"Part 2 in my Technical Deep dive into Streaming Games as a service Via PSNOW. Covering the performance both in frame-rates, Input variance and loading games, strap on and buckle up."
Last edited:
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Playstation Now: Performance Analysis on PC and Console | How bad is it? Part 2
- Thread starter Alucardx23
- Start date
"Part 2 in my Technical Deep dive into Streaming Games as a service Via PSNOW. Covering the performance both in frame-rates, Input variance and loading games, strap on and buckle up."
That table doesn't make for good reading. You wouldn't buy a TV with >40ms lag so not sure why you'd want to to add yet another 50-80ms lag in addition to your TV lag and game lag. It's depressing.
Can't see xCloud or Stadia being much better though tbh but won't be long to find out. Also depends on your connection and location of course.
Can't see xCloud or Stadia being much better though tbh but won't be long to find out. Also depends on your connection and location of course.
OP
OP
Why is that the case? I used to have a 48ms TV, now I have a 21ms one.
TVs over 40ms feel really laggy. Certainly wouldn't be in the list of recommendations in the gaming display thread, I'm really not saying anything controversial there.
You kind of prove the point here although one night buy a TV at 40ms you made a conscious decision to gain 20ms on your next TV purchase.
By comparison the numbers in the table are an order of magnitude larger, hopefully this can improve and Google and MS can do better.
OP
OP
I think that the point here is that there are a lot of people that either have bad TVs with high input lag or TV's that are capable of low input lag with bad settings that produce high latency. There is room for improvement. I was surprised when I found out that my old TV had 48ms of lag. I played and enjoyed all of the PS360 era games there.
This is too dependent on where someone lives atm so this test basically just applies to NX's location from the closest data center.
I know for a fact U4 wasn't that bad when I tried it the other day because I most certainly would notice near 200ms although it can spike to ~150ms during some moments when the stream lags which is rare for me at me location thankfully.
When I tried Google's streaming beta the experience was pretty much the same as PSNow but with higher resolution and less compression.
Sony definitely need to upgrade the experience soon with competitors at their heels.
I know for a fact U4 wasn't that bad when I tried it the other day because I most certainly would notice near 200ms although it can spike to ~150ms during some moments when the stream lags which is rare for me at me location thankfully.
When I tried Google's streaming beta the experience was pretty much the same as PSNow but with higher resolution and less compression.
Sony definitely need to upgrade the experience soon with competitors at their heels.
But those people are now going to be adding another 50-80ms of lag on top of their already high lag in that case. There's a very obvious tipping point for even the most casual gamer where games no longer feel fun at all, and you end up feeling totally detached from the game. They may not even realise it's due to lag, just that it doesn't feel fun and responsive to play.
This is certainly true. Would be interesting for him to do a ping test to the center to give us an idea of how much lag is physics and how much is the service itself.
OP
OP
An order of magnitude is around 10 times larger, so that term doesn't apply here. The point is that I did played and enjoyed the PS360 era games on that TV and now I have something even better. Going forward we will see that number go down even more, specially once we have 120hz TVs.
The best way to get good data on these streaming services is a publication would need to perform tests from various regions and just report the average experience for each general location which might be pretty tough to do. Ideally the tests would be done weekend nights to account for network congestion too.
edit: he also needs a better tv, just yikes at 60ms+ for Doom.
edit: he also needs a better tv, just yikes at 60ms+ for Doom.
Last edited:
OP
OP
Sure, that is why I shared the table, so people can see the difference. It is obviously the case that on PS Now you will get more latency compared to playing the game locally. Too bad that the PS4 Pro hardware is not available as we could all see how unlocking the framerate on God of War considerably reduces latency. A lot of people have an enjoyable experience on PS Now and going forward with PS5 hardware and support for 120fps, should be even better. Doom on PS Now for example has lower latency that other games you can play and enjoy locally.
Last edited:
Not yet, I think they are still exploring how they would work together.
I hope I get into the xCloud test this month to see what it's like for myself to stream but I suppose we wont get any videos like this until its officially released.
Last edited:
OP
OP
You could do something like that, but as you mentioned with the better TV, people need to be aware of their local setup, because it is just as important. There will be people reporting a bad experience on services like Stadia, because they are playing connected to a router that is plugged into another router, that would have an enjoyable experience otherwise.
OP
OP
Awesome stuff. But wait, are you talking about motion flow here (Frame interpolation) or is your TV capable of receiving and displaying a 120hz signal? Those two cases are very different.
My LG C9 OLED receives and displays a 120hz signal. I play games at 120hz all the time via PC.
Newer Samsung and Sony TVs display 120hz too.
My point is only, 120hz TVs are here already.
OP
OP
Awesome stuff again, that means your TV can display a new frame every 8.3ms, compared to 16.6ms for 60fps or 33.3ms for 30fps. Sure, they have been here for a while now, but they are by no means mainstream. Starting next gen we should see them become a lot more popular.
Best thing I learned thanks to this:
God of War on Pro at 60fps: 59.5 ms latency
God of War on PS4 at 30fps: 138.7 ms latency.
People who own the game on Pro, don't listen to the "stable 30fps above all" fanatics. Play the game with the framerate mode even if it dips. This is not only about framerate.
God of War on Pro at 60fps: 59.5 ms latency
God of War on PS4 at 30fps: 138.7 ms latency.
People who own the game on Pro, don't listen to the "stable 30fps above all" fanatics. Play the game with the framerate mode even if it dips. This is not only about framerate.
In practice using the service is not like that though. It's a case where the numbers may look bad on paper, but doesn't really register so badly in practice. Unless you are OCD'ing and constantly comparing local games to streaming games, or "watching out for it", almost every Now session I've played has not elicited the sentiments above. (Given I have a good Internet connection).
Actually in some games (here Red faction) there is less lag on PSNow compared to PS4.
Which other games? Or are we only talking about games via the infamously laggy PS2 emulator?
In practice it felt awful both times I've tried it via the 14 day trials. I'm actually surprised the lag numbers aren't higher.
So i was planning on getting a month of PSNOW to beat new GOW, and i wonder if it will have lag when i download it to my console?
If you want to be exhaustive and make a definitive statement, you can't ignore the results that don't fit your narrative. It can be interesting to know that this PS2 game (and maybe others) has less lag using PS Now.
OP
OP
Yes, the cool part is that based on his numbers, running the PS4 Pro version in the cloud with unlocked framerate, would give a lower input latency than God of War running on the vanilla PS4 locally.
God of War PS4 Pro at 60fps = 59.5ms of latency
God of War PS4 Vanilla = 138.7ms of latency
God of War PS4 Pro at 60fps + 50ms of network latency = 109.5ms of latency
Exactly. And this is not limited only to God of War, but all games.
This confirms (as if it never been confirmed, people simply refused the idea) that 60fps is factually better than 30fps.
This is a good rule of thumb regardless, imho. An unstable 60fps is better than stable 30fps (unless that unstable 60fps dips all the way down to that 35-40fps no-man's-land).
Red Faction is not a native PS4 game, is my point. But fair enough.
You did say "some games" but failed to give any further examples beyond Red Faction.
Indeed, but this difference is obvious just when switching back and forth between the two modes. Same with games like The Last Of Us Remastered and Hellblade. The 30fps mode feels laggy and as soon as you move to 60fps/unlocked it feels super responsive.
Games feel more responsive on Xbox One X when VRR is enabled too.
That said, it's a big reason why I invested in a gaming PC this year, games feel so much more responsive compared to their console counterparts.
And all this is why I'm a bit grumpy when it comes to game streaming. Seeing that additional lag just at a time where I'm enjoying less lag via my C9 low-latency TV and my gaming PC, well it's a bit depressing seeing gaming potentially going backwards with more lag again.
My brief experience with Now in 2016 wasn't very good.
I was OK with the lag but constantly being told I would be disconnected in 30 seconds was frustrating.
I was OK with the lag but constantly being told I would be disconnected in 30 seconds was frustrating.
If the numbers are correct, this could be due to buffering techniques. Some methods of triple buffering add between 2 and 3 frames of lag, so the input lag might be getting multiplied by a factor of 2-3.
I never actually considered buffering previously. I always thought the difference would be 16ms, but thinking about it, it actually makes sense that the issue would be compounded by buffering and vsync.
OP
OP
The framerate only reflects how many new frames you are getting per second. The input lag (measured from the time the game engine receives the input, until you see a frame that reflects the button you pressed) is dependent on several factors. For example, a double buffered game vs a triple buffered one, the same game compared using these two methods will show more input lag on the triple buffered one, as it will have a higher number of frames in queue that don't reflect the button you pressed. Other factors like how fast is the game logic running will also affect how fast you see the results of a button press, so if the game logic is running at a higher rate than the game framerate you will see something like what happens on God of War. These are just a few examples, but there is a lot more going on the background of a game engine than people think.
Below you can see some examples of how games with the same framerate have very different input lags. This is one of my points when I say that there is a lot of room to grow for cloud gaming on regards to latency. Just see the difference from Call of Duty Infinite to the multiplayer mode on Killzone Shadowfall.
Last edited:
Like I said before and I'll say it again, ps nows streaming is garbage, and that's the main point of the service, to stream ps3 games, because of a lack of bc. Yes I can play and download ps4 games now but they're old ass games that were already super cheap years ago. The best they have is God of War which has been like $19.99. Until they add new releases from 1st party day one or even third party day 1 or soon after release like Devil May Cry 5 on gamepass, it just isn't a good experience you should pay for.
No it'll probably be years before we see that in action if the deal is actually ever made and followed through with, remember they were only talking about a partnership
More than enough people using it without issue and having fun with it. It's fine if it's not for you, but it's most definitely for a lot of others
OP
OP
I fixed it for you.
PlayStation Now has 700,000 subscribers
Last edited:
Yikes at the hot takes.
According to the PSNOW official site it wont let you connect if your latency is greater than 80ms.
I would like to see some actual ping times to PSNOW servers. Hopefully I can snag an address out of my router shortly so I can ping and tracrt.
According to the PSNOW official site it wont let you connect if your latency is greater than 80ms.
I would like to see some actual ping times to PSNOW servers. Hopefully I can snag an address out of my router shortly so I can ping and tracrt.
Exactly my experience as well. And I have tried out some titles like Virtua Fighter 5.
Why not? If the games are the ones they like, can be played locally, and the streaming is fine on their end...
People open threads about GP every week, we could have the same for psnow too now
OP
OP
Don't know why you would say that. Right here you can see examples of people saying that they like the service. We already knew that PS Now had more latency compared to playing the same game locally, the point is that a lot of people can still enjoy it even with a higher latency. Going forward things will continue to improve and we will see even lower latency numbers, that will attract a larger number of people.
I tried PSNow again last week.
I have the same feeling I get trying every one of these services. It's amazing how well it works but once or twice every couple minutes I feel something be slightly off, a movement not quite as fast as I expected or a button press.
Basically the very fundamentals of the game I'm playing break slightly here or there.
A technical achievement, but no thanks.
(wired gigabit fiber)
I have the same feeling I get trying every one of these services. It's amazing how well it works but once or twice every couple minutes I feel something be slightly off, a movement not quite as fast as I expected or a button press.
Basically the very fundamentals of the game I'm playing break slightly here or there.
A technical achievement, but no thanks.
(wired gigabit fiber)