On June 28 2017 07:42 chocorush wrote: How are you so sure that that's how it works? By your logic, there would be states where you are blocking but not moving backwards, because movement IS affected by the input buffer. States in the game are pretty much always tied to the animation frames in games where the engine is tied to the framerate.
And even then, input delay on movement is also terrible in a game like tekken where you need to side step and duck on reaction. Extra delay on attacks is terrible because there are so many strings that you need to interrupt and punish. You've given no reason to explain why this delay makes the core experience better (offline, competitive play), yet continue to bash people for not understanding why the hardships of developers.
That's just how netcode works because it's not possible to have a constant connection, you have to poll regularly and make it "look" real time as much as possible.
Anyway, about the "blocking having priority aspect", I didn't specifically say that's how Tekken works (and I said the same in an earlier post). But it is a common technique used in rollback netcode. Your job with netcode is to give the illusion of being as responsive as possible. When your playing a FPS and you click the mouse, it immediately does a graphic on your screen and shows a bullet travelling quickly, but that is shown on your side before it reaches the server. The server will only notify you on hit (ever notice in some games you hit an enemy and see blood but they don't die?) and those bullets are actually just rays and the bullets are for show.
Well, the only way it's "better" is delay is used in netcode to give an extra buffer to ensure you won't have any harsh auto corrections. But assuming they balance the startup and recovery frames properly, it should be very subtle, and in theory can be quick enough that it's not noticable.
Yes it does affect movement. But if I attack you and you try to evade, how would you differentiate if it was the natural amount of time it takes your character to go through startup frames on the moving animation, or was it the 2 frames that were added from input delay? Experience from earlier Tekken games doesn't answer that question either, because the amount of frames it takes to startup may have changed between iterations.
And developers are not stupid. The games are built upon frame data, and they balance it around that. They know the exact amount of delay in the game. All the abilities are exactly as hard to block as the developers want them to be. The balancers are balancing it at that amount of response time. If you are not able to avoid an attack fast enough, it's not a situation where your problem is the input delay. The problem you have is with the design of the game.
Even in SF5, people complained that the input delay made it so you can't react. Well they patched the input delay to a reasonable amount. And guess what? You still can't react. It's how the game was intentionally designed. And I do think it sucks, Akuma was one of my mains and he used to be built around playing evasively/reactionary, and it's not even possible anymore. My prob isn't with the input delay, it's an issue with the design.
I don't believe I've "bashed" anyone...? I'm just trying to explain how it works, because if you look in to the subject you can see most of the complaints are caused by misconceptions of how everything comes together, so I'm explaining.
The increased input lag affects offline play, not just online play. Do you even play Tekken? The game is almost the exactly the same for the last 15 years, and it's just the last 2 games that have increased the input lag. Please explain how the game is more balanced. This isn't a case of the devs knowing better what's for the game, it's their inability to split netplay considerations from offline play.
On June 28 2017 11:35 chocorush wrote: The increased input lag affects offline play, not just online play. Do you even play Tekken? The game is almost the exactly the same for the last 15 years, and it's just the last 2 games that have increased the input lag. Please explain how the game is more balanced. This isn't a case of the devs knowing better what's for the game, it's their inability to split netplay considerations from offline play.
Yes I play Tekken. I never said it was more balanced? I'm not sure that anyone has even tested if online and offline play has any differences or not. Same as I'm sure noone tested blocking vs attacking (as I'm sure most are not familiar enough with netcode). But it is typical in the industry these days that usually they want the game to feel consistent between online and offline.
The main point I'm actually sharing is what I bolded in the above post. They balanced the frame data for every single attack in the game around those numbers of input lag. Every ability is EXACTLY as hard to avoid as they want it to be. If it feels like the differences are causing issues where you can not avoid attacks like you used to, the reason is not the input lag. The reason you are upset is their design & balancing.
The frame data hasn't changed for most moves since tekken 5. It's so hard to believe that the real reason is because they really thought hard about the balance of every move.
On June 29 2017 09:54 chocorush wrote: The frame data hasn't changed for most moves since tekken 5. It's so hard to believe that the real reason is because they really thought hard about the balance of every move.
I haven't tried to compare them side to side. But if that is truly the case, then apparently they have decided the 1.7 frame difference was not significant enough to affect anything. Because if they truly did believe that it is an issue, their balancing team would have addressed it.
Not to mention, Harada himself has been very upset with the complaints about lag, saying people have been playing the same version for 2 years and have never had an issue until now.
Tekken 7 in arcades runs natively on PC's. It IS the definitive version, because Namco clearly doesn't give a shit about console players. The PS4 is not the version people have been playing for 2 years, and Koreans and Japanese have always complained that something feels wrong the few times they played in America for the pre-launch tournaments.
There is nothing to defend here other than it is possible that there is just something wrong with the PS4 as a piece of hardware, as the codebase is probably almost identical. You don't seem to be interested in analyzing Tekken critically. If you seriously played the game, you would know that the game is chock full of interactions between 16 and 25 frames, and this is a threshold where being 2-4 frames slow actually does have a measurable impact on gameplay. None of this is new. Tekken in it's current form has been mostly unchanged since Tekken 5 as far as general strategy and frame data, and it's bullshit to think that throwing sand in the gears by adding 2-4 frames of lag is good for the game.
What puzzles me, and what I'm trying to figure out is why the PC would have less lag and be more stable than the PS4. I have not seen anyone complain about offline play on the PS4, and supposedly it have the same fixed input lag as online. So if offline play is fine it means that the delays that the PS4 players are experiencing when online must be significantly exceeding the 7 frames of base input delay. Has this number been confirmed?
7 frames is around 120ms, so the lag issues that PS4 users are seeing are probably "ping spikes" of varying length in the 200+ms region. Harada have dismissed it being about network code and I see no reason to not trust him. He may not be a networking expert, but he works in the field and I'm sure he knows many experts who have briefed him. And shouldn't the PC have the same issues if this was about bad code. I doubt the network code is any different for the two versions, and supposedly the battles are run on P2P, so you can't blame it on PS4 running on poor servers.
So unless I've overlooked something, it's not the networking code or anything that has anything to do with the actual application. And then the only thing I can think of is that the PS4 players for some reason have on average, noticeably weaker internet connections. Is there something wrong with the networking components of the PS4? I highly doubt that, but you never know. I haven't heard much from Xbox1 players, but they seem to be having similar problems, which would dispel the faulty component theory, but actually I'm not sure if they have it any better.
The best explanation I can think of is that a huge amount of PS4 players still haven't plugged in their internet cables and are playing on wifi, leading to slower speeds on average (probably not an issue for most), but more importantly the occasional packet loss, which would lead to lag spikes.
I can't get a good PS4 connection with people wired an hour drive away (Madison to Milwaukee). The netplay is not good. To be honest, it's not that much better on PC either, but the problem is more to do with the junk infrastructure we have in America compared to Japan where they probably did most of their testing.
People will complain about the netplay because that's what most console players in america are playing on. The input lag discussion is specifically an issue with offline competitive players, and there is a noticeable difference. The PS4 version is playable, but it is absolutely an inferior version to play on.
On June 29 2017 21:51 chocorush wrote: Tekken 7 in arcades runs natively on PC's. It IS the definitive version, because Namco clearly doesn't give a shit about console players. The PS4 is not the version people have been playing for 2 years, and Koreans and Japanese have always complained that something feels wrong the few times they played in America for the pre-launch tournaments.
There is nothing to defend here other than it is possible that there is just something wrong with the PS4 as a piece of hardware, as the codebase is probably almost identical. You don't seem to be interested in analyzing Tekken critically. If you seriously played the game, you would know that the game is chock full of interactions between 16 and 25 frames, and this is a threshold where being 2-4 frames slow actually does have a measurable impact on gameplay. None of this is new. Tekken in it's current form has been mostly unchanged since Tekken 5 as far as general strategy and frame data, and it's bullshit to think that throwing sand in the gears by adding 2-4 frames of lag is good for the game.
First off, the PS4 version (with the highest increase) has been measured at 1.7 frames difference... not 2-4 or anything like that.
Second, no one has ever said it was better for the game either. Not sure why you keep arguing against comments that were never said. I reiterated my words twice to make the point clear, and both of those times did not say it was "better", I specifically said that it was designed exactly to the specifications that the developers wanted, and everything is exactly as easy or hard to defend as they designed it to be. If it's how the developers designed it to be, how could it be any "better" or "worse"? I said the problem you have is with the "DESIGN' and not the "INPUT LAG".
You keep arguing and trying to shoot down peoples credibility, yet your ruining your own credibility, by arguing points that nobody even said, as well as continuously exaggerating the amount of lag higher & higher - from "128ms ADDITIONAL" when it was actaully 27ms difference, to now exaggerating by saying "2-4 frames of lag" when it's actually been measured as a 1.7 frame increase from previous games.
Where's the honesty, man? Is this an actual discussion, or are u just trolling/arguing for the hell of it?
On June 30 2017 00:21 L1ghtning wrote: What puzzles me, and what I'm trying to figure out is why the PC would have less lag and be more stable than the PS4. I have not seen anyone complain about offline play on the PS4, and supposedly it have the same fixed input lag as online. So if offline play is fine it means that the delays that the PS4 players are experiencing when online must be significantly exceeding the 7 frames of base input delay. Has this number been confirmed?
It's likely just the sacrifices they made for optimizing on PS4 with forced vsync. Similar issues happened on SF5. There's a rumor going around (I believe started by ChrisG) saying that being on PS4 is the reason. But this isn't true, for example the GG Xrd input delay is lower than the competition and is natively on PS4, and uses the same game engine.
Shader versions are a likely culprit as well. If you were to emulate the same exact options as PS4 uses on PC, it would probably give the same amount of input delay. But with the expanded options PC has, you can make the game run smoother. Some may think that using higher quality shaders will make the game run worse, but on newer hardware, higher quality shaders actually run BETTER. This is why with some games, you will get a crap frame rate on low graphics, and/or introduce microstutter, while medium or high might run smoother FPS.
Sadly, T7 didn't learn much from SF5's release. I hoped they would with the 2 year delay. But, japanese game developers don't exactly have the best reputation for being adaptable...
Also, note that no one has even compiled test data on the PC version. It's mostly heresay at this point. The only tests publicized so far were on PS4/PS4Pro/XB1, and while some have described it as "significant difference", the difference was actually only about 1 frame (15ms IIRC). The difference between PS4 version and past PS4 versions of Tekken, is 1.7 frames.
We would need multiple tests on the PC version to know exactly how it matches up - vsync on/off, AA options, etc etc. Similar to SF5.
On June 30 2017 00:38 chocorush wrote: I can't get a good PS4 connection with people wired an hour drive away (Madison to Milwaukee). The netplay is not good. To be honest, it's not that much better on PC either, but the problem is more to do with the junk infrastructure we have in America compared to Japan where they probably did most of their testing.
People will complain about the netplay because that's what most console players in america are playing on. The input lag discussion is specifically an issue with offline competitive players, and there is a noticeable difference. The PS4 version is playable, but it is absolutely an inferior version to play on.
Regarding the PC version primarily, IDK about this one either. I live in Los Angeles, and my primary training partner is Toronto, Canada, and we play regularly on PC without issues. It's really hard to be in the same region and logistically be at a larger distance than we are! If it can hold up for us, it should def be able to hold up for you at that range.
It actually runs better than SF5 does for us. If you are having issues with someone just 1 hour away on a wired connection, you guys need to troubleshoot your connections and/or do some traceroutes to see what the exact problem is.
As long as your ISP isn't providing you shit service, tour results should definitely be reasonable. No reason that you should have over ~45-50 ping at that distance at the maximum.
That said, the game does have a lot of weirdness with connections and NAT issues. It's the first game I've actually had to port forward in probably 3-4 years. Which is pretty sad in 2017 tbh.
I can assist on how to troubleshoot your connection probs if you'd like.
I've explained this before. The 1.7 frames you mentioned is the difference between PS4 and PC with VSync On. Technologically speaking, V-Sync inherently increases input lag, so 2-4 frames of lag is conservative if you are comparing to the PC version with V-Sync off. Depending on the implementation, it can increase it even more. Is there even a point in arguing over this?
You keep saying this is designed. What is the point of designing the game to lag more on PS4 over the PC (The PC version being the original version)? This is almost certainly not introduced by design. Maybe it was an effect that they decided to live with, but you can't possibly make the argument that they balanced the game around this extra lag.
On July 04 2017 07:05 chocorush wrote: I've explained this before. The 1.7 frames you mentioned is the difference between PS4 and PC with VSync On. Technologically speaking, V-Sync inherently increases input lag, so 2-4 frames of lag is conservative if you are comparing to the PC version with V-Sync off. Depending on the implementation, it can increase it even more. Is there even a point in arguing over this?
You keep saying this is designed. What is the point of designing the game to lag more on PS4 over the PC (The PC version being the original version)? This is almost certainly not introduced by design. Maybe it was an effect that they decided to live with, but you can't possibly make the argument that they balanced the game around this extra lag.
You are mistaken (or, based on what I mentioned in the last post, exaggerating to try to support your point). The 1.7 frames difference is between previous Tekken games (Such as TTT2) and T7.
As mentioned earlier, PC version has not had any solid tests publicized. The only testing AT ALL has been from a twitter user @noodalls, who put it at 105.8ms with vsync on. He put the PS4 version at 120.5ms. This is UNDER the number you keep giving out. And this puts the PC version basically on par with XB version. You can see for yourself: https://twitter.com/noodalls
Also, as mentioned, if you emulate the PS4 video settings you would likely have similar outcome, and again as mentioned, there's a good chance it's to do with varied shader versions. Newer shader versions actually run faster (especially if we discuss the ones created in the PS4 era) and keep in mind PS4 does not even run on DirectX, while Windows does.
To be clear (and VERY specific) - according to Noodalis's tests with his setup - including some numbers that were previously not included: TTT2 input lag: 100ms (according to previous tests) T7 PS4 Pro: 120.5ms T7 PS4 : 120ms T7 PS4 @720: 119.9ms T7 PS4 ONLINE: 113.6ms T7 PC: 105.8ms (he does NOT specify options for this one) T7 X1: 104.2ms
So where exactly are these "3-4 added frames" you keep mentioning....? The biggest difference is 16ms. 16ms is slightly UNDER 1 frame!
And btw, the numbers showing that online has a different number, is confirmation for what I described eariler about netcode (which you argued with). This shows that the game does indeed handle inputs differently in different cirumstances.
Again, I ask you the question: Where is the honesty? Are you just trolling/arguing for the hell of it? Because I linked you all these things already, are you are still saying misleading information, and giving numbers that consistently don't match up to those provided as evidence. Which means you are either intentionally exaggerating, or you have been arguing without even looking at any information provided.
When comparing versions of tekken 7 I've consistently been estimating 2-4 frames estimating on the one frame difference you mentioned + the effect of Vsync, which is shockingly... at least 2 frames of lag and usually more. You are choosing to think that I am lying about something.
On July 04 2017 08:37 chocorush wrote: When comparing versions of tekken 7 I've consistently been estimating 2-4 frames estimating on the one frame difference you mentioned + the effect of Vsync, which is shockingly... at least 2 frames of lag and usually more. You are choosing to think that I am lying about something.
Seriously? You have been saying things about 3+ frames "difference" and mentioning in different ways about how that makes it feel worse since the beginning?
If you have established a baseline, 3 frames of lag is when you start noticing it, and feels noticeably worse than 3 frames less
...
Online is going to feel different from offline. It's stupid to tie the two together.
...
Tekken 7 in arcades runs natively on PC's. It IS the definitive version, because Namco clearly doesn't give a shit about console players. The PS4 is not the version people have been playing for 2 years, and Koreans and Japanese have always complained that something feels wrong the few times they played in America for the pre-launch tournaments.
...
this is a threshold where being 2-4 frames slow actually does have a measurable impact on gameplay.
Meanwhile, the arcade version, as you said runs on PC's, but the difference between the truth and what your saying is that there was NOT a difference of 2-4 frames. Now your claiming it was vsync, but this isn't true. Because the arcade version DOES have vsync! All the tournaments run with vsync on, and it's the standard, to a point that vsync off wasn't even tested. Furthermore, the better system you have, the less viable of an option vsync is, because on strong systems the tearing gets so bad it can cause microstuttering. On my laptop shutting vsync works fine, but on my home PC it's unreasonable.
And we can confirm that online functions different from offline anyway, which was another part of your complaints.
Regardless of how you look at it, in the "definitive version" in the arcades is WITH vsync on, and that version is below your "2-4 frames difference", as well as below your "3 frames where you start noticing it"... that's the reality, at the arcade, in tournaments, and its within that difference from past Tekken games, IF YOU PUT VSYNC ON at home. .
I will admit to having a false memory because you quoted a post where I said 2-4 frames of lag, but you literally asked me where I got the 3-4 frame number and I explained how you get there with the V-Sync adjustment. You're being pretty unreasonable here if you're trying to suggest that I'm pulling these numbers out of thin air.
I'm not really sure where you're getting the idea that V-Sync was a set standard on arcade. I've never seen configuration options. V-Sync was definitely not always implemented though in the past. Microstutter isn't really an issue because the framerate is capped at 60fps with or without v-sync on, and arcade cabinets had standardized hardware of generally higher quality than a generic pc+monitor setup.
If top Koreans and Japanese are feeling the difference, I'm inclined to think you're reaching that 3 frame threshold, but I'm willing to concede this point if presented evidence. But to say that the extra lag on PS4 is by design? That just doesn't make sense.
I saw a thread on tekken zaiabatsu forum that stated turning vsync off on pc do reduce the input lag by 2-4 compared to the ps4. At the same time, there has been no "true" test on the pc yet as far i know. The guy that tested the ps4 input lag and other games will probably test it later, he twittered that he had new gear to test stuff with. Yes he did test pc a bit but with lesser gear or something.
Still waiting for a true test of pc, right?
Further, i have also read that the monitor on your pc and your tv on your ps4 can matter as well.
Now for the love of god, stop to talk about this issue any further since what has been going on is going in circles alot of the time. The last 20 comments has been total shit. Go PM if it is so important to accuse someone of lying when he clearly isnt. Btw, go PM anyway unless you have something new and constructive to say.
Input lag its just even 1 frame is smtg, not a lot, if you can get ur game to have only 1 frame of lag thats perfect, always better, but you need a good screen too (a CRT is nice^^). If the game is great and you can play it with no lag, thats perfect! particularly important for a fighting game ofc, frame advantages matter and eh, 1 frame more can make difference between hit or miss right? afaik its known fact that Vsync causes input lag by definition so you want to turn that off for playing games few frames of lag you want to get rid of that