reason to let 3 or 4 frames every few tens of thousands of frames make you chose a particular, lower frame rate especially when doing so won't get rid of the problem anyway.
Devil's advocate, and I have no knowledge of how Remedy's engine in a general sense or Control in a specific sense works when it comes to framerate limits or frame pacing, but you get better frame pacing if you run vsync'd and scan out only on even or odd frames. The reason games have bad frame pacing when locked to 30fps is because they don't do this. So if you miss a frame @33ms, and you *should* be updating in times that are multiples of 33.333. So if you take more than 66.666ms you would get a frame at 100ms (or 99.999999999999.... to be more accurate). If you are running at refresh rate, or targeting it anyway, you won't get frame time spikes that are unnaturaly high. Another solution is to not lock your output to odd or even frames, but then you may have situations where you output 30 frames in a second but have 66ms frames tossed into the mix and have the perception of a lower frame rate. Or turn off vsync and deal with torn frames, and probably bad frame pacing as well.
 
Devil's advocate, and I have no knowledge of how Remedy's engine in a general sense or Control in a specific sense works when it comes to framerate limits or frame pacing, but you get better frame pacing if you run vsync'd and scan out only on even or odd frames. The reason games have bad frame pacing when locked to 30fps is because they don't do this. So if you miss a frame @33ms, and you *should* be updating in times that are multiples of 33.333. So if you take more than 66.666ms you would get a frame at 100ms (or 99.999999999999.... to be more accurate). If you are running at refresh rate, or targeting it anyway, you won't get frame time spikes that are unnaturaly high. Another solution is to not lock your output to odd or even frames, but then you may have situations where you output 30 frames in a second but have 66ms frames tossed into the mix and have the perception of a lower frame rate. Or turn off vsync and deal with torn frames, and probably bad frame pacing as well.

Control does seem to have the odd frame pacing issue in the normal sense, but it also has those transitional bits where the game can go from 16 ms (or 33 in graphics mode) down to 100 ms and also see a kind of "micropause". That's unique to Xbox - even the higher overhead PC seems to avoid this. I've got the odd idea about what it might be, but like you I don't know either. Fun to speculate though! :D

Fun fact that blew my tiny mind: 9.99999 recurring = 10. I was older than I should have been when I found that out. :(
 
Well yes (and no) - I'm actually agreeing that Control plays better (or at least as good when stutter is fixed) on PS5 in all modes - it's a wash.

I believe Alex has drawn the conclusion that because of the photo mode showing such a good performance gain (well, to the expected level of gap) that 'if dropped to 1080p XSX could run RT capped at 60'.

I can see why he has concluded this from the photo mode, but in the same breath we have the 2 consoles in gameplay neck and neck. So my conclusion (and I'm no expert - Alex knows a lot more than me!) is that something in the gameplay element of the game is dragging the XSX back...as such if both were dropped to 1080p 60 with RT I'd personally expect a similar performance.

From memory (in this tread) I believe Alex said the texture streaming is off in photo mode and others have said the CPU is not really utilised at all in photo mode...I would ass-u-me if those were being utilised the graphs would level off to a similar figure (from my ass) of 30-50?. We've seen examples where the photo mode varies wildly (from 32 to 60+) so you can imagine such a varying FPS would be jarring and hence the 30FPS cap.

I guess the main question is, has the PS5 got something that in game boosts performance (eg IO/cache scrubbers) or is the XSX being held back due to immature dev kits (or whatever the technical term is)...I'd guess the latter, this photo mode (for me) cements the XSX potential and the paper specs show it shouldn't be any more bottle-necked than PS5 (unless I missed something).

Does this mean that the Xbox Series X can have a much better looking cutscenes later on vs the PS5? I assume that plenty of high quality cinematic real time rendered scenes don't use the CPU as much?
 
Does this mean that the Xbox Series X can have a much better looking cutscenes later on vs the PS5? I assume that plenty of high quality cinematic real time rendered scenes don't use the CPU as much?

It means it can have marginally better looking realtime cutscenes or framerate, maybe, depending on workload.:)
 
We can't draw such a conclusion, because the game has a capped framerate. That's precicely why Alex used the Photomode, because it takes the CPU out of the picture and shows the GPU power.

Who knows this game could be running at a higher average if the FPS was unlocked on Xbox than PS5 with the caveat of having heavier drops.
How can it run on a higher average of the frame rates where unlocked it it already drops frames the locked 60 to 50s in both machines during gameplay. Photomode can't be used as a benchmark because it doesn't represent gameplay and as u said it takes the cpu out of the picture so what was the point. You can't draw conclusions on photomode when the cpu is out of the picture because you need both a cpu and gpu to play

If you look at the frame time graph on the photo you attached, you can see that it's going from a steady 33ms, to several frames at 100 ms, back to a steady 33ms. That's not normal frame to frame variance (and the game is triple buffered no less), so it's almost certainly hitching caused by some other element of the game / development environment and not the GPU's ability to render regular, in-gameplay frames.

This kind of hitching will be present in the engine on Xbox whatever your target frame rate is. It's just an issue that the Xbox versions of the game have at this time, for some reason. Even PC seems to fare better in this regard.

It doesn't tell you how the renderer will perform above 30 fps at either 1440p or 1080p in RT mode. For the other 99.999% of frames that aren't stuck at 100ms for some obscure reason, the best way we have of inferring RT mode performance is to push as many RT frames as we can with the frame cap off. If we do that, it's clear that the XSX is a bit better at maintaining higher frame rates, in the general case, in RT quality mode.

In gameplay, I would expect some of the heavier full screen alpha effects to hit the PS5 a little less hard, so there's probably that to take into account, but then again at a lower 1080p XSX would probably be pretty okay in that regard.

But those sharp, sudden drops to 100ms followed a full recovery are probably going to be there whatever frame rate you target on XSX / XSS. There's no reason to let 3 or 4 frames every few tens of thousands of frames make you chose a particular, lower frame rate especially when doing so won't get rid of the problem anyway.
Df said it themselves that both consoles drop frames during gameplay particularly during combat and the Xbox drops even more because of the irregular stutters that happen, so saying that the series x could do 1080 60 with raytracing better than a ps5 just because of photo mode is ridiculous since both consoles perform the same during gameplay already. What difference would 1080p make. He should have instead said series x will hold above 60fps during phototmode at 1080p.
 
Status
Not open for further replies.
Back
Top