Digital Foundry Article Technical Discussion [2022]

Status
Not open for further replies.
Somehow i doubt that a GTX750Ti/7870/HDD is the baseline on pc now, or becoming that way.



Not the whole picture i think, RT is a visual feature that is generally more dominant in the pc space, head-started on pc years before the consoles launched. Same for hw/ai accelerated reconstruction-tech. Fast-er loading did exist on the pc aswell in select titles (doom eternal had no loading on pc at all).
you have to look at it as a feature perspective and not a power perspective.
DX12 cards start at 7750 series levels of cards. Anything before that there is no support. A lot of older powerful cards that can run dota, and counterstrike, Wow etc all exist and do well with those older titles, and so moving them to DX12 just makes their systems obsolete.

I think you're looking at just visual compute power. That's not the same thing as leveraging the APIs to their fullest extent.
 
How would you accomplish that?
?Use this same procedure, but install the game to the external Seagate card.
?Move that card over to a Series S and play from there.

I don't have either console, so I don't know how they work with their installations.


Nevermind, this was dumb.
I forgot that you still needed a disc in a drive for the game to verify and play.
Although I do wonder what would happen if you tried this. Does Series S say "insert disc"?
 
Nevermind, this was dumb.
I forgot that you still needed a disc in a drive for the game to verify and play.
Although I do wonder what would happen if you tried this. Does Series S say "insert disc"?
I think Xbox allows for digital licenses to validate disc installs, so this could be possible if you could launch offline (install and delete the XSS version online, then go offline and attach the external drive?) and/or cancel any SmartDelivery update.
 
How would you accomplish that?
?Use this same procedure, but install the game to the external Seagate card.
?Move that card over to a Series S and play from there.

I don't have either console, so I don't know how they work with their installations.
Same way. Install the game onto am external HDD of a Xbox one, than take that drive and plug it into a series x console in offline mode.
That might work, but I guess we get more or less the same result: A 4k image with lower res assets but cranked up shadow resolution etc.
 
I think Xbox allows for digital licenses to validate disc installs, so this could be possible if you could launch offline (install and delete the XSS version online, then go offline and attach the external drive?) and/or cancel any SmartDelivery update.
This has been true from Xbox One. There are physical games I own that get added to Gamepass and I no longer have to put in the disc (until they get removed from GP, of course). You can also install games using the cell phone Xbox app that you don't own, from anywhere in the world (this is an assumption as I haven't tested this from all places in the world myself), to your Xbox at home, and it will install the full game and updates, but won't play it unless you put the disc in, or buy it digitally of course. The license is interchangeable between physical and digital versions in my experience.

How this would work in practice doing the Elden Ring switcheroo, I don't know. Also, I know DF have a theory about the game ramping up quality settings based on some unseen factor, and that the game visually resembles the One X version of the game, but could it be that the version on the disc isn't just the One X version. It just has to be Gen9Aware to have exclusive modes, right? So while the PS4Pro version of PS5 is hard limited to PS4Pro settings regardless, perhaps the OneX version on Series isn't just the OneX version, hence why some settings are ramping higher. I would be curious what the file info says for the unpatched disc version says under Console Type vs what the fully patched version says, and if you could run the disc version from a HDD or external SSD, and how the load times compare.
 
That might work, but I guess we get more or less the same result: A 4k image with lower res assets but cranked up shadow resolution etc.

But on a Series S, if it was reading an unpatched version of the game from the Xbox One disc, it should run the Xbox One S settings. Isn't that the default behavior of the Series S unless there is additional code (a downloaded upgrade) present?
 
Is it possible to have a res counter like the fps counter to see at what res each version is running when there are framedrops ?
 
this one seems to hit where we think they should hit.
ps5 running holding framerate when rop heavy, and increased shadow resolution
xsx running ahead on resolution as a result of additional compute bandwidth, but falling apart when rop heavy and decreased shadow quality.
This does not really make sense. Lower framerate but higher resolution in those situations. Resolution should also be ROP bound in those situations.
I guess the shadow quality is more a bug than anything else. Same quality than on Series S ...

Like so often, we don't really see a hardware difference, just a software-based difference.
 
this one seems to hit where we think they should hit.
ps5 running holding framerate when rop heavy, and increased shadow resolution
xsx running ahead on resolution as a result of additional compute bandwidth, but falling apart when rop heavy and decreased shadow quality.
Do they even say the the XSX usually runs at higher resolution? Minimum resolution could be lower on PS5 due to Pro version min resolution. We already saw this happen in those games. This could also explain the PS5 having a better fps if res can drop lower.
 
Do they even say the the XSX usually runs at higher resolution? Minimum resolution could be lower on PS5 due to Pro version min resolution. We already saw this happen in those games. This could also explain the PS5 having a better fps if res can drop lower.
Tighter bounds usually means a higher average resolution. Not always, but with the way things are done today, they measure the lowest resolution when the frame rate is lowest.
 
Tighter bounds usually means a higher average resolution. Not always, but with the way things are done today, they measure the lowest resolution when the frame rate is lowest.
Like you said, not always. We already saw this. Anyways we'll know more about that with VGTech precise findings like: min, max, usual and frame by frame resolution comparison.

EDIT: from the article they actually never looked for a min res on XSX. We have to wait for more data here. Apparently they stressed tested PS5, not XSX.
 
Last edited:
This has been true from Xbox One. There are physical games I own that get added to Gamepass and I no longer have to put in the disc (until they get removed from GP, of course). You can also install games using the cell phone Xbox app that you don't own, from anywhere in the world (this is an assumption as I haven't tested this from all places in the world myself), to your Xbox at home, and it will install the full game and updates, but won't play it unless you put the disc in, or buy it digitally of course. The license is interchangeable between physical and digital versions in my experience.

How this would work in practice doing the Elden Ring switcheroo, I don't know. Also, I know DF have a theory about the game ramping up quality settings based on some unseen factor, and that the game visually resembles the One X version of the game, but could it be that the version on the disc isn't just the One X version. It just has to be Gen9Aware to have exclusive modes, right? So while the PS4Pro version of PS5 is hard limited to PS4Pro settings regardless, perhaps the OneX version on Series isn't just the OneX version, hence why some settings are ramping higher. I would be curious what the file info says for the unpatched disc version says under Console Type vs what the fully patched version says, and if you could run the disc version from a HDD or external SSD, and how the load times compare.

They could have designed the Xbox one version to be Xbox series aware so that people updating to the Series X or S aren't necessarily forced to buy the Series port just to enjoy higher resolution or better IQ. Just a good will gesture thats not terribly expensive to add since they are supporting both Xbox consoles as well as the PC.
 
Whenever they can get their game running acceptably under DX12. Developers should have a specific reason to use lower level APIs given the extra complexity and knowledge required. Using it just because has not been a benefit to anyone. It's made games worse.

Making games is not an academic exercise it’s a business. Devs don’t have the resources to build out a DX12 game and test across a bevy of hardware and use cases. All for the purpose of experimentation.

The only way to become verse in well performing DX12 games is to put out DX12 games and learn from the experience. PCs aren’t consoles. There are 100s of configurations of hardware running a plethora software which cannot be replicated in a test environment. At least not one that whose costs would be tenable for devs.

Devs are experiencing growing pains with DX12 but if you expect game devs to push hardware and stay on a cutting edge path then the current situation is unavoidable.
 
Like you said, not always. We already saw this. Anyways we'll know more about that with VGTech precise findings like: min, max, usual and frame by frame resolution comparison.

EDIT: from the article they actually never looked for a min res on XSX. We have to wait for more data here. Apparently they stressed tested PS5, not XSX.

Not sure if it was clear in the article - I stress tested both systems. PS5 drops resolution a little more readily (ex: 1800p when Series X is at 2016p) but is at a full 4K plenty of the time.
 
This does not really make sense. Lower framerate but higher resolution in those situations. Resolution should also be ROP bound in those situations.
I guess the shadow quality is more a bug than anything else. Same quality than on Series S ...

Like so often, we don't really see a hardware difference, just a software-based difference.
Depends where the bottleneck is. The ROPS are double pumped for colour meeting most of the requirements of keeping up. But depth and stencil rops are 1/2 of ps5.

 
Status
Not open for further replies.
Back
Top